Retail EU Study

The economic impact of
modern retail on
choice and
innovation in the
EU food sector
final report
Competition
The economic impact of modern
retail on choice and innovation in
the EU food sector
Final report
European Commission
The economic impact of modern retail on choice and

innovation in the EU food sector
Final report
Report by:
EY
Cambridge Econometrics Ltd.
Arcadia International
September 2014
The information and views set out in this study are those of the authors and do not necessarily
reflect the official opinion of the Commission. The Commission does not guarantee the accuracy of
the data included in this study. Neither the Commission nor any person acting on the Commission’s
behalf may be held responsible for the use which may be made of the information contained
therein.
More information on the European Union is available on the Internet: http://europa.eu

More information about Competition Policy is available on: http://ec.europa.eu/competition
Cataloguing data can be found at the end of this publication.
Luxembourg: Publications Office of the European Union, 2014
© European Union, 2014

Reproduction is authorised provided the source is acknowledged.
ISBN 978-92-79-40324-8
doi: 10.2763/77405
4
The economic impact of modern retail on choice and innovation in the EU food sector
Table of Contents
1. Executive summary ......................................................................................21
1.1. Objectives of the study ............................................................................21
1.2. Methodology ..........................................................................................21
1.3. Background: a strong development of modern retail across Europe, a new
landscape for EU consumers ................................................................................25
1.4. Evolution of choice: choice offered to consumers has notably increased in a
majority of MS...................................................................................................26
1.5. Evolution of innovation: a steady stream of innovation was made available to
EU consumers; however the number of innovations declined between 2008 and 2012 28
1.6. Evolution of concentration: concentration of retailers and suppliers showed
different trends depending on the MS, the product category and the level of analysis
(local or national) ..............................................................................................31
1.7. Conclusions regarding factors driving choice ..............................................33
1.8. Conclusions regarding factors driving innovation.........................................36
2. Reminder of objectives .................................................................................39
2.1. Objectives of the study ............................................................................39
2.2. Motivations behind study .........................................................................39
2.3. Structure of the final report .....................................................................41
2.4. Limitations of this report .........................................................................42
2.5. Different tasks of the study ......................................................................43
3. Background of the study ...............................................................................45
3.1. Europe retail sector in brief ......................................................................45
3.2. Recent evolutions in the grocery retail sector in the EU ...............................48
3.3. Macro evolutions impacting the grocery retail sector in the EU .....................56
4. Scope, measures and methodology ................................................................65
4.1. Selection of MS ......................................................................................65
4.2. Selection of time period ...........................................................................70
4.3. Selection of 105 consumer shopping areas (CSAs) ......................................72
4.4. Representativeness of the sample that was selected ...................................75
4.5. Selection of product categories .................................................................77
4.6. Method for data extrapolation (supermarkets and discounters) .....................78
4.7. Measures defined for the study .................................................................80
4.8. Database construction .............................................................................91
5. Descriptive statistics from data analysis ..........................................................95
5.1. Introduction ...........................................................................................95
5.2. Question 1: How has choice in the EU food sector evolved over time and across
MS? 95
5.3. Question 2: How has innovation in the EU food sector evolved over time and
across MS? ..................................................................................................... 113
5
5.4. Question 3: How have the a priori drivers of retail and supplier concentration
evolved over time and across MS? ..................................................................... 129
5.5. Question 4: How have the other a priori drivers of choice and innovation
evolved over time and across MS? ..................................................................... 151
6. Econometric analysis scope and methodology ................................................ 183
6.1. General specification ............................................................................. 183
6.2. Econometric issues ............................................................................... 184
6.3. Economic importance and statistical significance....................................... 185
7. Characteristics of the data set and its implications for the econometric analysis . 186
7.1. Dataset construction and availability ....................................................... 186
7.2. Sample selection .................................................................................. 186
7.3. The scope of the data set used in the econometric analysis ........................ 187
7.4. Implications of the sample selection process ............................................ 194
8. Main features of choice and innovation indicators in the sample data set ........... 197
8.1. Choice ................................................................................................. 197
8.2. Innovation ........................................................................................... 200
9. Results of the econometric analysis .............................................................. 204
9.1. Introduction ......................................................................................... 204
9.2. Summary of results for drivers ............................................................... 204
9.3. Retail concentration .............................................................................. 212
9.4. Supplier concentration .......................................................................... 218
9.5. Measure of imbalance between retailers and suppliers at national level ....... 222
9.6. Private labels ....................................................................................... 224
9.7. Product category turnover ..................................................................... 227
9.8. General economic drivers: unemployment ............................................... 230
9.9. General economic drivers: GDP per capita/Retail business expectations ....... 231
9.10. General economic drivers: population and population density ..................... 235
9.11. Shop characteristics: size, format and the opening of a new shop in the same
local area ........................................................................................................ 236
10. Accounting for changes over time in selected shops ........................................ 238
10.1. Examples of the impacts of the drivers in five shops ................................. 239
10.2. Examples of the impacts of the drivers in five CSAs .................................. 247
11. Annexes .................................................................................................... 253
11.1. Annex A: Illustration of © Mintel GNPD launch types................................. 253
11.2. Annex B: Descriptive statistics ............................................................... 280
11.3. Annex C: Design of the econometric analysis ........................................... 378
11.4. Annex D: The data sets ......................................................................... 385
11.5. Annex E: Econometric estimation issues .................................................. 385
11.6. Annex F: Results of the econometric analysis ........................................... 387
6
7
Figures and tables
Figure 1: Evolution of the market share of modern retail compared to total edible
grocery market (2000 - 2011)................................................................................47
Figure 2 : Evolution of the European food retail (in number of outlets) by type of shop
(2000-2011) ........................................................................................................49
Figure 3: Evolution of the European food retail sales area (in thousands of m²) by type of
shop (2000-2011) ................................................................................................49
Figure 4: Evolution of the combined market shares of the top 5 retailers C(5) per MS
(2000 - 2011) ......................................................................................................50
Figure 4: Domestic share of EU grocery sales for top ten retail groups ........................55
Figure 5: Compound annual growth rate of GDP per capita in EU 27 ...........................57
Figure 6: Average final consumption expenditure of households for food and non-
alcoholic beverages across EU 27 (% of the total expenditure) ...................................57
Figure 7: Compound annual growth in the share of final consumption expenditure of
households of food and non-alcoholic beverages per MS (% CAGR) ............................58
Figure 8 : Proportion of key household expenditures compared to the total household
expenditure for EU-27 (2003-2011) ........................................................................59
Figure 9: Edible grocery proportion (in %) of total retail sales in EU 27 between 2004 and
2012 ...................................................................................................................59
Figure 10: Compound annual growth rate in EU retail markets (2006 to 2012) .............60
Figure 11: Compound annual growth in unemployment rate (in %) across EU 27 between
2004 and 2012 .....................................................................................................61
Figure 12: Compound annual growth in percentage of population at risk of poverty after
social transfers (2004-2012) ..................................................................................62
Figure 13 : Top 5 major impact factors on grocery purchase choice in 2011 .................62
Figure 14: Representativeness of sample vs EU27 population by standard of living
categories............................................................................................................77
Figure 15: Representativeness of sample vs EU27 population by type of living zone ......77
Figure 16: Database construction – per MS and at consolidated level ..........................93
Figure 17: 2004-2012 data set: Total number of shops in Member State (local level) -
average CAGR across all modern retail shop types (source: EY analysis based on ©
Nielsen Trade Dimensions) ....................................................................................96
Figure 18: 2008-2012 sample: Total number of shops in CSAs by Member State (local
level) - average CAGR across all modern retail shop types (source: EY analysis based on
© Nielsen Trade Dimensions) .................................................................................97
Figure 19: 2004-2012 data set: Total number of shops in CSAs by CSA type of living
(local level) - average CAGR across all modern retail shop types (source: EY analysis
based on © Nielsen Trade Dimensions) ...................................................................97
Figure 20 : 2004-2012 data set: Total number of shops in CSAs by CSA GDP
segmentation (local level) - average CAGR across all modern retail shop types (source:
EY analysis based on © Nielsen Trade Dimensions) ..................................................98
Figure 21 : 2004-2012 data set: Total EAN codes by CSA type and GDP range (local
level) and average annual growth rate across 23 product categories (source: EY analysis
based on © Nielsen Opus) .....................................................................................98
8
Figure 22: 2004-2012 data set: Total EAN codes by Member State (local level) and
average annual growth rate across 23 product categories (source: EY analysis based on
© Nielsen Opus) ...................................................................................................99
Figure 23: 2008-2012 sample: Total EAN codes by Member State (local level) and
average CAGR across 23 product categories (source: EY analysis based on © Nielsen
Opus) ............................................................................................................... 100
Figure 24: 2004-2012 data set: Total EAN codes by shop type (local level) (source: EY
analysis based on © Nielsen Opus) ....................................................................... 100
Figure 25: 2008-2012 sample: Total EAN codes by shop type (local level) (source: EY
analysis based on © Nielsen Opus) ....................................................................... 101
Figure 26: 2004-2012 data set: Total number of EAN codes by product category (local
level) - average CAGR across 6 MS (source: EY analysis based on © Nielsen Opus) .... 102
Figure 27: 2008-2012 sample: Total number of EAN codes by product category (local
Figure 28: 2004-2012 data set: Total number of pack sizes by CSA type and GDP range
(local level) - average CAGR across 23 product categories (source: EY analysis based on
© Nielsen Opus) ................................................................................................. 103
Figure 29: 2004-2012 data set: Total number of pack sizes by Member State (local level)
- average CAGR across 23 product categories (source: EY analysis based on © Nielsen
Opus) ............................................................................................................... 104
Figure 30: 2008-2012 sample: Total number of pack sizes by Member State (local level)
Opus) ............................................................................................................... 104
Figure 31: 2004-2012 data set: Total number of pack sizes by product category (local
Figure 32: 2008-2012 sample: Total number of pack sizes by product category (local
Figure 33: 2004-2012 data set: Total number of pack sizes by shop type (local level)
(source: EY analysis based on © Nielsen Opus) ...................................................... 106
Figure 34: 2008-2012 sample: Total number of pack sizes by shop type (local level)
(source: EY analysis based on ©Nielsen Opus) ....................................................... 106
Figure 35: Number of suppliers by CSA type and GDP range (local level) – average CAGR
across 23 product categories (source: EY analysis based on © Nielsen Opus) ............ 107
Figure 36: 2004-2012 data set: Total number of suppliers by Member State (local level)
– average CAGR across 23 product categories (source: EY analysis based on © Nielsen
Opus) ............................................................................................................... 108
Figure 37: 2008-2012 sample: Total number of suppliers by Member State (local level) –
Opus) ............................................................................................................... 108
Figure 38: 2004-2012 data set: Total number of suppliers by product category (local
level) – average CAGR across 6 MS sample (source: EY analysis based on © Nielsen
Opus) ............................................................................................................... 109
Figure 39: 2008-2012 sample: Total number of suppliers by product category (local
Opus) ............................................................................................................... 110
Figure 40: 2004-2012 data set: Total number of suppliers by shop type (local level)
9
Figure 41: 2008-2012 sample: Total number of suppliers by shop type (local level)
Figure 42: 2004-2012 data set: Evolution of number of EAN codes (local level) – across
23 product categories and 6 MS sample (source: EY analysis based on © Nielsen Opus)
........................................................................................................................ 114
Figure 43: 2008-2012 sample: Evolution of number of EAN codes (local level) – across 23
product categories and 9 MS sample (source: EY analysis based on © Nielsen Opus) .. 114
Figure 44 : 2004-2012 data set: total number new EAN codes by CSA type and GDP
range (local level) (source: EY analysis based on © Nielsen Opus) ........................... 115
Figure 45: 2004-2012 data set: Evolution of innovations (new EAN codes) by MS (local
level) – average CAGR across 23 product categories (source: EY analysis based on ©
Nielsen Opus) .................................................................................................... 115
Figure 46: 2008-2012 data set: Total number of innovations (new EAN codes) by MS
(local level) – average CAGR across 23 product categories (source: EY analysis based on
© Nielsen Opus) ................................................................................................. 116
Figure 47: 2008-2012 data set: Evolution of innovations (new EAN codes) by shop type
(local level) –6 MS sample (source: EY analysis based on © Nielsen Opus) ................ 116
Figure 48: 2006-2012 sample: Evolution of innovations (new EAN codes) by product
category (local level) –across 6 MS sample (source: EY analysis based on © Nielsen
Opus) ............................................................................................................... 118
Figure 49: 2008-2012 data set: Total number of innovations (new EAN codes) by product
category (local level) – average CAGR across 9 MS sample (source: EY analysis based on
© Nielsen Opus) ................................................................................................. 119
Figure 50 : 2004-2012 data set: Proportion of types of innovations by MS (local level)
(source: EY analysis based on © Mintel GNPD and © Nielsen Opus) ......................... 120
Figure 51: 2004-2012 data set: Proportion of types of innovations by shop type (local
level) – average % across 6 MS sample (source: EY analysis based on © Mintel GNPD
and © Nielsen Opus) .......................................................................................... 120
and © Nielsen Opus) .......................................................................................... 120
Figure 53: 2004-2012 data set: Proportion of innovations by type for cereals (local level)
– average % across 6 MS sample (source: EY analysis based on © Mintel GNPD and
©Nielsen Opus) .................................................................................................. 121
Figure 54: 2004-2012 data set: Proportion of innovations by type for cheese (local level)
©Nielsen Opus) .................................................................................................. 121
Figure 55: 2004-2012 data set: Proportion of innovations classified as “new products” by
© Mintel GNPD (local level) – average % across 23 product categories and 6 MS sample
Figure 56: 2004-2012 data set: Proportion of innovations by type for canned vegetables
(local level) – average % across 6 MS sample (source: EY analysis based on © Mintel
GNPD and © Nielsen Opus) ................................................................................. 122
Figure 57: 2004-2012 data set: Proportion of innovations by type for chocolate (local
and © Nielsen Opus) .......................................................................................... 123
Figure 58: 2004-2012 data set: Proportion of innovations classified as “new
variety/range extension” by © Mintel GNPD (local level) – average % across 23 product
10
categories and 6 MS sample (source: EY analysis based on © Mintel GNPD and © Nielsen
Opus) ............................................................................................................... 123
Figure 59: 2004-2012 data set: Proportion of innovations by type for mineral water (local
and © Nielsen Opus) .......................................................................................... 124
Figure 60: 2004-2012 data set: Proportion of innovations by type for edible oil (local
and © Nielsen Opus) .......................................................................................... 124
Figure 61: 2004-2012 data set: Proportion of innovations classified as “new packaging”
by © Mintel GNPD (local level) – average % across 23 product categories and 6 MS
sample (source: EY analysis based on © Mintel GNPD and © Nielsen Opus)............... 125
Figure 62: 2004-2012 data set: Proportion of innovations by type for ready-cooked meals
Figure 63: 2004-2012 data set: Proportion of innovations by type for starters/pizzas
Figure 64: 2004-2012 data set: Proportion of innovations classified as “new formulation”
sample (source: EY analysis based on © Mintel GNPD and © Nielsen Opus)............... 126
Figure 65: 2004-2012 data set: Proportion of innovations by type for baby food (local
and © Nielsen Opus) .......................................................................................... 127
Figure 66: 2004-2012 data set: Proportion of innovations by type for tea (local level) –
average % across 6 MS sample (source: EY analysis based on © Mintel GNPD and ©
Nielsen Opus) .................................................................................................... 127
Figure 67: 2004-2012 data set: Proportion of innovations classified as “relaunch” by ©
Mintel GNPD (local level) – average % across 23 product categories and 6 MS sample
Figure 68: comparative map of HHI modern retail across Europe (2004 - 2012) ......... 129
Figure 69: 2004-2012 data set: Retail concentration HHI per MS by retail group sales
area (local level) - 4 MS sample (source: EY analysis based on © Nielsen Trade
Dimensions) ...................................................................................................... 132
Dimensions) ...................................................................................................... 133
Figure 71: 2004-2012 data set: Retail concentration HHI per CSA type by retail group
sales area (local level) - 4 MS sample (source: EY analysis based on © Nielsen Trade
Dimensions) ...................................................................................................... 134
Dimensions) ...................................................................................................... 135
sales area (local level) – average CAGR across 4 MS sample (source: EY analysis based
on © Nielsen Trade Dimensions) .......................................................................... 136
11
Figure 75: Supplier concentration HHI by market share per product category (national
level) – average across 14 MS sample – first set of categories (source: EY analysis based
on © Euromonitor International) .......................................................................... 140
level) – average across 14 MS sample – second set of categories (source: EY analysis
based on © Euromonitor International) ................................................................. 141
level) – average CAGR across 14 MS sample (source: EY analysis based on ©
Euromonitor International) .................................................................................. 142
Figure 78: 2004-2012 data set: Supplier concentration by MS across 23 product
categories (local level based on HHI) (source: EY analysis based on © Nielsen Opus) . 143
categories (local level based on HHI) (source: EY analysis based on © Nielsen Opus) . 143
Figure 80: 2004-2012 data set: Assortment concentration HHI by share of EANs per
product category (local level) – average across 6 MS sample – first set of categories
product category (local level) – average across 6 MS sample – second set of categories
product category (local level) – average CAGR across 6 MS sample (source: EY analysis
based on © Nielsen Opus) ................................................................................... 146
Figure 83: Measure of imbalance HHI at procurement level per product category (national
level) – average across 14 MS – first set of categories (source: EY analysis based on ©
Planet Retail and © Euromonitor International) ...................................................... 148
level) – average across 14 MS – second set of categories (source: EY analysis based on
© Planet Retail and © Euromonitor International) .................................................. 149
level) – average CAGR across 14 MS (source: EY analysis based on © Planet Retail and
© Euromonitor International) ............................................................................... 150
Figure 86: Growth in total number of modern retail outlets in the EU 27 (national level) -
CAGR (source: EY analysis based on © Planet Retail) .............................................. 152
Figure 87: 2004-2012 data set: Total number of modern retail shops across CSAs by
shop type (local level) – across 4 MS sample (source: EY analysis based on © Nielsen
Trade Dimensions).............................................................................................. 153
shop type (local level) – average CAGR across 4 MS sample (source: EY analysis based
Figure 89: Growth in hypermarket outlets in the EU 27 (national level) - CAGR (source:
EY analysis based on © Planet Retail) ................................................................... 154
Figure 90: Growth in supermarket outlets in the EU 27 (national level) - CAGR (source:
Figure 91: Growth in discount store outlets in the EU 27 (national level) - CAGR (source:
Figure 92: 2004-2012 data set: Growth of hypermarkets by CSA type (local level) –
CAGR across 4 MS sample (source: EY analysis based on © Nielsen Trade Dimensions)
........................................................................................................................ 157
12
Figure 93: 2004-2012 data set: Growth of supermarkets by CSA type (local level) –
........................................................................................................................ 157
Figure 94: 2004-2012 data set: Growth of discount stores by CSA type (local level) –
........................................................................................................................ 158
Figure 95: 2004-2012 data set: Average sales area for hypermarkets by MS (local level)
– CAGR for 4 MS sample (source: EY analysis based on © Nielsen Trade Dimensions) 159
Figure 97: 2004-2012 data set: Average sales area of hypermarkets per MS (national
level) – in m² for 4 MS sample (source: EY analysis based on © Planet Retail) ........... 160
Figure 99: 2004-2012 data set: Average sales area for supermarkets by MS (local level)
Figure 101: 2004-2012 data set: Average sales area of supermarkets per MS (national
level) – in m² for 6 MS sample (source: EY analysis based on © Planet Retail)........... 162
Figure 103: 2004-2012 data set: Average sales area for discount stores by MS (local
level) – CAGR for 4 MS sample (source: EY analysis based on © Nielsen Trade
Dimensions) ...................................................................................................... 163
Dimensions) ...................................................................................................... 163
Figure 105: 2004-2012 data set: Average sales area of discount stores per MS (national
Figure 107: Progression in % points of private label market share from 2004 to 2012 for
14 MS sample (national level) - average across 23 product categories (source: EY
analysis based on © Euromonitor International) ..................................................... 166
Figure 108: 2004-2012 data set: Progression in % points of private label EAN share from
2004 to 2012 for 6 MS sample (local level) - average across 23 product categories
Figure 109: 2004-2012 data set: Proportion of private label EAN for 6 MS sample (local
level) - average across 23 product categories (source: EY analysis based on © Nielsen
Opus) ............................................................................................................... 167
Opus) ............................................................................................................... 168
13
Figure 112: Percentage of private label sales share by product category - average across
14 MS (national level) (source: EY analysis based on © Euromonitor International) .... 169
Figure 113: 2008-2012 data set: Proportion of private label EAN by product category
(local level) - average across 9 MS sample (source: EY analysis based on © Nielsen
Opus) ............................................................................................................... 171
Figure 114: 2004-2012 data set: Progress in % points of private label EAN share from
2004 to 2012 (local level) - average across 6 MS sample (source: EY analysis based on ©
Nielsen Opus) .................................................................................................... 172
Nielsen Opus) .................................................................................................... 173
Figure 116: 2004-2012 data set: Product category turnover for 6 MS sample (national
level) - average CAGR across 23 product categories (source: EY analysis based on ©
Figure 118: 2004-2012 data set: Product category turnover (national level) – in M €
across 6 MS sample – first set of categories (source: EY analysis based on © Euromonitor
International)..................................................................................................... 175
across 6 MS sample – second set of categories (source: EY analysis based on ©
Figure 120: 2004-2012 data set: Population Size in CSAs by Member State (local level) -
average CAGR for 6 MS sample (source: EY analysis based on Eurostat) ................... 177
Figure 122: 2004-2012 data set: Population Density in CSAs by Member State (local
level) - average CAGR for 6 MS sample (source: EY analysis based on Eurostat) ........ 178
Figure 124: 2004-2012 data set: Unemployment Rate in CSAs by Member State (local
Figure 126: 2004-2012 data set: GDP per capita in CSAs by Member State (local level) -
Figure 128: 2004-2012 data set: Evolution of the proportion of income spent on food and
non-alcoholic beverage by Member State (national level) - CAGR for 6 MS sample
(source: EY analysis based on Eurostat) ................................................................ 181
Figure 129: 2008-2012 data set: Consumption of food and non-alcoholic beverage by
Member State (national level) - CAGR for 9 MS sample (source: EY analysis based on
Eurostat) ........................................................................................................... 181
Figure 130: EU28 retail business expectations and GDP growth (Source: Eurostat) ..... 182
Figure 131: Retail business expectations in France, Poland and Spain (source: Eurostat)
........................................................................................................................ 182
14
Figure 132: Distribution of supplier concentration (HHI – brand only by sales market
share at national level) for the 23 product categories in each country in 2012 (source: EY
analysis based on © Euromonitor) ........................................................................ 191
Figure 133: Distribution of measure of imbalance for the 23 product categories in each
country in 2012 (source: EY analysis based on © Planet Retail and © Euromonitor) ... 193
Figure 134: Distribution of shops by C5 concentration measure at banner level (long data
set) (source: analysis based on © Nielsen Trade Dimensions sales area data)............ 195
Figure 135: Distribution of shops by HHI concentration measure at banner level (long
data set) (source: analysis based on © Nielsen Trade Dimensions sales area data) .... 196
Figure 136: Average number of EAN codes per shop and per product category, presented
by shop type (long data set) (source: analysis based on © Nielsen Opus. Data are for
first period in each year) ..................................................................................... 197
Figure 137: Average number of EAN codes per shop and per product category in
hypermarkets in Member States (long data set) (source: analysis based on © Nielsen
Opus. Data are for first period in each year) ......................................................... 198
Figure 138: Average number of EAN codes per shop by product category (source:
analysis based on © Nielsen Opus. Data are for first period in each year) .................. 199
Figure 139: Average number of EAN codes per shop in hypermarkets in selected Member
States in 2012, presented by product category (source: analysis based on © Nielsen
Opus) ............................................................................................................... 200
Figure 140: Average number of new EAN codes per shop and per product category,
presented by shop type (long data set) (source: analysis based on © Nielsen Opus. Data
are for first period in each year) ........................................................................... 200
Figure 141: Average number of new EAN codes per shop and per product category in
hypermarkets in selected Member States (long data set) (source: analysis based on ©
Nielsen Opus. Data are for first period in each year) .............................................. 201
Figure 142: Average number of new EAN codes per shop by product category (long data
set) (source: analysis based on © Nielsen Opus. Data are for first period in each year)
........................................................................................................................ 202
Figure 143: Average number of new EAN codes per shop in hypermarkets in selected
Member States in 2012, presented by product category (long data set) (source: analysis
based on © Nielsen Opus) ................................................................................... 203
Figure 144: Choice in variety of EAN codes in the sampled shops versus national retail
concentration (source: analysis based on © Nielsen Opus and © Planet Retail. Data are
for first period in each year 2004, 2006, 2008, 2010 and 2012) ............................... 213
Figure 145: New EAN codes (innovation) versus national retail concentration (source:
analysis based on © Nielsen Opus and © Planet Retail. Data are for first period in each
year 2006, 2008, 2010 and 2012) ........................................................................ 215
Figure 146: Choice in variety of EAN codes versus local retail concentration by shop type
in 2004 and 2012 (source: Analysis based on © Nielsen Opus and © Nielsen Trade
Dimensions. Data are for first period in each year and cover Italy, Spain, France,
Portugal and Poland.) .......................................................................................... 216
Figure 147: Opus innovations versus local retail concentration by shop type in 2004 and
2012 (source: Analysis based on © Nielsen Opus and © Nielsen Trade Dimensions. Data
are for first period in each year and cover Italy, Spain, France, and Portugal.) ........... 217
Figure 148: New EAN codes (innovation) versus local retail concentration, all shops and
years (source: analysis based on © Nielsen Opus and © Nielsen Trade Dimensions. Data
are for first period in each year of 2006, 2008, 2010 and 2012, and, in the left-hand
chart, cover Italy, Spain, France and Portugal) ....................................................... 218
15
Figure 149: Choice in variety of EAN codes versus national supplier concentration by
product category, 2008 (source: analysis based on © Nielsen Opus and © Euromonitor
International. Data are for first period in the year and cover Italy, Spain, France, Portugal
and Poland) ....................................................................................................... 220
Figure 150: Opus innovations versus national supplier concentration by product
category, 2008 (source: analysis based on © Nielsen Opus and © Euromonitor
International. Data are for first period in the year and cover Italy, Spain, France,
Portugal.) .......................................................................................................... 221
Figure 151: New EAN codes (innovations) versus the ratio of retailer to supplier
concentration (source: analysis based on © Nielsen Opus, © Planet Retail and ©
Figure 152: Choice and the private label share by shop type .................................... 225
Figure 153: Innovation and the private label share by shop type .............................. 227
Figure 154: Choice in variety of EANs versus national product category sales turnover in
2010 period 1 in four Member States (source: analysis based on © Nielsen Opus and ©
Figure 155: New EAN codes (innovations) versus national product category sales
turnover in 2010 period 1 in four Member States (source: analysis based on © Nielsen
Opus and © Euromonitor International) ................................................................ 229
Figure 156: New EAN codes (innovations) versus unemployment rate (source: analysis
based on © Nielsen Opus and Eurostat. Innovation data are for first period in each year
2004, 2006, 2008, 2010, 2012) ........................................................................... 231
Figure 157: Choice in variety of EAN codes versus GDP per capita (source: analysis based
on © Nielsen Opus and Eurostat. Choice data are for first period in each year 2004,
2006, 2008, 2010, 2012) .................................................................................... 233
Figure 158: Opus innovations versus retailer business expectations (source: analysis
based on © Nielsen Opus and Eurostat), 2006, 2008, 2010, 2012 ............................ 234
Figure 159: Choice in variety of EAN codes and population density, 2008-12 ............. 235
Figure 160: Opus innovations and population density, 2008-12 ................................ 236
Figure 161: Contribution of drivers accounting for change in total choice and innovation
2006-12 in a hypermarket in Italy ........................................................................ 242
2006-12 in a hypermarket in France ..................................................................... 243
2006-12 in a supermarket in Spain ....................................................................... 244
2006-12 in a hypermarket in Poland ..................................................................... 245
2006-12 in a supermarket in Portugal ................................................................... 246
Figure 166: Change in choice (product variety) offered by sample hypermarkets in
consumer shopping areas, 2004-2012 (source: analysis based on © Nielsen Opus) .... 247
Figure 167: Change in innovation (total new EAN codes) offered by sample hypermarkets
in consumer shopping areas, 2004-2012 (source: analysis based on © Nielsen Opus). 248
Figure 168: Contribution of drivers accounting for change in product variety 2006-12 in
five CSAs ........................................................................................................... 249
Figure 169: Contribution of drivers accounting for change in total innovations 2006-12 in
five CSAs ........................................................................................................... 251
16
17
Table 1: Market share (in edible grocery sales) of the top 10 retailers in EU (2000 -
2011)..................................................................................................................51
Table 2: Selection of main international buying groups in EU .....................................53
Table 3: Variety of situations in 9 MS vs. EU 27 .......................................................67
Table 4: Banner coverage in shop sample across MS ................................................68
Table 5: Coverage of largest retail groups in Europe ................................................70
Table 6: Scope of selected measures at procurement (national) level .........................70
Table 7: Study samples by MS and time period coverage – descriptive statistics (source
EY analysis) .........................................................................................................71
Table 8: List of regions where consumer shopping areas are located ...........................73
Table 9: Number of CSA in relation to population size ...............................................75
Table 10: Number of CSA per type of living zone and standard of living category .........76
Table 11: Comparison of proportion of CSA vs proportion of EU27 population ..............76
Table 12: Selection of 23 product categories ............................................................77
Table 13: Extrapolation of discounters.....................................................................79
Table 14: Extrapolation of supermarkets .................................................................79
Table 15: Maximum travel times for defining a given shop’s catchment area ................85
Table 16: Summary of findings on evolution of choice ...............................................95
Table 17: Retail group HHI by sales market share, for modern retail only (national level)
(source: EY analysis based on © Planet Retail) ...................................................... 131
Table 18: Supplier concentration HHI (national level) by market share per product
category – average across 23 sample product categories (source: EY analysis based on ©
Table 19: Supplier concentration by product categories and by MS – CAGR 2004-2012
(source: EY analysis based on © Euromonitor International) .................................... 139
Table 20: Number of situations of imbalance HHI across 23 product category sample
(source: EY analysis based on © Planet Retail and © Euromonitor) .......................... 147
Table 20: Private label sales share (national level) averaged across 23 product category
sample (source: EY analysis based on © Euromonitor International) ......................... 165
Table 21: Evolution of private label market share from 2004 to 2012 (national level) -
average across 6 MS sample (source: EY analysis based on © Euromonitor International)
........................................................................................................................ 169
Table 22: 2004-2012 data set: Proportion of private label EAN by product category (local
level) - average across 6 MS sample (source: EY analysis based on © Nielsen Opus) .. 170
Table 23: The two data sets used in the econometric analysis ................................. 187
Table 24: Retail group HHI by sales market share in modern retail (national level)
(source: EY analysis based on © Planet Retail) ...................................................... 188
Table 25: Supplier HHI – brand only by sales market share (national level), averaged
across 23 product category sample (source: EY analysis based on © Euromonitor) ..... 190
(source: EY analysis based on © Planet Retail and © Euromonitor) .......................... 192
Table 27: Private label percentage share by sales (national level), averaged across 23
product category sample (source: EY analysis based on © Planet Retail) ................... 194
18
Table 28: Summary of econometric results for key drivers: choice ............................ 207
Table 29: Summary of econometric results for key drivers: innovation ...................... 210
Table 30: Key to the figures showing the contribution of drivers to change in choice and
innovation ......................................................................................................... 241
Table 31: Correlations between choice variables (long data set) ............................... 380
Table 32: Correlations between innovation variables (long data set) ......................... 380
Table 33: Correlations between national and local supplier concentrations (long data set)
........................................................................................................................ 380
Table 34: Correlations between national and local retail concentrations (long data set)
........................................................................................................................ 381
Table 35: Correlations between selected measures of national and local retail
concentrations (long data set) .............................................................................. 382
Table 36: Variables and alternative indicators ........................................................ 383
Table 37: Country and shop coverage in short and long data sets ............................. 385
Table 38: Results - Product Variety ....................................................................... 390
Table 39: Results - Product Size Variety ................................................................ 394
Table 40: Results - Product Supplier Variety .......................................................... 398
Table 41: Results - Product Price Variety ............................................................... 402
Table 42: Results - Opus Innovations .................................................................... 408
Table 43: Results - New Products ......................................................................... 413
Table 44: Results - New Packaging ....................................................................... 417
Table 45: Results - New Formulation ..................................................................... 421
Table 46: Results - New Range extensions ............................................................. 425
19
Abbreviations
CA Catchment area
CAGR Compound annual growth rate
Cx Concentration of x market players
CSA Consumer shopping area
DG COMP Directorate-General for Competition
EAN European article number (now international article
number)
ERRT European Retail Round Table
EU European Union
Eurostat Statistical office of the European Union
GDP Gross domestic product
GNPD Global New Products Database (© Mintel Group Ltd)
HHI Herfindahl–Hirschman Index
HICP Harmonised index of consumer prices
HM Hypermarket
HD Hard discounter
INT Intermediate (Eurostat rural/urban typology)
€M Millions of Euro
MS Member state of the European Union
NCA National competition authorities
NFC Near field communication
NUTS Nomenclature of Territorial Units for Statistics
PR Predominantly rural (Eurostat rural/urban typology)
PU Predominantly urban (Eurostat rural/urban typology)
QR Quick response code
R&D Research & Development
SKU Stock-keeping unit
SM Supermarket
SME Small and Medium Enterprises
VAT Value added tax
20
1. Executive summary
EY, together with Arcadia International and Cambridge Econometrics has been awarded a
contract by DG COMPETITION of the European Commission as a result of a call for
tenders published in the Official Journal on 19 December 2012. DG COMP commissioned
a study on the economic impact of modern retail on choice and innovation in the EU food
sector. The study has been conducted between May 2013 and September 2014.
The full report is available at the following address:
http://ec.europa.eu/competition/publications/reports/
The executive summary is available in French at the following address:
http://ec.europa.eu/competition/publications/reports/retail_study_ex_fr.pdf
1.1. Objectives of the study

The main objectives of the study are the following:
 measure the evolution of choice and innovation over the last decade in the
modern retail food sector; and
 identify the main drivers of choice and innovation, measure their evolution
over the last decade, and their economic impact on choice and innovation.
1.2. Methodology
A combination of tools and methods has been adopted:
 Literature review;
 Collaborative workshops with experts to define a framework of analysis for choice
and innovation;
 Collection of data from a broad range of sources;
 Setting up of an extensive database compiling the sources,
 Statistical analyses describing the evolution of choice, innovation and the potential
drivers;
 Econometric analyses aiming to assess the impact of drivers on choice and
innovation;
 Six case studies bringing complementary information on product categories and
Member States (MS) not covered by the statistical analyses.
The concepts of choice and innovation ha ve been defined and their
potential drivers identified
Two types of choice are addressed in the study:
 Food choice has been defined as the product assortment available on retail
shelves, measured by the number of EAN codes1 in shops, and also by the variety
of packaging sizes, the variety of prices, and the variety of alternative suppliers.
 Shop choice has been defined as the number of shops to which a consumer has
access within a normal distance (consumer shopping area2).
Innovation for this study exclusively refers to product innovation3. Product innovation is
measured both in terms of the number of innovations introduced on shelves in a given
1
European Article Numbering bar code. Excluding promotions.
2
Consumer shopping areas are local areas that include all the modern retail shops to
which a consumer could reasonably travel to do their regular grocery shopping, based on
travel distances that are set according to the type of area (rural, intermediate, urban).
21
Executive summary
period and the associated types of innovation: new product, range extension,
packaging, new formulation, relaunch.
Consultations with experts and a literature review identified a list of key potential
drivers of choice and innovation:
 Concentration of modern retailers: national (procurement) level and local level
 Concentration of suppliers: national (procurement) level and local level
 Measure of imbalance in the market between modern retailers and suppliers (the
relative concentration of modern retailers and suppliers in the national market)
 Shop type
 Shop size
 New shop opening
 Socio-economic characteristics, including Gross Domestic Product (GDP) per
capita, retailers’ business expectations, population size and density,
unemployment and food consumption
 Private label share (at the local level and at national level)
 Product category turnover, i.e. sales market size in each product category
 Region / Member States characteristics including access to finance, legal
environment, pricing regulation, public health regulation and tax regulations.
An extensive database has been set up according to a sampling
strategy seeking to maximise geog raphical scope, product category
and time period
The identification of relevant and consistent data sources has been an important step of
the study. The choice of data sources was based on their availability, their level of
reliability for each indicator and their alignment to the definitions of choice and
innovation. The main objective was to maximise the geographical scope, the product
category coverage and the time period coverage. An extensive database that
integrates all gathered data has been developed. The study covers the largest data
sample available on choice and innovation at the local level.
3
Other types of innovation are excluded: process innovation (efficiency to drive down
costs), technology innovation (e.g. automation in distribution centres or logistics
operations) or concept innovation (e.g. new types of shopping experiences).
22
United Kingdom
Number of MS
Czech Republic
Luxembourg
Netherlands
Indicators source
Lithuania
Denmark
Germany
Romania
Portugal
Hungary
Slovenia
Slovakia
Bulgaria
Belgium
Sweden
Estonia
Finland
Austria
Ireland
Cyprus
Poland
Greece
France
Malta
Latvia
Spain
Italy
Evolution of choices 2004-2012
Shop choices (2004-2012) ©Nielsen trade dimension n n n n 4
Shop choices (2008-2012) ©Nielsen trade dimension n n n n n n 6
Product variety, price variety, size variety (2004-2012) ©Nielsen Opus n n n n n n 6
Product variety, price variety, size variety (2008-2012) ©Nielsen Opus n n n n n n n n n 9
Evolution of innovations 2004-2012
Number of innovations (2004-2012) ©Nielsen Opus n n n n n n 6
Number of innovations (2008-2012) ©Nielsen Opus n n n n n n n n n 9
Categories of innovations (2004-2012) ©Mintel GNDP n n n n n n 6
Categories of innovations (2008-2012) ©Mintel GNDP n n n n n n n n n 9
Evolution of concentration
Retail concentration at national level (Retail group & banner
©Planet retail n n n n n n n n n n n n n n n n n n n n n n n n n n 26
level) - 2004-2012 - C5 / HHI
Retail concentration at local level - C5 / HHI (2004-2012) ©Nielsen trade dimension n n n n 4
Retail concentration at local level - C5 / HHI (2008-2012) ©Nielsen trade dimension n n n n n n 6
Supplier concentration at national level - 2004-2012 ©Euromonitor n n n n n n n n n n n n n n 14
Supplier concentration at local level - 2004-2012 ©Nielsen Opus n n n n n n 6
Measure of imbalance (national level only) - 2004-2012 ©Planet retail, ©Euromonitor n n n n n n n n n n n n n n 14
Evolution of other a priori drivers

Macroeconomic data (GDP, population, unemployment, etc.) Eurostat n n n n n n n n n n n n n n n n n n n n n n n n n n n 27
Shop types at national level - 2004-2012 ©Planet retail n n n n n n n n n n n n n n n n n n n n n n n n n n 26
Shop type, shop size - 2004-2012 ©Nielsen trade dimension n n n n 4
Shop type, shop size - 2008-2012 ©Nielsen trade dimension n n n n n n 6
Private label share (national level) - 2004-2012 ©Euromonitor n n n n n n n n n n n n n n 14
Private label share (local level) - 2004-2012 ©Nielsen Opus n n n n n n 6
Private label share (local level) - 2008-2012 ©Nielsen Opus n n n n n n n n n 9
Product category turnover at national level - 2004-2012 ©Euromonitor n n n n n n n n n n n n n n 14
Econometric analysis
Impact of drivers on choice and innovation (2004-2012) Consortium computation n n n n n 5
Impact of drivers on choice and innovation (2008-2012) Consortium computation n n n n n n n 7
Coverage of case studies
Case studies Consortium analysis n n n n n n 6
As shown in the table, a decision was taken to establish two data sets (a long period
over 2004-2012 and a shorter period over 2008-2012 for which more data is available)
so that a wider range of Member States could be included.
Choice and innovation have been quantitatively measured at a local level across 23
product categories and 343 shops in 9 Member States. This selection of product
categories covers a broad spectrum of fresh, ambient, frozen food / non-processed, less-
processed and processed food products sold through self-service. The 343 shops sample
include the three shop types regarded as making up modern retail (hypermarkets
>=2 500 m² ; supermarkets – 400 to 2499 m², discount stores characterised by limited
assortment, mainly composed of private labels and a low cost market strategy). They are
located in 105 consumer shopping areas (CSA), which have been selected to be
representative of a variety of living area types (rural, intermediate and urban) and
economic prosperity levels (low, medium, high GDP per capita) found in the EU 27.
At national level, we have been able to measure the evolution of modern retail and
supplier concentration in 14 Member States from 2004 to 2012. At local level, because of
limited availability of data, concentration has been measured in a more limited sample of
4 (2004-2012) to 6 MS (2008-2012).
Econometric analysis identifying the correlation between the observed evolution of
choice and innovation and their drivers covers the period 2004 to 2012 across 5 key
Member States (France, Italy, Poland, Portugal and Spain) and 296 shops. The scope has
been enlarged to 7 Member States and 337 shops for the short term period (2008-2012)
including Belgium and Hungary.
The data set available for the econometric analysis has certain characteristics that should
be noted when considering the results because of the possibility of biases introduced by
the nature of the sample:
 the Member States included in the econometric analysis are mainly those with
light or moderate modern retail concentration at national level;
 the Member States included in the econometric analysis cover a wide range of
situations with regard to supplier concentration and measure of imbalance at
national level.
23
Executive summary
Six case studies bringing additional qualitative information

The case studies complement the descriptive and econometric analysis to bring
qualitative and complementary information to six selected product categories: three fresh
non-barcoded products (apples in France, tomatoes in Belgium, fresh pork in Germany),
and three barcoded products (olive oil in Spain, cheese in the Netherlands and milk in
Finland). The objective of the case studies was two-fold: to be able to measure choice
and innovation for fresh products that are sold without an EAN code and therefore not
included in the econometric analysis, and to capture the specificities regarding choice and
innovation for barcoded products that are closer to the agricultural level of the food
supply chain.
The report on case studies is available at the following address:
http://ec.europa.eu/competition/publications/reports/retail_study_cases_en.pdf
24
1.3. Background: a strong development of modern retail across Europe,

a new landscape for EU consumers
Over the past decade, the retail landscape has evolved for EU
consumers due to a combination of different factors
The period covered by the study is characterized by the 2008 economic crisis which has
had significant impacts on consumer purchasing power. Seeking lower prices has
become a key priority for EU consumers. In addition, changes in household composition,
the trend towards an ageing population, increased interest in new health issues (food
intolerances, allergies, food-related diseases, overweight and obesity) and increased
environmental awareness have had an impact on the grocery retail market in Europe,
with the growth of specific product categories (fresh products, organic food, gluten-free
products, etc.). The desire of more convenient products has become an increasingly
important consideration for consumers leading to a number of innovations (ready
prepared meals, easy opening cans, etc.). Edible grocery sales have remained stable
over the last 8 years.
The period is characterized by a strong development of modern retail across the EU:
from 2004 to 2012, modern retail’s share of total grocery sales increased in 24 Member
States. It has been evident in new shop openings and increased floor space. Discount
stores have experienced the strongest growth in number of outlets and floorspace over
the past decade: they have increased their sales areas by 81% between 2000 and 2011
across the EU, whereas the total sales areas of hypermarkets increased by 46% and that
of supermarkets by 26% between 2000 and 2011.
The largest modern retail groups have expanded and increased their market share in
many Member States. At pan-European level, the top 10 European food retailers
accounted for a 26% market share in 2000, compared to 31% in 2011.
Finally, the market share of private label products has increased across most product
categories in Europe. Key reasons for this likely include a perception among consumers
that these products offer good value for money, the opportunity of higher margins for
retailers, and a profitable way for manufacturers to make use of spare capacity.
25
Executive summary
1.4. Evolution of choice: choice offered to consumers has notably

increased in a majority of MS
Choice available to consumers in local shops increased in terms of the
number of alternative products 4, the number of different brand
suppliers and the number of modern retail shops ; the increase was
greater during 2004-2008 than 2008-2012
Choice in alternative products, measured at a local level, has increased on
average by 5.1% annually from 2004 to 2012 in the shops sampled in the CSAs covered
by the study. During the pre-crisis period (2004-2008) the annual growth rate was
higher (7.9%) than during the crisis period since 2008 (2.4%).
16,0%
14,0%
12,0%
10,0%
8,0%
6,0% CAGR(04 - 08)
4,0%
2,0% CAGR(08 - 12)
0,0% CAGR(04 - 12)
2004-2012 sample: Evolution of number of EAN codes (local level) by CSA type and GDP range
(source: EY analysis based on © Nielsen Opus). CAGR: compound annual growth rate; PR: Predominantly
rural; PU: predominantly urban; IN: intermediate; ‘low’, ‘medium’ and ‘high’ refer to the level of GDP per
capita.
Choice in alternative products on the shelves of shops increased in all 9 MS of the

sample, with the highest growth seen in Poland (+8.3% on average annually), and the
lowest in Italy (+3.2%).
Starting from (and remaining at) much lower levels, discounters registered the
strongest growth in the number of alternative products with +8.0% annually on
average compared to +5.2% on average for hypermarkets and +3.6% for supermarkets.
Choice in alternative products at local level increased across all product categories
over the 2004 2012 period when considering the sample as a whole, but there were
significant variations across product categories. Across all CSAs, the product categories
where the number of alternative products increased the most were notably
ham/delicatessen, cereals, cheese, ready-cooked meals and starters/pizzas, all
registering around annual growth of 6% over 2004-2012; on the other hand,
butter/margarine and fruit juice registered the lowest annual growth of around 2%.
The variety of product sizes offered on modern retailers’ shelves, also increased
across all CSAs, Member States, product categories and shop types. As with choice in
alternative products, annual growth was notably higher during 2004-2008 (annual
growth of between 2.1% in Italy and 8.6% in Spain) than after 2008 (between 1.2% in
Italy and 4.1% in Belgium).
Evolutions of choice in product sizes differed considerably across the sampled product
categories. Cereals, coffee, edible oil and mineral water experienced the highest growth
over the decade, whilst desserts, frozen vegetables, cheese and butter/margarine
4
Measured by the EAN codes available on the shelves of retailers’ shops.
26
registered the lowest growth rates. Growth contracted for three product categories
(cheese, frozen vegetables, and ham/delicatessen) over the crisis period.
From 2004 to 2012, there was an overall contraction in the range of prices5 available to
consumers within a given product category. It is the only choice measure where a
negative overall trend was observed over the decade under study.
The number of brand suppliers for which products were offered on shop shelves within
a given product category increased on the whole from 2004 to 2012. Like other measures
of choice presented above, trends varied across consumer shopping areas, product
categories and shop types. Choice in brand suppliers available in modern retailers’ shops
increased over time in all Member States, ranging from 1.7% annual growth in Italy to
6.4% in Spain over the 2004-2012 period. The trend over the pre-crisis period was more
positive (between 2.1% in Belgium and 9.9% in Poland) than that of the crisis period
(between -0.8% in France and 6.8% in Belgium).
Notable variation in supplier choice was observed across the analysed product categories.
Choice in brand suppliers increased the most from 2004 to 2012 in cereals,
ham/delicatessen, chocolate and soft drinks. The product categories experiencing the
lowest growth over the same period were butter/margarine, coffee and frozen
vegetables. The total number of suppliers declined for two product categories (frozen
vegetables, and baby food) over the crisis period.
Variations in supplier choice were observed across the three shop types, with an annual
growth of +4.1% for hypermarkets on average between 2004 and 2012, +4% for
discounters, + 2.1% for supermarkets.
Choice measured by the number of shops that consumers have access to in their
consumer shopping areas increased between 2004 and 2012 by 1.6% annually, on
average. The annual growth was higher (1.8%) during the 2004-2008 period than after
2008 (1.3%).
Looking at living area types, during the pre-crisis period, annual growth in the number of
shops registered in ‘predominantly rural’ areas (3.6%) was twice the rate seen in
‘intermediate’ (1.8%) and ‘predominantly urban’ areas (1.7%). By comparison, the crisis
period saw lower annual growth rates across all types of living areas, and the trend
reversed, with ‘predominantly urban’ (1.6%) seeing higher growth than ‘predominantly
rural’ (1.5%), while ‘intermediate’ registered the lowest growth rate (0.8%).
5
The price data in Nielsen Opus contained many inconsistencies which could only be
partially corrected, leading to a less robust analysis on price variety.
27
Executive summary
1.5. Evolution of innovation: a steady stream of innovation was made

available to EU consumers; however the number of innovations
declined between 2008 and 2012
Innovations (number of new EANs) continue d to be developed and
made available to consumers in the EU, but the number of innovations
declined after 2008
The number of innovations6 increased pre-crisis between 2006 and 2008 (+3.8%
annually) but this trend was reversed during the crisis period with falls registered
between 2008 and 2010 (-1.2%), as well as 2010 and 2012 (-5.3%). The share of
innovations in the total number of products decreased steadily from 43% in 2006 to 30%
on average in 2012.
140 000
45 041 40 434
46 111
120 000
42 779 30%
100 000 31%
40%
80 000 43%1
Total new EANs
60 000 Total EANs
40 000 Total EANs removed
20 000
-
2004 2006 2008 2010 2012
1
Share of new EAN codes in the total number of EAN codes available on the shelves of modern retailers in 2006
2004-2012 sample: Evolution of the number of EAN codes (local level) – across 23 product
categories in 302 shops sampled in 91 CSAs in 6 MS (source: EY analysis based on © Nielsen Opus -
Be, Fr, It, Pl, Pt, Sp).
The experience with regard to the number of new EAN products made available in shops
varied across different types of CSA. The strongest growth in the pre-crisis period was in
more prosperous rural areas and less prosperous urban areas; during the crisis, the
number of innovations only increased in less prosperous urban areas.
25,0%
20,0%
15,0%
10,0%
5,0%
0,0% CAGR(06 - 08)
-5,0% CAGR(08 - 12)
-10,0%
-15,0% CAGR(06 - 12)
2004-2012 sample: total number new EAN codes by CSA type and GDP range (local level) (source:
EY analysis based on © Nielsen Opus)
When aggregating data from the sampled shops by Member States, the number of
innovations increased over the period only in Poland, Spain, and to a lesser extent in
6
Measured by analysis of the EAN codes available on the shelves of retailers’ shops.
28
Belgium; it contracted (particularly since 2008) in Italy and France, and to a lesser
extent in Portugal. Only in Spanish CSAs was positive growth in innovations registered
both pre-crisis and during the crisis period.
Trends in innovations varied greatly across the sampled product categories. Across the
sampled shops as a whole, only three product categories (baby food, cereals,
starters/pizzas) registered notable positive annual growth over 2006-2012, another three
(chocolate, soft drinks, yoghurt) remained stable, and the remainder registered negative
annual growth over this period. The categories where the growth in new products
contracted the most were mineral water (-6.8%), canned vegetables (-4.9%) and fresh
pre-packaged bread (-4.3%).
The fastest growth in the pre-crisis period was observed in discount stores and
hypermarkets, whilst the trend for innovations in supermarkets was stable. After
2008, the trend remained positive but slowed down in discount stores while the number
of innovations declined in both hypermarkets and supermarkets.
Types of innovation have changed from 2006 to 2012; innovations
focused on new packaging have become considerably more common
over time in most Member States in the analysed sample
Trends in the types of innovative products on offer at local level varied across the
Member States. In France, Spain and Italy, and to a lesser extent in Portugal and Poland,
there has been a trend towards more new packaging innovations as a proportion of total
innovations at the expense of new products and range extension products. On average
across all MS in the sample, new packaging innovations represented approximately 30%
of total innovations in 2012 compared to approximately 6% in 2004. By contrast, the
shares of new varieties and range extensions have decreased from 40% in 2004 to 30%
in 2012.
100%
80%
60% Relaunch
40% Range extension

20% Formula
0% Packaging
Product
2004-2012 sample: Proportion of innovations by MS (local level) (source: EY analysis based on ©

Mintel GNPD and © Nielsen Opus)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Product Packaging Formula Range extension Relaunch
29
Executive summary
2004-2012 sample: Proportion of innovations by product categories (local level) (source: EY

analysis based on © Mintel GNPD and © Nielsen Opus)
The trend towards an increasing proportion of packaging innovations was observed

across the three product categories with the highest growth in new products over the last
decade: cereals, baby food and starters/pizzas.
For both canned vegetables and fresh pre-packaged bread there was a decline in the
proportion of new products. In the case of canned vegetables, new products have been
replaced by range extensions, while for pre-packaged bread both range extensions and
packaging innovations became more important. In the case of mineral water, the
proportion of new products did not change, but here range extensions have been
replaced by packaging innovations.
30
1.6. Evolution of concentration: concentration of retailers and suppliers

showed different trends depending on the MS, the product category
and the level of analysis (local or national)
Trends in modern retailers’ concentration varied in different Member
States
In the edible grocery market as a whole including modern retail stores as well as smaller
independent and traditional stores, a clear trend towards greater retailers’ concentration
has been observed during the period in 22 of 26 MS, pulled by the development of
modern retail.
The picture is more mixed when the focus is on modern retail, where there were two
opposite trends in the concentration of modern retailers among the MS over 2004-2012.
Although the largest modern retail groups have increased their market shares at the pan-
European level, the growth of modern retailers who had a small market share in 2004 or
were not even present (like discounters in some MS) led to a decrease in modern
retailers’ concentration at national level over 2004-2012 in 16 of 26 MS7. But in the 10
other MS modern retailers’ concentration increased.
In the 14 MS sample, representing 85% of the EU population, concentration of modern
retailers increased in 7 MS (Czech Republic, Germany, Finland, Portugal, Poland, Spain
and United-Kingdom) and decreased in the other 7 MS (Belgium, Denmark, France,
Hungary, Italy, Netherlands, Romania).
3501 < < 4000 1501 < < 2000
1001 < < 1500

3001 < < 3500
2501 < < 3000 500 < < 1000
2001 < < 2500 N/A
2004-2012 HHI Modern Retail across EU MS sample at a national level
7
Data is not available for Malta, and Croatia was not in the EU in 2012. Modern retailers’
concentration is based on the Herfindahl–Hirschman Index (HHI), calculated as the sum
of the squares of the food market shares of each modern retail group and expressed as a
value between 0 and 10,000.
31
Executive summary
At local level in terms of share of floorspace, retail concentration decreased by -1.1%

annually on average over the 2004-2012 period. The decrease was the same for the pre-
crisis and crisis periods. As mentioned above in shop choice evolution, this is mainly due
to the increase in number of outlets.
Concentration of brand suppliers tended to increase at national level
from 2004 to 2012 across most MS and product categories, while
assortment concentration decreased at local level, particularly
between 2004 and 2008
At national level, supplier concentration increased for 20 of the 23 product
categories and 13 of the 14 sampled Member States. Concentration increased more on
average during the pre-crisis period (when 22 of 23 product categories became more
concentrated) than after 2008 (when 17 of 23 product categories became more
concentrated).
The product categories with the highest average concentration levels between 2004 and
2012 were frozen ready cooked meals, baby food, cereals and coffee. The categories with
the lowest average concentration levels were ham/delicatessen, cheese and fresh pre-
packed bread.
At local CSA level, the trend in assortment concentration8 changed over the
period 2004-2012. During the 2004-2008 period, assortment concentration decreased
in all 6 MS9 in the sample by -1.3% annually on average and for most product categories
(15 out of 23). After 2008, the decrease in assortment concentration slowed down
reaching -0.4% annually on average; concentration increased in two MS (France and
Portugal, averaged across product categories) and 10 product categories (average across
MS).
A wide range of situations in measure of imbalance between modern
retailers and suppliers has been observed depending on the MS and
the product category.
The balance of the relationship between suppliers and modern retailers was measured at
the procurement level, i.e. at national level, considering that negotiations mainly take
place at national level. Analyses of situations by product category and Member States
attest that they are equal numbers of situations in favour of retailers as they are
situations in favour of suppliers.
At national level, modern retail groups are concentrated to a greater extent than brand
suppliers in 6 out of 14 MS for the majority of product categories (for example: in
Finland, retailers are more concentrated than suppliers for 21 out of 23 product
categories). In the other 8 MS, suppliers are more concentrated than modern retailers
for the majority of analysed product categories (for example in Hungary: suppliers are
more concentrated than modern retailers in 17 product categories out of 23). For 12
product categories, modern retailers are more concentrated than suppliers in a majority
of the 14 analysed MS, whereas suppliers are more concentrated than modern retailers in
a majority of MS for 11 product categories. For instance, baby food and cereals suppliers
are more concentrated than modern retailers in most MS in the sample, whereas the
opposite is the case for cheese, ham or bread.
8
Assortment concentration is a measure of supplier concentration at local level reflecting
the share of EANs in a specific product category that each brand supplier has on the
shelves of retailers' shops. It is affected by retailers' assortment decisions to stock
certain products.
9
Belgium, France, Italy, Poland, Portugal, Spain
32
1.7. Conclusions regarding factors driving choice

In addition to the shop types and shop size which have an obvious
impact on choice, economic prosperity and product category turnover
have been favourable factors for choice
The main drivers were found to be the GDP per capita of the region in which the shop is
located, national turnover in the product category, certain shop characteristics (format,
floorspace) and the presence of a new shop opening in the local area: these all had
positive impacts on choice.
The level of prosperity in the region in which shops are located, measured by GDP
per capita, had a strong positive impact on most choice indicators except product price
variety10. The reduction in choice growth between 2008 and 2012 is in part associated
with the decline in GDP per capita across Member States. Prosperous areas may
encourage retailers to extend product choice to secure higher spend by customers.
Product category turnover at national level was found to show a strong positive
relationship with all measures of choice (product variety, product size variety and product
supplier variety in particular), except product price variety. Obviously, product categories
with high sales turnover are also those where there is a greater commercial potential,
and therefore where suppliers focus on product development and retailers have high
expectations on turnover, ultimately accounting for a wide variety of products on offer.
Evidence was found, as expected, that hypermarkets offered the most choice and
discounters the least, and that, for any given format, larger shops offered more choice.
The impacts of the drivers that measure indicators that relate directly to retailers and
suppliers were mostly small.
There is no evidence that the concentration of modern retailers has
been an economic driver of choic e
Econometric analyses found very little evidence of a relationship between modern
retailer concentration (at either local or national level) and the level of choice made
available to consumers, but the countries in the sample did not include those with the
highest levels of national modern retailer concentration.
Some case studies suggested that the structure of modern retail can have a positive
effect on choice. For instance, in the case of tomatoes in Belgium, the high concentration
of modern retail has not prevented intense competition amongst modern retailers to
select, source and propose the most attractive range of products. For olive oil in Spain,
modern retailers have used increased choice to attract and retain customers.
Competition in the form of a new shop opening in the local area
improves the choice offered in existing shops
Evidence was found that shops that experienced the opening of a new shop within a
distance close enough for the new shop to be regarded as a competitor tended to offer
somewhat more choice.
There is no evidence that the concentration of suppliers has been an
economic driver of choice
The impact of supplier concentration on choice was found to be negligible.
10
The price data in Nielsen Opus contained many inconsistencies which have been
partially corrected.
33
Executive summary
The impact of the measure of imbalance between modern retailers and

suppliers was generally found to be consistent with those found
separately for retail concentration and supplier concentration
Consistent with the separate findings for retailer concentration and supplier
concentration, there was very little evidence that the measure of imbalance between
modern retailers and suppliers had an impact on choice.
The impact of private labels on the evolution of choice was found to be
positive but small
We did not find evidence generally that a larger share of private labels (at national or
local level) curbed choice. If anything, at least up to a moderate level, it is associated
with slightly more choice, except in the case of the range of product prices where a
larger share of private labels in a given product category and shop was associated with
slightly less choice. However, beyond a certain level (which varies depending on the
product category) it appears that a higher share of private labels is associated with less
product variety.
34
Summary of econometric results for key drivers: choice
Product supplier Product price

Low Product variety Product size variety
Driver variety variety
Dim.
Sign Signif. Import. Sign Signif. Import. Sign Signif. Import. Sign Signif. Import.
Modern retail concentration
Procurement (national) level  .. .. .. .. .. .. .. .. ..   
Local level .. .. ..   .. .. .. .. .. .. ..
Supplier concentration at procurement (national) level .. .. ..   .. .. .. .. .. .. ..
Imbalance between modern retailers and suppliers at procurement level .. .. .. ?  .. .. .. .. ?  

Private labels
National level .. .. .. .. .. ..   ..   ..

Local level   ..   ..   ..   ..
Product category turnover (sales) at procurement (national) level            ..
New shop opening in the local area         ..   ..
General economic drivers
Unemployment   ..   ..   ..   
GDP per capita          ?  ..
Population .. .. .. .. .. .. .. .. .. ..   ..
Population density            ..
Shop type   not   not   not ?  not
app. app. app. app.
Shop floor space            ..
The ‘Low Dim column shows :
 where the indicator varies only over time and countries, so that there are few observations from which to draw conclusions.
The ‘Sign’ column shows
 positive impact (when the driver increases in value)
 negative impact (when the driver increases in value)
? where the sign varies according to whether the parameter is estimated over the long or short data sets
If an estimate was found to be statistically significant at 5% level or lower, the ‘Signif.’ column shows:
 significant at 5% level
 significant at 1% level
For statistically significant drivers, the ‘Economic importance’ (Import.) column shows the scale of impact of the driver on the dependent variable when the driver is
increased by one standard deviation above its mean value (both based on the sample used for econometric estimation). The symbols used are:
 an impact of more than 5%
 an impact of more than 10%
Where a driver is not statistically significant or economically important according to these thresholds, this is denoted by the symbol ‘..’
35
Executive summary
1.8. Conclusions regarding factors driving innovation

The main drivers were found to be level of employment of the region in which the shop is
located, measure of retailers’ business expectations, the national turnover in the product
category, certain shop characteristics (format, floorspace) and the presence of a new
shop opening in the local area: these all had positive impacts on innovation.
Concentration of modern retailers and suppliers also had either positive or negative
impacts on innovation.
As for choice, shop type and shop size had an obvious impact on
innovation
Evidence was found, as expected, that hypermarkets offered the largest number of
innovations and discounters the least, and that, for any given format, larger shops
offered a larger number of innovative products. The impact of being a discounter,
relative to being a hypermarket, was greater for innovation than for choice, suggesting
that the (narrower) range offered by discounters tended to focus on less innovative
products.
Some general economic drivers have had strong impact on innovation
The rate of unemployment in the region was found to have a generally important
negative impact on innovation. The risk associated with difficult economic times or areas
of higher unemployment may discourage suppliers from developing innovations and
retailers from offering new innovative products at these times and in these locations.
A measure of retailers’ business expectations was found to have a large positive
impact for some measures of innovation, suggesting that a positive macro business
environment encourages both suppliers to develop product innovations, and retailers to
list them. However, it should be noted that this measure has relatively few observations
because it varies only across time and Member States.
Product category turnover at national level was found to show a positive
relationship with some types of innovation. Product categories with high sales turnover
are also those where suppliers are more likely to develop innovations. The relationship
may be negative in the short period due to the effect of the crisis, whereby suppliers may
invest less in research and development or limit their marketing expenditures despite
product categories continuing to grow in size.
Some evidence was found that g reater concentration among modern
retailers at a local level was associated with less innovation
At local level, the estimated impact of modern retail concentration on most innovation
measures was negative but only statistically significant for some indicators or time
periods; the only clear negative impact that was found was for new packaging,
suggesting a tendency for a higher number of innovations to be found when
concentration is low.
At national level, modern retail concentration seems to have had various impacts on
innovation. However, there are only a very small number of observations for national
modern retail concentration because it varies only across time and Member States. Also,
the Member States in the sample did not include those with the highest level of national
modern retailer concentration.
Competition in the form of a new shop opening in the local area
stimulates some improvement in the innovation offered in existing
shops
The presence of a new shop opening in the local area was associated with more new
product innovations available to consumers in existing shops.
36
Greater concentration among suppliers at procurement level was

associated with less innovation
For some of the indicators of innovation, a negative impact was found for supplier
concentration, consistent with the principle that pressure to innovate is stronger when
competition is stronger (concentration is lower).
The impact of the measure of imbalance between modern retailers and
suppliers was generally found to be consistent with those found
separately for retail concentratio n and supplier concentration
A larger imbalance away from suppliers and towards modern retailers was generally
found to be associated with more innovation, reflecting in particular the finding that
greater supplier concentration was associated with less innovation. But the Member
States in the sample did not include those with the highest level of national retailer
concentration.
There was little evidence of an impact of private labels on the
evolution of innovation
We did not find evidence generally that a larger share of private labels (at national or
local level) curbed innovation: the impacts were generally found to be negligible in size
even when statistically significant. However, beyond a certain level (which varies
depending on the product category) it appears that a higher share of private labels is
associated with fewer innovative products being offered.
The economic importance of the drivers was generally larger for
innovation than for choice, although results were not consistent
across different innovation measures
In particular, a different result was often found for the number of new packaging
innovations compared with the other measures. Results also varied substantially between
the long and short data sets, suggesting that behaviour changed during the recession
due to other factors apart from those captured by the drivers included in the model
applied in this study.
Other drivers identified from the case studies
Evidence from the case studies suggested that, for fresh non-barcoded products, the key
driver of positive innovation evolution over the past decade was the organisation of the
supply chain. For tomatoes in Belgium, seed houses have initiated new product
development, thanks to increased research and development effort; whilst for apples in
France, Club Organisations have been the key factor in creating the conditions for
introducing new breeds.
37
Executive summary
Summary of econometric results for key drivers: innovation
New range
Low Opus innovations New products New packaging New formulations
Driver extensions
Dim.
Sign Signif. Import. Sign Signif. Import. Sign Signif. Import. Sign Signif. Import. Sign Signif. Import.
Modern retail concentration
Procurement (national) level           ?     
Local level 
.. .. .. .. .. ..    .. .. .. .. .. ..
Supplier concentration at procurement level

   .. .. .. ?        
Imbalance between modern retailers and suppliers at
procurement (national) level
       ?        
Private labels
National level

?   .. .. .. .. .. .. .. .. .. .. .. ..
Local level 
  .. .. .. .. .. .. .. .. .. ..   ..
Product category turnover (sales) at procurement
.. .. .. ?      ?     
(national) level
New shop opening in the local area .. .. ..    .. .. .. .. .. .. .. .. ..
General economic drivers
Unemployment

         .. .. ..   
Retailer business expectations     ?           
Population .. .. .. .. .. .. .. .. ..    .. .. ..
Population density .. .. .. .. .. ..       .. .. ..
not not not not not
Shop type 
 
app.
 
app.
 
app.
 
app.
 
app.
Shop floor space               

The ‘Low Dim column shows
For statistically significant drivers, the ‘Economic importance’ (Import.) column shows the scale of impact of the driver on the dependent variable when the driver is
increased by one standard deviation above its mean value (both based on the sample used for econometric estimation). The symbols used are:
Where a driver is not statistically significant or economically important according to these thresholds, this is denoted by the symbol ‘.. ‘.
38
Reminder of objectives
2. Reminder of objectives
This chapter presents a reminder of the motivations and objectives of the present
study and the work that has been undertaken.
2.1. Objectives of the study

In late 2012, the European Commission announced that, following calls by
stakeholders, it would commission a study to assess the impact of recent
developments in the European food retail sector on consumer welfare. A call for
tenders was published in the Official Journal on 19 December 2012.
This study intends to look at the economic impact of changes in the food modern retail
sector on consumer welfare. In particular, it aims to identify the impact of retail
concentration and supplier concentration at the procurement market on choice and
innovation and provide robust empirical evidence on this impact. The study goes
beyond retailer and supplier concentration to assess other factors, such as shop type
and size, and socio-demographic characteristics, to account for structural differences
that may influence choice and innovation. On the other hand, the study does not focus
on the developments impacting food manufacturers, for instance, volatility of food
commodity prices, energy costs, food safety regulations, globalisation, tax issues.
In more specific objective, the study aims to:
 Provide detailed results, key facts and in-depth analysis on the evolution of
choice and innovation over the last decade in the EU food retail sector and
analyse the trends;
 Identify and qualify the main drivers of choice and innovation and measure
their evolution over the last decade;
 Verify whether retailer and supplier concentration at both procurement market
and local levels, and the ratio between these two concentration measures are
important factors impacting choice and innovation;
 Qualify other factors such as shop type, shop size, private labels success, and
socio-demographic characteristics; and measure their impact on choice and
innovation.
The following tasks have been performed to address these objectives:
 Definition of choice and innovation and identification of the a priori drivers
through workshops and expert views;
 Descriptive analyses of the evolution of choice and innovation, and the a priori
drivers;
 Econometric analyses to identify the correlation between the observed
evolution and their drivers;
 Six case studies in order to complement and complete any quantitative data
gaps and serve as illustrative examples in support of statistical and
econometric analysis.
The results of the case studies are available at the following address:
http://ec.europa.eu/competition/publications/reports/retail_study_cases_en.pdf
2.2. Motivations behind study

The impact of modern retail developments on the EU food supply chain has been a
controversial subject over the past several years.
39
The conomic impact of modern retail on choice and innovation in the EU food sector
 On one hand, modern retail is acknowledged as being a very competitive

sector, with a positive impact on consumer prices, given that, arguably,
competition between retailers limits food price increases11.
 On the other hand, claims have been made for action from different actors
(including, consumer and manufacturer organisations) stating that choice and
innovation are being jeopardised by retailers’ practices, and the growth of
private labels may increase the power of retailers vis-à-vis their suppliers and
lead to a deterioration of choice and innovation, therefore impacting consumer
welfare. However, these claims have not been sufficiently substantiated. 12
Turbulence in commodities markets beginning in 2007 and subsequent consumer price

trends revealed by EU market monitoring reports 13 on the food supply chain has
brought the retail sector under intense scrutiny over the past few years. 14 However, it
is worth noting that the issue of the increasing concentration of the retail sector has
long been on Europe’s political agenda15. Recently, a number of consumers and
supplier organisations as well as competition authorities16 have argued that increasing
buyer power in retail is giving rise to practices which may have detrimental effects on
the welfare of both consumers and, on the procurement side, producers and suppliers.
The European retail industry has questioned this scrutiny, citing the lack of empirical
evidence to support these claims and has tended to point out that trading relations are
very complex and the arguments concerning abuses of buyer power simplistic and
unsubstantiated17. Opponents have also pointed out that while producers and
suppliers may be facing challenging prospects, these problems are more directly linked
with the need to restructure the sector than buyer power abuses. 18
11 Commission Staff Working Document (2009), “Competition in the food supply chain”
12 See, e.g. DG ENTR (2011), “The impact of private labels on the competitiveness of the European food supply chain”.
EuroCommerce, “Own brands: Increasing consumer choice and driving innovation”, December 2010. ERRT, “Retail and innovation”,
ERRT contribution to the 2nd European Commission Workshop on the Retail Action Plan.
13 For more details see, Commission communication on a better functioning food supply chain in Europe (COM(2009) 59)
14 See, inter alia, Resolution of the European Parliament of 26 March 2009 on ‘food prices in Europe’; Commission communication on
“A better functioning food supply chain in Europe” (COM(2009) 591); EP, Report on a more efficient and fairer retail market
(2010/2109(INI)); EP, Report on fair revenues for farmers: A better functioning food supply chain in Europe (A7-0225/2010); High
Level Forum for a Better Functioning Food Supply Chain, draft report, October 2012; Commission Communication on “Setting up a
European Retail Action Plan” (COM(2013) 36 final); Green Paper on Unfair Trading Practices in the Business-to-Business food and
non-food supply chain in Europe, (COM(2013) 37 final)
15 See e.g., “Buyer Power and its Impact on Competition in the Food Retail Distribution Sector of the European Union”. EC, 1999
16 See e.g., Press release “Copa Cogeca welcomes European Commission plan to improve functioning of food supply chain, but
argues more action is vital”. Available at: http://pr.euractiv.com/pressrelease/copa-cogeca-welcomes-commission-plans-improve-
functioning-food-supply-chain-arguesmor?page=44; Press Release “FCA study shows that daily consumer goods trade uses its
buying power in several ways that are questionable for competition”. Available at: http://www.kilpailuvirasto.fi/cgi-
bin/english.cgi?luku=news-archive&sivu=news/n-2012-01-10
17 EuroCommerce Position Paper on the “Green Paper on Unfair Trading Practices in the Business-to-Business food and non-food
supply chain in Europe”. April 2013.
18 Press Release “EP vote on Bové report misrepresents the realities of food supply chain”. Available at:
http://pr.euractiv.com/pr/ep-vote-bov-report-misrepresents-realities-food-supply-chain-90100
40
Whilst quantitative evidence to date is insufficient to robustly support these claims, a

number of stakeholders across Europe have argued that buying power of retailers
and/or suppliers has a negative effect on choice and innovation for consumers and, in
the long term, the competitiveness of the European food supply chain itself. According
to these arguments, the asymmetrical bargaining power has a negative effect on
incentives to invest and/or innovate at different levels of the food supply chain for
smaller actors.
Drawing the link between the effects of buyer power abuses on suppliers and long
term impacts on consumer welfare, a 2012 report by Consumers International, a world
confederation of consumer groups, claimed that the abuse of buyer power is widely
and routinely practised against suppliers and that, if not immediately, certainly over
time, such abuse will inevitably damage consumers too. The report claims that the
downward pressure on supply prices threatens their viability and, in the long term,
threatens choice, innovation and quality for consumers 19.
A 2012 report from the Spanish NCA also draws this link, saying that while strong
bargaining power has an immediate positive effect on consumer prices, in the long
term, the greater bargaining power [of retailers] may reduce the capacity and
incentives for suppliers to invest and innovate. If suppliers expect they will not be able
to capture an adequate portion of the overall profits, they will have less incentive to
spend on capacity and innovation20. Similarly, a report by the Finnish NCA also states
that one consequence of the further consolidation of the grocery retail sector may be
the narrowing of consumer choice and…drying up of innovation21.
Nevertheless, the anecdotal empirical evidence that does exist cannot currently
support these claims. For instance, a 2011 report by the Swedish NCA which examined
the effects of the exercise of buyer power by retailers in the value chain concluded
that there was no empirical evidence as to whether it had an impact on innovation and
on the range of products offered by food producers 22.
2.3. Structure of the final report

This report provides detailed answers to the study questions as developed in the
analytical framework as follows:
Descriptive statistics:
 How has choice in the EU food sector evolved over time and across MS?
 How has innovation in the EU food sector evolved over time and across MS?
 How have the a priori drivers of retailer and supplier concentration evolved
over time and across MS?
 How have the other a priori drivers of choice and innovation evolved over time
and across MS?
 To what extent are the drivers of choice and innovation associated with each
other over time and across MS?
Econometric analysis:
19 “The relationship between supermarkets and suppliers: What are the implications for consumers?”. Consumer International, 2012
20 “Report on the relations between manufacturers and retailers in the food sector”. Comision Nacional de la Competencia, 2012.
21 ‘Study on trade in Groceries: How does buyer power affect the relations between trade and industry’. Finnish NCA, 2012.
22 Konkurrensverket, Mat och marknad — från bonde till bord, April 2011
41
 How have the a priori drivers of retailer and supplier concentration impacted
upon choice and innovation?
 How have the other a priori drivers of choice and innovation impacted upon
choice and innovation?
Prior to responding to each of the study questions addressed in the descriptive

statistics and econometric analysis sections, the report provides a background
literature study into the modern retail sector in Europe and the evolution of the main
characteristics over the last decades. This chapter serves as an introduction to the
subsequent findings. In addition, a chapter on methodology is provided, with the
objective of providing methodological and scope aspects of the study.
2.4. Limitations of this report

It is important to outline a number of general limitations, which need to be taken into
consideration.
 Sources: this study draws on data and calculations generated from a wide
range of statistical databases. For each measure the most reliable data source
has been sought. The quality of the statistical databases however has not been
verified.
 Data availability: the approach used was to identify and select CSAs strongly
influenced by the availability of modern retail shop assortment data through
©Nielsen Opus. Given that the approach only considers shops that have been
audited twice per year (in summer and winter) from 2004 to 2012, the shops in
the sample tend to be located in areas of strong competition. The assortment
of a given shop is audited by Nielsen at the request of a competitor.
 Scope: the time scope and geographical scope addressed in this report is highly
dependent on data availability. Analyses address the largest possible scope of
MS and longest time period to the best possible extent. However, some
measures cover only a narrow scope due to data limitations, e.g. choice in
shops has been measured on 4 MS between 2004 and 2012, because of the
limited availability of © Nielsen Trade Dimensions data on the long time period.
In addition, in terms of product scope, fresh products are not covered in the
descriptive statistics and econometric analysis as they are not barcoded
products that enable choice and innovation to be measured. As a result, a
selection of fresh products has been addressed in the case studies.
 Comparability: this study has sought to maximise the sample size for each
variable and driver being measured. As a result, the scope of Member States
can differ according to the specific measure in question. Caution should
therefore been made when comparing results across different measures.
Furthermore, some results presented at MS level reflect results across the
selected CSAs within that MS – as a result the findings do not represent all
situations in the sample MS.
 Innovation: the definition of innovation from an operational perspective for this

study is the introduction of a new EAN code. The Consortium team has not
sought to qualify what should and should not be considered a genuine
innovation. Therefore, the number of innovations in this study is synonymous
with the number of new EANs that appear in the assortments across the
42
sample (with the exception of EANs identified as promotions, which have been
excluded), whilst the different categories of innovation have been identified
through applying data from © Mintel GNPD. Two different sources have been
used for innovation, and therefore the absolute numbers according to each
source cannot be reconciled.
 Price data: We have found some © Nielsen Opus product price data to be
inconsistent in terms of units and currency across shops and time periods.
Where possible incorrect data has been removed from calculations, however
given the volume of data, the removal of all inconsistent prices cannot be
ensured.
2.5. Different tasks of the study
2.5.1. Task 1
Task 1 was completed through the submission of the first progress report in July 2013.
Over the first months of the study, the key study concepts were refined and
operationalized through internal discussions, workshops, and the organisation of an
online focus group with external individual experts, which enabled the consortium to
identify relevant literature to be used, to develop and validate definitions of the key
concepts of choice and innovation, as well as address their operationalization and
measurement. A list of a priori drivers was established, which has set the foundation
for the descriptive statistics, and have been applied therefore to the econometric
analyses.
In addition, the key questions the study poses were reviewed, broken down into
constituent sub-questions and a data source mapping was conducted to ensure the
coherence of the study’s approach, the robustness of the data collection strategy and
the efficient articulation of the various data collection tools, namely the quantitative
and qualitative analyses. The development of this analytical framework has served as
a guide throughout the different stages of the study.
Furthermore, the approach was refined for the following methodological aspects, in
order to best align the types of data available with the representation of a variety of
different living and retail situations in Europe:
 Sampling approach and validation of the selection of MS, regions and CSA;
 Precise definition of Catchment Areas (CA) and CSA
 Selection of product categories;
 Selection of timeframe;
 Identification of additional data providers and final selection, study of data
limitations and preparation for data purchase.
Finally, in light of data limitations, the gaps identified and the comprehensive
analytical framework constructed, the case study approach has been refined to take
these data needs into account.
2.5.2. Task 2
Following the acceptance of the Task 1 First Progress Report, the Consortium
proceeded with the purchase of the agreed data sources. During August to September
2013, significant work was undertaken acquiring the various data sources, reviewing
their quality, consolidating them into a database by MS, and running queries to enable
descriptive statistics to be produced. The process of integrating different sources
presented many challenges due to the differing nature of each data source and the
high complexity and extensive quantity of the data.
43
During this process of data consolidation and treatment, it became apparent that the
data provided for the analyses at shop level and CSA level was incomplete in terms of
time periods covered across all selected MS. As a result, database construction could
not commence, as having the full set of data per MS (all time periods and all shops) is
an important prerequisite. In light of the data gaps, the scope of the study was
completely revised, in terms of MS, geographical regions, and time scope. In parallel,
a number of methodological points were revisited following preliminary analyses
undertaken – concerning specifically drive time rules for CSAs and CAs. A revised
scope of MS, mixed periods of analysis, CSAs and shops was validated with DG COMP
on 29 October 2013, which is detailed in the following section of this report addressing
scope and methodology revisions.
The complete second progress report was submitted to DG Competition on 28 March
2014. This report provided descriptive statistics on the evolution trends of choice,
innovation and their a priori drivers over the 2004 to 2012 period.
2.5.3. Task 3
Following completion of the construction of the data set for all indicators in early 2014,
the Consortium proceeded with the econometric analysis compiled in an Interim
Report. The Consortium held an internal workshop to compare conclusions from the
descriptive analysis (Task 2) and the preliminary conclusions from the econometric
analysis in March 2014, and a workshop with DG Competition was held in April 2014.
Comments from these workshops were incorporated in subsequent rounds of
econometric analysis.
2.5.4. Task 4
The Consortium proceeded to the completion of 6 case studies. The objective of the
six case studies was primarily to complement the main findings of the econometric
analysis with qualitative inputs. They covered key fresh products (both EAN and non-
EAN) in specific markets and certain areas that cannot be addressed due to data
limitations, but which are considered essential in order to have a more representative
picture of the impact of modern retail on choice and innovation. They enabled the
consortium to observe through concrete examples the reality of the trends analysed
from the econometric analyses, and to understand how and why the drivers impact
choice and innovation.
2.5.5. Tasks 5&6

Task 5 and 6 consisted of the drafting of conclusions by the Consortium and the
presentations that will take place with the DG COMP Food task force. A conference in
October 2014 prepared by the Consortium and DG COMP aims to present the final
results of the study to major European stakeholders.
44
Background of the study
3. Background of the study

This section presents an overview of important characteristics of the modern retail
food sector in Europe, and an overview of the key macro trends impacting the
evolution of this sector. The objective of this section is to provide contextual
background to the motivations of the study.
3.1. Europe retail sector in brief

According to the latest figures, the overall retail sector represents 4.3% of the Gross
Value Added in the EU economy 23, over 8% of employment24 and 3.7 million SMEs25.
This sector interacts with both an upstream procurement market and a downstream
consumer market. In the upstream market, retailers provide producers and suppliers
with critical access to millions of final consumers through their distribution channels,
as well as a number of different parallel services, such as logistics and product
merchandising depending on the sector. In the downstream market, retailers offer
customers access to an assortment of products, as well as information on those
products through advertising or staff. It is important to take into consideration the two
different dimensions of the retail sector, in order to properly situate it in the broader
supply chain and fully understand the role it plays in the European economy.
Furthermore, retailers are key players in the functioning of the Internal Market,
allowing consumers to access goods from many different MS. Thus, the functioning of
the food supply chain has important ramifications on consumers (given that
approximately 13 % of their household expenditure is spent on food 26), as well as the
functioning of a number of other essential economic sectors, such as agriculture, the
food processing industry and retailers. Taken as a whole, the food supply chain
generates value added of €715 billion per year, almost 6% of the EU GDP. 27
3.1.1. Definition of modern grocery retail

The scope of the present study concerns specifically modern grocery retailing,
covering hypermarkets, supermarkets and discount stores. Modern grocery retail sales
account for 54% (in 2012, same in 2004) of the total edible grocery sales in the EU28.
Edible grocery sales represent 42% (in 2012, 43% in 2004) of total EU retail sales29.
The distinction between ‘modern retail’ and what is sometimes referred to as
‘traditional retail’ can be difficult to define. As mentioned in the tender specifications
for this study, the ‘modern retail’ concept refers to distribution channels that emerged
over the past 30 to 40 years. Traditional retail can be characterised largely as small,
23 http://ec.europa.eu/internal_market/retail/
24 Commission staff working document on “Retail services in the Internal Market” (SEC(2010) 807) - Eurostat, National Accounts
Statistics, 2007 (share of NACE G52 of total Gross Value Added at basic prices) and for the EU27. Employment data are from EU
KLEMS, 2007, and for the EU 25.
25 European Commission retail market monitoring report “Towards more efficient and fairer retail services in the internal market for
2020” (COM(2010)355 final)
26 Eurostat, (TSDPC520), Average final consumption expenditure of households for food and non-alcoholic beverages across EU 27
(% of the total expenditure)
27 High Level Forum for a Better Functioning Food Supply Chain. Draft report, October 2012
28 EY analysis based on © Planet Retail
45
independent and often family-owned businesses with non-organised distribution

channels. From the 1950s onwards, this structure has gradually given way to larger
and more diverse store formats offering a wider assortment of goods, sometimes
highly integrated into a sophisticated supply chain, with ownership concentrated in a
small number of national or international retail groups.
It is challenging to put forth a single definition for modern retail, as retail is very
culture-specific, and its development has been highly influenced by contextual
elements unique to each market, as diverse as local legislation, consumer behaviour,
geographic characteristics and the structure of pre-modern channels of distribution. A
number of different characteristics often associated with modern retail include:
 Group of shops with the same banner integrated in a network
 Shop size and format (hypermarkets (>= 2 500m²), supermarkets (400 -
2 499m²), discounters (all sizes))
 Assortment of goods offered (the number of Stock-keeping units - SKUs,
different product categories)
 Self-service formats
 Technology and equipment
 General business practices (logistic, marketing…)
To illustrate the heterogeneous definitions modern retail can represent, one can
consider the development of various retail formats often associated with modern
retail. The lifecycles of these formats can vary drastically from one market to another.
Thus, whilst the development of hypermarkets may be indicative of modernity in
certain markets, notably the emerging markets, in others, the rising share of
discounter stores is a much more apt indicator for tracing the evolution of retail over
the past decades.
The definition of modern retail adopted by this study takes into account size (sales
area) and shop type, and thus indirectly assortment and different organisational
models. Thus, modern grocery retail is defined in this study as including hypermarkets
(>= 2 500m²), supermarkets (400 - 2 499m²) and discount shops (all sales area
sizes).
Consequently, this definition disregards independent and traditional shops, as well as
“new modern retail”, including e-commerce, drive-through markets, frozen food
shops, organic food shops, fresh product shops, and very small supermarkets
(<400m²) such as convenience stores. The “new modern retail” formats only
represents a small share of edible grocery sales – in 2012, grocery e-commerce
represents 1.2%, frozen food stores 0.7%, health food stores 0.1% and convenience
stores 4.5% (slightly up from 3.6% in 2004) - for a combined total of 6.5%. 30
Therefore the exclusion of these formats from the study will have limited impact on
the trends observed in the grocery retail sector in the EU over the last 10 years.
Finally, it is important to mention that whilst one of the important distinguishing
trends of modern retail has been the diversification of the assortment offered, both in
the edible and non-edible offering, the scope of this study will specifically look at the
edible assortment and thus disregard other durable and non-durable goods commonly
sold in modern food formats, as well as other service markets, such as banking and
mobile phones, in which large retailers have increasingly begun to enter.
46
3.1.2. Prevalence of modern grocery retail in the EU

Modern retail, as defined by the current study, constitutes a large portion of the
grocery retail market in most of the EU 27 markets (greater than 70% in 12 MS, and
greater than 50% in 19 MS). Conversely, traditional retail still makes up a majority of
the edible grocery market in many newer MS, with Romania and Bulgaria topping the
list of traditional edible grocery retail market share (20% to 30% of modern retail in
2011).
However, these less mature markets are also the most dynamic in terms of the growth
of modern retail and the productivity gains being achieved. Whilst the share of modern
retail in the edible grocery market increased by only 5.7% points in France and 0.7%
points in Germany from 2004 to 2012, this growth reached 20.9% in Romania and
18.9% in Poland over the same period31.
Figure 1 illustrates the share of modern retail grocery sales as a proportion of the total
edible grocery market in 2000 and 2011.
Figure 1: Evolution of the market share of modern retail compared to total edible
grocery market (2000 - 2011)
100%
90%
80%
70%
62%
60%
50% 44%
40%
30%
20%
10%
0%
2000 2011
Source: EY analysis based on © Planet Retail, 2011
This figure presents an opportunity to draw the reader’s attention to a number of

caveats arising from the definition of modern retail adopted for this study. Firstly, the
full extent of modern retail is slightly understated using the definition that has been
adopted, because it does not include convenience stores, which account for 4.5% of
the edible grocery market in the EU, and up to 10% in the UK and in Ireland 32.
Secondly, in some countries (such as Italy) the grocery retail market is highly
fragmented and regionalised with a predominance of traditional stores (including fruit
31 Estimates based on © Planet Retail
47
and vegetable market, non-branded neighbourhood stores, butchers and bakers)33.

For this reason, the share of formats defined as modern retail is lower in these
countries.
These specific situations however have not had a significant impact on the study, as
modern retail is predominant in the majority of the MS that form the study sample (6
of the 9 MS have a modern retail share of greater than 60%). The remaining 3 MS
enable the study to address a variety of retail situations that reflect the situation in
Europe as a whole.
3.2. Recent evolutions in the grocery retail sector in the EU

A number of key trends can be identified in the development of the retail sector in the
EU over the past two decades. The descriptions hereunder focus specifically on how
these trends have impacted the grocery retail sector.
These trends, covered in the following sections, include:
 Increase in number of shops and commercial sales areas;
 Changes in retail structuration;
 Development of private labels;
 Expansion of retail groups as they look for new sources of growth;
 Development of technologies in parallel with structured back-office
organisations.
3.2.1. Increase in number of shops and commercial sales areas

Many modern retailers have increased their number of shops and commercial sales
areas, as illustrated in Figure 3, changing the commercial landscape for a majority of
consumers in the EU:
 Hypermarkets (>= 2,500 m2) sales areas have increased by more than 46%
between 2000 and 2011 in EU, with 2,681 additional outlets (523 of which
opened in the UK).
 Supermarkets (400 to 2,499 m²) are the leading food shop type in Europe with
44 177 m² of food sales area in 2011. Between 2000 and 2011, the total sales
area increased by close to 26%, with 5,326 additional outlets
 Discounters have increased their sales areas by 81% between 2000 and 2011
across the 27 MS, with 12,778 additional outlets.
48
Figure 2 : Evolution of the European food retail (in number of outlets) by type of shop
(2000-2011)
Total = 105 117
Hypermarkets: + 72% 6 372

Total = 84 332
3 691
39 887
Discount stores: + 47%
27 109
53 532 Supermarkets: + 10% 58 858
2 000 2 011
Supermarkets Discount stores Hypermarkets
Source: EY analysis based on © Planet Retail

Figure 3: Evolution of the European food retail sales area (in thousands of m²) by type
of shop (2000-2011)
Total = 96 391k sqm

Total surface : + 44%
Total = 67 069k sqm

23 146
Hypermarkets: + 46%
15 876 29 068
Discount stores + 81%
16 093
Supermarkets : + 26% 44 177

35 101
2000 2011
Supermarket Discount store Hypermarkets
The evolution of shop types potentially has had a major impact on choice of food
products for consumers, as different shop types have different assortments:
 Hypermarkets typically have the broadest assortment (20,000 Stock-Keeping
Units (SKU) is a common figure for food products) among all food shops
because of their superior shelf-space.
 Supermarkets typically sell 5,000 to 10,000 different food SKUs.
 Discounters have the narrowest assortment, typically between 1,000 and 2,000
SKUs.
The assortment offered by grocery retailers is highly related to the shop format they
are operating. In the first self-service supermarkets of Europe, it was the breadth of
food products offered that was one of the major symbols of the advent of modern
retail. Hypermarkets have sufficient sales area to offer a large choice of non-food
products, in an effort to diversify revenue streams and stock shelves with high margin
non-grocery items to offset low-margin staples. Some hypermarket and superstore
49
formats provide additional services from restaurants and cafés to beauty salons
(Tesco) and banking (Asda, Sainsbury).
Discount stores have experienced the highest growth in terms of outlets over the past
decade. Whilst this format has proven successful throughout Europe, it is in Germany,
where discounters captured more than 33% of the edible grocery market share in
2012, that the format has most profoundly changed the landscape of grocery retail.
The basic concept behind the discounter format is to provide consumers with highly
competitive prices, but a limited assortment. The limited assortment as well as the
polyvalent function of staff, has led to considerable economies, allowing the format to
remain competitive.
3.2.2. Overall concentration of the total edible grocery market.

As a whole, the concentration of the total edible grocery market has increased over
the period of observation, mainly due to the growth of modern retailers.
The evolution of market shares for the top retailers has differed across the various MS
over time; however the overall trend is towards increasing concentration at national
level. For instance the figure below shows that the top 5 retailer market share at
national level (not necessarily the same 5 in each MS) exceeds 60% in 13 MS,
representing 52.8% of the EU population in 2011, in comparison with 8 MS in 2000,
representing 38.4% of the population.
Figure 4: Evolution of the combined market shares of the top 5 retailers C(5) per MS
(2000 - 2011)
C(5) > 80%
60% < C(5) < 80%
40% < C(5) < 60%

2011
2000
20% < C(5) < 40%
C(5) < 20%
Data not available
Source: Planet Retail

Source : PlanetRetail – “Modern retail” equals the following categories : hypermarkets, hypermarkets & superstores, superstores,
supermarkets, supermarkets & neighbourhood stores, discount stores and discount superstores
Source : PlanetRetail – “Modern retail” equals the following categories : hypermarkets, hypermarkets & superstores, superstores,
supermarkets, supermarkets & neighbourhood stores, discount stores and discount superstores
3.2.3. Structuration of modern retailers

Three major trends have been observed at EU level across MS in retailers’
structuration:
 Increasing share for modern retailers in terms of the total edible grocery
market
50
 Increasing market shares for 10 top retailers through either organic growth or
acquisitions
 Organisation of retailers in buying groups and alliances
Each of these characteristics will be presented in the analysis below.

Increasing share of modern retail
Modern retailers have expanded and increased their influence in most EU MS, but the
situation remains heterogeneous across the different MS, as illustrated in Figure 1. In
comparison to other sectors, retail concentration in the market for total edible grocery
(including all types of shops) is relatively low (though also increasing). The market
share of modern retail (limited to hypermarkets, supermarkets and discount stores)
has generally increased in mature MS, but also in MS where modern retail has
developed more recently. Modern retail share has increased in 24 MS, whilst only
slightly decreased in 2 MS (Germany and Sweden), over the period 2000 to 2011.
Malta has been excluded from the analysis due to a lack of available comparable data.
Member States can generally be grouped into three categories:
1) MS where modern retail had already developed before 2000 (at least 65% of
total food market) and has maintained this high modern retail market share –
this includes for instance Germany, Finland, UK, Sweden, France, Luxembourg
or Austria.
2) MS where modern retail has developed particularly since 2000 and is
significantly higher in 2011 than 2000 (at least 20 percentage points) – this
includes for instance the Netherland, Estonia, Belgium, Ireland, Spain,
Portugal, Hungary or Slovakia
3) MS where modern retail has developed significantly but remains relatively low
compared to other MS (less than 50% of total food market in 2011) – this
includes Latvia, Poland, Cyprus, Bulgaria or Romania.
It is interesting to note that some MS with the lowest modern retail share in 2011
(e.g. Romania, Bulgaria, Latvia and Slovakia) have experienced the greatest growth
over the following decade, whilst other MS with a similarly low share in 2000 (such as
Italy and Greece) have not witnessed the same extent of growth over the last decade.
Increasing market shares for top 10 European retailers
At the pan-European level, the top 10 European retailers accounted for 26% market
share in 2000, compared to 30.7% in 2011, representing an increase of +4.7 points
(see
Table 1 below). It is interesting to note that, whilst some retailers increased their
market shares and others lost market share during this period, the top 10 European
retailers have remained the same.
Table 1: Market share (in edible grocery sales) of the top 10 retailers in EU (2000 -
2011)
Top 10 in 2000 Top 10 in 2011
% EU % EU
Edible grocery Edible grocery
Company market Company market
banner sales (€ M) banner sales (€ M)
share share
Carrefour 44 441 5.2% Schwarz Group 50 059 4.7%
ITM (Intermarché) 27 308 3.2% Carrefour 49 267 4.5%
Rewe Group 23 355 2.6% Tesco 40 310 3.8%
Tesco 23 034 2.7% Edeka 37 031 3.4%
51
Top 10 in 2000 Top 10 in 2011
% EU % EU
Edible grocery Edible grocery
Company market Company market
banner sales (€ M) banner sales (€ M)
share share
Edeka 21 654 2.5% Aldi 33 529 3.1%
Aldi 21 268 2.5% Rewe Group 32 324 3.0%
Ahold 15 811 1.9% Auchan 23 378 2.2%
Schwarz Group 15 471 1.8% ITM (Intermarché) 22 668 2.1%
Auchan 15 234 1.8% Leclerc 22 509 2.1%
Leclerc 14 311 1.7% Ahold 19 851 1.9%
TOTAL 221 889 26.0% TOTAL 330 926 30.7%
The two past decades have been marked by a number of important joint-ventures,
mergers and acquisitions in the retail sector, either to penetrate new markets, or
consolidate positions on domestic markets, particularly in the face of increased
international competition. As an example, in 1999, Carrefour’s merger with Promodes
created Europe’s largest and the world’s second largest retailer to Wal-Mart. Other
notable mergers at this time involved Rewe/Meinl in Austria in 1999 and Makro/Metro
in 1998. However not all merger bids have been successful. In 1996 the European
Commission prohibited the proposed merger between Kesko and Tuko in Finland which
would have created a company with a national market share of 60%. This said,
merger and acquisition plans are tending to be increasingly scrutinised by national and
European competition authorities.
Organisation in buying groups and alliances
Increasing concentration can also be seen at the procurement level, through the
development of buying groups. Buying groups are essentially a type of retail
purchasing alliance, at a regional, national or international level. In essence, a buying
group is an organization created by several shops or retailers with the aim of
improving their purchasing conditions as well as enhancing their market
competitiveness compared to other types of retail players. 34
Buying groups, or procurement organisations, have existed since the 1930s but they
have developed particularly since the 1980s-1990s, a period which has witnessed the
rise of cross-border alliances. The aim of cross-border groups is particularly to
strengthen the retailers’ bargaining power through higher volumes to reduce
purchasing costs, for the procurement of large international brands or for private
labels.
Several types of buying groups have emerged, which differ by their scope and
organisation:
 Regional buying groups: group several shops operating in the same
geographical area
 National buying groups operating at national level for one or several banners or
retail groups
 International buying groups operating for one single retail group across several
geographies or several retail groups operating in different countries.
34 Bălan, Carmen, The Alliances of European Retailers and their effects in the field of marketing and supply chain, The Romanian
Economic Journal, 2007
52
All these types of buying groups aim to strengthen the retailers’ bargaining power
through higher volumes to reduce purchasing costs, for the procurement of
manufacturer brands or for private labels.
Specifically relating to private labels, as an example, in 2010 French retailer Auchan
and Metro Cash and Carry formed a purchasing collaboration to expand their
respective private label businesses. Under the agreement, Auchan granted Metro Cash
and Carry (wholesale arm of Germany's Metro Group) access to its private label
supplier network in order for Metro to gain better buying prices.35
Retailers have also created international alliances to respond to the increased
internationalisation of suppliers. The main international buying groups in Europe are
presented in the Table 2 below.
Table 2: Selection of main international buying groups in EU
Buying Countries of operation

Selected Members
Group in Europe
Ahold, Booker, Dansk Supermarked, Delhaize, Esselunga, Hagar,

AMS 22
Jeromimo Martins, Kesko, Migros, Morrisons, Systeme U, Uniarme
EMD 20 Axel Johnson, Casino, Mercator, Musgrave Group, Norgesgruppen
Core 18 Colruyt, Conad, Coop, Rewe Group
Agenor/Alidis 8 Edeka, Eroski, ITM
Bloc 4 Cactus, Louis Delhaize,…

Retailers face challenges today in forming/joining buying groups for several reasons:
 Commercial sensitivity surrounding purchasing decisions and sharing of
information (purchasing conditions are confidential and sharing of information
is limited by law). As an example, Coopernic recently dissolved in 2013 and
was replaced by a new group called “Core”, which excluded the founding
member of Coopernic Leclerc, due to “insurmountable differences concerning
the future form and strategic focus of the group”36 according to a Rewe official.
The new alliance “Core” therefore comprises Colruyt, Conad, Coop and Rewe
Group.
 Buying groups impose a certain degree of centralization, and not all retailers
have the same approach in this area (independents are often resistant to
centralization) and even the largest groups are favouring flexible local
arrangements, with the exception of hard discounters which seem to be more
centralized.
3.2.4. Steady development of private labels

Private labels, sometimes referred to as retail ‘own-brands', are goods for which
retailers directly contract manufacturers to produce and then sell under their own
brands. These products are typically sold as lower cost alternatives to major national
and international brands, although retailers also develop upscale private label
products.
35 http://www.lsa-conso.fr/auchan-developpe-des-mdd-pour-metro,116793
36 © Planet Retail, “COOPERNIC members exclude LECLERC from new alliance”, 9 September 2013
53
Globally, penetration of private labels is high in Europe, where they can exceed 40%
market share in countries such as Switzerland and the UK, compared with an average
in the US of 18% market share in 201137. Offering lower prices is potentially a key
reason why the market share of private label has increased in grocery sales over the
years – with price being a primary concern of European consumers. Finally, the
market share of private label products has increased across most product
categories in Europe. Key reasons for this likely include a perception among
consumers that these products offer good value for money, the opportunity of higher
margins for retailers, and a profitable way for manufacturers to make use of spare
capacity.
Private labels are increasingly being seen by retailers as important tools for building
client loyalty and strengthening banner image. Thus, beyond generic and ‘mimic’
private labels, which are designed to provide low-cost alternatives or directly compete
with manufacturer’s brands, retailers have increasingly developed high quality private
labels brands that compete side by side with manufacturer’s brands or specifically
positioned product ranges, such as organic.
3.2.5. Geographic expansion as a new path of growth

Grocery retail groups in Europe have become increasingly international over the past
two decades. Whilst this may be taken as a given today, it is a fairly recent
phenomenon for the retail sector. Compared with other industries, retail is still fairly
anchored in domestic markets. Even among the most internationalised grocery
retailers, very few have succeeded in surpassing the 50% mark for turnover in foreign
countries.38
This internationalisation can be attributed to several factors. As retail markets in more
developed MS are seeing their growth rates stabilise, retailers have increasingly begun
to expand outside of their home markets and export their business models to markets
that possibly offer higher sales growth potentials. In addition, the enlargement of the
EU and the Single Market has facilitated this expansion in opening up new markets to
Western European retailers. These new markets tend to offer stronger economic
growth and historically lower levels of competition. In particular, newer MS from
Central and Eastern Europe have been popular targets for expansion for Western
European groups over the past decade39. As a result, whilst modern retail has
increased rapidly in these countries, very few of the top grocery retailers in these
countries are locally based40.
However, there are examples of restrictions put in place to limit the influence of
foreign retailers. In Hungary in January 2012, a decision was made to ban for the
three subsequent years the construction of retail outlets of greater than 300m². In
addition, new hypermarkets exceeding 10,000 m² require planning permission from
local magistrates. The aim of the policy is to support the development of the small-
37 Private labels 2013 : The Global Grocery Trends to Watch. © Planet Retail
38 Sandberg, Erik, "The retail industry in Western Europe - Trends, facts and logistics challenges". 2010, Department of Management
and Engineering, Linkoping University
39 International expansion has been particularly pronounced among French (Carrefour, Auchan) and German (Aldi) retailers, both
MS boating grocery retailers among the top ten worldwide. On the other hand, retail groups in newer MS have not enjoyed the same
level of expansion.
40 Notable exceptions include the Maxima group, a highly successful Lithuanian based retailer that is one of the biggest and most
successful in the Baltic market, with a 35% share of the edible grocery market in its home market (© Planet Retail)
54
scale retail sector; however it has had significant impacts on the expansion plans of
Schwarz Group, Tesco and Aldi.
Nevertheless, this geographical expansion has been accomplished through a number
of different models:
 The acquisition of local retailers (e.g. Jeronimo Martins acquisition of
Biedronka in Poland in 1997)
 Joint ventures with local retailers or investors (e.g. Ahold has been present in
Portugal via its joint venture with local retailer Jeronimo Martins since 1992)
 The setting-up of own subsidiaries (e.g. Casino in Brazil, Colombia and
Thailand)
 Franchising (e.g. Carrefour Poland continuing to expand smaller shop formats
via franchises)
Figure 5: Domestic share of EU grocery sales for top ten retail groups
100% 95%
89% 90%
92% 91%
71% 73%
82%
65%
61% 61% 61%
68%
62% 49%
57%
50% 2002
41% 42%
2012
For most European retailers, domestic markets remain their main market. However, as
illustrated in Figure 5, one can observe the general decrease in the importance of
home markets for top European retailers in terms of the domestic share of European
grocery banner sales, due to international expansion, with the notable exception of
ITM and Edeka, which, following strategic reorientations, have made the decision to
focus on consolidating their respective home markets. Furthermore, some retail
groups have focused their growth strategies on non-European markets. Whilst less
mature markets in the EU offer the advantages of geographical and relative cultural
proximity, the developing world can offer even more growth opportunities for
European retailers. For example, whilst the figure above shows Carrefour’s domestic
sales have remained stable over the last decade as a proportion of European grocery
sales (62%), the graph does not show that in terms of international sales, domestic
sales only represent 43% in 2012 compared to 51% in 2002. Similarly, for Tesco,
which has pursued only relatively modest expansion in Europe, the UK market still
represented only 51% of total grocery banner sales in 2012. It is important to mention
that, compared to global manufacturers, retailers benefit less from synergies through
international expansion. Each country uses specific logistics, purchasing organisations,
back-offices.
55
3.2.6. Development of technology in retail

Sophisticated technology now pervades modern retail and has played an important
role in the rationalisation of the sector and the efficiency gains achieved. Some
technological advances are readily apparent to consumers in shops, such as self-
service checkouts, price-check scanning machines, and electronic shelf labels that
allow store managers to respond instantaneously to fluctuations in prices.
Furthermore, the recent development of NFC (near field communication) wireless
technology has the opportunity to impact modern retailers, since shoppers can scan
items into their basket and then checkout by simply tapping their NFC phone to a
reader attached to the retailer's cash register.
Other technological advances are apparent outside the shop, such as retail grocery
price comparison websites (e.g. www.mysupermarket.co.uk), and the emergence of
digital walls in subway stations where groceries each have a QR code that the shopper
can scan with a smartphone camera, adds to its shopping list, pay using the phone
and have the groceries delivered.
Finally, whilst it does not fall within the scope of this study, e-commerce has also
developed significantly in the grocery retail sector over the past decade. This sector
has caught on with consumers and developed in a number of markets, most notably
the UK, France and Spain.41 E-commerce food sales are still marginal across Europe,
however, in developed markets such as the UK, online grocery sales currently account
for nearly 3% of total food sales.42 A new e-commerce concept, the “Drive”, has been
developed over the last few years in some MS, especially in France. It enables
consumers to order their food products on a website and then pick them up in store or
in alternative locations.
3.3. Macro evolutions impacting the grocery retail sector in the EU

This section highlights a number of key macro developments over the last decade
which have potentially impacted the grocery retail sector and specifically modern
retail.
3.3.1. Evolution of broader retail market and consumption due to stunted

purchasing power
Declining GDP per capita since 2008
Since 2008, the economic crisis in Europe has arguably had a number of impacts on
purchasing power and therefore food consumption trends and modern retail market
development. Figure 6 below demonstrates the relatively lower growth of GDP per
capita since 2008, compared to the 2004-2008 period, resulting from the economic
crisis. Growth has slowed in 2008-2012 compared to 2004-2008 for all MS, and
negative growth was registered in the 2008-2012 period by 9 MS.
41 ‘The future of online grocery in Europe’. McKinsey, 2012.
42 © Planet Retail
56
Figure 6: Compound annual growth rate of GDP per capita in EU 27
14%
12%
10%
8%
6%
4%
2%
0%
EU 27
France
Latvia
Czech Republic
Austria
Romania
Finland
Denmark
Germany
Spain
Ireland
Greece
Italy
Hungary
Poland
Bulgaria
Cyprus
Netherlands
Luxembourg
Slovenia
Slovakia
Belgium
Portugal
Estonia
Lithuania
Malta
Sweden
United Kingdom
-2%
-4%
-6%
CAGR (04-08) CAGR (08-12)
Source: EY analysis based on Eurostat, [nama_gdp_c]
Consumers are spending on average 13% of total expenditure on

food and non-alcoholic beverage
The proportion of household expenditure spent on food has been impacted by the
state of the economy. As demonstrated in Figure 7, after decreasing from 2004 to
2006, the average share of household expenditure on food and non-alcoholic
beverages across the EU 27 is on a gradual upward trend since 2006. This means that
consumers are spending more of their income on food and beverage. It may explain
the recent focus of consumers on value-for-money products, in order to save money
on what may be considered staple purchases.
Figure 7: Average final consumption expenditure of households for food and non-
alcoholic beverages across EU 27 (% of the total expenditure)
14
% of final consumption spent on food
12,9 13
and non-alcoholic beverages
13 12,8 12,8
12,5
12
11
10
2004 2006 2008 2010 2012
Source: EY analysis based on Eurostat, (TSDPC520)
57
As can be seen from Figure 8, the situations have been quite heterogeneous across
MS over this period43. Whilst growth since 2008 has been positive in 15 MS (notably in
Portugal, Ireland, Latvia and Cyprus), consumers are spending a smaller proportion on
food over recent years in 8 MS (notably Malta, Poland and Luxembourg).
Figure 8: Compound annual growth in the share of final consumption expenditure of
households of food and non-alcoholic beverages per MS (% CAGR)
3%
2%
1%
-1%
-2%
-3%
-4% CAGR (2004-2008)

CAGR (2008-2012)
-5%
-6%
Source: EY analysis based on Eurostat, (nama_co3_c)
This said, the main household expenditure item remains housing, water, electricity
and gas which has increased steadily over the last decade (from 21.2% in 2003 to
23.8% in 2011), as shown in Figure 9. This increase places further pressure on the
available budget for groceries.
43 Bulgaria, Greece, Lithuania and Romania not represented on graph due to incomplete data
58
Figure 9 : Proportion of key household expenditures compared to the total household

expenditure for EU-27 (2003-2011)
25,0%
20,0%
15,0%
10,0%
5,0%
0,0%
2003 2004 2005 2006 2007 2008 2009 2010 2011
Housing, water, electricity, gas Food and non alcoholic beverages

Transport
Source: EY analysis based on Eurostat, [nama_co3_c]
Stable edible grocery share of b roader retail market

In the EU, grocery sales account for a large proportion of retail sales, as shown in
Figure 10; however this proportion has remained rather stable over the past decade,
experiencing a slight reduction between 2006 and 2010.
Figure 10: Edible grocery proportion (in %) of total retail sales in EU 27 between 2004
and 2012
44% 43%
42%
41% 41% 42%
42%
40%
38%
36%
34%
32%
30%
2004 2006 2008 2010 2012
The growth of edible grocery retail sales is relatively stable compared to other retail
sectors over the last 8 years (see Figure 11). Markets that have grown significantly
over recent years include Leisure & Entertainment and Home, Garden and Automotive.
59
Figure 11: Compound annual growth rate in EU retail markets (2006 to 2012)
8%
6%
4%
2%
0%
-2%
-4%
Source: EY anaysis based on © Planet Retail
3.3.2. Socio-economic evolutions in the EU impacting food consumption

patterns
Three particular characteristics have potentially had an impact on evolutions in the
grocery retail sector:
 Unemployment rates have risen
 Consumers are increasingly seeking low prices
 Household composition evolution impacts consumption habits
 New consumer needs have appeared
Each of these characteristics will be addressed in the section below.
Unemployment rates have risen since 2008

Since the economic crisis that began in 2008, the unemployment rates in nearly all MS
have risen. Figure 12 below shows the compound annual growth rate of the
unemployment rate over the pre-crisis (2004-2008) and crisis (2008-2012) periods.
The unemployment rate has increased between 2008 and 2012 in 25 MS (to the
greatest extent in Greece and Cyprus), and has only decreased in Germany, whilst in
Luxembourg it has remained stable.
This compares with the pre-crisis (2004-2008) period, where unemployment only
increased in 5 MS (Ireland, Spain, Hungary, Portugal and the UK), and was either
stable or decreased in all other MS.
60
Figure 12: Compound annual growth in unemployment rate (in %) across EU 27

between 2004 and 2012
40%
30%
20%
10%
0%
EU 27
Austria
France
Latvia
Hungary
Romania
Finland
Denmark
Spain
Poland
Germany
Ireland
Greece
Italy
Cyprus
Netherlands
Slovakia
Belgium
Bulgaria
Czech Republic
Estonia
Malta
Portugal
Slovenia
Lithuania
Luxembourg
Sweden
United Kingdom
-10%
-20%
CAGR (2004-2008)
CAGR (2008-2012)
-30%
Source: EY analysis based on Eurostat, [lfst_r_lfu3rt]
A greater proportion of the population is at risk of poverty and is

seeking lower prices as a priority for consumers
The purchasing power of many EU consumers is under pressure due to slow economic
growth and fiscal tightening. Figure 13 shows the growth in the percentage of the
population at risk of poverty both during the pre-crisis and crisis periods44. Whilst
growth in this indicator was negative in 18 MS between 2006 and 2009 (positive only
in 5 MS), from 2009 to 2012 the trend was reversed, with 19 MS registered increased
growth and only 5 MS with negative growth.
44 2006-2009 chosen as pre-crisis and 2009-2012 as crisis periods due to lack of data from 2004.
61
Figure 13: Compound annual growth in percentage of population at risk of poverty

after social transfers (2004-2012)
8,0%
3,0%
EU 27
Czech Republic
France
Denmark
Hungary
Finland
Germany
Greece
Spain
Cyprus
Latvia
Austria
Italy
Poland
Malta
Netherlands
Slovenia
Slovakia
Belgium
Luxembourg
Portugal
Estonia
Lithuania
Sweden
United Kingdom
-2,0%
-7,0%
CAGR (2006-2009)
CAGR (2009-2012)
-12,0%
Source: EY analysis based on Eurostat, [ilc_peps01]
As a consequence, price has become the number one criterion in shop selection for
consumers across all kinds of goods, including food (See Figure 14).
Figure 14 : Top 5 major impact factors on grocery purchase choice in 2011 45
Source: Nielsen Global Survey of Grocery Purchase Impact, Q1 2012
45
Package labelling entails the information contained on the label as well as it
presentation (user friendly,…)
62
The Nielsen Global Survey shows that the increasing cost of food is affecting 81% of
respondents in Europe, with more than half of respondents indicating that rising food
prices are having a major impact on choice of grocery purchases.
As a result, many consumers have changed their shopping behaviours, such as
showing preferences for products in multi-packs, family sized/economical size packs,
and lower cost private label products.
Household composition impacting consumpt ion habits
The last decade has seen a change in the household’s composition. In 2011, the most
common household type in the EU 27 was the single person living alone (31.4%). 46
This increase in single households or households with tight incomes has seen a trend
towards smaller portion sizes and packaging (single servings) to meet consumer
needs. For example, Auchan in France offers beef steak in small 80g packs to respond
to the needs of single household consumers. 47
Moreover, there is an increasing trend towards an ageing population, due particularly
to the retirement of baby boomers, a birth rate decrease and an increase in life
expectancy, which has potentially impacted consumption habits. In 2012, 17.8% of
the EU population was in the 65+ category, up from 15.6% in 200048.
Finally, an increase in the participation rate of women in the workforce (from 54.3% in
2001 to 58.5% in 2011)49, may account for changes in grocery retail, in terms of store
formats to address time-constrained consumers, and products requiring less
preparation and cooking time, such as ready-prepared meals or quick meal solutions.
New consumer preferences impacting grocery choice

A number of new consumer preferences have gained influence over the recent decade,
and have had an impact on the grocery retail market in Europe, including:
 Health consciousness changing food trends
 The development of ethnic food to reflect changing demographics and needs
 Increase in environmental awareness
 Focus on convenience for time-constrained consumers
Over recent years interest in issues connected to health has grown among consumers,
impacting choice and final food consumption. There is a better awareness of food
intolerances, allergies, food-related diseases, overweight and obesity, resulting in a
more educated and aware consumer, as well as the growth of specific product
categories such as gluten-free food (double-digit growth in Europe). As an example, in
2013 Ahold-affiliated Swedish grocer ICA launched a line of gluten-free food products
under the new private label ICA Glutenfri.50 In France, the government launched a law
in March 2007 that ensures that food and drink manufacturers include health
46 Eurostat, European social statistics, 2013 edition, ISSN 1977-7930
47 © Planet Retail, Euro crisis transforms shopping behaviour, 2012
48 Eurostat, European social statistics, 2013 edition, ISSN 1977-7930
49 Eurostat, 2012
50 © Planet Retail
63
messages when promoting their products on all broadcast and print advertising. The
aim is to encourage consumers to eat a more balanced diet. Both suppliers and
retailers have to include one of four health messages in their advertisements. If
companies do not mention a health messages they can be fined up to 1.5% of their
advertising budget.
With the globalization trend, food products from all corners of the world have become
more widely available. In 2010 UK grocery retailer Tesco introduced seven new ethnic
food ranges due to popular demand for world foods. Over the previous 12 months,
Tesco had doubled its world food ranges to include 3,000 products, and claimed the
sector had grown by 35% from 2009 to 2010.51
Moreover, environmental awareness has become an increasing concern, due to the
consequences of environmental degradation and pollution, and this has had an impact
on consumer choices. There has been a strong development in the offering of bio
products. Indeed the recent development of the organic food sector is due to the
environmental friendly farming systems required to benefit from the “organic” label.
Leclerc launched in 2011 an eco-friendly Conso Responsable label. The label now
covers more than 450 food and non-food SKUs, including entry price, private labels
and national brands. Items under the label are considered more environmentally
friendly in terms of ingredients, manufacture, packaging, transportation and
biodegradability.52 Furthermore, in 2013 Carrefour relaunched its organic product
range in France.
Finally, consumer interest in convenience aspects has become more important in
recent decades. Convenience refers to optimizing time and energy spent through the
private household’s meal production chain, i.e., during shopping, storage, preparation,
eating and disposal. This focus has driven the development of a number of retail
innovations over the last years, including the drive format, self-service check-outs,
ready prepared meals and prepared cuts of meat, such as for giros or goulash.
51 © Planet Retail
52 © Planet Retail
64
Scope, measures and methodology
4. Scope, measures and methodology

This section presents the final scope of data, indicators and measures for the study as
well as specific approaches applied. Its objective is to provide an introduction to the
subsequent section on descriptive statistics.
4.1. Selection of MS
The selection of MS was designed to be representative of a broad variety of situations
in the EU 27, taking into account different levels of retail concentration, private label
share, and in an attempt to cover a broad EU population. The scope of MS selected
covers 9 MS: the representativeness of the sample is presented in Table 3 below.
The selection of MS covered in the study varies according to the scope of analysis
(local level or procurement (national) level), and the different variables and drivers
being measured, in an effort to maximise the size and representativeness of the
sample of MS for each measure.
The table below synthetises the geographic coverage for each of the drivers, choice
and innovation components.
65
United Kingdom
Number of MS
Czech Republic
Luxembourg
Netherlands
Indicators source
Lithuania
Denmark
Germany
Romania
Portugal
Hungary
Slovenia
Slovakia
Bulgaria
Belgium
Sweden
Estonia
Finland
Austria
Ireland
Cyprus
Poland
Greece
France
Malta
Latvia
Spain
Italy
Evolution of choices 2004-2012
Shop choices (2004-2012) ©Nielsen trade dimension n n n n 4
Shop choices (2008-2012) ©Nielsen trade dimension n n n n n n 6
Product variety, price variety, size variety (2004-2012) ©Nielsen Opus n n n n n n 6
Product variety, price variety, size variety (2008-2012) ©Nielsen Opus n n n n n n n n n 9
Evolution of innovations 2004-2012
Number of innovations (2004-2012) ©Nielsen Opus n n n n n n 6
Number of innovations (2008-2012) ©Nielsen Opus n n n n n n n n n 9
Categories of innovations (2004-2012) ©Mintel GNDP n n n n n n 6
Categories of innovations (2008-2012) ©Mintel GNDP n n n n n n n n n 9
Evolution of concentration
Retail concentration at national level (Retail group & banner
©Planet retail n n n n n n n n n n n n n n n n n n n n n n n n n n 26
level) - 2004-2012 - C5 / HHI
Retail concentration at local level - C5 / HHI (2004-2012) ©Nielsen trade dimension n n n n 4
Retail concentration at local level - C5 / HHI (2008-2012) ©Nielsen trade dimension n n n n n n 6
Supplier concentration at national level - 2004-2012 ©Euromonitor n n n n n n n n n n n n n n 14
Supplier concentration at local level - 2004-2012 ©Nielsen Opus n n n n n n 6
Measure of imbalance (national level only) - 2004-2012 ©Planet retail, ©Euromonitor n n n n n n n n n n n n n n 14
Evolution of other a priori drivers

Macroeconomic data (GDP, population, unemployment, etc.) Eurostat n n n n n n n n n n n n n n n n n n n n n n n n n n n 27
Shop types at national level - 2004-2012 ©Planet retail n n n n n n n n n n n n n n n n n n n n n n n n n n 26
Shop type, shop size - 2004-2012 ©Nielsen trade dimension n n n n 4
Shop type, shop size - 2008-2012 ©Nielsen trade dimension n n n n n n 6
Private label share (national level) - 2004-2012 ©Euromonitor n n n n n n n n n n n n n n 14
Private label share (local level) - 2004-2012 ©Nielsen Opus n n n n n n 6
Private label share (local level) - 2008-2012 ©Nielsen Opus n n n n n n n n n 9
Product category turnover at national level - 2004-2012 ©Euromonitor n n n n n n n n n n n n n n 14
Econometric analysis
Impact of drivers on choice and innovation (2004-2012) Consortium computation n n n n n 5
Impact of drivers on choice and innovation (2008-2012) Consortium computation n n n n n n n 7
Coverage of case studies
Case studies Consortium analysis n n n n n n 6
66
4.1.1. Local consumer shopping area level

Local CSA level refers to analysis undertaken and measures at the level of a shop in
defined CSAs. This analysis is intended to enable analysis of certain phenomena at the
region level and uncover local differences in trends. The final list of MS for which
choice, innovation and their drivers will be analysed at the local CSA level is presented
in Table 3 below, which illustrates the variety of situations covered across these
selected 9 MS in terms of modern retail share, evolution of retail concentration,
market share per shop type, population size, economic prosperity and private label
share.
Table 3: Variety of situations in 9 MS vs. EU 27
M o de r n r e t a i l m a r ke t
sh a r e & E v o lu t i on o f T y p e o f sh o p
P o p u la t i on s i ze & Share of
M e m be r C (5 ) r e t a i l ma r ke t sh a r e o f
e con o m i c p r o sp e r it y in p r i va t e l a b e l
State con ce n t r a t i on (e d i b le t h e f oo d ma r ke t
2011 in 2 0 0 9
gr o ce r y ) : 2 0 0 0 v s in 2 0 1 1
2011
83% food market 11% HM

Population : 11 million
B (Belgium) C(5) 2000 : 51% 55% SM 27%
GDP/capita : 33.7 k€/year
C(5) 2011 : 76% 17% HD

CZ (Czech Population : 10.5 million
C(5) 2000 : 27% 14% SM 24%
Republic) GDP/capita : 14.7 k€/year
C(5) 2011 : 48% 12% HD

Population : 5.6 million
DK (Denmark) C(5) 2000 : 61% 49% SM 22%
C(5) 2011 : 83% 31% HD
Population : 65.1 million 28%
FR (France) C(5) 2000 : 59% 31% SM
C(5) 2011 : 66% 8% HD
60% food market 25% HM Population: 10.0 million

H (Hungary) C(5) 2000 : 29% 23% SM GDP/capita : 10.1 k€/year 20%
C(5) 2011 : 36% 12% HD

Population: 60.7 million
IT (Italy) C(5) 2000 : 20% 20% SM 15%
GDP/capita: 26.1 k€/year
C(5) 2011 : 24% 5% HD

PL (Poland) C(5) 2000 : 7% 9% SM 14%
C(5) 2011 : 26% 14% HD

PT ( Portugal) C(5) 2000 : 42% 39% SM 25%
C(5) 2011 : 60% 11% HD

E (Spain) C(5) 2000 : 34% 41% SM 31%
C(5) 2011 : 47% 9% HD
Modern retail share: Total outlets Population: 258.2 million

Min: 37% (Italy, 6% HM (51% of EU27)
TOTAL IN Poland) 60% SM
Share of private
SCOPE – Max : 87% (Denmark) 35% HD GDP/capita :
labels
analysis at Min : 9 697€/year (Poland)
Min : 14%
shop level Top 5 concentration Total outlets Max : 43 024 €/year
Max : 31%
Min : 24 % (Italy) 51 810 (49% of (Denmark)
Max: 83% (Denmark) EU27) Med : 23 125 €/year
Modern retail share : Total outlets Population : 500.5 million

Min: 5% 6 372 HM (6%)
Share of private
Max : 90% 58 858 SM (56%) GDP/capita :
labels
TOTAL EU 27 39 887 HD (38%) Min:5 255 €/year (Bulgaria)
Min : 12%
Top 5 concentration Max : 83 311 €/year
Max : 43%
Min : 21% Total outlets (Luxembourg)
Max: 87% 105 117 Med : 24 537 €/year
Sources: EY analysis based on © Planet Retail, Eurostat, and © Nielsen
67
The scope of this study covers MS representing a broad diversity across time and
space in the modern retail EU food sector, satisfying the following criteria:
 Coverage of 51% of the EU 27 population (72% when including case studies)
 Diversity in population size (from 5.6 million to 65.1 million)
 Diversity in GDP per capita (from 9 697€/year to 43 024 €/year)
 Differentiation in the market share of modern retail (3 MS above 75%, 4 MS
between 55% and 75% and 2 MS between 35% and 40%
 Types and size of shops, size of MS, private label market share (14%-31%);
 Variety in concentration of top 5 retailers (24% to 83%)
 A variety of banners in local CSAs, represented in Table 5.
Table 4: Banner coverage in shop sample across MS
Member State Banners in the sample Retail group Shop type in sample 53
Aldi Aldi Hard discounter
Champion Carrefour Supermarket
Belgium Colruyt Colruyt Supermarket
Cora Louis Delhaize Hypermarket
Lidl Schwarz Group Hard discounter
Kaufland Schwarz Group Hypermarket
Czech Republic
Tesco Tesco Hypermarket
Fotex Dansk Supermarked Supermarket
Netto Dansk Supermarked Hard discounter
Denmark
Rema1000 Reitan Hard discounter
SuperBrugsen Coop Danmark (FDB) Supermarket
Aldi Aldi Hard discounter
Auchan Auchan Hypermarket
Carrefour Carrefour Hypermarket
Carrefour Market Carrefour Supermarket
Casino Casino Hypermarket
Cora Louis Delhaize Hypermarket
Dia Dia Hard discounter
Géant Casino Hypermarket
Hyper U Système U Hypermarket
ITM Hyper ITM (Intermarché) Hypermarket
France
Supermarkets (<2500m²),
ITM Super ITM (Intermarché)
hypermarkets >2500m²)
Leader Price Casino Hard discounter
Leclerc Leclerc Hypermarket
Netto ITM (Intermarché) Hard discounter
Simply Market Auchan
Super U Système U
Auchan Auchan Supermarket
CBA CBA Supermarket
Interspar SPAR (Austria) Supermarket
Hungary
Penny Rewe Group Hard discounter
Tesco Tesco Supermarket
Auchan Auchan Hypermarket
Italy
Bennet Bennet Hypermarket
53
In the study the distinction between hypermarkets and supermarkets is based on the sales
areas definitions - supermarkets from 400 m² to 2499 m², and hypermarkets 2500 m² and
greater. Planet Retail categorisation has been used to distinguish discount stores from
supermarkets, with the defining criteria being the assortment / SKUs stocked.
68
Member State Banners in the sample Retail group Shop type in sample 53
Carrefour Market Carrefour Supermarket
Conad Conad Supermarket
Coop Coop Italia Supermarket
DOK Carrefour Supermarket
EMI Selex Commerciale Supermarket
Esselunga Esselunga
Eurospin Eurospin Hard discounter
Famila Selex Commerciale
Galassia Galassia Hypermarket
Il Gigante Il Gigante Hypermarket
Iper Finiper Hypermarket
Iperal Agora' network SCARL Hypermarket
Ipercoop Coop Italia Hypermarket
Ipersimply Auchan Hypermarket
Iperspar Despar servizi Hypermarket
Maxisconto Supermarket
Pam PAM Supermarket
Panorama PAM Hypermarket
Penny Rewe Group Hard discounter
Supermac Supermarket
U2 Finiper Supermarket
Biedronka Jerónimo Martins Hard discounter
Carrefour Express Carrefour Supermarket
Kaufland Schwarz Group
Poland Leclerc Leclerc Hypermarket
Netto Dansk Supermarked Hard discounter
Real Metro Group Hypermarket
Simply Market Auchan Supermarket
Tesco Tesco Hypermarket
Continente Sonae
Continente Modelo Sonae Supermarket
Jumbo Auchan
Portugal
Leclerc Leclerc
Mini Preco Carrefour Hard discounter
Pingo Doce Jerónimo Martins Hypermarket
Ahorramas ahorramas Supermarket
Alcampo Auchan Hypermarket
Caprabo Caprabo Supermarket
Carrefour Carrefour
Carrefour Planet Carrefour Hypermarket
Spain Dani Supermarket
Dia Carrefour Hard discounter
Eroski Eroski Supermarket
Maxi Dia Carrefour Hard discounter
Mercadona Mercadona Supermarket
Source: EY analysis based on ©Nielsen Opus and © Planet Retail
69
In addition, the largest retail players in the EU 27 are present in the selected MS, with
the exception of EDEKA that operates only in Germany, as shown in Table 5.
Table 5: Coverage of largest retail groups in Europe
EU 27 market share
Retailer MS in sample where retailers operate (in 2011)
(2011)
Schwarz Group 4.7% All 9 Member States
Carrefour 4.5% Belgium, France, Italy, Poland, Portugal, Spain
Tesco 3.8% Czech Republic, Hungary, Poland
Edeka 3.4% N/A
Aldi 3.1% Belgium, Denmark, France, Hungary, Poland, Portugal, Spain
Rewe Group 3.0% Czech Republic, Hungary, Italy
Auchan 2.2% France, Hungary, Italy, Poland, Portugal, Spain
ITM (Intermarché) 2.1% Belgium, France, Poland, Portugal
Leclerc 2.1% France, Italy, Portugal, Spain, Poland
Ahold 1.9% Belgium, Czech Republic
Source : EY analysis based on © Planet Retail
4.1.2. Procurement (national) level

Procurement (national) level refers to analysis undertaken and measures at the level
of a Member State. This analysis is intended to complement local level analysis by
measuring certain financial indicators that are not possible at local level, and provide
the macro view across MS.
In addition to the 9 MS which form the scope of the study in terms of the evolution of
choice and innovation within a sample of shops in defined CSAs, an additional scope of
MS has been addressed for a selection of a priori drivers at procurement (national)
level.
In the context of analysis undertaken to observe trends in retail concentration,
supplier concentration and measure of imbalance, the following additional scope has
been analysed:
Table 6: Scope of selected measures at procurement (national) level
Measure Scope Comments
All EU MS with the exception of Malta, for which

Retail concentration measured by
26 MS insufficient data is available, and Croatia, which was not a
grocery sales market share
member of the EU at the commencement of this study
Supplier concentration measured 9 MS of scope at local level – Belgium, Czech Republic,

by grocery sales market share Denmark, France, Hungary, Italy, Poland, Portugal,
Spain; 3 additional MS in scope of case studies – Finland,
14 MS Germany, Netherlands; and 2 additional MS that
Measure of imbalance measured represent specific cases of interest – Romania, due to its
by grocery sales market share very low modern retail share, and United Kingdom, due to
the importance of convenience stores
4.2. Selection of time period

The study seeks to cover a significant timeframe and frequency across MS where
modern retail is well established, and where the development of modern retail has
70
been more recent. In addition, time periods have been selected to cover both pre-
crisis (2004-2008) and crisis (2008-2012) periods.
These principles led to a selection of 5 years of data, covering biennial periods (2004,
2006, 2008, 2010, and 2012). In terms of frequency, for each year selected, the
study analyses data from November and June. November was selected as it captures
significant end of year sales, and June to observe the significant summer sales.
Following data issues encountered during the study and in an effort to optimise the
utility of data available at the local CSA level, a decision was made to establish two
shop samples for analysis:
1) a shop sample covering the full period 2004-2012 to be able to observe longer
term trends both before and during the crisis (10 periods of observation), and
2) a larger shop sample covering the period 2008-2012 in order to increase the
sample size in countries and areas where longer term data is not available and
to enable focused analysis on the crisis period (6 periods of observation).
The scope of the two shop samples is presented in the Table 7 below.
Table 7: Study samples by MS and time period coverage – descriptive statistics
(source EY analysis)
Final scope of time periods
2004-2012 2008-2012
MS
No. of CSAs with at Number of No. of CSAs with at Number of
least 2 shops shops least 2 shops shops
Spain 15 42 15 42
France 31 131 31 131
Italy 25 80 25 83
Poland 10 2454 11 29
Portugal 8 19 8 19
Hungary N/A N/A 9 24
Czech Republic N/A N/A 1 255
Belgium 2 656 3 9
Denmark N/A N/A 2 457
TOTAL 91 30258 105 34359
54
In the long data set, Poland was omitted from analysis that included local retail concentration as a driver
because of the absence of the required © Nielsen Trade Dimensions data to calculate this measure.
55
No econometric analysis possible for Czech Republic on short data set due to absence of the required ©
Nielsen Trade Dimensions data
56
No econometric analysis possible for Belgium on long data set due to absence of the required © Nielsen
Trade Dimensions data
57
No econometric analysis possible for Denmark on short data set due to absence of the required © Nielsen
Trade Dimensions data
58
296 shops in total for econometric analysis
59
337 shops in total for econometric analysis
References:
71
By combining the long full period timeframe and short timeframe, observations on
choice and innovation can be made for a total of 343 shops situated in 105 CSAs. Of
this sample, 302 shops corresponding to 91 CSAs can be observed over the long full
period timeframe.
Due to data limitations, the full set of shops outlined above cannot be addressed in the
econometric analysis, which requires data to be available for every driver in every
time period. As a result, 296 shops in total (across 5 MS) are analysed over the short
period and 337 shops (across 7 MS) over the long period for the econometric analysis.
4.3. Selection of 105 consumer shopping areas (CSAs)

The identification of a sample of modern retail shops is central to this study, as this is
the most detailed level where the evolution of choice, innovation, and their a priori
drivers is analysed.
In order to observe the evolution of choice and innovation, a sample of shops has
been selected from areas that are most representative of the different local consumer
markets in Europe. The approach rests on the fact that consumers have access to a
variety of local shops in which they make food purchases on a regular basis.
In summary, the definition of the CSA approach involved:
 Identification inside each selected MS of a selection of consumer shopping
areas that are representative of different consumer living environments in
Europe across two criteria:
o type of living areas categorised by [large city, medium city, small city
and rural zone]; and
o GDP per capita categorised by [low, medium-,medium+, high]).
 Regions (Eurostat NUTS 3) and cities/towns were categorised based on these

two criteria, and the number of CSAs for each category were determined
proportionally to the situation at EU level, to ensure that the selections in each
MS closely represent the most prevalent situations in the EU.
 For each of the representative cities/towns (also by size and GDP per capita)
within the regions, a central point for the CSA was determined: the city hall.
The geographical perimeter of the CSA is defined by the travel time between
the central point and outer limit of the area. The isochrone radius of shop
accessibility differs based on the retail density which usually depends on the
size of the city. Based on retail studies60 and sensitivity analysis, we defined:
o 15 minutes travel time for large cities;
o 20 minutes travel time for medium and small cities, and;
o 25 minutes for a rural zone.
 Finally, within each CSA a selection of a sample of shops was made, within
which choice and innovation can be observed, based on data availability.
“More pros and cons of merger control (2012); Competition Commission (2000),
“Supermarkets – a report on the supply of groceries from multiple stores in the UK”,§2.53;
Competition Commission (2008), “The supply of Groceries in the UK Market Investigation”, §4.145
60
Such as Sørgard, L “Merger screening in markets with differentiated products”, 2012
72
Table 8 presents the full list of regions where CSAs are located.
Table 8: List of regions where consumer shopping areas are located
Urban / Population % unemploy- GDP per Population % of MS Total %

Selected NUTS 3 regions
Rural type (2011) ment (2011) capita (2010) Density Population Pop.
Arr. de Bruxelles-Capitale PU 1 148 100 16,9 54 700 7 131 10%
Belgium Arr. Arlon PR 59 000 6,2 20 500 186 1% 15%
Arr. Charleroi PU 429 100 11,7 21 400 773 4%
Czech Rep Jihomoravský kraj IN 1 165 000 7,2 18 300 166 11% 11%
Sydjylland PR 715 900 7,7 29 900 82 13%

Denmark 22%
Københavns omegn PU 518 800 8,1 42 500 1 516 9%
Haut-Rhin IN 753 400 7,6 21 800 214 1%
Territoire de Belfort IN 143 600 8,3 26 200 236 0%
Hauts-de-Seine PU 1 585 400 8,6 75 700 9 028 3%
Seine-Saint-Denis PU 1 534 600 8,6 28 700 6 497 2%
Val-de-Marne PU 1 338 600 8,6 34 900 5 463 2%
Val-d'Oise PU 1 180 800 8,6 26 100 948 2%
Yvelines PU 1 420 300 8,6 31 800 622 2%
Corrèze PR 244 400 7,8 20 600 42 0%
Haut-Rhin IN 753 400 7,6 21 800 214 1%
Meurthe-et-Moselle IN 733 600 10,4 22 800 140 1%
Meuse PR 194 300 10,4 16 600 31 0%
Moselle IN 1 047 000 10,4 19 600 168 2%

France 33%
Vosges PR 380 400 10,4 19 400 65 1%
Loire-Atlantique PU 1 297 900 8,8 26 200 190 2%
Maine-et-Loire IN 794 500 8,8 21 000 111 1%
Côtes-d'Armor PR 598 700 7,4 19 100 87 1%
Vendée PR 642 600 8,8 21 300 96 1%
Finistère IN 908 300 7,4 20 900 135 1%
Ille-et-Vilaine IN 999 900 7,4 25 200 148 2%
Loiret IN 658 800 8,7 25 300 97 1%
Loir-et-Cher PR 331 500 8,7 20 800 52 1%
Haute-Saône PR 240 700 8,3 16 800 45 0%
Gironde PR 1 467 400 9,3 26 100 147 2%
Pas-de-Calais IN 1 464 500 12,8 18 300 220 2%
Budapest PU 1 736 900 8,8 34 900 3 308 17%
Pest IN 1 241 300 8,8 13 900 194 12%
Hungary Fejér PR 425 900 9,3 14 000 98 4% 46%
Bács-Kiskun PR 523 600 10,6 10 300 62 5%
Borsod-Abaúj-Zemplén IN 681 500 16,7 9 900 94 7%
Cosenza PR 734 800 12,7 15 800 111 1%
Italy Reggio nell'Emilia IN 532 200 5,3 29 800 232 1% 29%
Brescia IN 1 260 700 5,8 32 300 264 2%
73

Como PU 597 000 5,8 32 300 464 1%
Cremona IN 364 400 5,8 32 300 206 1%
Mantova PR 416 400 5,8 32 300 178 1%
Milano PU 3 172 900 5,8 32 300 2 010 5%
Pavia IN 550 000 5,8 32 300 186 1%
Alessandria PR 440 800 7,6 26 600 124 1%
Cuneo PR 593 500 7,6 26 600 86 1%
Torino PU 2 304 600 7,6 26 600 337 4%
Foggia PR 639 900 13,1 16 300 92 1%
Lecce IN 815 400 13,1 16 300 296 1%
Bologna IN 995 400 5,3 29 800 269 2%
Modena IN 703 000 5,3 29 800 262 1%
Ferrara PR 359 800 5,3 29 800 137 1%
Parma IN 443 700 5,3 29 800 129 1%
Firenze IN 1 000 900 6,5 26 700 285 2%
Pisa IN 418 800 6,5 26 700 171 1%
Perugia IN 673 100 6,5 22 800 106 1%
Lecco PU 340 700 5,8 32 300 418 1%
Wałbrzyski IN 684 000 10,6 11 500 164 2%
Miasto Warszawa PU 1 704 300 7,9 46 100 3 297 4%
Radomski IN 625 500 7,9 11 400 109 2%
Rzeszowski IN 620 700 12,4 12 200 175 2%

Poland 16%
Wrocławski PU 564 000 10,6 13 900 88 1%
Łódzki PU 383 800 9,3 12 700 174 1%
Lubelski IN 717 600 10,3 13 400 170 2%
Miasto Kraków PU 758 400 9,3 23 100 2 319 2%
Cávado IN 410 100 13,0 15 100 329 4%
Grande Lisboa PU 2 043 800 14,1 32 700 1 485 19%
Grande Porto PU 1 287 100 13,0 19 900 1 580 12%
Portugal Médio Tejo PR 220 400 10,3 15 400 96 2% 48%
Oeste PR 362 300 10,3 15 700 163 3%
Algarve PR 451 100 15,6 20 300 90 4%
Baixo Mondego PR 332 100 10,3 19 800 161 3%
Almería IN 691 600 30,1 19 300 79 1%
Madrid PU 6 378 500 16,3 31 600 801 14%
Castellón / Castelló IN 591 200 24,0 23 400 89 1%
Valencia / València PU 2 514 900 24,0 22 700 235 5%

Spain 62%
A Coruña IN 1 124 500 17,3 23 700 143 2%
Granada IN 912 500 30,1 17 200 72 2%
Pontevedra IN 947 100 17,3 21 300 212 2%
Asturias IN 1 054 100 17,8 22 700 100 2%
74

Granada IN 912 500 30,1 17 200 72 2%
Sevilla PU 1 882 300 30,1 19 600 136 4%
Barcelona PU 5 366 600 19,2 28 400 695 12%
Madrid PU 6 378 500 16,3 31 600 801 14%
Sources: EY analysis based on Eurostat, at NUTS 2 level for unemployment, and for GDP in Italy.
The final selection of CSAs was made in order to provide the best possible
representativeness of EU 27 population characteristics in terms of population size,
diversity of standard of living and type of living – illustrations of representativeness for
each of these characteristics are provided below.
4.4. Representativeness of the sample that was selected
4.4.1. Population size

Firstly, the allocation of CSAs amongst MS in the study sample has sought to take into
account the relative population size compared to the other MS in the scope. Table 9
below shows the proportion of population in each MS relative to the number of CSA
that have been selected.
Table 9: Number of CSA in relation to population size
Population size Proportion of total 9 MS Number of CSA

MS Proportion of CSA
(million in 2011) population (2008-2012 period)
Belgium 11,0 4% 3 3%
Czech Republic 10,5 4% 1 1%
Denmark 5,6 2% 2 2%
France 65,2 25% 31 30%
Hungary 10,0 4% 9 9%
Italy 60,7 24% 25 24%
Poland 38,5 15% 11 10%
Portugal 10,6 4% 8 8%
Spain 46,2 18% 15 14%
Total 258,2 100% 105 100%
There are some slight discrepancies for some MS, i.e. Poland and Portugal, due to
shop data limitations and to ensure better representativeness of the different types of
living and GDP per capita. However, on the whole the number of CSA per MS closely
reflects the relative population of each MS.
75
4.4.2. Type of living zone and standard of living

Taking into consideration the prevalence of the different types of living zones and
standard of living in each of the MS, the following table demonstrates the number of
CSAs that have been selected under each category. Type of living is denoted by (PR =
Predominantly Rural; INT = Intermediate; PU = Predominantly Urban) and GDP per
capita by (low, medium-, medium+, high). A total of 105 CSA have been selected and
allocated as shown in Table 10.
Table 10: Number of CSA per type of living zone and standard of living category
GDP/Capita Low Medium - Medium + High Total
Type of living Number of CSA Number of CSA Number of CSA Number of CSA Number of CSA
Predominantly
8 8 3 4 23
Rural (PR)
Intermediate
13 9 13 7 42
(INT)
Predominantly
5 8 15 12 40
Urban (PU)
TOTAL 26 25 31 23 105
Source: EY analysis
This selection above provides a solid coverage of the EU27 diversity in type of living
and standard of living. The extent of representativeness is illustrated in Table 11,
which compares the proportion of the EU27 population that corresponds to each of the
previously mentioned categories of type of living and standard of living with the
proportion of CSAs that correspond to the same categories.
Table 11: Comparison of proportion of CSA vs proportion of EU27 population
GDP/Capita Low Medium - Medium + High Total
Type of living EU27 CSA EU27 CSA EU27 CSA EU27 CSA EU27 CSA
Predominantly
11% 8% 6% 8% 4% 3% 2% 4% 23% 22%
Rural (PR)
Intermediate
10% 12% 10% 9% 9% 12% 7% 7% 35% 40%
(INT)
Predominantly
4% 5% 9% 8% 12% 14% 17% 11% 42% 38%
Urban (PU)
TOTAL 25% 29% 25% 24% 25% 30% 26% 22% 100% 100%
Sources: EY analysis based on Eurostat
In terms of GDP per capita representativeness, as the EU27 population has been
broken down by quartiles, each quartile refers to 25% of the population. In our CSA
selection, low GDP per capita accounts for exactly 25% of our selection; medium –
GDP per capita accounts for 26%; medium + GDP per capita accounts for 30%; and
high GDP per capita represents 22% of CSA (see Figure 15). In conclusion, the
selection of CSAs closely resembles the EU 27 average in terms of real GDP per capita.
76
Figure 15: Representativeness of sample vs EU27 population by standard of living

categories
30%
25%25% 25%24% 25% 26%
22%
EU27
Sample
Low Medium- Medium+ High

In terms of type of living, observations at the total EU27 population level are well
aligned to the selection of CSA. For Predominantly Rural areas, there is only 1%
discrepancy, for Intermediate zones 5%, and for Predominantly Urban, only 4%
difference (see Figure 16).
Figure 16: Representativeness of sample vs EU27 population by type of living zone
40% 42%
38%
35%
23% 22%
EU27
Sample
PR IN PU
In conclusion, the selected sample of 105 CSA closely resembles the situation across
EU 27 in terms of population size, real GDP per capita level, and type of living zones
(urban vs rural). The choice of geographic zones based on the above criteria ensures
that the study addresses a variety of situations faced by consumers in the EU.
4.5. Selection of product categories

The study on choice and innovation in local CSAs covers 23 product categories with
EAN barcodes. Only packaged and processed products with fixed weights are included
in the quantitative analysis, since non-packaged unprocessed products without a fixed
weight are not monitored regularly by panel databases. The selected 23 product
categories represent a range of turnover characteristics, including key HICP products.
Table 12: Selection of 23 product categories
Family Product category HICP Description

Edible oil  Olive oil, aromatic oil, other oil
Salted biscuits, natural/salted seeds, popcorn, stackers,

Savoury snacks
mixed bags, crisps
Savoury
Peas, green beans, spinach, bean sprouts, mushrooms,
Grocery Canned
lentils, beans, corn, asparagus, mixed vegetables, canned
vegetables
vegetable salad
Powder milk, ready cooked meals for babies, mashed

Baby food
potatoes, soups, baby drinks, baby flour, dry food
77
Family Product category HICP Description

White chocolate, black/milk filled chocolate bars, natural black
Chocolate  /milk chocolate bars, chocolate candies
Instant coffee, soluble/liquid coffee, chicory, arabica coffee,

Coffee  robusta/mixed coffee, decaffeinated coffee
Sweet Tea  Bags/loose/soluble tea, infusions
Grocery
Cereals  Ready cereals, cereals to cook, muesli
Cereal bars, wafers, afternoon snacks, sweet/dry biscuits,

Biscuits pastry cook biscuits, chocolate biscuits, mixed biscuits,
cookies
Mineral water  Carbonated water, plain water, aromatic water
Beverage Fruit juice  Fruit juices, vegetable/mixed drinks
Carbonated fruit drinks, carbonated soft drinks, liquid

Soft drinks
tea/coffee, tonic/ginger, cola, energy drinks
Ready-cooked
Frozen cooked meals
meals
Savoury Salty pastry (pizza, quiche), other frozen starters, aperitifs,

Starters/Pizzas
Frozen quick catering, stock and soup, sauces, dressings, spices
Single vegetables, potatoes, mixed vegetables, cooked

Frozen Vegetables
vegetables, mushrooms, mashed potato
Sweet
Ice-cream Ice cream
Frozen
Yoghurt Pasteurised yoghurt, yoghurt
Desserts Fresh desserts, fresh/white cheese, ready to eat desserts
Cheese, salty cheese/feta, mozzarella, hard cold cheese

Fresh (Gouda, Edam), old soft cheese, young soft cheese, parsley
Dairy Cheese  cheese, goat cheese, grated cheese, fondue cheese, hard
cooked cheese
Pasteurised/natural milk, natural fresh milk, fermented milk,

Milk  aromatic milk
Butter/Margarine Butter, fat to fry, low fat/spread butter, margarine
Fresh pre-packed
Fresh
bread
 Fresh pre-packed bread
Non
Dairy Dry sausage, chorizo, other cooked meat, ham, sausage,
Ham/Delicatessen paté, potted minced pork, gallantine, salted meats (bacon),
foie gras, cooked chicken meat, high prepared chicken
Due to data limitations, two product categories were removed from the initial list of
25: Fresh Prepacked Salad, as it could not be isolated from its product family in the
data source, and Eggs, because issues were faced by the data provider in obtaining
consistent and complete data across the selection of MS in the scope of the study.
4.6. Method for data extrapolation (supermarkets and discounters)

In order to increase the number of shops per CSA, number of CSAs and the
representativeness of shop types for analysis of choice, a method of extrapolating
discount store audit data from shops of the same banner, of similar size and in
comparable regions, was developed.
Where a discount store was present in the CSA according to © Nielsen Trade
Dimensions, but was not audited in © Nielsen Opus, the product assortment for a
78
proxy shop has been identified based on © Nielsen Opus data for other shops of the
same banner and size. This extrapolation technique works only where discounters
have a very similar assortment nationally in shops of similar size. The process for
including HD proxies is outlined per MS in Table 14.
Table 13: Extrapolation of discounters
MS Banner Process 2004-2012 Process 2008-2012
Lidl Used 1 time (in 1 CSA) Used 2 times (once in 2 CSA)

Belgium
Aldi Used 1 time (in 1 CSA) Same as 2004-2012
Czech Rep. N/A N/A N/A
Netto N/A Used 1 time (in 1 CSA)

Denmark
REMA 1000 N/A Used 1 time (in 1 CSA)
Leader Price Used 1 time (in 1 CSA) Same as 2004-2012
Lidl Used 1 time (in 1 CSA) Same as 2004-2012
France Dia Used 1 time (in 1 CSA) Same as 2004-2012
Aldi Used 1 time (in 1 CSA) Same as 2004-2012
Netto Used 1 time (in 1 CSA) Same as 2004-2012
Penny N/A Used 2 times (once 2 CSA)

Hungary
Lidl N/A Used 3 times (once in 3 CSA)
Penny N/A Used 1 time (in 1 CSA)
Italy Eurospin N/A Used 1 time (in 1 CSA)
Lidl N/A Used 1 time (in 1 CSA)
Lidl Used 1 time (in 1 CSA) Used 2 times (once in 2 CSA)
Poland Biedronka Used 1 time (in 1 CSA) N/A
Netto Used 1 time (in 1 CSA) N/A
Lidl Used 3 times (once in 3 CSA) Same as 2004-2012

Portugal
Mini-Preco Used 3 times (once in 3 CSA) Same as 2004-2012
Dia Used 1 time (in 1 CSA) Same as 2004-2012
Spain Maxi-Dia Used 1 time (in 1 CSA) Same as 2004-2012
Lidl Used 1 time (in 1 CSA) Same as 2004-2012
TOTAL HD proxies used 19 28
Source: EY analysis
Furthermore, the Consortium investigated whether extrapolation could be appropriate

for supermarkets in addition to discounters. Our analysis found that extrapolation was
only appropriate for two supermarket banners in the sample – Colruyt in Belgium, and
Mercadona in Spain. Based on the © Nielsen Opus data available for these two
banners, a process of extrapolation was determined, illustrated in Table 14 below.
Table 14: Extrapolation of supermarkets
MS Banner Process for 2004-2012 Process for 2008-2012
Belgium Colruyt Used 2 times (once in 2 CSA) Used 3 times (once 3 CSAs)
Spain Mercadona Used 3 times (once in 3 CSA) Same as 2004-2012
TOTAL SM proxies used 5 6
Source: EY analysis
79
4.7. Measures defined for the study
4.7.1. Choice
Food choice has been a subject of wide research and has been studied from various
perspectives. A number of different interpretations of the expression ‘food choice’
exist. In global terms, ‘food choice’ is mainly described as ‘the selection of foods for
consumption (i.e. by consumers), which results from the competing, reinforcing and
interacting influences of a variety of factors. These range from the sensory,
physiological and psychological responses of individual consumers to the interactions
between social, environmental and economic influences, and include the variety of
foods and the activities of the food industry to promote them’. In the context of this
study ‘food choice’ refers to the product assortment at retail level (i.e. what is
available on the shelves). In other terms, food choice at retail is defined as (all other
things equal) the variety of products that are made available to the consumer in a
particular product category.
Several components of “food choice” (or the variety of products) have been measured
through this study. These include:
 choice amongst different shops within a given consumer’s shopping area
 choice amongst alternative products available within a product category
(represented by the total number of different EAN codes);
 choice in the variety of prices of products within a product category;
 choice in the variety of packaging sizes within a product category;
 choice amongst alternative suppliers available on shelves.
Each of these five components of choice has been analysed at the CSA level through
the following indicators and measures.
Shop variety
Shop variety refers to the number of shops per shop type a consumer has access to in
its CSA.
Using © Nielsen Trade Dimensions data, for each CSA, and once per year (for 2004,
2006, 2008, 2010, 2012), we have counted the number of shops (per shop type)
situated within the defined boundaries of the CSA.
Whilst in reality a consumer may only make grocery purchases in 1 to 5 shops, this
measure rather reflects the full choice of shops available to a consumer based on the
maximum time/distance he or she would be willing to travel. As an illustration, some
CSA contain greater than 400 shops – whilst these are all within proximity of the
consumer, they will tend to only shop in 1 to 5 of the closest shops.
It is important to highlight some data limitations for these calculations. Measures of
shop variety are possible over the full time period from 2004-2012 for France, Spain,
Italy and Portugal (4 in total), and over 2008-2012 for these MS plus Hungary and
Belgium (6 in total). For Czech Republic, no data on the evolution of shop variety is
possible as only figures for 2012 are available, and for Poland, only 2010 and 2012 are
available. No such calculations are possible for Denmark, as © Nielsen Trade
Dimensions data does not exist for this country. As a consequence, figures for these
MS are not included in aggregated results.
Product variety
Product variety refers to the total number of different products (measure by unique
EAN codes in the © Nielsen Opus data) offered on the shelves of each shop in the
study sample.
80
For each shop, we have counted the unique EANs per product category in the summer
period (2004, 2006, 2008, 2010, 2012) and in the winter period (2004, 2006, 2008,
2010, 2012). An average figure for each year has been calculated by adding the
winter and summer total and dividing by two. When figures are presented per CSA,
the total number of unique EANs is recalculated so that the same EAN appearing in
two or more different shops within the same CSA is counted only once.
Product price variety
Product price variety refers to the range of prices offered to consumers within each
product category in a given shop, measured using © Nielsen Opus data.
For each shop, we have calculated the standard deviation of prices in a given product
category divided by the mean of prices for that product category, for summer (2004,
2006, 2008, 2010, 2012) and for winter (2004, 2006, 2008, 2010, 2012). An average
figure for each year has been calculated by adding the winter and summer total and
dividing by two. When figures are presented per CSA, the variety of prices is
recalculated so that the same price appearing in two or more different shops within
the same CSA is counted only once.
It is important to note that we have found some © Nielsen Opus product price data to
be inconsistent in terms of units and currency across shops and time periods. Where
possible incorrect data has been removed from calculations, however given the
volume of data, the removal of all inconsistent prices cannot be ensured.
Product size variety
Product size variety refers to the range of different product sizes offered to consumers
within each product category in a given shop, measured using © Nielsen Opus data.
For each shop, we have counted the number of different product sizes observed per
product category, for summer (2004, 2006, 2008, 2010, 2012) and for winter (2004,
2006, 2008, 2010, 2012). An average figure for each year has been calculated by
adding the winter and summer total and dividing by two. When figures are presented
per CSA, the total number of product sizes is recalculated so that the same pack size
appearing in two or more different shops within the same CSA is counted only once.
Product supplier variety
Product supplier variety refers to the number of different suppliers of EANs in the
assortment offered to consumers within each product category in a given shop,
measured using © Nielsen Opus data.
For each shop, we have counted the number of different suppliers observed per
product category, for summer (2004, 2006, 2008, 2010, 2012) and for winter (2004,
2006, 2008, 2010, 2012). An average figure for each year has been calculated by
adding the winter and summer total and dividing by two. When figures are presented
per CSA, the total number of suppliers is recalculated so that the same brand supplier
appearing in two or more different shops within the same CSA is counted only once.
© Nielsen Opus data does not distinguish between Private Label suppliers, and
therefore “Private Label” products count as one supplier in each retail banner.
4.7.2. Innovation
The concept of innovation in food products is complex and multi-dimensional and it
has been the subject of a large amount of theoretical and empirical literature. There
are different scopes and typologies of innovation. It can also be considered as a
controversial concept as there are discussions between stakeholders on “real” and
“false” innovations.
81
Regarding the scope, this study is exclusively concerned with product innovation and
does not take into account manufacturing process innovation, neither supply chain nor
logistics innovation.
To measure product innovation, the method adopted addresses both the number of
innovations (defined as new EAN codes in the data set) and the type of innovation.
Two separate sources are used to measure these two indicators, and therefore
absolute numbers of innovations for each of the sources cannot be reconciled.
Where an EAN code appears in a shop that was not present two years earlier this
represents, in principle, an innovation. The source for the number of innovations is ©
Nielsen Opus. In terms of the type of innovation, we have applied the methodology of
© Mintel Global New Products Database, which is among the most comprehensive
consumer product database worldwide, including specific categories for food and drink.
It categorises new products into the following groups:
 New product
 New variety/range extension
 New packaging
 New formulation
 Relaunch
The different measures of innovation are presented below:
Number of innovations
The number of innovations refers to the total number of EAN codes present in the
assortment of a given period for a given shop that were not present in the same
sample 24 months previously, measured using © Nielsen Opus data. In this case, the
new EAN is counted as an innovation for the period when it appeared in the
assortment.
For each shop, we have counted the number of new EAN codes observed in each
period for summer (2004, 2006, 2008, 2010, 2012) and for winter (2004, 2006, 2008,
2010, 2012). An average figure for each year has been calculated by adding the
winter and summer total and dividing by two. When figures are presented per CSA,
the number of innovations is recalculated so that the same innovation appearing in
two or more different shops within the same CSA is counted only once.
The number of new EAN codes is an indicator of innovation, but does not seek to
assess the quality or extent of innovation in the product in question. The reason why
© Nielsen Opus data has been used for the total number of innovations as opposed to
© Mintel GNPD is because the latter source has broadened its coverage of innovation
over recent years across product categories, which would result in bias through
overestimating the number of innovative products. As a consequence, © Mintel GNPD
is only used to identify the different types of innovation present in sample shops, by
matching the EAN codes between the two databases sources.
Categories of innovation
The five categories of new products as defined by © Mintel GNPD61, and used for this
study, are as follows. Each product can be classified as one type of innovation only.
 New product: assigned when a new range, line, or family of products is
encountered. This launch type is also used if a brand that already exists on
GNPD, in one country, crosses over to a new sub-category.
61
An illustration of the different categories of innovation according to GNPD is provided in Annex A.
82
 New variety/range extension: used to document an extension to an

existing range of products on the GNPD
 New packaging: determined by visually inspecting the product for changes,
and also when terms like New Look, New Packaging, or New Size are written on
pack.
 New formulation: determined by visually looking for key terms on pack like
New Formula, Even Better, Tastier, Now Lower in Fat, New and Improved,
Great New Taste, Now With…, or Better …. We cannot assume that a product is
newly reformulated unless it is clearly stated on pack or we know from
secondary sources that this is the case.
 Relaunch: a product should have the launch type relaunch when: there is
some wording to the effect that the product has been relaunched on the
packaging or the product does not exist on the database but there is secondary
source information (such as from a press release, magazine, trade show,
website or a shop display) that the product has been relaunched. Key phrases
to look out for include “previously or formerly known as…” and “new name”. If
a product meets the criteria for the new packaging launch type and for the new
formulation launch type, then the relaunch launch type should be selected.
New EAN codes in the © Nielsen Opus data have been matched with EAN codes in
theGNPD database, in order to determine the different types of innovation present in
the sample. Due to different coverage levels of GNPD and across time and countries,
only the categorisation of innovation has been provided through GNPD, whilst the
number of innovations has been determined through shop assortment data. The
absolute number of innovations according to each different source cannot be
reconciled, since there are more innovations identified through © Nielsen Opus data
than through Mintel GNPD.
Therefore the descriptive statistics provide the proportion of each type of innovation,
overall and by product category, in relation to the total count of innovations identified
through the matching method described above.
4.7.3. Retail concentration

Retail concentration refers to the market share of modern retail groups in their
respective markets. For this study, a number of indicators of retail concentration have
been measured, both at procurement (national) level and at the local level.
It is important to clarify how retail concentration is measured at the procurement
(national) level. Measuring retail concentration at the buying group level would enable
the impact of buying alliances on choice and innovation to be determined. However, in
reality, procurement organisations and buying alliances are a complex phenomenon.
Procurement organisations exist at pan-European, national and regional level and their
scope of purchasing depend on the given shop, product category and whether it
concerns a branded product or private label. The key source on procurement
organisations is © Planet Retail. A thorough analysis of this database has revealed the
complexity of procurement organisations in Europe, as we have found references to
several procurement organisations for a given banner and retail group. Furthermore,
information is not available on the split and scope and volume of products and brands
purchased by each (proportion purchased centrally vs locally, for example), as these
arrangements tend to be confidential and informal.
Considering information is incomplete and complex, the Consortium proposed to
express retail concentration at procurement level in terms of the retailer group and
banner market shares at national level only. Thus retail concentration will not be
measured at procurement organisation level.
83
At procurement (national) level

There are several dimensions of analysis at the procurement (national) level. Firstly
the same two measures have been adopted:
 C(5) ratio: sum of market shares of top 5 retailers (banners and retail groups)
 Herfindahl–Hirschman Index (HHI): sum of the squares of the market shares of
all market players (banners and retail groups), expressed as a value between 0
and 10,000
At the procurement (national) level, “market share” has been measured using data
from © Planet Retail, both in terms of:
 Edible grocery market sales share: referring to company sales of food, drink &
tobacco as a percentage of consumer spend on food, drink & tobacco.
 Modern retail edible grocery market sales share: referring to edible grocery
sales share (see above) but for modern retail groups only.
Therefore, 8 different indicators of retail concentration at the procurement (national)
level have been measured for this study.
 C(5) ratio for banners measured in terms of edible grocery market sales share
 C(5) ratio for retail groups measured in terms of edible grocery market sales
share
 C(5) ratio for banners measured in terms of modern retail edible grocery
market sales share
 C(5) ratio for retail groups measured in terms of modern retail edible grocery
market sales share
 HHI for banners measured in terms of edible grocery market sales share
 HHI for retail groups measured in terms of edible grocery market sales share
 HHI ratio for banners measured in terms of modern retail edible grocery
market sales share
 HHI for retail groups measured in terms of modern retail edible grocery market
sales share
For the purposes of presenting the descriptive statistics in the following
chapter, at procurement (national) level, retail concentration is measured by
HHI for retail groups in terms of food modern retail market sales share.
Comparisons with other retail concentration indicators are made if the differences in
results provide relevant information.
At local consuming shopping area level

There are several dimensions of analysis at this local level. Firstly two different
measures have been adopted:
 C(5) ratio: sum of market shares of top 5 retailers
all market players, expressed as a value between 0 and 10,000
84
At the local level, retailers market share are not available. Therefore, “market share”
has been measured both in terms of:
 The share of sales area of each retailer in a given shop’s catchment area (CA)
(at banner and retail group level)62
 The share of total number of shops of each retailer in a given shop’s CA (at
banner and retail group level)
A retail concentration statistic has been generated for each shop based on the creation
of “catchment areas” (reflecting retailer competition for the shops within the sample,
using © Nielsen Trade Dimensions, which provides a full listing of modern retail stores
and their key characteristics.). Therefore, a CA has been defined for each and every
shop within our data sample that falls within the geographical perimeter of the
previously defined CSAs.
The outer limits of each CA (presented in Table 15 below) have been determined
through:
 Review of practice in competition cases over the last decade 63
 Analysis of a wide range of time scenarios, to ensure an adequate number of
competitors per shop, and a reasonable differentiation between Predominantly
Urban (PU), Intermediate (INT) and Predominantly Rural (PR).
Table 15: Maximum travel times for defining a given shop’s catchment area
Shop type Predominantly Urban (PU) Intermediate (INT) Predominantly Rural (PR)
Hypermarket 10 min 15 min 20 min
Supermarket 5 min 7.5 min 10 min
Discounter 5 min 7.5 min 10 min
Source: EY analysis
Whilst there are some exceptions, there are generally more competitor shops in PU
areas, than in INT and PR areas, which is logical. A conversion rule has been
developed in order to translate travel time to distance64.
Therefore, 8 different indicators of retail concentration at the local level have been
measured for this study.
 C(5) of retail banners by share of sales area: a value for each shop (and
each year) is calculated by summing the combined sales area of the five
62
The merits of this measure have been highlighted in ECB (2011), "Structural features of distributive
trades and their impact on prices in the euro area".
63
See Bundeskartellamt, Case “B2-33/07 Tengelmann/EDEKA”; DG COMP merger cases “COMP/M.5677 –
Schuitema/ Super de Boer”, “COMP/M.5790 – Lidl/Plus Romania/Plus Bulgaria”, “COMP/M.6847 –
Triton/Suomen Lähipkaupaa”, “COMP/M.5134 –Spar/Plus Hungary”, “COMP/M.1684 –Carrefour/Promodes”,
“COMP/M.991 – Promodes/Casino”
64
For each shop in sample in a given CSA, the distance and drive time between a combination of at least
10 shops has been calculated (testing the equation, Distance = a x Drive time), in order to identify an
appropriate conversion rate between time and distance for each CSA. As a result, the translation of time to
distance is different for each of the 105 CSAs. For each CSA, an R-squared value indicates the goodness of
fit in the data. On the whole, the results of the analysis have been positive. The minimum R squared value
is 83%, whilst 99% has been achieved for a number of CSA, indicating a very strong (near perfect) fit
between drive time and distance.
85
banners that have the largest share of sales area in the given shop’s CA, and
dividing by the total combined sales area of all shops in the CA.
 C(5) of retail banners by share of total number of shops: a value for each
shop (and each year) is calculated by adding the number of shops for the five
banners that have the largest share of shops in the given shop’s CA, and
dividing by the total number of shops in the CA.
 C(5) of retail groups by share of sales area: a value for each shop (and
each year) is calculated by summing the combined sales area of the five retail
groups that have the largest share of sales area in the given shop’s CA, and
dividing by the total combined sales area of all shops in the CA.
 C(5) of retail groups by share of total number of shops: a value for each
shop (and each year) is calculated by adding the number of shops for the five
banners that have the largest share of shops in the given shop’s CA, and
dividing by the total number of shops in the CA.
 HHI of retail banners by share of sales area: a value for each shop (and
each year) is calculated by summing the squares of the combined sales area of
each banner in the given shop’s CA.
 HHI of retail banners by share of total number of shops: a value for each
shop (and each year) is calculated by summing the squares of the total number
of shops for each banner in the given shop’s CA.
 HHI of retail groups by share of sales area: a value for each shop (and
each year) is calculated by summing the squares of the combined sales area of
each retail group in the given shop’s CA.
 HHI of retail groups by share of total number of shops: a value for each
shop (and each year) is calculated by summing the squares of the total number
of shops for each retail group in the given shop’s CA.
chapter, at local level, retail concentration is measured by HHI of retail
groups by share of sales area. Comparisons with other retail concentration
indicators are made if the differences in results provide relevant information.
4.7.4. Supplier concentration

Supplier concentration refers to the market share of modern retail suppliers (or brand
owners) in their respective markets. For this study, a range of indicators of supplier
concentration have been measured, both at the procurement (national) level and at
local level. The procurement (national) level measure provides an indication of the
concentration of market shares of suppliers in a Member State as a whole; while the
measure at local level reflects rather the concentration of suppliers in the assortment
on shop shelves, which is impacted by shop decisions to stock certain products and
not others. In order to distinguish these two concepts, in the descriptive statistics, the
local level measure will be referred to as “assortment concentration”.
There are several dimensions of analysis at the procurement (national) level. Firstly
the same two measures have been adopted:
 C(5) ratio: sum of market shares of top 5 suppliers
86

all suppliers, expressed as a value between 0 and 10,000
At the procurement (national) level, “market share” has been measured using data
from © Euromonitor International, both in terms of:
 Full market: grocery market sales for identified suppliers (global brand owners)
based on Retail/Off-trade market (measured by year-on-year exchange rate at
current prices). This calculation looks at the market shares of all identified
manufacturer brand suppliers, compared to the whole market, covering brand
suppliers, artisanal suppliers, other smaller local suppliers and private label
suppliers.
 Brand only: grocery market sales for identified suppliers (global brand owners)
based on Retail/Off-trade market (measured by year-on-year exchange rate at
current prices). This calculation looks at the market shares of all identified
manufacturer brand suppliers, compared to the full branded market, covering
brand suppliers, artisanal suppliers and other smaller local suppliers, but
excluding private label suppliers.
Note: the category “bread” in © Euromonitor International database covers a wider
range of products than the category “fresh prepacakged bread” as defined by ©
Nielsen Opus used in the study.
Therefore, 4 different indicators of supplier concentration at the procurement
(national) level have been measured for this study.
 C(5) ratio for full market: measured by the addition of grocery market shares
of the top 5 identified suppliers, calculated on full market grocery share
(including private label, other & artisanal suppliers) each year and for each of
the 23 product categories.
 C(5) ratio for brand only market: measured by the addition of grocery market
shares of the top 5 identified suppliers, calculated on brand only market
grocery share (including other & artisanal suppliers but excluding private
labels) each year and for each of the 23 product categories.
 HHI for full market: measured by the sum of the squares of grocery market
shares of all identified suppliers, calculated on full market grocery share
(including private label, other & artisanal suppliers) each year and for each of
the 23 product categories.
 HHI for brand only market: measured by the sum of the squares of grocery
market shares of all identified suppliers, calculated on brand only market
grocery share (including other & artisanal suppliers but excluding private
labels) each year and for each of the 23 product categories.
chapter, at procurement (national) level, supplier concentration is measured
by HHI for brand only market, since negotiations at procurement level occur
differently for brand versus private label suppliers. Comparisons with other
supplier concentration indicators are made if the differences in results provide relevant
information.
87
At local consumer shopping area level

There are several dimensions of analysis at this local level. Firstly two different
measures have been adopted:
 C(5) ratio: sum of market shares of top 5 suppliers
all suppliers, expressed as a value between 0 and 10,000
At the local level, “market share” has been measured by “assortment share”, i.e. the
proportion of EAN codes in the assortments at shop level for each supplier.
Therefore, 2 different indicators of assortment concentration at the local level have
been measured for this study.
 C(5) ratio measured in terms of the combined share of EAN codes in the
assortments at shop level for the top 5 suppliers each year (2004, 2006, 2008,
2010, 2012).
 HHI measured in terms of the sum of the squares of the share of EAN codes in
the assortments at shop level for all suppliers each year (2004, 2006, 2008,
2010, 2012).
chapter, at local CSA level, supplier concentration is measured by HHI in
terms of the squares of the share of EAN codes in the assortments at shop
level for all suppliers.
As a reminder, © Nielsen Opus data does not distinguish between Private Label
suppliers, and therefore “Private Label” products count as one supplier in each banner.
4.7.5. Measure of imbalance

The measure of imbalance refers to the ratio of retail concentration divided by supplier
concentration, to obtain a measure of balance between retailers and suppliers. For this
study, a selected range of indicators have been measured. Local level and
procurement level indicators have been measured for comparison purposes, however
it should be noted that it is at procurement level where the relationship between
suppliers and retailers is most appropriately measured.

The measure used is the Herfindahl–Hirschman Index (HHI): sum of the squares of
the market shares of all retailers or suppliers, expressed as a value between 0 and
10,000.
The numerator in the calculation is retail concentration, measured by:
 HHI for retail groups measured in terms of modern retail edible grocery market
sales share: sum of squares of modern retail grocery market shares for all
retail groups each year (2004, 2006, 2008, 2010, 2012).
The denominator in the calculation is supplier concentration, measured by:
 HHI for brand only market65 : measured by the sum of the squares of grocery
market shares of all identified suppliers, calculated on full market grocery share
65
In the econometric analysis, both this measure and one that uses the ‘full-market’
HHI for supplier concentration were examined.
88
(excluding private label, but including other & artisanal suppliers) each year
(2004, 2006, 2008, 2010, 2012) and each of the 23 product categories.
The equation for the calculation is: Measure of imbalance = log (HHI retailers/HHI
suppliers). This measure of imbalance entails the following advantages:
 Same calculation method can be applied regardless of the precise market
situation, i.e. regardless of whether retailers or suppliers are more
concentrated
 Outcome is centred around 0, which is easier for the reader to interpret
 Symmetry is preserved: the outcome of the statistic is the same regardless as
to whether for instance the retailer HHI is twice as high as the supplier HHI, or
vice versa.
An average measure of imbalance figure is calculated by taking the average of all 14
measures of imbalance calculated separately.
At local consumer shopping area level
The measure used is the Herfindahl–Hirschman Index (HHI): sum of the squares of
the market shares of all retailers or suppliers, expressed as a value between 0 and
10,000.
The numerator in the calculation is retail concentration, measured by:
 HHI of retail groups by share of sales area: a value for each shop (and for each
year 2004, 2006, 2008, 2010, 2012) is calculated by summing the squares of
the combined sales area of each retail group in the given shop’s CA.
The denominator in the calculation is supplier concentration, measured by:
 HHI measured in terms of the sum of the squares of the share of EAN codes in
the assortments at shop level for all suppliers each year (2004, 2006, 2008,
2010, 2012).
Once again, the equation for the calculation is: Measure of imbalance = log (HHI
retailers/HHI suppliers).
4.7.6. Shop type

Shop type refers to the type of modern retail store: either a hypermarket (>2500 m²),
supermarket (between 400m² and 2,499 m²) or discount store (all shop sizes).
Two different measures have been adopted for shop type, one at the local level, and
the other at the procurement (national) level.
At procurement (national) level, using © Planet Retail we have summed the total
number of shops per shop type per year in each Member State.
At local CSA level, using © Nielsen Trade Dimensions we have counted the number of
shops for each type that fall within the boundaries of the CSAs in the sample. A figure
is provided for each year (2004, 2006, 2008, 2010, 2012). Due to © Nielsen Trade
Dimensions data restrictions on shop type, the 2004-2012 data set covers 4 MS,
whilst the 2008-2012 sample covers 6 MS.
4.7.7. Shop size

Shop size refers to the sales area dedicated to edible grocery in modern retail shops.
We have identified the sales area of all shops in the sample of CSAs and CAs, through
using © Nielsen Trade Dimensions. Sales area dedicated to edible grocery has been
calculated based on the following assumptions:
 Hypermarkets = Food sales area is 50% of total sales area
89
 Supermarkets = Food sales area is 80% of total sales area

 Discount stores = Food sales area is 90% of total sales area
Due to © Nielsen Trade Dimensions data restrictions on shop type, the 2004-2012
data set covers 4 MS, whilst the 2008-2012 sample covers 6 MS.
4.7.8. Private label share

Private label share refers to the market share of private label products compared to all
edible grocery products available. Two different measures have been adopted, one at
the local level, and the other at the procurement (national) level.
At procurement (national) level, private label share refers to the proportion of private
label sales compared to total edible grocery retail sales per product category over
time, using © Euromonitor International as the source. It is calculated for each
product category, by summing the private label component of market size in millions
of euros for Retail/Off-trade (retail channels) at retail selling price (using year-on-year
exchange rate at current prices).
At local CSA level, private label share refers to the proportion of private label EAN per
product category per shop over time, using © Nielsen Opus as the source. It is
calculated for each product category, by summing the total EAN codes identified as
Private Label and dividing by the total number of EAN in that given product category,
for summer (2004, 2006, 2008, 2010, 2012) and for winter (2004, 2006, 2008, 2010,
2012). An average figure for each year has been calculated by adding the winter and
summer total and dividing by two.
4.7.9. Product category turnover

Product category turnover refers to retail sales per product category over time. For
each product category, it is measured by market size in millions of euros for
Retail/Off-trade (retail channels) at retail selling price (using year-on-year exchange
rate at current prices), using © Euromonitor International as the source. No quality
check of © Euromonitor International data has been performed.
4.7.10. Socio-demographic characteristics

The study looks at the evolution of a number of socio-demographic statistics, in an
effort to observe any impact they may have on choice and innovation. Six separate
indicators have been gathered for the study.
 Population size: in millions of people, measured at the NUTS 3 level for each
CSA using Eurostat data (demo_r_d3avg)
 Population density: measured at the NUTS 3 level for each CSA using Eurostat
data (demo_r_d3dens)
 Unemployment rate, measured at the NUTS 2 level for each CSA using Eurostat
data (lfst_r_lfu3rt)
 GDP per capita, measured by GDP at current prices (Purchasing Power
Standard per inhabitant ) at NUTS 3 level for each CSA using Eurostat data
(nama_r_e3gdp) with the exception of 2012, for which data comes from
Banque de France based on European Commission estimates (AMECO)
 Final consumption expenditure of households on food and non-alcoholic
beverages at national level using Eurostat data (nama_co3_c)
 Retailers’ business expectations at the national level using Eurostat data
(ei_bsrt_m_r2)
Population size, population density, unemployment rate and GDP per capita figures
represent the average across the CSAs within the sample MS. Figures for final
consumption expenditure and retailers’ business expectations relate to values for the
MS as a whole.
90
The index “Retailers’ business expectations” is a forward-looking index which

measures the expected business situation of firms operating in the retail industry
(NACE Rev. 2) over the coming three months. The measure used is a monthly
seasonally adjusted balance; the last month of each half i.e. June for H1 and
December for H2 has been used to capture retailers’ expectations of future business
activity for the periods for which Opus data have been gathered.
Where there are data gaps in Eurostat for certain regions, appropriate proxies have
been defined. For example, where 2012 data is not available 2011 has been used as a
proxy. Where NUTS 3 data is not available, the NUTS 2 figure has been used. No
quality check of Eurostat data has been performed.
4.7.11. Region/MS characteristics

Region/MS characteristics refer to a range of qualitative factors that may influence
choice and innovation. For this study, factors that have been addressed include the
legal environment, shop opening hours and pricing policies. Qualitative analysis of key
characteristics in each Member State is addressed in the relevant sections of the
descriptive statistics.
4.8. Database construction

Analysis of choice and innovation in 23 product categories across a broad sample of
modern retail grocery stores in Europe over the period of eight years has required the
compilation a significantly large amount of data (11 million records in total), acquired
from multiple providers and consolidated in a number of SQL Server databases. One
SQL Server database has been produced for each Member State that forms part of the
study (9 in total) and a consolidated file brings all results together in one central
repository.
4.8.1. Key principles and sources

Data was acquired on individual products (identified by EAN), and therefore for each
individual product stocked in each shop in our sample, in each time period, there is a
unique record in the database for that country. This has resulted in exceptionally large
datasets, of approximately 11 million records in total.
Calculations have been performed within each database based on the indicators and
measures defined in the scoping of the study. The analysis that has subsequently been
produced in the form of Excel output files is a result of export queries from the
database. An appropriate selection of these results has been presented in the present
section on descriptive statistics.
There are six key sources of data that have been compiled to produce the databases,
which are detailed below:
 © Nielsen Opus – this is the “anchor” data set from which the databases have
been constructed. © Nielsen Opus contains data for each EAN present on the
shelves of shop that was audited at a particular point in time over the eight
year period. There is one record for each individual product in each period.
Data from © Nielsen Opus was received in the form of either one or two Excel
files for each shop in our sample (343 shops in total). These individual files
were then combined into a single SQL Server database which could be queried
to perform checks, make necessary calculations and obtain results.
 © Nielsen Trade Dimensions – this data source provided a list of all shops
within a given geographic area. It includes information on the shop type, shop
91
size, banner, group, address and location (GPS coordinates). This information
was required for analysing the choice of shops for consumers in a given CSA,
the evolution of shop type and shop size in CSAs, and for retailer competition
(retail concentration at local level) for each of the sampled shops.
 © Mintel Group Ltd – Mintel GNPD is an online data repository which

catalogues and categorises new products using a range of techniques. We
downloaded extractions of © Mintel GNPD data for all time periods, countries
and product categories in the study, and matched the data with the EAN codes
observed on shop shelves. Each new product identified in GNPD is classified as
a particular type of innovation, according to consistent GNPD methodology.
 Eurostat – this is the statistical information system of the European Union

(EU), aiming to harmonise approaches across the Member States. Data has
been extracted to provide the socio-demographic statistics for each of the
NUTS 3 areas (closest equivalent to CSAs/CAs) and Member States in the
study, as well as the initial definition of the CSAs.
 © Euromonitor International– this online data was used to gather data on

supplier market shares and private label market shares for each of the product
categories across the scope of MS as well as data on market size per product
category (product category turnover). It enabled the calculation of supplier
concentration statistics at national level, and was integrated into the shop level
database to provide an alternative measure of supplier concentration and
private label share.
 © Planet Retail – this online data source provided extensive data and analysis
on edible grocery and modern retail groups to enable the calculation of retail
concentration measures at national level. It also provided a key source of
qualitative information on the evolution of MS and regional characteristics.
4.8.2. Process for consolidation of sources

Due to the significant volumes of data and the number of similar databases to be
created, the databases were created in series. To ensure consistency, the queries
required to build the databases were written once, and reused for each subsequent
country database. The queries also served as the basis for the overall database.
The diagram below (Figure 17) shows how the data sources were combined to create
each database. The databases are at EAN level, and so © Nielsen Opus forms the
basis. There is one line in the databases for each record in © Nielsen Opus. Once the
individual Excel files (received for each of the 343 shops in the sample) were
consolidated into the database, checks were carried out on the quality and
completeness of the data. Where errors or omissions were found, the project team
liaised with the data provider in order to address these issues.
With © Nielsen Trade Dimensions, a large amount of pre-processing was carried out
before it could be used in the database. Firstly extensive data cleansing techniques
were employed to ensure data was consistent, as formats and quality of presentation
differed across the MS in the sample. A macro was developed by the project team to
determine whether the shops were located within the boundaries of our CSAs. Finally,
for each shop in our © Nielsen Opus sample, we used the geo-coordinates to identify
which competitor shops were located within the boundaries of the CA. © Nielsen Trade
Dimensions data was joined to © Nielsen Opus data based on period and CSA, whilst
the competing shops (CA) are joined to each given sampled shop.
92
Using the list of competing shops, retail concentration analysis was calculated for each
given shop in the sample, in line with the measures that had been defined. This was
joined into the main database at the level of each shop.
Calculations were performed on all records within the database, to show when a given
EAN first appeared in the sample. Each of these is subsequently classified as a new
product in © Nielsen Opus. GNPD is used to assess innovations, and is joined to a
particular EAN number. A join is only made if the record is in the same period as
when the EAN first appeared in the sample. For example, if a product was stocked in
some shops from 2006, but in others from 2008, it would be flagged as a GNPD
innovation along with the appropriate classification in 2006, but not in 2008. This
represents the fact that the consumer could have purchased the product previously,
albeit not in that particular shop. Clearly because of the nature of innovation,
innovative products can only be identified from 2006 onwards (new products in 2004
cannot be identified, as 2002 data is not available for comparison).
Eurostat was joined to the main database using CSA and time period, whilst ©
Euromonitor International and © Planet Retail are joined using period, MS and product
category.
Database
Figure construction
17: Database – per MS–and
construction perat
MSconsolidated level
and at consolidated level
Join by period, MS and

Nielsen Trade Join by period and CSA product category
Euromonitor
Dimenesions
Nielsen Opus Join by period, MS

and product
Join by period (EAN level) category
and EAN
GNPD Database Planet Retail

calculations
Join by
period
and
CSA Shop level
Eurostat
database
Output tables for
descriptive
statistics
Whilst the data is held at product (EAN) level in the database, the outputs required
are at much “higher” levels so as to be useful for presentation in the descriptive
statistics.
For each Member States, we produced a shop level summary database for use in the
econometrics. For each shop, period and product category combination there is a
single record in these shop level databases. This allows testing of econometric
equations on a consistent basis. Outputs are counted (as in the case of EANs),
counted unique (as in the case of suppliers) or grouped (as in the case of
demographics). For example, for a given shop/period/product category combination,
93
there may be 500 records in the database (in other words, 500 EANs), 10 unique
suppliers within those 500 records and only one population figure (as the shop is
always within the same CSA).
Output tables for the descriptive statistics are similar, but are calculated on a number
of different levels. For example, to show the number of EANs available in a CSA, it is
necessary to count unique records tagged to that CSA. It is not possible to retrieve
this information from the shop level database, and demonstrates why the data must
be held at EAN level.
4.8.3. Limitations of the database

Like any database of a similar nature, the quality of the outputs produced is
dependent upon the quality of the input data. The following limitations should be
borne in mind when considering the results of analyses:
 Quality of pricing data has not been audited by the provider. Where identified,
corrections to erroneous prices have been made, however this process has not
been exhaustive. The study team has presented outputs using this information
in Choice 3, but would suggest caution when interpreting these results.
 For Denmark there is no © Nielsen Trade Dimensions data, so retail
concentration and Choice 1 could not be calculated for the entire study period.
Similarly, © Nielsen Trade Dimensions is only available from 2008 onwards for
Belgium, from 2010 onwards for Poland and from 2012 for Czech Republic.
These limitations have therefore restricted the sample for which retail
concentration, shop type, shop size and Choice 1 results are available.
 In some cases totals cannot be reconciled when comparing different measures.
This is not so much a limitation, but an important explanatory note to ensure
the integrity of the results is not questioned. For example, the sum of shops
split by CSA type or GDP segmentation will not equal the total number of
shops. This is because some CSAs overlap, and shops are therefore located in
more than one CSA.
 GNPD may not capture all innovations due to its sampling methodology.
Coverage across particularly Member States in the sample has improved over
time, so as a result, GNPD has been used primarily as the source enabling
innovations to be categorised by type.
 In Hungary, there is a notable number of missing data points for ready cooked
meals in individual shops. In many cases this can be observed in the
descriptive statistics, so whilst this data has been included where it is available,
these results should be interpreted with caution. The results on this product
category for Hungary however are not expected to significantly modify the
overall observations for the category across the sample as a whole.
94
Descriptive statistics from data analysis
5. Descriptive statistics from data analysis
5.1. Introduction
This section provides a description of the evolution of choice, innovation and the a priori
drivers identified in the first phase of this study. Choice and innovation have been
observed in a sample of shops located in CSAs that represent a broad range of living
situations of EU citizens. The evolution of the a priori drivers has been measured either
at the CSA (local) level based on the sample of MS and shops selected, or alternatively
used national statistical databases and sources to provide a measurement at the
procurement (national) level.
5.2. Question 1: How has choice in the EU food sector evolved over
time and across MS?
5.2.1. Introduction
This section illustrates the results of analysis for each of the five components of choice
across the selected sample of shops and consuming shopping areas in the EU.
5.2.2. Summary of findings

Choice in shops and products for consumers, on the whole, has increased over the past
decade in the EU. This increasing trend was generally higher in the pre-crisis period
(between 2004 and 2008) and has slowed since 2008.
Choice has increased for four of the five components, as illustrated in Table 16 below.
The only measure where a decreasing trend has been observed is in the variety of prices
across a given product category, which despite an increase in the pre-crisis period,
contracted during the crisis period of 2008-2012.
Table 16: Summary of findings on evolution of choice
Component 2004-2008 2008-2012 2004-2012 Trend
+ for all sample Member States

Choice in shops + + + ++ in rural areas
+ for all shop types

Choice in
alternative ++ + + + for all sample product categories
products
+ for all shop types (++ discount stores)
+ for all sample Member States (++ for

Choice in Spain)
++ + +
packaging sizes + for all sample product categories + for
all shop types

Choice in
+ for all sample product categories
alternative ++ + +
suppliers + for supermarkets, ++ for hypermarkets
and discount stores
95
Choice in prices + -- - - overall across Member States,

per product
category66 +in discount stores, - in hypermarkets
and supermarkets.
+ Positive CAGR; - Negative CAGR; ++ CAGR is twice as much as average growth value; -- CAGR is twice as
less as average growth value
The choice trends for each of these components are developed in the sections below.
5.2.3. Findings by component of choice

Choice in shops
Overall choice in shops for consumers living in the sample CSAs increased over the past
decade (2004-2012) by 1.6% annually. During the pre-crisis period (2004-2008) the
growth was higher (1.8%) than during the crisis period since 2008 (1.3%).
The trends in choice are relatively similar across the sample of MS, however with a few
exceptions. In Figure 18 & Figure 19 below, the MS that experienced the highest growth
in the pre-crisis period (sample of 4 MS) was Portugal (5.9%) followed by Italy (1.9%).
During the crisis period from 2008-2012 (sample of 6 MS), the highest growth was in
Belgium (1.9%); followed by France (1.8%). The MS that experienced the least growth
is Italy, followed by Portugal (both saw 1.0% or less growth during the 2008-2012
period).
Figure 18: 2004-2012 data set: Total number of shops in Member State (local level) -
average CAGR across all modern retail shop types (source: EY analysis based on ©
Nielsen Trade Dimensions)
6,0%
5,0%
4,0%
3,0%
2,0%
1,0%
0,0%
France Italy Portugal Spain Total across 4 MS
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
66
Results to be considered with caution: inconsistency found in data.
96
Figure 19: 2008-2012 sample: Total number of shops in CSAs by Member State (local
level) - average CAGR across all modern retail shop types (source: EY analysis based on
© Nielsen Trade Dimensions)
2,5%
2,0%
1,5%
1,0%
0,5%
0,0%
Belgium France Hungary Italy Portugal Spain TOTAL
CAGR(08 - 12)
Variations in choice of shops were observed in the different types of living areas and GDP
segmentation. As shown in Figure 20 below, during the pre-crisis period the annual
growth registered in predominantly rural areas (3.6%) was twice as high as the growth
in intermediate (1.8%) and predominantly urban areas (1.7%). By comparison the crisis
period saw lower annual growth rates across all types of living areas, with predominantly
urban (1.6%) seeing higher growth than predominantly rural (1.5%), with intermediate
registering the lowest growth rate (0.8%). The same trend was observed in both the
2004-2012 and 2008-2012 samples. It should be remembered, however, that the
number of shops in predominantly rural represent only 10% of the total number of shops
–vs 8% in 2004.
Figure 20: 2004-2012 data set: Total number of shops in CSAs by CSA type of living
(local level) - average CAGR across all modern retail shop types (source: EY analysis
based on © Nielsen Trade Dimensions)
4,0%
3,5%
3,0%
2,5%
2,0%
1,5%
1,0%
0,5%
0,0%
Predominantly urban Intermediate Predominantly rural
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
Similarly, low GDP areas observed highest growth rates in terms of number of shops
over the period, but their absolute numbers represented only 1.6 to 2.2% of the total
number of shops of the sample.
97
Figure 21 : 2004-2012 data set: Total number of shops in CSAs by CSA GDP
segmentation (local level) - average CAGR across all modern retail shop types (source:
EY analysis based on © Nielsen Trade Dimensions)
9,0%
8,0%
7,0%
6,0%
5,0%
4,0%
3,0%
2,0%
1,0%
0,0%
High Medium+ Medium Low
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
Choice in alternative products

Overall choice in alternative products has increased over the past decade in the CSAs of
the sample by 5.1% annually. During the pre-crisis period (2004-2008) the annual
growth was higher (7.8%) than during the crisis period since 2008 (2.4%).
The growth trend in choice quite similar across CSA types, ranging from 3.6% to 9.3%
over the period; the highest growth was seen in the sampled shops in the less
prosperous predominantly urban CSAs.
Figure 22 : 2004-2012 data set: Total EAN codes by CSA type and GDP range (local level)
and average annual growth rate across 23 product categories (source: EY analysis based
on © Nielsen Opus)
0,76 1,26 Annual

1,76 growth of total2,76
2,26 number 3,26
of EAN by3,76
CSA type4,26 4,76
10% 300000
8% 250000
200000
6%
150000
4%
100000
2% 50000
0% 0
Predominantly urban Intermediate Predominantly rural TOTAL
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12) 2004 Values
98
0,76 Annual growth

1,76 2,76of total number
3,76 of EAN by GDP
4,76 5,76
10% 250000
8% 200000
6% 150000
4% 100000
2% 50000
0% 0
Low Medium Medium + High TOTAL
When aggregating the sampled shops by MS, the trends in choice differ across the
sample of MS. As shown in Figure 23 and Figure 24 below, the MS that experienced the
highest growth in the pre-crisis period were Poland, Spain and France. During the crisis
period (sample of 6 MS), the highest growth was in Belgium67, followed by Portugal,
Spain and Poland. The MS that has experienced the slowest growth is Italy, followed by
France (both saw less than 2% annual growth from 2008-2012). The highest growth
rates were experienced in the MS with the lowest number of EAN codes in 2004.
Figure 23: 2004-2012 data set: Total EAN codes by Member State (local level) and
average annual growth rate across 23 product categories (source: EY analysis based on
© Nielsen Opus)
50 000 14%
12%
40 000 10%
8%
30 000 6%
4%
20 000 2%
0%
10 000 Belgium France Italy Poland Portugal Spain Total
EAN
- codes
2004 2006 2008 2010 2012 across
all 6 MS
Belgium France Italy
Poland Portugal Spain
CAGR(04 - 08) CAGR(08 - 12)
67
Due to the small sample size in Belgium, this growth appears to be influenced by the shops selected in the
sample and the trend is abnormal compared other MS. As a consequence, this result should be interpreted with
caution. Small sample size may also affect results relating to Czech Republic and Denmark.
99
Figure 24: 2008-2012 sample: Total EAN codes by Member State (local level) and
Opus)
7%
60000
6%
5%
40000 4%
3%
2%
20000 1%
0%
0
2008 2010 2012
Belgium
Czech Republic
Denmark
France CAGR(08 - 12)
Hungary
Variations in choice trends have been observed across the three shop types. Choice has
grown differently across the shop types over the past decade. As shown in Figure 25
below, for the sample of 6 MS, during the pre-crisis period, choice in discount stores
grew at the fastest rate, followed closely by hypermarkets, with supermarket registering
the lowest annual growth rate. During the crisis period, growth for discount stores
slowed slightly, and for supermarkets moderately, however the annual growth for
hypermarkets decreased to a greater extent (just over 2% for 2008-2012 compared to
8% for 2004-2008).
Figure 25: 2004-2012 data set: Total EAN codes by shop type (local level) (source: EY
analysis based on © Nielsen Opus)
2004-2012 data set: Total number of EAN

2004-2012 data set: Total EAN codes by shop type - average CAGR across
codes by shop type - across 23 23 product categories and 6 MS
product categories and 6 MS 10%
100 000 9%
90 000 8%
80 000 7%
70 000 6%
60 000 5%
50 000 4%
40 000 3%
2%
30 000
1%
20 000
0%
10 000
Discount Stores Hypermarkets Supermarkets
0
2004 2006 2008 2010 2012 CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
Discount Stores
Hypermarkets
100
These trends are amplified when looking at the 2008-2012 shop sample covering 9 MS,
as shown in Figure 26 below. As can be seen in Figure 26 below, during the crisis period
the growth in discount stores has far exceeded both supermarkets and hypermarkets, to
be mitigated by the fact that discount stores are also shops where the variety of EAN
codes is the lowest.
Figure 26: 2008-2012 sample: Total EAN codes by shop type (local level) (source: EY
analysis based on © Nielsen Opus)
2008-2012 sample: Total EAN codes 2008-2012 sample: Total number of EAN
by shop type - across 23 product codes by shop type - average CAGR across
categories and 9 MS 23 product categories and 9 MS
120 000
7%
100 000 6%
80 000 5%
60 000 4%
3%
40 000
2%
20 000
1%
0
0%
2008 2010 2012
Discount Stores Hypermarkets Supermarkets
Discount Stores
Hypermarkets CAGR(08 - 12)
Supermarkets
Choice in products per product category

Choice has increased across all product categories over the past decade; however
evolutions vary significantly across product categories, illustrated by figures below.
The 2004-2012 data set covering 6 MS in Figure 27 below shows that 6 of 23 product
categories registered greater than 6% compound annual growth over 2004-2012
(notably ham/delicatessen, cereals and cheese), whilst 5 of 23 did not exceed 4%
annual growth (with butter/margarine the lowest at 2%).
101
Figure 27: 2004-2012 data set: Total number of EAN codes by product category (local
level) - average CAGR across 6 MS (source: EY analysis based on © Nielsen Opus)
14% 8000
12% 7000
6000
10%
5000
8%
4000
6%
3000
4%
2000
2% 1000
0% 0
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12) 2004 values
The 2008-2012 shop sample covering 9 MS in Figure 28 below shows that 7 of 23

product categories saw greater than 3% annual growth from 2008-2012 (notably cereals
and ham/delicatessen), and another 8 of 23 registered growth of less than 2% (the
lowest being frozen vegetables). Trends across the two samples are generally similar for
the majority of product categories. In contrast to what was observed before, the growth
intensity has no link with the initial value: ham/delicatessen and cheese observed the
strongest growth rates in spite of the high level of the 2004 value in terms of number of
EAN codes, placing them among the largest product categories.
Figure 28: 2008-2012 sample: Total number of EAN codes by product category (local
6%
5%
4%
3%
2%
1%
0%
CAGR(08 - 12)
102
Choice in product size

The variety of different product sizes available within each product category, on the
whole, has observed a positive trend, amounting to 3.5% over the period, with higher
growth during the pre-crisis period (5%). Predominantly urban and/or medium GDP
range areas both experienced higher growth rates than other areas.
Figure 29: 2004-2012 data set: Total number of pack sizes by CSA type and GDP range
(local level) - average CAGR across 23 product categories (source: EY analysis based on
© Nielsen Opus)
Annual growth of total number of pack sizes by CSA types

0,76 1,26 1,76 2,26 2,76 3,26 3,76 4,26 4,76
6,0% 12000
5,0% 10000
4,0% 8000
3,0% 6000
2,0% 4000
1,0% 2000
0,0% 0
Predominantly urban Intermediate Predominantly rural TOTAL
Annual growth of total number of pack sizes by GDP

0,76 1,26 1,76 2,26 2,76 3,26 3,76 4,26 4,76 5,26 5,76
7,0% 12000
6,0% 10000
5,0% 8000
4,0%
6000
3,0%
2,0% 4000
1,0% 2000
0,0% 0
When aggregating the sampled shops by MS, for the 2004-2012 data set covering 6 MS
in Figure 30 below, choice has increased over time in all MS, ranging from 1.6% to 6.1%
over the 2004-2012 period. Similarly to trends observed in other indicators, the trend
over the pre-crisis period was more positive (between 2.1% and 8.6%) than that of the
crisis period (between 1.2% and 4.1%). The most significant annual growth has been
observed in Spain, followed by France and Poland. Italy, on the other hand, registered
the lowest growth level. The growth levels do not seem to be correlated to the initial
number of pack sizes in 2004.
103
Figure 30: 2004-2012 data set: Total number of pack sizes by Member State (local level)
Opus)
10% 1 000
9% 900
8% 800
7% 700
6% 600
5% 500
4% 400
3% 300
2% 200
1% 100
0% -
Belgium France Italy Poland Portugal Spain Total number
of pack sizes
across 6MS
For the 2008-2012 sample covering 9 MS in the Figure 31 below, the growth levels
largely followed the situation in the 2004-2012 data set. Portugal, Spain and Belgium
accounted for the strongest growth, while Hungary contracted slightly. Of the other MS
not covered in the 2004-2012 data set, Denmark recorded growth of 3.3% whilst the
Czech Republic’s growth rate was 0.5%. It is important to note that results relating to
these two latter MS are based on a limited number of observations.
Figure 31: 2008-2012 sample: Total number of pack sizes by Member State (local level)
Opus)
4,0% 1200
3,5%
1000
3,0%
2,5%
800
2,0%
1,5% 600
1,0%
400
0,5%
0,0%
200
-0,5%
-1,0% 0
Belgium Czech Denmark France Hungary Italy Poland Portugal Spain Total
Republic number
of pack
sizes
across 9
MS
CAGR(08 - 12) 2008 values
Choice evolutions vary significantly across the sample product categories, illustrated by
figures below. Figure 32 below, covering the 2004-2012 data set across 6 MS, shows
26% of product categories registered greater than 4% compound annual growth over
2004-2012, whilst 17% saw very growth that did not exceed 2%. The figure also
104
demonstrates that the positive evolution was much greater over the pre-crisis period
than the crisis period. In fact, growth contracted for three product categories (cheese,
frozen vegetables, and ham/delicatessen) over the crisis period.
Figure 32: 2004-2012 data set: Total number of pack sizes by product category (local
14%
12%
10%
8%
6%
4%
2%
0%
-2%
-4%
-6%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
The 2008-2012 shop sample across 9 MS in Figure 33 below, shows 50% of product
categories saw greater than 2% annual growth from 2008-2012, and another 30%
registered growth of less than 2%. Three product categories registered a contraction in
choice (frozen vegetables, ham/delicatessen and ready-cooked meals). Trends across
the two samples are generally similar for the majority of product categories.
Figure 33: 2008-2012 sample: Total number of pack sizes by product category (local
10%
8%
6%
4%
2%
0%
-2%
-4%
CAGR(08 - 12)
Variations in the choice trend have also been observed across the three shop types.
When looking at the 2004-2012 shop sample covering 6 MS in Figure 34 below, choice
105
has grown differently across shop types. During the pre-crisis period, choice in discount
stores grew at the fastest rate, followed closely by hypermarkets, with supermarkets
much lower. During the crisis period, growth for discount stores slowed, for
supermarkets growth levels remained constant, and the growth for hypermarkets
decreased to a greater extent (1.5% for 2008-2012 compared to 5% for 2004-2008).
Figure 34: 2004-2012 data set: Total number of pack sizes by shop type (local level)
(source: EY analysis based on © Nielsen Opus)
2004-2012 data set: Total number of

pack sizes by shop type - across 23 2004-2012 data set: Total number of pack
product categories and 6 MS sample sizes by shop type - annual growth rate
across 23 product categories and 6 MS
2 000 7% sample
1 800
6%
1 600
5%
1 400
1 200 4%
1 000 3%
800 2%
600
1%
400
0%
200
Hard Discounters Hypermarkets Supermarkets
0
2004 2006 2008 2010 2012
CAGR(04 - 08) CAGR(08 - 12)
Hard Discounters
Hypermarkets CAGR(04 - 12)
Supermarkets
These trends slightly differ across shop types when looking at the 2008-2012 shop
sample covering 9 MS. As can be seen in Figure 35 below, during the crisis period the
growth for discount stores was greater than for hypermarkets, however supermarkets
registered the highest growth rate.
Figure 35: 2008-2012 sample: Total number of pack sizes by shop type (local level)
(source: EY analysis based on ©Nielsen Opus)
2008-2012 sample : Total number of pack 2008-2012 sample: Total number of

sizes by shop type - across 23 product pack sizes by shop type - annual
categories and 9 MS sample growth rate across 23 product
3%
3 000 categories and 9 MS sample
3%
2 000 2%
1 000 2%
0 1%
2008 2010 2012 1%
Hard Discounters Hypermarkets
0%
Supermarkets Hard Hypermarkets Supermarkets
Discounters
CAGR(08 - 12)
106
Choice in suppliers
The number of different suppliers within each product category, on the whole, has
observed a positive trend.
The total number of different suppliers increased by 3.9% annually on average across
the sampled shops and CSAs. During the pre-crisis period (2004-2008) the annual
growth was higher (5.8%) than during the crisis period since 2008 (2%).
More than for other choice measures, differences among CSAs were noticed: less
prosperous predominantly rural areas experienced the lowest growth in the number of
suppliers available (0.4%) whereas predominantly urban areas with medium range of
GDP experienced the highest growth.
Figure 36: Number of suppliers by CSA type and GDP range (local level) – average CAGR
across 23 product categories (source: EY analysis based on © Nielsen Opus)
Number of unique suppliers by CSA type

8,0%
6,0%
4,0% CAGR(04 - 08)
2,0% CAGR(08 - 12)
0,0% CAGR(04 - 12)

Predominantly Intermediate Predominantly TOTAL
rural urban
Number of unique suppliers by GDP range

8,0%
6,0%
CAGR(04 - 08)
4,0%
CAGR(08 - 12)
2,0% CAGR(04 - 12)
0,0%
When aggregating the data for the sampled shops by MS, for the 2004-2012 data set as
shown in Figure 37 below, choice increased over time in all MS, ranging from 1.7% to
6.4% annual growth over the 2004-2012 period. Similarly to trends observed in other
indicators, the trend over the pre-crisis period was more positive (between 2.1% and
9.9%) than that of the crisis period (between -0.8% and 6.8%). The most significant
growth was observed in Spain, followed by Poland and Portugal. Italy, on the other
hand, registered the lowest growth at 1.7%.
107
Figure 37: 2004-2012 data set: Total number of suppliers by Member State (local level)
– average CAGR across 23 product categories (source: EY analysis based on © Nielsen
Opus)
12,0% 1800
10,0% 1600
1400
8,0%
1200
6,0% 1000
4,0% 800
600
2,0%
400
0,0% 200
-2,0% 0
Belgium France Italy Poland Portugal Spain Total brand
suppliers
across all 6
MS
For the 2008-2012 9 MS sample, as shown in Figure 38 below, the growth levels largely
reflect observations for the 2004-2012 data set. Spain, Belgium, Portugal and Hungary
accounted for the strongest growth, while France contracted. Of the MS not covered in
the 2004-2012 data set, Denmark and Czech Republic recorded growth of 1.5% and
0.6% respectively. It is important to note however that the results for these two latter
MS are based on a limited number of observations.
Figure 38: 2008-2012 sample: Total number of suppliers by Member State (local level) –
Opus)
6,0% 2500
5,0%
2000
4,0%
3,0% 1500
2,0%
1,0% 1000
0,0%
500
-1,0%
-2,0% 0
Belgium Czech Denmark France Hungary Italy Poland Portugal Spain Total
Republic brand
suppliers
across all
CAGR(08 - 12) 2008 values 9 MS
Choice evolutions once again vary to a large extent across the sample product
categories, illustrated by figures below.
The 2004-2012 6 MS sample in Figure 39 below shows that 35% of product categories
registered greater than 4% compound annual growth over 2004-2012, whilst three
108
product categories (butter/margarine, coffee, frozen vegetables) saw growth that did not
exceed 2%. The figure also demonstrates a more significant positive evolution over the
pre-crisis period than the crisis period. The number of suppliers fell in two product
categories (frozen vegetables, and baby food) over the crisis period.
Figure 39: 2004-2012 data set: Total number of suppliers by product category (local
Opus)
10%
8%
6%
4%
2%
0%
-2%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
In the 2008-2012 9 MS sample in Figure 40 below, three product categories (cereals,

fresh pre-packaged bread, and fruit juices) saw greater than 5% annual growth from
2008-2012. Three product categories registered a contraction in choice (ready-cooked
meals, starters/pizzas and ice cream). Trends across the two samples are generally
similar for the majority of product categories, with some exceptions (such as frozen
vegetables, and baby food).
109
Figure 40: 2008-2012 sample: Total number of suppliers by product category (local
Opus)
7%
6%
5%
4%
3%
2%
1%
0%
-1%
-2%
CAGR(08 - 12)
Variations in the choice trend have also been observed across the three shop types.
When looking at the 2004-2012 6 MS sample, as shown in Figure 41 below, choice grew
differently across the shop types. During the pre-crisis period, choice in discount stores
grew at the fastest annual rate, followed closely by hypermarkets, with supermarkets
the lowest. During the crisis period, the annual growth for discount stores and
hypermarkets slowed significantly, and for supermarkets annual growth levels halved.
Figure 41: 2004-2012 data set: Total number of suppliers by shop type (local level)
2004-2012 data set: Total number of

2004-2012 data set: Total suppliers suppliers by product category - average
by shop type - across 23 product annual growth rate across 23 product
categories and 6 MS sample 8% categories and 6 MS sample
6 000 7%
5 000 6%
4 000 5%
3 000 4%
2 000
3%
1 000
2%
0
1%
2004 2006 2008 2010 2012
0%
Hard Discounter Hypermarket Hard Discounter Hypermarket Supermarket
Supermarket CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
110
These trends differ across shop types when looking at the 2008-2012 9 MS sample. As
can be seen in Figure 42 below, during the crisis period the growth for discount stores
was much lower than for hypermarkets and supermarkets, with hypermarkets
registering the highest growth rate.
Figure 42: 2008-2012 sample: Total number of suppliers by shop type (local level)
2008-2012 sample: Total suppliers 2008-2012 sample: Total number of

by shop type - across 23 product suppliers by product category - average
categories and 9 MS sample annual growth rate across 23 product
8 000 2,5% categories and 9 MS sample
7 000
2,0%
6 000
5 000
1,5%
4 000
3 000 1,0%
2 000
1 000 0,5%
0
2008 2010 2012 0,0%
Hard Discounter Hypermarket Supermarket
Hard Discounter Hypermarket
Supermarket CAGR(08 - 12)
Choice in price per product category

The variety of different prices within a given product category provides an additional
indicator of choice per product category. Unlike the other components of choice, the
overall evolution is negative. It is important, however, to emphasise that product price
data in © Nielsen Opus has been found to be inconsistent in terms of units and currency
across shops and time periods. Nielsen has confirmed this observation. As a result, the
findings below should be considered with caution.
Across all shop types, choice in price increased in the pre-crisis period, but contracted to
a greater extent in the crisis period, resulting in a decline in the spread of prices over the
period 2004-2012. The decline during the crisis period has been confirmed through the
results for the 2008-2012 sample.
When aggregating data for the sampled shops by Member States, during the pre-crisis
period, spread of prices increased most notably in Spain and Italy, and slightly in
Portugal and France. On the other hand, the spread decreased most notably in Belgium,
and to a lesser extent in Poland. During the crisis period, the spread of prices was steady
or decreased in all but one MS, Czech Republic, where it grew annually by 1.2%.
Results vary across the different shop types:
 For discount stores, in terms of the 2004-2012 data set covering 6 MS, the
spread of prices has grown in discount stores particularly over the 2008-2012
period. This trend is particularly due to growth observed in Poland and France,
and may be due to the entry into the assortment of national brands at generally
higher prices. Similar trends are observed for the 2008-2012 sample covering 9
MS.
 In relation to hypermarkets, the spread of prices contracted slightly over 2004-
2012: whilst moderate growth of 3.4% was observed from 2004-2008, a
contraction of 3.5% was noted over the 2008-2012 period. Spain and Italy are
the MS that experienced the largest growth but also the most significant
111
contractions. The contraction during the crisis period has been confirmed through
the results for the 2008-2012 sample.
 For supermarkets, in terms of the 2004-2012 data set, the opposite trend has
been observed, in that the contraction for the pre-crisis period was greater than
for the crisis period. Belgium and Spain account for the largest contractions in the
pre-crisis period, whilst Belgium contracted the most during the crisis period
along with France, Italy and Portugal, which contract by between 1% and 2%.
To assess whether the decrease in price range observed since 2008 was due to any
particular identifiable factor, an analysis of the average price for the most common
product size per product category per MS was performed.
The way in which the average price across a given product category has evolved over
time differs across MS and product categories. The most common trend is an increase in
the average price over time, which is logical given retail price inflation, however when
looking at individual MS and product categories, several relevant trends become
apparent.
In Belgium and Poland, 16 of 23 product categories experienced an observable increase
in average prices for the most common package size within the given product category,
whilst no product categories (all sizes included) saw an observable increase. The
remaining 7 product categories showed no obvious trend.
In France, 13 of 23 product categories experienced an observable increase in average
prices for the most common package size within the given product category, whilst the
average price of desserts and ice cream was lower in 2012 compared to 2004.
On the other hand in Spain and Portugal, only 9 of 23 product categories experienced an
observable increase in average prices over time. In addition, in Spain there were 5 of 23
categories (cereals, cheese, chocolate, fresh pre-packaged bread, and yoghurt) where
prices remained relatively stable over the 2004-2012 period; whilst in Portugal, there
were 5 of 23 product categories (butter/margarine, edible oil, fresh pre-packaged bread,
tea and yoghurt) where the average price increased steadily until 2008, and then in
2010 and 2012 stabilised or decreased.
Finally, in Italy a range of situations was encountered: average prices increased for 10 of
23 product categories, increased between 2004 and 2008 then subsequently stabilised or
decreased since 2010 for 4 of 23 product categories (edible oil, ice cream, mineral water
and starters/pizzas), whilst decreased for baby food.
Therefore, no conclusions can be drawn on the overall direction of the price
range contraction.
112
5.3. Question 2: How has innovation in the EU food sector evolved over
time and across MS?
5.3.1. Introduction
This section illustrates the results of analysis of innovation across the selected sample of
shops and consuming shopping areas in the EU.
5.3.2. Summary of findings

The growth in innovation for consumers, on the whole, has slowed over the past decade
in the EU. The trend was positive pre-crisis between 2006 and 2008, but reversed during
the crisis period (2008-2012). The fastest growth in the pre-crisis period was observed
in discount stores and hypermarkets, whilst the innovation trend in supermarkets was
stable. Between 2008 and 2012, the growth trend slowed in discount stores, and the
number in innovations fell in both hypermarkets and supermarkets.
In terms of the trends in the types of innovative products on offer, over the 2004-2012
period there has been an increasing trend towards new packaging and away from new
varieties and range extension products in France, Spain, Portugal and Italy. On average
across all MS in the sample, new packaging innovations represent approximately 30% of
total innovations in 2012 compared to approximately 6% in 2004. This compares to new
varieties and range extensions, whose share has decreased from 40% in 2004 to 30% in
2012.
However the trend towards increased new packaging innovations as a proportion of the
total was not observed in Poland, Hungary, Belgium, Czech Republic and Denmark,
where new product innovations have accounted for a growing proportion of total
innovations over time. It is important to note however that these results are based on a
limited number of observations.
The innovation trends corresponding to each of the categories are developed in the
sections below.
5.3.3. Findings by component of innovation

Overall number of innovations
The growth rate of the number of innovations (new EANs) has slowed over the past
decade, when analysing the two sets of data (2004-2012 and 2008-2012). As shown in
Figure 43 below, innovation increased between 2006 and 2008 (46,111 innovations in
2008 compared to 42,779 innovations in 2006), but declined between 2008 and 2010
(45,014 innovations in 2010), as well as 2010 and 2012 (40,434 innovations in 2012).
Therefore, whilst innovations are still being offered to consumers, their number is
declining, and represents a lower proportion of the overall number of products available
(figures in the circles).
113
Figure 43: 2004-2012 data set: Evolution of number of EAN codes (local level) – across
140 000
40 434
46 111 45 041
120 000
100 000 42 779
31%
30% Total new EANs
40%
Total EANs
80 000 43%
Total EANs removed
60 000
40 000
20 000 30% Share of new EANS in
the total number of
-
EANs
2004 2006 2008 2010 2012
Annual growth in the number of new EANs

6% 3,8%
4%
2%
0%
-2%
-1,2%
-4%
-6%
-5,3%
Total new EANs
CAGR(06 - 08) CAGR(08-10) CAGR(10-12)
In the 2008-2012 period in the 9 MS sample, as shown in Figure 44 below, there were
58,824 innovations in 2010 compared to 52,005 in 2012.
Figure 44: 2008-2012 sample: Evolution of number of EAN codes (local level) – across
180 000
160 000
140 000
40% 34%
120 000
100 000 Total EANs
80 000 Total EANs removed
60 000 Total new EANs
40 000 30% Share of new EANS in
20 000 the total number of
- EANs
2008 2010 2012
114
The experience with regard to the number of new EAN products made available in shops
varied across different types of CSA. The strongest growth in the pre-crisis period was
in more prosperous rural areas, prosperous predominantly rural areas and less
prosperous urban areas; during the crisis, the number of innovations only increased in
less prosperous urban areas..
Figure 45 : 2004-2012 data set: total number new EAN codes by CSA type and GDP
range (local level) (source: EY analysis based on © Nielsen Opus)
25,0%
20,0%
15,0%
10,0%
5,0%
0,0% CAGR(06 - 08)
-5,0% CAGR(08 - 12)
-10,0%
-15,0% CAGR(06 - 12)
Figure 46 and Figure 47 below confirm that, in spite of a general increase of EAN codes
in CSAs of each MS, the share of innovative products tended to decrease. For the 2004-
2012 6 MS sample below, the proportion of innovations (new EAN codes) dropped from
47% to 36% of total EAN products in average. In terms of CAGR over the period, growth
in innovation has been positive over 2006-2012 in Poland, Spain, and to a lesser extent
in Belgium; whereas the number of innovations fell in Italy and France, and to a lesser
extent in Portugal.
Figure 46: 2004-2012 data set: Evolution of innovations (new EAN codes) by MS (local
level) – average CAGR across 23 product categories (source: EY analysis based on ©
Nielsen Opus)
Proportion of new EAN codes by MS Average CAGR across 23 product

60% categories by MS
50% 20%
15%
40% 10%
30% 5%
0%
20% -5%
10% -10%
-15%
0%
2006 2008 2010 2012
Belgium France Italy

Poland Portugal Spain CAGR(06 - 08) CAGR(08 - 12) CAGR(06 - 12)
For the 2008-2012 9 MS sample below, the crisis period is highlighted. Only Belgium
registered positive growth from 2010 to 2012. On the other hand, Czech Republic and
Italy recorded notable negative growth in innovations. It is important to note however
that results for Belgium and Czech Republic are based on a limited number of
observations.
115
Figure 47: 2008-2012 data set: Total number of innovations (new EAN codes) by MS
(local level) – average CAGR across 23 product categories (source: EY analysis based on
© Nielsen Opus)
10%
5%
0%
-5%
-10%
-15%
CAGR(10 - 12)
-20%
-25%
-30%
-35%
Belgium Czech Denmark France Hungary Italy Poland Portugal Spain
Republic
Variations in the innovation trend have also been observed across the three shop types.
In all types of shops, proportion of innovation (new EAN codes) fell over the period
2006-2012 from 42% to 32% of all EAN codes (in average) in a context of increasing
number of overall EAN codes. When looking at the CAGR, in discount stores, innovations
grew throughout 2006-2012 (+4.3%), however growth slowed between 2008 and 2012
(+2.6%). For supermarkets, the number of innovations fell over 2006-2012 (-2.7%),
modestly over 2006-2008 (-0.7%) but more significantly over 2008-2012 (-3.6%).
Finally for hypermarkets, despite growth from 2006 to 2008, the overall trend from 2006
to 2012 is a slight contraction of -0.7% annual growth. These trends can be observed in
Figure 48 and Figure 49 below representing the 2004-2012 6MS sample– the same trend
was observed for the 2008-2012 9 MS sample.
Figure 48: 2008-2012 data set: Evolution of innovations (new EAN codes) by shop type
(local level) –6 MS sample (source: EY analysis based on © Nielsen Opus)
Proportion of new EAN codes by Average CAGR across 23 product categories

shop type 10% by shop type
50% 8%
40% 6%
30% 4%
20%
2%
0%
10%
-2% Hard Hypermarkets Supermarkets
0% Discounters
2006 2008 2010 2012 -4%
-6%
Hard Discounters
Hypermarkets
Supermarkets CAGR(06 - 08) CAGR(08 - 12) CAGR(06 - 12)
116
Number of innovations per product category

The proportion of innovations in total number of products tends to decrease over the
period for all categories of products, falling from 48 to 35% of the total number of EAN
codes available in average across all product categories between 2006 and 2012.
In terms of CAGR over the period, evolutions vary significantly across the sample
product categories, illustrated by figures below.
In the 2006-2012 sample, as shown in Figure 49 below, only three product categories
(baby food, cereals, ham/delicatessen) registered notable positive annual growth over
the period, another three (chocolate, soft drinks, yoghurt) were stable, and the
remainder registered negative annual growth over this period. The categories where the
growth in new products contracted the most are mineral water (-6.8%), canned
vegetables (-4.9%) and fresh pre-packaged bread (-4.3%).
117
Figure 49: 2006-2012 sample: Evolution of innovations (new EAN codes) by product
category (local level) –across 6 MS sample (source: EY analysis based on © Nielsen
Opus)
Proportion of new EAN codes by product category

80%
70%
60%
50%
40%
30%
20%
10%
0%
2006 2008 2010 2012
20%
Average CAGR by product category
15%
10%
5%
0%
-5%
-10%
-15%
CAGR(06 - 08) CAGR(08 - 12) CAGR(06 - 12)
118
The 2008-2012 9 MS sample in the figure below confirms the recent negative trend. All
but one product category (chocolate) saw negative growth over 2008-2012, the most
significant being baby food, fresh pre-packaged bread, tea and yoghurt.
Figure 50: 2008-2012 data set: Total number of innovations (new EAN codes) by product
category (local level) – average CAGR across 9 MS sample (source: EY analysis based on
© Nielsen Opus)
2%
0%
-2%
-4%
-6%
-8%
-10%
-12%
-14%
CAGR(10 - 12)
Overall evolution by type of innovation

Trends have been observed in the evolution of the different types of innovation. On the
whole, there has been a trend toward more new packaging, and away from new varieties
and range extensions particularly in France, Spain, Portugal and Italy. On the other
hand, new product innovations account for an increasing share over time in CSAs in
Poland, Hungary, Belgium, Czech Republic and Denmark 68 . In most recent years, there
has also been an increase in relaunches compared to other types of innovations. These
trends have been generally observed across all shop types and are illustrated in Figure
52 and Figure 53 below. Although new packaging used to represent a small proportion of
the overall number of innovation (less than 5% in 2004), it represents the majority of
innovations in our sample in 2012.
68
Although these results are based on a limited number of observations
119
Figure 51 : 2004-2012 data set: Proportion of types of innovations by MS (local level)

(source: EY analysis based on © Mintel GNPD and © Nielsen Opus)
100%
90%
80%
70%
60% Relaunch
50% Range extension
40%
Formula
30%
Packaging
20%
10% Product
0%
and © Nielsen Opus) 2004-2012 sample: Proportion of new product innovations
2004 2006 2008 2010 2012

100%
90%
80%
70% Relaunch
Range extension
60%
Formula
50% Packaging
40% Product
30%
20%
10%
0%
Discounter Supermarket Hypermarket Discounter Supermarket Hypermarket Discounter Supermarket Hypermarket Discounter Supermarket Hypermarket Discounter Supermarket Hypermarket
and © Nielsen Opus) 2008-2012 sample: Proportion of new product innovations
2008 2010 2012

100%
90%
80%
70% Relaunch
Range extension
60%
Formula
50% Packaging
40% Product
30%
20%
10%
0%
Discounter Supermarket Hypermarket Discounter Supermarket Hypermarket Discounter Supermarket Hypermarket
Type of innovation per product category

Evolutions in the types of innovation are heavily dependent on the product category in
question. Below an illustrative sample of product categories are presented to
demonstrate the trends in each innovation type.
120
There has been a declining trend in innovations classified as new products from 2004 to
2012; it has gone from the most common innovation type in 2004 to second place (after
range extension) in 2012. Cereals and cheese are two representative examples of this
trend, as illustrated below. In both examples, new packaging has increased its share.
Figure 54: 2004-2012 data set: Proportion of innovations by type for cereals (local level)
©Nielsen Opus)
100%
90%
80%
70%
Relaunch
60%
Range extension
50%
Formula
40% Packaging
30% Product
20%
10%
0%
2004 2006 2008 2010 2012
Figure 55: 2004-2012 data set: Proportion of innovations by type for cheese (local level)
©Nielsen Opus)
100%
90%
80%
70% Relaunch
60% Range extension
50%
Formula
40%
Packaging
30%
Product
20%
10%
0%
2004 2006 2008 2010 2012
Innovations classified as new products have become less common across all shop types.
In 2012, new products as a proportion of total innovations were highest in
hypermarkets, followed by discount stores. Very similar trends were observed in the
results of the 2008-2012 9 MS sample.
121
Figure 56: 2004-2012 data set: Proportion of innovations classified as “new products”
sample (source: EY analysis based on © Mintel GNPD and © Nielsen Opus)
60%
50%
40% 2004
2006
30%
2008
2010
20%
2012
10%
0%
Discount stores Supermarkets Hypermarkets
New variety/range extensions increased in terms of total innovations from 2004 to 2006,
but between 2006 and 2012 lost share to below 2004 levels. Nevertheless it has gone
from the second most common type of innovation in 2004 to the most common in 2012.
The increase from 2004 to 2006 is best represented by canned vegetables, whilst the
loss in share in more recent years is illustrated by chocolate.
Figure 57: 2004-2012 data set: Proportion of innovations by type for canned vegetables
GNPD and © Nielsen Opus)
100%
90%
80%
70%
Relaunch
60%
Range extension
50%
Formula
40%
Packaging
30%
Product
20%
10%
0%
2004 2006 2008 2010 2012
122
Figure 58: 2004-2012 data set: Proportion of innovations by type for chocolate (local
and © Nielsen Opus)
100%
90%
80%
70%
Relaunch
60%
Range extension
50%
Formula
40%
Packaging
30%
Product
20%
10%
0%
2004 2006 2008 2010 2012
Innovations classified as new variety/range extensions have followed a simple trend

across all shop types. In 2004, they accounted for between 40% and 43% of innovations
across the three shop types, but by 2012, the proportion has reduced to between 30%
and 35%. In 2012, the proportion of new variety/range extensions is highest in
hypermarkets. Very similar trends were observed in the results of the 2008-2012 9 MS
sample.
variety/range extension” by © Mintel GNPD (local level) – average % across 23 product
categories and 6 MS sample (source: EY analysis based on © Mintel GNPD and © Nielsen
Opus)
60%
50%
40% 2004
2006
30%
2008
2010
20%
2012
10%
0%
New packaging has grown the most relative to other innovations, and is the third most
common type of innovation (after new variety/range extension and new product). Its
noticeable growth has best been exemplified by mineral water and edible oil, as
illustrated in Figure 60 below.
123
Figure 60: 2004-2012 data set: Proportion of innovations by type for mineral water
100%
90%
80%
70%
Relaunch
60%
Range extension
50%
Formula
40%
Packaging
30% Product
20%
10%
0%
2004 2006 2008 2010 2012
Figure 61: 2004-2012 data set: Proportion of innovations by type for edible oil (local
100%
90%
80%
70%
Relaunch
60%
Range extension
50%
Formula
40%
Packaging
30% Product
20%
10%
0%
2004 2006 2008 2010 2012
Innovations classified as new packaging have become more common across all shop
types. The increase has been most noticeable in discount stores, followed by
supermarkets. In 2012, new packaging as a proportion of total innovations was highest
in supermarket, followed by discount stores. The same trends were observed in the
124
Figure 62: 2004-2012 data set: Proportion of innovations classified as “new packaging”
sample (source: EY analysis based on © Mintel GNPD and © Nielsen Opus)
35%
30%
25%
2004
20% 2006
2008
15%
2010
10%
2012
5%
0%
New formulations account for a very small proportion of innovations, and despite
increasing their share from 2004 to 2010, this trend has been reversed from 2010 to
2012. This type of innovation has been most common in ready cooked meals and
starters/pizzas.
Figure 63: 2004-2012 data set: Proportion of innovations by type for ready-cooked
meals (local level) – average % across 6 MS sample (source: EY analysis based on ©
Mintel GNPD and © Nielsen Opus)
100%
90%
80%
70%
Relaunch
60%
Range extension
50%
Formula
40% Packaging
30% Product
20%
10%
0%
2004 2006 2008 2010 2012
125
Figure 64: 2004-2012 data set: Proportion of innovations by type for starters/pizzas
100%
90%
80%
70%
Relaunch
60%
Range extension
50%
Formula
40%
Packaging
30% Product
20%
10%
0%
2004 2006 2008 2010 2012
Innovations classified as new formulations have followed a similar trend across all shop
types. The peak in 2010 was most evident in discount stores, and the reversal from
2010 to 2012 of the historical increasing trend was noticed across all shop types.
According to the sample, new formulations are generally less common in hypermarkets
than in discount stores and supermarkets. Very similar same trends were observed in
the results of the 2008-2012 9 MS sample.

formulation” by © Mintel GNPD (local level) – average % across 23 product categories
and 6 MS sample (source: EY analysis based on © Mintel GNPD and © Nielsen Opus)
12%
10%
8% 2004
2006
6%
2008
2010
4%
2012
2%
0%
Relaunches have been the least common of innovation types but have increased
proportionally in most recent years. The particular growth from 2010 to 2012 is most
evident in baby food and tea, as illustrated in the Figure 66 and Figure 67 below.
126
Figure 66: 2004-2012 data set: Proportion of innovations by type for baby food (local
100%
90%
80%
70% Relaunch
60%
Range extension
50%
Formula
40%
Packaging
30%
Product
20%
10%
0%
2004 2006 2008 2010 2012
Figure 67: 2004-2012 data set: Proportion of innovations by type for tea (local level) –
average % across 6 MS sample (source: EY analysis based on © Mintel GNPD and ©
Nielsen Opus)
100%
90%
80%
70%
Relaunch
60%
Range extension
50%
Formula
40%
Packaging
30%
Product
20%
10%
0%
2004 2006 2008 2010 2012
Innovations classified as relaunches followed a relatively similar trend. Discount stores

accounted for the majority of the proportion in 2004, whilst supermarkets saw a
significant increase in proportion in 2012 from historically low levels. The same trends
were observed in the results of the 2008-2012 9 MS sample.
127
Figure 68: 2004-2012 data set: Proportion of innovations classified as “relaunch” by ©

Mintel GNPD (local level) – average % across 23 product categories and 6 MS sample
(source: EY analysis based on © Mintel GNPD and © Nielsen Opus)
6%
5%
4% 2004
2006
3%
2008
2010
2%
2012
1%
0%
128
5.4. Question 3: How have the a priori drivers of retail and supplier
concentration evolved over time and across MS?
5.4.1. Introduction
This section illustrates the results of analysis of retail concentration, supplier
concentration and the measure of imbalance both at the procurement (national) level
and the local CSA level. The procurement level indicator provides an appropriate
measure of the interactions between retailers and suppliers at banner and group level.
The local level indicator provides an indication of the level of retailer competition within a
given CSA.

Summary of findings
As the figure below illustrates, the evolution of retail concentration, measured by the
HHI in modern retail across EU 27 MS between 2004 and 2012 is varied.
Figure 69: comparative map of HHI modern retail across Europe (2004 - 2012)
3501 < < 4000 1501 < < 2000
1001 < < 1500

3001 < < 3500
2501 < < 3000 500 < < 1000
2001 < < 2500 N/A
At national level in terms of sales market share, modern retail concentration decreased
annually throughout 2004 to 2012 in 16 of 26 EU MS. This is generally due to the
changes in market shares among the main retailers in many Member States, amplified
by the growth of retailers who detained a small market share in 2004 or even were not
present, like hard discounters.
129
In the 14 MS sample, retail concentration increased in 7 MS (Finland, Germany,

Portugal, United Kingdom, Czech Republic, Spain and Poland) and decreased in 7 MS
(Netherland, Denmark, Belgium, Romania, France, Hungary and Italy).
At local level in terms of retailer share of sales area in the 4 MS sample (France, Italy,
Portugal and Spain), retail concentration decreased by -1.1% annually on average over
the 2004-2012 period. The decrease was the same for the pre-crisis and crisis periods.
This trend was confirmed in the results of the 2008-2012 data set, composed of 6 MS
(Belgium and Hungary in addition to the 4 MS sample), where retail concentration fell
annually by -1.3%.
Retail concentration at national level: findings by Member States
When considering all EU Member States, the highest concentration levels over the last
decade were in Finland (HHI of 3935 in 2012), Latvia (3443 in 2012), Sweden (3305 in
2012) and Cyprus (2878 in 2012, however down from 6530 in 2004); whilst the lowest
levels were seen in Italy (1170 in 2012), Hungary (1229 in 2012) and France (1419 in
2012). It is worth noting that the population size of the top 12 MS is lower in general
than the MS where retail concentration was at more moderate levels in 2012. As an
example, only the Netherlands has a population greater than 10 million inhabitants in
this list and the top 5 account for a combined population size of less than 20 million
inhabitants.
Notable increases in concentration were seen in Poland (830 in 2004 to 1580 in 2012,
CAGR of 8.4%) and Czech Republic (1200 in 2004 to 1780 in 2012); whilst Bulgaria
(2940 in 2004 to 1910 in 2012), Cyprus and Slovenia (3180 in 2004 to 2020 in 2012)
saw the most notable decreases. These trends are presented in the three figures that
follow.
130
Table 17: Retail group HHI by sales market share, for modern retail only (national level)
(source: EY analysis based on © Planet Retail)
Member State 2004 2006 2008 2010 2012 CAGR (04-12)
Finland* 2881 3736 3751 3862 3935 4,0%
Latvia 3076 3460 3590 3244 3443 1,4%
Sweden 3417 3261 3386 3359 3305 -0,4%
Cyprus 6529 4049 3634 3572 2879 -9,7%
Luxembourg 3499 3343 2998 2704 2730 -3,1%
Austria 2261 2263 2615 2598 2617 1,8%
Lithuania 2795 2282 2451 2525 2543 -1,2%

69
Netherlands* 2972 2893 2279 2043 2478 -2,2%
Ireland 2581 2511 2451 2294 2381 -1,0%
Denmark* 2373 2481 2458 2385 2320 -0,3%
Estonia 2981 2522 2308 2246 2225 -3,6%
Slovakia 1659 1772 1964 2035 2127 3,2%
Belgium* 2120 2060 1990 2000 2020 -0,6%
Slovenia 3182 2838 2216 2077 2015 -5,6%
Germany* 1384 1620 1653 1927 1957 4,4%
Bulgaria 2943 2047 1959 1646 1907 -5,3%
Portugal* 1681 1652 1830 1888 1901 1,5%
Romania* 2302 1572 1394 1361 1880 -2,5%
United Kingdom* 1748 1745 1793 1817 1811 0,4%
Czech Republic* 1199 1387 1690 1701 1779 5,1%
Spain* 1334 1422 1686 1735 1701 3,1%
Greece 1707 1648 1681 1603 1682 -0,2%
Poland* 825 926 1228 1353 1580 8,4%
France* 1533 1528 1492 1482 1410 -1,0%
Hungary* 1250 1243 1308 1198 1229 -0,2%
Italy* 1299 1220 1188 1192 1170 -1,3%
*In the 14 MS for which further analyses are conducted, retail concentration is only
slightly increasing during the period, pulled by Poland, Germany, Finland and Czech
Republic.
69
Netherlands encountered major changes in the last ten years: the major retailer
market share decreased, another retailer left the national market, and a discounter
increased its market share from 2 to 16% when another important stakeholder left the
market.
131
Retail Concentration at local level: general trends

Retail concentration at local level is based on the 4 MS analysis over the 2004-2012
period and completed by the 6 MS analysis over the 2008-2012 period. The results show
a relative de-concentration trend of retailer at local level, based on the HHI. Apart from
Spain, Belgium, Portugal and to a minor extent Italy, the trends at a local level are
similar to those observed at national level.
Retail Concentration at local level: findings by Member State
At local level using © Nielsen Trade Dimensions data, the level and evolution of retail
concentration has differed moderately across the sample MS. During the pre-crisis period
(sample of 4 MS), the most significant de-concentration was in Portugal (-3.8%),
followed by Italy (-1.3%), whilst other MS were relatively stable. By comparison, during
the crisis period, the greatest de-concentration was in Belgium (-5.0%), followed by
Spain (-2.8%). During the crisis period (6 MS sample), Portugal was the only MS in the
sample that became more concentrated (0.5%). In terms of the level of retail
concentration, France, Italy and Portugal were consistently the highest over the decade,
followed by Belgium, Spain and Hungary.
Dimensions)
3000
2500
2000
1500
1000
500
0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
France Italy Portugal Spain
132
Dimensions)
3000
2500
2000
1500
1000
500
0
2008 2010 2012
Belgium France Hungary Italy Portugal Spain
Retail concentration at local level: findings by consumer shopping

area type
At local level, retail concentration decreased slightly over the 2004-2012 period. The
level of retail concentration differs across the various types of living areas.
The types that have been most concentrated over the past decade are primarily high
GDP per capita areas in Predominantly Urban or Intermediate living zones. On the other
hand, predominantly rural areas had the lowest levels of retail concentration. These
trends for the 2004-2012 data set in Figure 72 below have been generally confirmed by
the results of the 2008-2012 data set, in the second figure.
133
Dimensions)
4000
3500
3000
2500
2000
1500
1000
500
0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
IN_High IN_Low IN_Medium IN_Medium+
PR_High PR_Low PR_Medium PR_Medium+
PU_High PU_Low PU_Medium PU_Medium+
134
Dimensions)
3000
2500
2000
1500
1000
500
0
2007 2008 2009 2010 2011 2012 2013
IN_High IN_Low
IN_Medium IN_Medium+
PR_High PR_Low
PR_Medium PR_Medium+
PU_High PU_Medium
PU_Medium+ Average per CSA type & GDP_range
In terms of evolution of retail concentration, it is not possible to draw general

conclusions for types of living areas, as the results in the 2004-2012 data set and 2008-
2012 vary significantly.
For example, for 2004-2012 in Figure 74 below, areas that are intermediate / low GDP
per capita and predominantly rural / high GDP per capita have experienced the greatest
decrease in retail concentration. Only predominantly rural / low GDP per capita areas
have slightly concentrated over this period.
On the other hand, the 2008-2012 data set, in Figure 75 below, shows that
predominantly rural / low GDP per capita areas have seen the largest decrease, following
by predominantly urban and intermediate areas.
135
on © Nielsen Trade Dimensions)
3,0%
2,0%
CAGR(04 - 08)
1,0%
0,0% CAGR(08 - 12)
-1,0% CAGR(04 - 12)

-2,0%
Average(04 - 08)
-3,0%
-4,0% Average(08 - 12)
-5,0% Average(04 - 12)

-6,0%
-7,0%
1%
0%
-1%
-2%
-3%
-4%
-5%
-6%
-7%
-8%
CAGR(08 - 12) Average(08 - 12)
5.4.3. Supplier Concentration

Summary of findings
Similar to retail concentration, trends in supplier concentration over the last time period
depend on the level of analysis. Two levels of supplier concentration have been
measured: supplier concentration at procurement or national level, measured by sales
market share (brand only), and concentration at local CSA level (or assortment
concentration) reflecting shop choices to stock certain suppliers. It is essential to
consider supplier concentration in terms of product categories as suppliers tend to focus
136
on one or a limited number of categories. Nevertheless, overall conclusions across

product categories are presented below in order to illustrate indicative average trends.
At procurement (national) level, data are available for 14 MS. Supplier concentration
increased over the 2004-2012 period across 13 of the 14 MS and for 20 of the 23
product categories. Concentration increased more during the pre-crisis period (22 of 23
product categories becoming more concentrated), than after 2008 (17 of 23 product
categories becoming more concentrated).
In terms of the level of supplier concentration at national level, the product categories
with the highest concentration levels over the last decade are baby food, frozen ready
cooked meals, cereals and coffee. The categories with the lowest concentration levels
are ham, bread and cheese. Regarding the evolution of supplier concentration at
procurement level, the product categories that have concentrated the most between
2004 and 2012 are Frozen pizzas/starters, butter/margarine and desserts. The product
categories where supplier concentration has decreased the most over this same period
are mineral water, soft drinks and biscuits.
At local CSA level, the trend in assortment concentration70 changed over the
period 2004-2012. During the 2004-2008 period, assortment concentration decreased
in all 6 MS71 in the sample by -1.3% annually on average and for most product
categories (15/23). After 2008, the decrease in assortment concentration slowed down
reaching -0.4% annually on average, with even a concentration trend in two MS (France
and Portugal) and 13 product categories becoming more concentrated. A wide range of
situations in measure of imbalance has been observed depending on the MS and the
product category. This trend was confirmed in the results of the 2008-2012 data set,
where average supplier concentration across the 23 product categories and 9 MS sample
fell by -0.3% (11 of 23 product categories becoming less concentrated).
At local level, the product categories with the highest concentration levels over the last
decade are baby food, fresh pre-packaged bread, frozen vegetables and ready cooked
meals. The categories with the lowest concentration levels are cheese, chocolate and
butter/margarine. Regarding the evolution of supplier concentration at local level, the
product categories that have de-concentrated the most between 2004 and 2012 are
chocolate, mineral water and tea. The product categories where supplier concentration
has increased the most over this same period are frozen vegetables, starters/pizzas and
savoury snacks. This result is directly linked with the increasing choice described in the
previous chapter: supplier concentration at local level measures the assortment available
on the shelves, equivalent to measuring choices in supplier.
Supplier concentration at national level
When considering the largest sample for which data was available, Denmark (HHI of
2840 in 2012, excluding private label), the Netherlands (2839 in 2012) and Finland
(2594 in 2012) have been the most concentrated on average over the past decade
across the 23 product categories, whilst Italy (1590 in 2012) and Germany (1359 in
2012) have been the least concentrated on average. In terms of the evolution over time,
supplier concentration has the most on average in Spain (1776 in 2004 to 2179 in
2012), Poland (1439 in 2004 to 1743 in 2012), and Czech Republic (1700 in 2004 to
2456 in 2012). On the other hand, supplier concentration has decreased on average in
Finland only (2792 in 2004 to 2594 in 2012).
70
Assortment concentration is the measure at local level which reflects the concentration
of suppliers in the assortment on shop shelves, which is impacted by shop decisions to
stock certain products and not others.
71
Belgium, France, Italy, Poland, Portugal, Spain
137
Table 18: Supplier concentration HHI (national level) by market share per product
category – average across 23 sample product categories (source: EY analysis based on
© Euromonitor International)
Rank Population CAGR (04-

Member State 2004 2006 2008 2010 2012
(m) 12)
1 Denmark 5.6 2433 2438 2706 2779 2840 2,0%
2 Netherlands 16.7 2575 2636 2926 2890 2839 1,2%
3 Finland 5.4 2792 2729 2768 2741 2594 -0,9%
4 Portugal 10.6 2123 2167 2289 2339 2427 1,7%
5 Belgium 11.0 2096 2240 2325 2397 2337 1,4%
6 Spain 46.2 1776 1914 1958 2018 2179 2,6%
7 France 65.2 1839 1955 1999 2123 2130 1,9%
8 Czech Republic 10.5 1700 1801 2042 2057 2057 2,4%
9 Hungary 10.0 1964 2107 2035 2055 2017 0,3%
10 United Kingdom 62.3 1717 1707 1715 1795 1766 0,4%
11 Romania 21.4 1751 1758 1721 1721 1747 0,0%
12 Poland 38.5 1440 1529 1648 1725 1743 2,4%
13 Italy 60.7 1407 1461 1500 1519 1590 1,5%
14 Germany 81.8 1202 1226 1268 1384 1359 1,5%
Calculations based on C5 are in line with these observations for all 14 MS.
Findings at national level by product category

There is significant variation in the concentration of suppliers across the different product
categories and Member States as illustrated in Table 19 below, questioning the relevance
of product categories analysis, based on an average for the 14 MS.
Note: Euromonitor bread category covers a wider range of products than fresh
prepackaged bread only.
138
Table 19: Supplier concentration by product categories and by MS – CAGR 2004-2012

(source: EY analysis based on © Euromonitor International)
United Kingdom
Average 14 MS
Czech Republic
Netherlands
Denmark
Germany
Romania
Hungary
Portugal
Belgium
Finland
France
Poland
Spain
Italy
1,1% 0,0% -2,0% -1,1% 0,5% -0,1% -1,1% 1,6% 0,2% 2,4% 2,3% -1,0% 1,0% 2,3% 0,4%
Baby food
Biscuits 0,5% -0,7% -1,1% -5,8% 3,7% -0,8% -3,0% 2,3% 4,1% 0,3% 6,8% 2,1% 5,0% 2,0% -0,2%
4,0% 18,4% -2,9% 2,1% -0,8% 1,3% 6,4% 8,3% -3,7% 9,4% -1,9% 25,6% 10,7% 8,0% 2,6%
Bread
-0,2% 0,8% 1,2% 6,2% 1,9% 1,1% -1,1% 3,7% 5,3% 9,9% 1,3% 3,0% -1,0% 8,2% 3,0%
Butter/margarine
7,7% -3,5% 8,3% -17,8% 2,0% 7,4% -6,7% 2,8% -0,6% 0,0% -5,8% 4,3% 5,0% 1,7% 0,5%
Canned vegetables
-1,3% 1,6% 3,0% 1,5% -0,4% -1,0% 1,4% 0,5% 0,5% -1,2% 4,5% -1,9% 1,0% -2,5% 0,4%
Cereals
1,7% 1,9% -0,9% 3,1% 2,6% 3,0% 2,2% 5,8% 6,0% 1,4% 5,9% 1,6% 5,8% 2,4% 2,0%
Cheese
1,2% 0,2% 0,9% 0,2% 2,3% 1,7% -1,1% 0,7% 1,5% -0,1% -0,7% -0,9% 1,9% 0,7% 0,5%
Chocolate
3,6% -0,3% 0,2% -6,1% 0,7% -0,7% -0,2% -1,3% 2,9% 2,6% 2,2% 3,0% 4,5% -4,8% 0,4%
Coffee
-2,6% 3,1% 94,6% -1,6% -4,7% 4,3% -1,3% 9,6% 5,5% 6,2% -4,2% 5,3% -4,3% -3,5% 3,2%
Desserts
-3,4% 3,8% 2,3% 1,5% 4,0% -8,3% -0,5% 7,2% 7,4% 9,7% 2,7% -1,0% 5,2% -2,8% 1,7%
Edible oil
2,7% 15,0% 6,7% 0,8% 2,4% 0,9% 9,1% 0,8% 3,1% -1,7% 4,3% 5,1% -3,2% 2,7% 4,6%
Frozen pizzas/starters
0,5%
Frozen ready cooked
meals 0,3% 5,5% 4,8% 0,8% -5,6% 3,2% -8,3% 4,0% 0,5% -0,3% -1,0% 0,0% -0,2% -0,3%
-1,2% 2,2% 5,4% -1,3% 2,4% 9,5% 1,4% 3,5% 0,0% 0,7% 0,6% -5,2% 7,5% -0,6% 1,8%
Frozen vegetables
-0,4% 2,0% -1,6% 2,8% 5,1% 6,8% 10,7% -0,2% -2,7% 3,0% 3,0% -3,3% 4,2% 8,5% 1,6%
Fruit Juices
2,6% 2,1% 9,2% 1,2% 2,8% 5,3% 13,0% 2,4% -1,1% 0,1% 1,6% -6,7% 2,8% 1,3% 2,8%
Ham
2,0% 1,0% 1,7% 2,8% 1,9% 0,5% -4,9% -9,3% 5,0% -1,2% 3,9% 0,0% 2,1% 1,4% 1,8%
Ice Cream
2,3% 3,9% -0,9% -3,0% 0,4% 9,6% -0,8% -3,2% 3,4% 9,7% 1,9% 1,7% 2,4% -5,3% 0,5%
Milk
Mineral water 1,2% 2,7% -0,2% 0,9% 2,9% -0,4% -3,0% 0,6% -9,0% 9,3% 1,2% 0,9% -1,0% -1,3% -0,5%
7,2% 0,9% 2,3% -2,6% 8,0% 3,4% 1,7% 2,9% 7,9% 2,0% 1,9% -1,5% 3,6% -0,9% 2,7%
Savoury snacks
Soft drinks -1,5% -0,2% 1,0% -2,5% 2,1% -2,0% -3,4% 2,7% -1,2% 1,3% 3,6% -1,4% 0,7% -0,7% -0,1%
4,3% -2,0% -1,1% 1,1% 0,2% 2,6% -0,4% 0,9% -4,1% 11,0% 1,8% 3,5% 0,6% -0,8% 0,6%
Tea
4,9% -3,7% -0,8% -1,8% 6,2% 0,3% 2,7% -0,6% -4,2% 6,3% 0,9% 1,4% 7,8% -0,9% 1,4%
Yoghurt
1,3%
Average 23 product
categories 1,4% 2,4% 2,0% -0,9% 1,9% 1,5% 0,3% 1,5% 1,2% 2,4% 1,7% 0,0% 2,6% 0,4%
139
The analysis below considers the 14 MS studied 14 separate markets, to reflect the fact
that procurement of FMCGs is done on a national basis (as results from the treatment of
FMCG procurement markets in competition cases). Figure 76 and Figure 77 presented
below are therefore the arithmetic average of all 14 supplier concentration HHI by
market share.
In terms of the level of supplier concentration at national level using © Euromonitor
International data, the product categories with the highest concentration levels over the
last decade across the 14 MS as a whole are frozen ready cooked meals, baby food,
cereals and coffee. Conversely, the categories with the lowest concentration levels
across the 14 MS as a whole are ham/delicatessen, cheese and bread.
level) – average across 14 MS sample – first set of categories (source: EY analysis based
on © Euromonitor International)
3 000
2 500
2 000
1 500
1 000
500
0
2004 2006 2008 2010 2012
Baby food Biscuits
Bread Butter/margarine
Canned vegetables Cereals
Cheese Chocolate
Coffee Desserts
Edible oil
Note: Bread category of EUROMONITOR covers a wider range of products than fresh
prepackaged bread.
140
level) – average across 14 MS sample – second set of categories (source: EY analysis
based on © Euromonitor International)
3 500
3 000
2 500
2 000
1 500
1 000
2004 2006 2008 2010 2012
Frozen pizzas/starters Frozen ready cooked meals

Frozen vegetables Fruit Juices
Ham Ice Cream
Milk Mineral water
Savoury snacks Soft drinks
Tea Yoghurt
Regarding the evolution of supplier concentration at procurement level, as shown in the

figure below, the product categories that concentrated the most over 2004-2012 across
the 14 MS as a whole are bread, desserts, starters/pizzas. The product categories where
supplier concentration decreased the most over this same period across the 14 MS as a
whole are mineral water, soft drinks and biscuits.
It is worth noting that supplier concentration occurred at a stronger level during the pre-
crisis period 2004-2008 (+1.9% on average) than after 2008 (+0,6% on average).
141
level) – average CAGR across 14 MS sample (source: EY analysis based on ©
Euromonitor International)
5,0%
4,0%
3,0%
2,0%
1,0%
0,0%
-1,0%
CAGR(04 - 12) Average(04 - 08) Average(08 - 12) Average(04 - 12)
Supplier concentration at local level

At the local level, in terms of assortment concentration (i.e. supplier share of EAN codes
in assortments at local level) the MS that have been most concentrated on average
across the 23 product categories over the last decade are Spain, France and Belgium.
The least concentrated on average have been Italy and Poland, in line with the
observations made at national level. The 2008-2012 data set confirms this situation, and
illustrates furthermore that Denmark is also very concentrated on average compared to
the other sample MS, and Hungary and the Czech Republic are amongst the least
concentrated MS on average. It is important to note however that results for Denmark
and the Czech Republic are based on a limited number of observations.
142

categories (local level based on HHI) (source: EY analysis based on © Nielsen Opus)
2400
2200
2000
1800
1600
1400
1200
1000
2004 2006 2008 2010 2012
Belgium France Italy Poland Portugal Spain
Regarding the evolution of supplier concentration over the last decade, Poland is the MS
that has de-concentrated the most on average since 2004, followed by Belgium and
Italy. In the 2008-2012 data set, Hungary is the MS that has de-concentrated the most
on average since 2008, followed by Belgium and Denmark. Portugal is the only MS
where suppliers have concentrated in both the 2004-2012 and 2008-2012 data sets, but
supplier concentration has also increased in Czech Republic since 2008. In general, MS
de-concentrated on average to a greater extent in the pre-crisis period than the crisis
period.
categories (local level based on HHI) (source: EY analysis based on © Nielsen Opus)
2500
2000
1500
1000
500
0
2008 2010 2012
Belgium Czech Republic Denmark

France Hungary Italy
Poland Portugal Spain
In terms of the level of assortment concentration at local level using © Nielsen Opus
data, as shown in Figure 81 and Figure 82 below the product categories with the highest
average concentration levels over the last decade across the 6 MS sample are baby food,
143
fresh pre-packaged bread, frozen vegetables and ready cooked meals. The categories
with the lowest average concentration levels across the 6 MS sample are cheese,
chocolate and butter/margarine. These situations presented below were confirmed in the
product category (local level) – average across 6 MS sample – first set of categories
4000
3500
3000
2500
2000
1500
1000
500
0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Baby food (ambient) Biscuits

Butter/margarine Canned vegetables
Cereals Cheese
Chocolate (Bar + Candies) Coffee
Dessert Edible oil
Frozen vegetables Average per product category
Fresh pre-packaged bread
144
product category (local level) – average across 6 MS sample – second set of categories
3000
2500
2000
1500
1000
500
0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Fruit juices (ambient) Ham/delicatessen
Ice cream Milk
Mineral water Ready-cooked meals
Savoury snacks Soft-drinks
Starters/pizzas Tea
Yoghurt Average per product category
Regarding the evolution of assortment concentration at local level, as shown in Figure 83

below, the product categories that have de-concentrated the most on average over
2004-2012 across the 6 MS sample are chocolate, mineral water and tea. The product
categories where supplier concentration has increased the most on average over this
same period across the 6 MS sample are frozen vegetables, starters/pizzas and savoury
snacks.
145
product category (local level) – average CAGR across 6 MS sample (source: EY analysis
based on © Nielsen Opus)
3%
2%
1%
0%
-1%
-2%
-3%
-4%
CAGR(04 - 12) Average(04 - 12)
The above situations and trends are mainly observed for the 2008-2012 9 MS sample.
146
5.4.4. Measure of Imbalance

Summary of findings
attest that they are equal numbers of situations in favour of retailers as they are
situations in favour of suppliers.
At national level, modern retail groups are concentrated to a greater extent than brand
suppliers in 6 out of 14 MS for the majority of product categories (for example: in
Finland, retailers are more concentrated than suppliers for 21 out of 23 product
categories). In the other 8 MS, suppliers are more concentrated than modern retailers
for the majority of analysed product categories (for example in Hungary: suppliers are
more concentrated than modern retailers in 17 product categories out of 23). For 12
product categories, modern retailers are more concentrated than suppliers in a majority
of the 14 analysed MS, whereas suppliers are more concentrated than modern retailers
in a majority of MS for 11 product categories. For instance, baby food and cereals
suppliers are more concentrated than modern retailers in most MS in the sample,
whereas the opposite is the case for cheese, ham or bread.
(source: EY analysis based on © Planet Retail and © Euromonitor)
Measure of imbalance 2004 2006 2008 2010 2012
In the 14 MS
Situations in favour of suppliers (MoI below 0) 168 175 165 173 162
Situations in favour of retailers (MoI above 0) 154 147 157 149 160
% of situations in favour of suppliers 52% 54% 51% 54% 50%
% of situations in favour of retailers 48% 46% 49% 46% 50%
A comprehensive view of the measure of imbalance at procurement level is provided in

the analysis below: firstly, trends by MS (averaged over 23 product categories), then by
product category (averaged over 14 MS).
The local level measure of imbalance is not presented in this section as the procurement
level is where the relationship between suppliers and retailers is most appropriately
measured.
Findings by Member State
Going through the 14 MS72, across all 23 sampled product categories over the period
2004-2012 diverse trends have been observed. At procurement level, the concentration
of retailers has grown to a greater extent than the concentration of suppliers for a
majority of product categories in Spain, Czech Republic, Finland and Germany. The
opposite trend has been observed in Belgium, France, Italy, Netherlands and Romania
where the concentration of suppliers has grown to a greater extent than the
concentration of retailers. Finally, in the UK, Hungary and Portugal, the ratio between
retail concentration and supplier concentration has remained fairly stable over the last
decade.
Of the 14 MS analysed, on average across 23 product categories, retail concentration is
higher than supplier concentration to the greatest extent in Finland and Romania. On the
72
Supplier concentration is only available for 14 MS, so as the MOI.
147
other hand, the level of supplier concentration is higher than retail concentration to the
greatest extent in France, Denmark and Hungary.
Findings by product category

The measure of imbalance at procurement level differs significantly over product
categories. Because the 14 MS cannot be considered as a single market (see approach
for supplier concentration above), on average, the product categories where retail
concentration exceeds supplier concentration most are fresh pre-packaged bread and
ham/delicatessen. On the other hand, across the 14 MS the categories where supplier
concentration exceeds retail concentration most are cereals, baby food, and savoury
snacks.
Figure 84: Measure of imbalance HHI at procurement level per product category
(national level) – average across 14 MS – first set of categories (source: EY analysis
based on © Planet Retail and © Euromonitor International)
1,50
1,00
0,50
0,00
2004 2006 2008 2010 2012
-0,50
Baby food Biscuits Bread

Butter/margarine Canned vegetables Cereals
Cheese Chocolate Coffee
Desserts Edible oil
148
(national level) – average across 14 MS – second set of categories (source: EY analysis
based on © Planet Retail and © Euromonitor International)
0,90
0,70
0,50
0,30
0,10
2004 2006 2008 2010 2012

-0,10
Frozen pizzas/starters Frozen ready cooked meals

Frozen vegetables Fruit Juices
Ham/Delicatessen Ice Cream
Milk Mineral water
Savoury snacks Soft drinks
Tea Yoghurt
In terms of evolutions of the measure of imbalance, the categories where the rate of
retail concentration growth has exceeded the rate of supplier concentration growth to
the greatest extent across the 14 MS as a whole are ready cooked meals, cheese and
desserts; whilst the categories where supplier concentration growth has exceed retail
concentration growth the most are edible oil, canned vegetables, starters/pizzas and
milk.
149
(national level) – average CAGR across 14 MS (source: EY analysis based on © Planet
Retail and © Euromonitor International)
15,0%
10,0%
5,0%
0,0%
-5,0%
-10,0%
-15,0%
-20,0%
-25,0%
-30,0%
-35,0%
CAGR(04 - 12) Average(04 - 12)
150
5.5. Question 4: How have the other a priori drivers of choice and
innovation evolved over time and across MS?
5.5.1. Introduction
In addition to concentration factors, a number of other a priori drivers of choice and
innovation have been analysed in this study. Drivers that are assessed in this section
include:
 Shop type: hypermarket, supermarket or discount store
 Shop size: shop sales area dedicated to grocery items
 Socio-demographic characteristics, including population size and density, GDP per
capita, unemployment rate and consumption of food and non-alcoholic beverage;
 Private label share: both the sales share of private label products, and their
proportion in shop assortments
 Product category turnover: the market size in terms of edible grocery sales of
each sample product category
 Retail business expectations
This section presents the evolution of these a priori drivers over the past decade.
5.5.2. Shop type

Summary of findings
In the EU 27 the most common shop formats over the past decade are supermarkets
(56% in 2012) and discount stores (38% in 2012), representing around 94% of modern
retail outlets in the EU. Hypermarkets represent the remaining 6% of modern retail
outlets. There has been growth in all modern retail shop types over the past decade,
with higher growth during the pre-crisis period (2.5%) than the crisis period (1.5%).
151
Figure 87: Growth in total number of modern retail outlets in the EU 27 (national level) -
CAGR (source: EY analysis based on © Planet Retail)
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
EU 27
Austria
Finland
France
Cyprus
Latvia
Romania
Denmark
Hungary
Italy
Netherlands
Poland
Bulgaria
Germany
Greece
Spain
Czech Republic
Ireland
Malta
Slovakia
Belgium
Slovenia
Estonia
Lithuania
Luxembourg
Portugal
Sweden
United Kingdom
-5%
CAGR (2004-2008) CAGR (2008-2012) CAGR (2004-2012)
In the sample of CSAs and MS, the most common shop format was supermarkets,
representing approximately 60% of all modern retail shops, followed by discount stores
(around 30%) and hypermarkets (around 10%). There has been growth in all shop types
over the past decade, with higher growth during the pre-crisis period than the crisis
period.
In terms of trends, during the pre-crisis period, the shop type that grew the most was
discount stores, closely followed by hypermarkets, with supermarkets registering lower
growth. During the crisis period, the growth of discount stores and hypermarkets was
similar, but notably lower than pre-crisis, and supermarket growth only fell marginally.
No noticeable differences to this trend were observed in the 2008-2012 data set.
152
shop type (local level) – across 4 MS sample (source: EY analysis based on © Nielsen
Trade Dimensions)
7000
6000
542 556
5000 491 517
466
4000
3173 3224 3296
3034 3145
3000
2000
1000 1718 1766 1877 1902

1577
0
2004 2006 2008 2010 2012
shop type (local level) – average CAGR across 4 MS sample (source: EY analysis based
3,5%
3,0%
2,5%
2,0%
1,5%
1,0%
0,5%
0,0%
Hypermarkets Supermarkets Discount stores
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
Findings by Member States

At national level using © Planet Retail data, trends in the growth of each shop type as a
whole within each MS have been different across MS.
Hypermarkets
Across the 27 EU MS, hypermarkets grew on the whole by 4.3% annually during the pre-
crisis period, and by 1.5% annually during the crisis period. Growth in hypermarkets has
been highest in Romania, Ireland, Cyprus, Lithuania and Greece; whilst growth has been
lowest or negative in Austria, Belgium, Denmark and Germany.
Across the sample MS over the past decade, hypermarkets also grew on the whole.
During the pre-crisis period, growth was highest in Portugal, followed by Italy, whilst it
153
was lowest in France. By comparison, during the crisis, growth of hypermarkets declined,
and only Portugal and Spain registered slightly positive growth, whilst Italy saw a
reduction in the number of hypermarkets between 2008 and 2012.
Portugal’s relatively higher growth can possibly be explained by a less restrictive law on
new large shop openings in March 2004. Other the other hand, Italy’s reduction in
hypermarkets since 2008 could be the result of administrative procedures hindering the
expansion of companies operating large-sized outlets. Obtaining authorisation for new
large-sized store openings in Italy is characterized by significant administrative
procedures designed to protect small shops. In spite of the important growth in
hypermarkets over the period, their overall number remains low in comparison with
other type of outlets (85 per MS in average in 2004 vs. 107 in 2012 whereas the
average number of supermarkets and discount stores are respectively 1632 and 1452 ).
Figure 90: Growth in hypermarket outlets in the EU 27 (national level) - CAGR (source:
EY analysis based on © Planet Retail)
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
-10%
-20%
CAGR (2004-2008) CAGR (2008-2012) CAGR (2004-2012)
Supermarkets
Supermarkets grew on the whole over the last decade, by 1.1% annually during the pre-
crisis period, and by 1.2% annually during the crisis period. Supermarkets accounted for
50% of total modern retail outlets in the EU 27 in 2004 compared to 47% in 2012.
Growth in supermarkets has been highest in Cyprus, Romania, Slovakia, and Estonia;
whilst growth has been lowest or negative in Denmark, Germany and the Netherlands.
Across the sample MS over the past decade, supermarkets also grew. During the pre-
crisis period, growth was highest in Portugal, followed by Italy, whilst it was lowest in
France and Spain. By comparison, during the crisis, growth of supermarkets declined
markedly. Growth remained relatively high in Portugal, and Spain saw higher growth in
the crisis period than the pre-crisis period. Highest growths correspond to MS where
initial values were low (under 100 outlets).
154
Figure 91: Growth in supermarket outlets in the EU 27 (national level) - CAGR (source:
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
Austria
France
Latvia
Romania
EU 27
Denmark
Finland
Greece
Hungary
Spain
Bulgaria
Cyprus
Ireland
Italy
Netherlands
Luxembourg
Poland
Slovakia
Slovenia
Belgium
Czech Republic
Germany
Portugal
Estonia
Lithuania
Sweden
United Kingdom
-5%
CAGR (2004-2008) CAGR (2008-2012) CAGR (2004-2012)
Discount stores
Discount stores grew the most of the modern retail shop types in the EU 27 over the
past decade, by 3.9% annually during the pre-crisis period, and by 1.9% annually during
the crisis period. Discount stores accounted for 47% of total modern retail outlets in the
EU 27 in 2004 compared to 50% in 2012. Growth in discount stores has been highest in
Slovenia, Romania, Ireland, and Latvia; whilst growth has been lowest or negative in
Austria and Greece.
Across the sample MS over the past decade, discount stores also grew the most of all
shop types. During the pre-crisis period, growth was highest in Italy, followed by
Portugal, whilst it was lowest in Spain. By comparison, during the crisis, growth of
discount stores declined. Growth remained relatively high in Portugal and Italy, and
Spain saw higher growth in the crisis period than the pre-crisis period. Highest growths
also correspond to MS where initial levels in 2004 were very low (under 100 outlets).
155
Figure 92: Growth in discount store outlets in the EU 27 (national level) - CAGR (source:
100%
80%
60%
40%
20%
0%
Austria
Finland
France
Latvia
Romania
Denmark
Poland
Germany
Greece
Hungary
Italy
Spain
Ireland
Netherlands
Portugal
Slovakia
Belgium
Czech Republic
Estonia
Luxembourg
Slovenia
Lithuania
Sweden
Grand Total
United Kingdom
-20%
-40%
CAGR (2004-2008) CAGR (2008-2012) CAGR (2004-2012)
Findings by consumer shopping area type

Variations in trends have been observed using © Nielsen Trade Dimensions data across
different CSA types in the sample of shops, CSAs and MS. For hypermarkets, the highest
growth was observed in intermediate / low GDP areas (due to very high growth in the
crisis period), whilst the lowest growth was observed in intermediate / high GDP areas.
Figure 93, Figure 94 and Figure 95 below present the trends for the 2004-2012 data set
covering 4 MS, as the 2008-2012 data set shows predominantly similar trends.
156
Figure 93: 2004-2012 data set: Growth of hypermarkets by CSA type (local level) –
16,0%
14,0%
12,0%
10,0%
8,0%
6,0%
4,0%
2,0%
0,0%
CAGR(04 - 12) CAGR(08 - 12) CAGR(04 - 08)

Average(04 -12) Average(08 -12) Average(04 -08)
For supermarkets, the highest growth was observed in intermediate / low GDP areas
(due to very high growth in the pre-crisis period) and predominantly rural / low GDP
areas, whilst the lowest growth was observed in intermediate / medium GDP areas.
Figure 94: 2004-2012 data set: Growth of supermarkets by CSA type (local level) –
13,0%
11,0%
9,0%
7,0%
5,0%
3,0%
1,0%
-1,0%
-3,0%
CAGR(04 - 12) CAGR(08 - 12) CAGR(04 - 08)

157
For discount stores, once again the highest growth was observed in intermediate / low
GDP areas and predominantly rural / low GDP areas (both due to very high growth in the
pre-crisis period), whilst the lowest growth was observed in intermediate / medium GDP
areas.
Figure 95: 2004-2012 data set: Growth of discount stores by CSA type (local level) –
25,0%
20,0%
15,0%
10,0%
5,0%
0,0%
-5,0%
CAGR(04 - 12) CAGR(08 - 12) CAGR(04 - 08)

5.5.3. Shop size

Summary of findings
In the sample of CSAs, the average shop size for 2 of the 3 modern retail formats has
increased on the whole over the past decade. This figure is confirmed by the average
size of shops in the EU 27 which grew by 75% over the last decade. Discount stores
have grown on average by 2% over the last decade, with higher growth in the pre-crisis
period (2.4%) than during the crisis period (1.5%). Supermarkets have grown on
average by 1.1% over the last decade, once again with higher growth in the pre-crisis
period (1.6%) than during the crisis period (0.6%). On the other hand, hypermarkets
have decreased on average by -0.5% over the last decade, with a higher decrease
during the crisis period (-0.8%) than the pre-crisis period (-0.1%). The 2008-2012
largely confirmed these trends however growth is stable for supermarkets in this sample.
Variations in trends have been observed across different MS. As shown in the figures
below, for hypermarkets, the decrease in average size has mainly been due to Portugal
and Spain, and to a lesser extent, France. Average shop size has only grown over the
past decade in Italy, and over the 2008-2012 period it grew in Belgium.
158
– CAGR for 4 MS sample (source: EY analysis based on © Nielsen Trade Dimensions)
1,5%
1,0%
0,5%
0,0%
-0,5%
-1,0%
CAGR(04 - 08)
-1,5%
CAGR(08 - 12)
-2,0% CAGR(04 - 12)
-2,5%
2,0%
1,5%
1,0%
CAGR(08 - 12)
0,5%
0,0%
-0,5%
-1,0%
-1,5%
-2,0%
-2,5%
The figures observed in the CSAs using © Nielsen Trade Dimensions closely reflect the
trends in the wider Member States, using © Planet Retail data. As shown in the Figure
98 and Figure 99 below, on the whole across this sample of MS, the same trend is
observed in Portugal, Spain, Hungary and Belgium. Average hypermarket size across
France has slightly increased whilst in Italy average size has slightly decreased.
159
level) – in m² for 4 MS sample (source: EY analysis based on © Planet Retail)
12000
10000
8000
6000
4000
2000
0
France Italy Spain Portugal
2004 2006 2008 2010 2012
14000
12000
10000
8000
6000
4000
2000
0
Belgium France Hungary Italy Spain Portugal
2008 2010 2012
Figure 100 and Figure 101 below illustrate that for supermarkets, the slight growth in
average size has been due to Portugal, Spain and Italy. Average shop size decreased in
France over this period, and for the 2008-2012 period it decreased in Belgium, France
and Hungary.
160
4%
3%
2%
1%
0%
-1%
-2%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
2,0%
1,5%
1,0%
0,5%
0,0%
-0,5%
-1,0% CAGR(08 - 12)
-1,5%
The figures observed in the CSAs show some differences with national sources. As
illustrated in Figure 102 and Figure 103 below, on the whole across this sample of MS,
the same trend is observed in Spain, Italy and Belgium. However average supermarket
size trends differ for France, Portugal and Hungary, due to the differences between the
MS as a whole and the CSAs selected.
161
2000
1800
1600
1400
1200
1000
800
600
400
200
0
2004 2006 2008 2010 2012
2000
1800
1600
1400
1200
1000
800
600
400
200
0
2008 2010 2012
As shown in Figure 104 below, for discount stores, the growth in average size has been
due to Italy and Spain. Average shop size grew the least in Portugal over this period. In
the 2008-2012 period, in Figure 105 below, the two MS that saw the highest growth
were Belgium and Hungary.
162
Dimensions)
5%
4%
3%
2%
1%
0%
-1%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
Dimensions)
4%
4%
3%
3%
2%
2%
1%
1%
0%
-1%
CAGR(08 - 12)
The figures observed in the CSAs above are fully reflected in the wider MS, using ©
Planet Retail. As shown in Figure 106 and Figure 107 below covering the MS as a whole,
average sales area of discount stores increased across all MS in the sample.
163
800
700
600
500
400
300
200
100
0
2004 2006 2008 2010 2012
900
800
700
600
500
400
300
200
100
0
2008 2010 2012
5.5.4. Private label share

Summary of findings
In this study, private label share has been measured both at national level, in terms of
sales market share for private labels products, and at local level, by share of private
label EANs on shop shelves.
At local level, there are a higher proportion of private label products on shop shelves,
and at national level the market share of private labels has increased. This is the case in
all sample MS and across all 23 sample product categories.
164

At procurement level, private label share ranges from 4.5% on average in Romania to
32.9% on average in Germany. There are 3 MS where private label share averaged
across the 23 product category sample exceeded 30% in 2012 – Germany, Spain, and
Portugal. There were an additional 3 MS with an average between 25% and 30% on
average – United Kingdom, Belgium and the Netherlands. On the other hand, there were
3 MS with less than 15% private label share on average – Romania, Poland and Czech
Republic.
Table 21: Private label sales share (national level) averaged across 23 product category
sample (source: EY analysis based on © Euromonitor International)
Rank Population CAGR (04-

Member State 2004 2006 2008 2010 2012
(m) 12)
1 Germany 81.797 30,48 32,18 33,38 33,54 32,93 1,0%
2 Spain 46.174 20,55 22,36 24,99 28,73 32,11 5,7%
3 Portugal 10.557 17,26 19,66 23,56 27,37 30,05 7,2%
4 United Kingdom 62.271 29,15 29,13 29,52 29,52 29,60 0,2%
5 Belgium 11.047 27,51 28,61 28,98 29,25 29,51 0,9%
6 Netherlands 16.693 23,75 24,87 25,37 26,76 27,86 2,0%
7 France 65.161 23,26 24,05 24,93 25,45 24,82 0,8%
8 Denmark 5.570 17,68 18,69 19,54 20,06 21,60 2,5%
9 Hungary 9.971 8,65 11,88 15,89 18,51 19,68 10,8%
10 Finland 5.388 13,61 15,11 16,01 17,60 19,01 4,3%
11 Italy 60.723 12,41 13,06 13,61 14,75 15,77 3,0%
12 Czech Republic 10.496 8,22 9,81 11,71 12,44 13,22 6,1%
13 Poland 38.534 5,48 6,26 6,80 7,98 11,20 9,3%
14 Romania 21.384 2,59 3,53 3,51 3,81 4,56 7,3%
Private label share has grown on average across all MS in the 14 MS sample. As shown
in Figure 108 below, progression in private label market share over the last decade in
terms of average percentage point growth across the 23 categories differed to a large
extent amongst MS. Highest growth was observed in Spain, Portugal and Hungary, whilst
the lowest growth was seen in the UK, France, Belgium and Romania.
165
Figure 108: Progression in % points of private label market share from 2004 to 2012 for
14 MS sample (national level) - average across 23 product categories (source: EY
analysis based on © Euromonitor International)
14%
12%
10%
8%
6%
4%
2%
0%
As an illustrative of a practical case: In France, the Loi de Modernisation de l'Economie

(Law on the Modernisation of the Economy (LME)) was passed in July 2008 to modernise
the French economy and encourage competition and commerce in France. The LME
allowed retailers to directly negotiate the terms and conditions of sale and prices of
suppliers, effectively enabling retailers to negotiate different prices for products with
manufacturers and to integrate back margins into the sales prices. This resulted in
decrease in the price gap between private labels and manufacturer brands, making
private labels less attractive for consumers. This has consequently impacted the market
share of private labels in France since 2009. Furthermore, the discount store format has
become less attractive to consumers relative to supermarkets and hypermarkets.
At local level, private label share has grown across all MS. Based on the sample and
irrespective of shop type, growth in private label EAN share over the last decade when
comparing 2012 with 2004 was highest in Poland, followed by Spain and Italy, whilst
France registered the lowest gain in private label EAN share. In all MS with the exception
of Portugal, private label growth was higher in the pre-crisis period than the crisis
period. These trends are presented in Figure 109 below.
166
10%
9%
8%
7%
6%
5%
4%
3%
2%
1%
0%
In terms of proportion of private labels compared to total EAN, Spain (43% in 2012) and
France (42% in 2012) have the highest average number across the 23 product
categories, whilst Poland (19% in 2012) has the lowest proportion, as illustrated below.
Opus)
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
2004 2006 2008 2010 2012
As can be seen in Figure 111 below, since 2008 the trend has been slightly different.
Highest average growth during this crisis period has been observed in Spain and
Portugal, followed by Italy, France and Belgium, whilst the lowest gain in terms of
percentage points was in Czech Republic.
167
7%
6%
5%
4%
3%
2%
1%
0%
Republic
In terms of proportion of private labels compared to total EAN during the 2008 to 2012
period in the sample shown in Figure 112 below, France and Spain still have the highest
number. On average across the 23 product categories, Denmark has 31% private label
share in 2012, while Hungary and Czech Republic have 20%. It is important to note
however that these latter results are based on limited observations.
Opus)
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
2008 2010 2012
Belgium Czech Republic
Denmark France
Hungary Italy
Poland Portugal
168

The market share of private labels is very different from one product to another. At
procurement level using © Euromonitor International data, the product categories with
the highest private label market share averaged across the sample of 14 MS were
ham/delicatessen, milk, frozen vegetables and canned vegetables. The product
categories where market share has grown the most on average over the past decade
and across MS are milk, savoury snacks, and edible oil, as illustrated below.
Figure 113: Percentage of private label sales share by product category - average across
14 MS (national level) (source: EY analysis based on © Euromonitor International)
60%
50%
40%
30%
20%
10% 2004
0% 2012
Milk
Cheese
Coffee
Edible oil
Ham
Bread
Fruit juice
Savoury snacks
Tea
Baby food
Biscuits
Fresh desserts
Frozen pizza/starters
Canned vegetables
Soft drinks
Cereals
Chocolate
Butter/margarine
Frozen vegetables
Frozen ready-made meals
Ice cream
Yoghurt
Mineral water
In terms of evolutions over time, private label market share has also grown on average
across all product categories in the sample of 6 MS, as illustrated below, despite
significant differences between categories when compared to the sample of 14 MS. Milk,
fresh pre-packaged bread, ready-cooked meals and ham/delicatessen saw the greatest
average increase in percentage point share over the decade, whilst average growth was
lowest for baby food and butter/margarine.
Table 22: Evolution of private label market share from 2004 to 2012 (national level) -
average across 6 MS sample (source: EY analysis based on © Euromonitor International)
Percentage of private label Progression

2004 2006 2008 2010 2012
EANs by product category (%) 2004-2012
Baby food (ambient) 1,2 1,3 1,5 1,9 2,3 1,0
Biscuits 17,8 19,6 21,2 22,7 24,1 6,2
Butter/margarine 7,0 7,4 7,8 7,9 8,7 1,7
Canned vegetables 16,3 17,1 18,6 19,9 21,5 5,2
Cereals 34,5 36,0 37,3 38,7 39,5 5,0
Cheese 14,2 15,8 18,0 19,2 20,4 6,3
Chocolate (Bar + Candies) 14,2 15,0 16,4 17,8 18,8 4,6
169

2004 2006 2008 2010 2012
Coffee 6,3 7,2 7,9 8,7 9,5 3,2
Dessert 9,0 9,6 10,1 11,0 12,5 3,5
Edible oil 17,0 18,5 19,9 21,1 23,2 6,2
Fresh pre-packaged bread 29,4 31,5 33,5 35,6 37,5 8,1
Frozen vegetables 24,5 26,1 28,2 29,0 29,9 5,4
Fruit juices (ambient) 26,5 29,0 31,3 32,1 32,7 6,3
Ham/delicatessen 32,1 35,6 37,5 39,0 39,6 7,4
Ice cream 19,3 20,2 21,1 22,3 23,1 3,8
Milk 32,8 35,7 37,8 41,0 42,5 9,6
Mineral water 9,6 10,4 11,5 12,6 13,7 4,1
Ready-cooked meals 26,4 27,9 29,7 31,4 34,0 7,6
Savoury snacks 9,2 10,3 11,8 13,6 14,2 5,0
Soft-drinks 16,7 18,1 20,4 22,0 22,8 6,0
Starters/pizzas 8,5 9,1 10,2 11,4 11,9 3,4
Tea 10,2 11,1 11,2 12,3 13,4 3,2
Yoghurt 10,1 10,8 11,9 13,1 15,0 4,9
At local level, the product categories with the highest proportion of private label EANs
were frozen vegetables (53% in 2012), ice cream (48%), desserts (48%) and ready
cooked meals (46%); whilst the lowest were baby food (12%), chocolate (22%) and tea
(27%). These trends are presented in Table 23 below.
Table 23: 2004-2012 data set: Proportion of private label EAN by product category (local
level) - average across 6 MS sample (source: EY analysis based on © Nielsen Opus)

2004 2006 2008 2010 2012
Baby food (ambient) 8% 8% 9% 11% 12% 4%
Biscuits 27% 28% 29% 32% 34% 6%
Butter/margarine 27% 27% 27% 29% 30% 3%
Canned vegetables 41% 42% 43% 43% 44% 3%
Cereals 40% 37% 37% 39% 41% 1%
Cheese 24% 26% 27% 31% 33% 9%
Chocolate (Bar + Candies) 19% 20% 21% 22% 22% 3%
Coffee 28% 28% 28% 29% 30% 2%
Dessert 42% 43% 45% 46% 48% 6%
Edible oil 27% 25% 24% 27% 28% 2%
Fresh pre-packaged bread 21% 26% 27% 28% 30% 9%
Frozen vegetables 45% 45% 47% 50% 53% 8%
Fruit juices (ambient) 37% 40% 41% 44% 44% 7%
170

2004 2006 2008 2010 2012
Ham/delicatessen 29% 30% 29% 32% 34% 4%
Ice cream 38% 40% 41% 46% 48% 9%
Milk 24% 26% 26% 30% 31% 7%
Mineral water 24% 27% 26% 26% 28% 4%
Ready-cooked meals 35% 38% 41% 44% 46% 10%
Savoury snacks 29% 30% 32% 33% 35% 6%
Soft-drinks 28% 30% 32% 32% 31% 3%
Starters/pizzas 35% 35% 37% 40% 42% 8%
Tea 22% 21% 21% 23% 27% 4%
Yoghurt 28% 31% 34% 36% 38% 11%
Similar to the 2004-2012 data set, in 2008-2012 as seen in Figure 114 below, frozen
vegetables (51% in 2012), desserts (46%) and ice cream (46%) have the highest
proportion of private label EANs, whilst baby food (10%) has the lowest proportion.
Figure 114: 2008-2012 data set: Proportion of private label EAN by product category
(local level) - average across 9 MS sample (source: EY analysis based on © Nielsen
Opus)
Percentage of private label EANs by

2008 2010 2012 Progression 2004-2012
product category (%)
Baby food (ambient) 8% 9% 10% 2%
Biscuits 28% 31% 32% 5%
Butter/margarine 26% 27% 28% 2%
Canned vegetables 41% 42% 43% 2%
Cereals 32% 34% 36% 4%
Cheese 25% 29% 31% 6%
Chocolate (Bar + Candies) 19% 21% 21% 1%
Coffee 27% 28% 28% 1%
Dessert 41% 44% 46% 5%
Edible oil 24% 27% 28% 4%
Fresh pre-packaged bread 30% 31% 33% 3%
Frozen vegetables 44% 48% 51% 6%
Fruit juices (ambient) 37% 40% 40% 3%
Ham/delicatessen 29% 31% 33% 4%
Ice cream 38% 44% 46% 8%
Milk 24% 28% 30% 5%
Mineral water 23% 24% 26% 3%
Ready-cooked meals 39% 42% 45% 5%
Savoury snacks 29% 31% 34% 4%
Soft-drinks 30% 31% 30% 0%
Starters/pizzas 35% 39% 42% 7%
171
Percentage of private label EANs by

2008 2010 2012 Progression 2004-2012
product category (%)
Tea 19% 21% 23% 4%
Yoghurt 32% 34% 36% 4%
In terms of evolutions over time, as shown in Figure 115, private label share has grown
across all product categories. The product categories that experienced the highest
percentage point growth from 2004 to 2012 in private label EANs were yoghurt and
ready cooked meals; whilst the categories registering the lowest level of percentage
growth were cereals, coffee and edible oil.
Nielsen Opus)
12%
10%
8%
6%
4%
2%
0%
Since 2008 the trend has been slightly different, as can be seen below. Highest
percentage point growth during the crisis period was observed in ice cream,
starters/pizzas and frozen vegetables, whilst lowest growth was in soft drinks, chocolate
and coffee.
172
Nielsen Opus)
9%
8%
7%
6%
5%
4%
3%
2%
1%
0%
5.5.5. Product category turnover

Summary of findings
Sales turnover for the sample product categories increased annually by 2.9% over the
2004-2012 period. Annual growth during the pre-crisis period (4.5%) was notably
greater than during the crisis (1.4%). 20 of the 23 product categories grew between
2004 and 2012: the only exceptions were mineral water, butter/margarine and edible
oil.
Variations in trends were observed across the sample MS. As can be seen below, annual
growth on average across the 23 sample product categories over the last decade was
highest in Poland, followed by Belgium, whilst the lowest annual growth levels were seen
in Portugal and Spain. During the crisis period, annual growth was negative in Poland,
Portugal and Spain. Highest annual growth during this period was seen in Belgium,
France, Czech Republic and Denmark.
173
16%
14%
12%
10%
8%
6%
4%
2%
0%
-2%
-4%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
4%
3%
2%
1%
0%
-1%
-2%
Republic
CAGR(08 - 12)
174

The largest product categories in terms of turnover are fresh pre-packaged bread73,
cheese and ham/delicatessen; whilst the smallest categories are starters/pizzas, tea,
ready cooked meals and cereals.
Growth rates vary significantly across product categories. As illustrated in Figure 119
and Figure 120 below, of the sample product categories, those that grew most over the
last decade across the 6 MS sample were coffee, ham/delicatessen and soft drinks. The
three product categories that contracted most were mineral water, butter/margarine and
edible oil. A larger number of product categories contracted across the 9 MS sample over
the 2008-2012 period, notably ready cooked meals, butter/margarine, milk, edible oil,
baby food and mineral water.
across 6 MS sample – first set of categories (source: EY analysis based on ©
7000
6000
5000
4000
3000
2000
1000
0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Baby food (ambient) Biscuits
Butter/margarine Canned vegetables
Cereals Cheese
Chocolate (Bar + Candies) Coffee
Dessert Edible oil
Frozen vegetables Average per product category
73
This category includes traditional and artisanal bread sold in retail, thus accounting for the high category
turnover figure
175
across 6 MS sample – second set of categories (source: EY analysis based on ©
6000
5000
4000
3000
2000
1000
0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Fruit juices (ambient) Ham/delicatessen

Ice cream Milk
Mineral water Ready-cooked meals
Savoury snacks Soft-drinks
Starters/pizzas Tea
Yoghurt Average per product category
No notable different trends were observed in the 2008-2012 9 MS sample.
5.5.6. Socio-demographic characteristics

Population size
In terms of population size, the zones of the selected CSAs in the MS saw growth over
the 2004 to 2012 period. Growth during the pre-crisis period exceeded the crisis period
in the CSA in France, Italy, Portugal and Spain; whilst for the CSA in Belgium population
growth was higher after 2008 and growth across both periods was marginal in Poland.
From 2008-2012, shown in Figure 122 below, the CSAs in Belgium saw the highest
growth of all MS, followed by Italy and Czech Republic.
176
average CAGR for 6 MS sample (source: EY analysis based on Eurostat)
2,5%
2,0%
1,5%
1,0%
0,5%
0,0%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)

Average(04 - 08) Average(08 - 12) Average(04 - 12)
1,8%
1,6%
1,4%
1,2%
1,0%
0,8%
0,6%
0,4%
0,2%
0,0%
Republic
CAGR(08 - 12) Average(08 - 12)
Population density
In terms of population density, observations varied across MS. As can be seen below,
during the pre-crisis period, density grew slightly in the CSAs in France, Italy, Poland
and Spain; whilst it decreased notably in the CSAs in Belgium and to a lesser extent in
Portugal. During the crisis period between 2008 and 2012, in Figure 124 below, the CSAs
in Belgium increased in density, as did CSAs in Spain, Czech Republic and Hungary;
whilst CSAs in France, Italy and Poland saw a decrease in population density.
177
level) - average CAGR for 6 MS sample (source: EY analysis based on Eurostat)
2,0%
1,5%
1,0%
0,5%
0,0%
-0,5%
-1,0%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)

2,0%
1,5%
1,0%
0,5%
0,0%
-0,5%
-1,0%
Republic
CAGR(08 - 12) Average(08 - 12)
178
Unemployment rate
In terms of the unemployment rate, the overall trend was towards higher unemployment
in the CSAs within the sample MS. As shown below, during the post-crisis period, the
unemployment rate increased across CSAs in all MS; whilst pre-crisis unemployment
rates decreased in the CSAs in Poland, France, and Italy. During the crisis period
between 2008 and 2012, in Figure 126 below, the largest increases in the
unemployment rate were in CSAs in Denmark, Spain, Portugal and Czech Republic.
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%
-20%
-25%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)

25%
20%
15%
10%
5%
0%
Republic
CAGR(08 - 12) Average(08 - 12)
179
GDP per capita

In terms of GDP per capita, the CSAs in the sample MS on average saw an increase in
GDP per capita on the whole over the past decade. As shown below, during the pre-crisis
period, average GDP per capita increased most in the CSAs in Poland and Spain, whilst a
small decrease was seen in average in the 3 CSAs in Belgium. During the crisis period
between 2008 and 2012, the largest increases in average GDP per capita were in the
CSAs in Denmark and Poland, whilst GDP per capita growth was negative in CSAs in
Spain, Portugal and Italy.
9%
8%
7%
6%
5%
4%
3%
2%
1%
0%
-1%
-2%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)

6%
5%
4%
3%
2%
1%
0%
-1%
-2%
Republic
CAGR(08 - 12) Average(08 - 12)
Final consumption on food and non-alcoholic beverage

In terms of final consumption on food and non-alcoholic beverage, a variety of trends
were observed across MS. As shown below, during the pre-crisis period, all MS saw a
180
decrease in the proportion of income spent on food and non-alcoholic beverage, however
during the crisis period this trend was reversed in Belgium, France, Portugal and Spain.
Only in Italy and Poland did the trend remain similar to the pre-crisis period. In the 2008
and 2012 sample, Czech Republic and Hungary saw an increase in the proportion of
income spent on food and non-alcoholic beverage, in line with the majority of MS.
Figure 129: 2004-2012 data set: Evolution of the proportion of income spent on food and
non-alcoholic beverage by Member State (national level) - CAGR for 6 MS sample
(source: EY analysis based on Eurostat)
3,0%
2,5%
2,0%
1,5%
1,0%
0,5%
0,0%
-0,5%
-1,0%
-1,5%
-2,0%
-2,5%
CAGR(04 - 08) CAGR(08 - 12) CAGR(04 - 12)
Figure 130: 2008-2012 data set: Consumption of food and non-alcoholic beverage by
Member State (national level) - CAGR for 9 MS sample (source: EY analysis based on
Eurostat)
3,0%
2,5%
2,0%
1,5%
1,0%
0,5%
0,0%
-0,5%
-1,0%
-1,5%
-2,0%
-2,5%
Republic
CAGR(08 - 12) Average(08 - 12)
181
Retail business expectations

Retail business expectations over the coming three months are heavily influenced by the
general state of the economy, but tend to be somewhat more volatile (reflecting the
speed with which confidence changes) than GDP growth, as Figure 131 shows.
Figure 131: EU28 retail business expectations and GDP growth (Source: Eurostat)
Retail business expectations is the value for the last month in the quarter. Both series are
seasonally adjusted.
The pattern across countries reflects the differences in national economic activity and the
extent to which this coincides with the EU average. Figure 132 illustrates the differences
for three countries, showing the differing trends in confidence in the period 2004-2008
and the more similar pattern from 2009.
Figure 132: Retail business expectations in France, Poland and Spain (source: Eurostat)
182
Econometric analysis scope and methodology
6. Econometric analysis scope and methodology

This section presents the final scope of data, indicators and measures for the
econometric analysis as well as specific approaches applied. Its objective is to provide an
introduction to the subsequent section on econometric results. A more comprehensive
description is provided in the annexes.
6.1. General specification

The objective of the econometric analysis is to analyse the historical evidence for the
impact of a priori drivers on each of choice and innovation. The analysis models the
behaviour of each shop and the selection of products that it offers, and seeks to explain
this with reference to various national and local drivers. It is important to note that this
differs from modelling the total assortment available to consumers from the shops to
which they have access, which would include the impact of a change in the number and
mix of types of shops in the local area. The number and mix of shops is examined and
reported in the descriptive analysis of this study.
The relationships of interest are expressed below:
[choice or innovation]s,p,t = f {
shop types,t
shop sizes,t
private label sharen/s,p,t
retailers' concentrationn/s,t
suppliers' concentrationn/s,p,t
[or imbalance (retailer vs supplier concentration)n/s,p,t]
socio-demographic indicatorsc,t
rural/urban categoryc or population densityc
product category turnovern,p,t
economic prosperity/macroeconomic conditionsc/n,t
Member Staten
product categoryp
yeary
seasonm
new competitor shop openings,t
}
where the indices used are:

c consumer shopping area
m month in the year (2nd quarter or 4th quarter),
n Member State
p product category
s shop
183
t time period (two per year, every second year)

y year
The indicators selected to measure choice, innovation and the drivers are discussed in
the annexes.
6.2. Econometric issues

Two main econometric issues arise in the estimation of relationships across these data
sets.
6.2.1. Unobserved heterogeneity among shops

This is the standard issue that arises with data where the unit of observation is an
individual (a shop, in this case). It considers the possibility that there is some difference
between the observed outcome for choice/innovation for different shops that is due to
something specific about the shop that is not already captured in the drivers. In a pure
cross section there is no way of identifying such effects, but in panel data (where
indicators are measured for the same shops over different time periods) it is
conventional to seek to use the information available for shops over time to detect such
(time-invariant) effects and thereby improve the estimates of the effects of the observed
drivers. Since the shops are a sample drawn from a wider population, in principle a
random effects specification is preferred if the data support this (Hausman test), but the
fixed-effects (within) estimator has also been calculated. In practice, in many cases the
two methods give results that are broadly similar in terms of the sign and relative size of
parameter estimates.
6.2.2. Spatial dependence

The literature on spatial econometrics identifies different kinds of spatial dependence
which call for different methods. Spatial dependence means the possibility that
outcomes in a shop are affected not just by the characteristics of that shop (including
the area/MS in which it lies) but by the behaviour of nearby shops and/or characteristics
of nearby areas. Moran’s I to test has been used for such dependence (in cross
sections).
In practice, the Hausman test is rejected in most specifications, but the fixed-effects
estimator is somewhat more vulnerable to spatial dependence, and both types of
estimator are reported in Annex F. Because of the results of the Hausman test, we give
priority to the fixed effects estimator when summarising conclusions about the impacts
of the drivers.
A particular form of spatial dependence arises when it is believed that the residuals
(which capture all the reasons for variation in the dependent variable that are not
accounted for by the drivers that have been included) could be ‘clustered’, that is related
to one another by geographical area. The shops in this study are located in common
consumer shopping areas and the possibility arises that there are unobserved (i.e. not
taken into account in the indicators that are included in the analysis) influences at the
local level that affect all shops in the same area. In that case the estimated standard
errors associated with each parameter estimate, which are used to assess whether it is
statistically significantly different from zero, would be underestimated if no allowance
were made for clustering. The results reported here use standard errors estimated on
the assumption of clustering at the CSA level so as to take a cautious approach to
reporting statistical significance of results. In many cases the parameter estimates that
are treated as statistically insignificant as a result of taking this approach are those that
are in any case so small as to be economically irrelevant.
184
Econometric analysis scope and methodology
6.3. Economic importance and statistical significance

The econometric analysis provides estimates of the impact of a driver (the parameter
estimate) and of the degree of uncertainty (due to random variation) associated with
this estimate (the standard error of the parameter estimate). If the parameter estimate
is considerably different from zero (measured by the number of standard errors), the
estimate is regarded as statistically significant: that is, if the model is correctly specified,
it is unlikely that the ‘true’ value of the parameter that is being estimated is zero.
However, an estimated impact can be statistically significant but small – too small to be
economically important. This data set has a large number of observations (shops x
product categories x time periods) and so the standard errors of the parameter
estimates are small, with the result that typically the parameter estimates are
statistically significant at the 1% level, including estimates that are small in absolute
magnitude. One therefore requires some means of assessing whether a given estimate
is large enough to be important. The dependent variables (choice and innovation) and
most of the drivers are represented in the equations in logarithmic form, which has the
benefit that the estimated impacts are ‘elasticities’, independent of the units in which
they are measured. In this kind of log-linear specification, a parameter estimate of, say,
, is interpreted as meaning that a 1% change in the value of the driver will lead to
approximately a % change in the dependent variable. If all the drivers were subject to
the same typical range of variation, the relative size of their elasticities could be used to
rank the importance of each driver’s impact. However, in practice variation in some
drivers is typically greater than in others. A driver whose typical variation in the sample
is, say, rarely more than 5% would need to be associated with a larger elasticity than
another driver whose typical variation is commonly more than 10% for them both to
have similar typical impacts on the dependent variable (calculated by multiplying the
elasticity by the variation).
The approach taken is therefore to vary each driver by an amount equivalent to one
standard deviation of its values in the data set and calculate the proportionate impact on
the dependent variable74.
In the present study, the variation being analysed is over shops (space), product
categories and time. Some indicators do not vary much over time but do vary over
space: for example, the population density of an area. Some do not vary much over
space, but do vary over product categories: for example, supplier concentration. Some
do not vary much over space or product categories, but do vary over time: for example,
national retail concentration. Therefore, we have measured one standard deviation in
each driver across all the dimensions in the data set and not, for example, simply over
time.
74
More precisely, we calculate what an equation predicts for the dependent variable
when all the drivers are set to their mean values (over the data set). We then, in turn,
increase each driver by an amount equal to one standard deviation of the values that it
takes in the data set and calculate the impact on the dependent variable, keeping all the
other drivers at their mean values. We express the impact of each such change as a
proportionate change in the dependent variable from the value predicted when all drivers
are set to their mean values.
185
7. Characteristics of the data set and its implications

for the econometric analysis
7.1. Dataset construction and availability

The data set for the econometric analysis is a subset of the data gathered for the study
and which formed the basis of the descriptive analysis presented in Chapter 4. It is a
subset because the econometric analysis requires data to be available for every indicator
(innovation, choice and every explanatory variable) for every observation (a given
product category in a given shop and time period), whereas descriptive analysis that
focuses on one indicator at a time can choose all the observations in the sample for
which data are available for each indicator in turn.
The key limitation compared with the full data set reviewed in Chapters 3 and 4 concerns
the measure of retail concentration at the local level. For some Member States,
Nielsen’s Trade Dimensions data were not available for the full time period 2004-2012.
Since this is the source of comprehensive information about the location, size and type
of shops in each area, it is required to construct the indicator that measures the degree
of retail concentration at the local level (or, in other words, the extent of local
competition faced by a given shop).
Two data sets were established, one running from 2004-2012 and the other from 2008-
2012 and, in each case, included all the Member States for which a full data set was
available for the period. Because data are available for the choice and innovation
measures for a substantial number of shops in Poland, we added Poland to the analysis
for equation specifications that use national rather than local retail concentration (which
is not available for Poland) as a driver.
7.2. Sample selection

As a reminder, through the Nielsen Opus data set, it was not possible to ensure an
entirely random selection of the shops in the data set. The shops that are included in
Opus are the ones that competitors have requested Nielsen to cover at any particular
time.
As a result, Opus has an over-representation of hypermarkets. The sample reflects this
over-representation: it has been mitigated by making special provision to choose
supermarkets and hard discounters where possible. Not all Member States have good
coverage in Opus, particularly going back over the past decade: the main use of Opus by
its customers is to gain insight into the current situation rather than a historical time
series. A complete time series of data for shops is required, and so the selection is
limited to those shops that have been included in Opus in every time period (twice per
year) in the years that have been chosen.
186
Characteristics of the data set and its implications for the econometric analysis
7.3. The scope of the data set used in the econometric analysis
The coverage of the data sets used for econometric analysis therefore comprises:
Table 24: The two data sets used in the econometric analysis
Long Data set No. of shops Short Data set No. of shops
(2004H1 - 2012H2) (2008H1 - 2012H2)
France 131 Belgium 9

Italy 80 France 131
Poland* 24 Hungary* 24
Portugal 19 Italy 83
Spain 42 Poland 29
Portugal 19
Spain 42
Total 296 Total 337
There are small differences in the selection of shops compared with the descriptive
analysis because of the requirement for the econometric analysis for data to be available
for every driver in every time period.
* Poland was omitted from analysis that included local retail concentration as a driver
because of the absence of the required Trade Dimensions data to calculate this measure.
Hungary was omitted from the analysis of innovations that covered the whole of
2008H1-2012H1 because Opus data were only available from 2008 onwards (and so the
first ‘innovation’ could only be detected in 2010).
Specificities of sample for econometri c analysis

Because the econometric analysis requires data to be available for all drivers that are
included in any given specification, econometric analysis on the long data set (2004-
2012) is limited to France, Italy, Spain and Portugal. Poland is included in the long
period econometric analysis only for those specifications that use national rather than
local retail concentration because the Nielsen Trade Dimensions data necessary to
calculate the local concentration measure are not available over that period. Over the
short term period (2008-2012), in addition to France, Italy, Spain, Portugal and Poland,
the sample covers Hungary and Belgium. Czech Republic and Denmark are not covered
in the econometric analysis, since there is insufficient data on retail concentration at the
local level. As a consequence, the findings of the econometric analysis predominantly
reflect the situations and evolution of drivers, choice and innovation in France, Italy,
Spain, Portugal and Poland, and to a lesser extent in Hungary and Belgium.
The analysis below reviews the extent to which Member State coverage of the
econometric analysis reflects the range of situations and trends found across the EU for
a number of key drivers at national level. Across the drivers of shop type, shop size,
product category turnover and socio-demographic characteristics, the Member States
included in the econometric analysis cover a broad variety of cases that are generally
found across the EU. The analysis below therefore focuses on concentration-related
drivers, namely retail concentration, supplier concentration, measure of imbalance, as
well as private label share.
187
Retail concentration at national level

There is a broad range of situations and evolutions in the EU over 2004-2012 regarding
the concentration of retailers at national level, as illustrated in Table 25 below75. The MS
that are included in the econometric analysis are highlighted in blue.
Table 25: Retail group HHI by sales market share in modern retail (national level)
(source: EY analysis based on © Planet Retail)
Rank Member State Population (m) 2004 2006 2008 2010 2012 CAGR (04-12)
1 Finland 5.388 2881 3736 3751 3862 3935 4,0%
2 Latvia 2.058 3076 3460 3590 3244 3443 1,4%
3 Sweden 9.449 3418 3261 3386 3359 3305 -0,4%
4 Cyprus 0.850 6530 4049 3634 3572 2879 -9,7%
5 Luxembourg 0.518 3499 3343 2998 2704 2730 -3,1%
6 Austria 8.423 2262 2263 2615 2598 2617 1,8%
7 Lithuania 3.030 2796 2282 2451 2525 2543 -1,2%
8 Netherlands 16.693 2972 2893 2279 2043 2478 -2,2%
9 Ireland 4.576 2582 2511 2451 2294 2381 -1,0%
10 Denmark 5.570 2374 2481 2458 2385 2320 -0,3%
11 Estonia 1.339 2981 2522 2308 2246 2225 -3,6%
12 Slovakia 5.398 1659 1772 1964 2035 2127 3,2%
13 Belgium 11.047 2120 2060 1990 2000 2020 -0,6%
14 Slovenia 2.052 3183 2838 2216 2077 2015 -5,6%
15 Germany 81.797 1384 1620 1653 1927 1957 4,4%
16 Bulgaria 7.348 2943 2047 1959 1646 1907 -5,3%
17 Portugal 10.557 1681 1652 1830 1888 1901 1,5%
18 Romania 21.384 2302 1572 1394 1361 1880 -2,5%
19 United Kingdom 62.271 1749 1745 1793 1817 1811 0,4%
20 Czech Republic 10.496 1199 1387 1690 1701 1779 5,1%
21 Spain 46.174 1335 1422 1686 1735 1701 3,1%
22 Greece 11.300 1708 1648 1681 1603 1682 -0,2%
23 Poland 38.534 826 926 1228 1353 1580 8,4%
24 France 65.161 1533 1528 1492 1482 1410 -1,0%
25 Hungary 9.971 1251 1243 1308 1198 1229 -0,2%
26 Italy 60.723 1299 1220 1188 1192 1170 -1,3%
As can be seen in the table above, the econometric analysis sample includes the four MS
with the light or moderate retail concentration levels in the EU in 2012 - Italy (HHI of
1170), Hungary (HHI of 1229), France (HHI of 1410) and Poland (HHI of 1580). On the
other hand, the econometric sample does not include any MS where retail concentration
levels are the highest - Finland (HHI of 3935 in 2012), Latvia (3443 in 2012), Sweden
(3305 in 2012) and Cyprus (2878 in 2012, however down from 6530 in 2004). This said,
the case of retail concentration impacting the Milk sector in Finland is addressed outside
of the econometric analysis through a specific case study.
75
MS ranked in descending order by 2012 HHI figures. Figures for Malta not provided due to insufficient data
188
The MS with the highest levels of retail concentration tend to be smaller in population
size: of the 12 MS with the highest levels, only the Netherlands has a population greater
than 10 million inhabitants in this list and the top 5 account for a combined population
size of less than 20 million inhabitants.
In relation to the evolution of retail concentration over time, the sample includes the MS
with the greatest increase in concentration over the 2004-2012 period, Poland (HHI of
826 in 2004 to 1580 in 2012). On the other hand, the econometrics sample does not
consider any of the MS where retail concentration decreased the most between 2004 and
2012 - Bulgaria (HHI of 2940 in 2004 to 1910 in 2012), Cyprus and Slovenia (HHI of
3180 in 2004 to 2020 in 2012). It does however include 3 MS where retail concentration
decreased over time, as explained in the paragraph below.
Of the other MS in the econometric sample, Belgium, which is only represented in the
short data set (2008-2012), is the MS with the highest retail concentration level in 2012,
in 13th place compared the EU27. Belgium has undergone an annual decrease of -0.6%
since 2004. Portugal is the next MS, in 17th place in the whole of the EU, having seen an
annual increase of 1.5% since 2004. Spain had 6 th lowest retail concentration HHI figure
in the EU in 2012, having increased by 3.1% annually since 2004. Poland still had the
4th lowest concentration level in 2012, despite a 8.4% annual increase since 2004.
Meanwhile, the bottom three MS, France, Italy and Hungary, have observed a slight
decrease in concentration from 2004 to 2012, with compound annual growth rates of -
1.0%, -0.2% and -1.3% respectively.
In conclusion, the scope of MS in the econometric analysis covers predominantly
situations of lower retail concentration; there is, however, a range of trends over time –
whilst the negative trends of France, Italy, Belgium and Hungary are prevalent, Poland
represents the MS with the highest growth in concentration levels, and Spain and
Portugal experienced above average increases. Areas of high concentration are not
covered in the econometric analysis, since these are predominantly in MS with smaller
population sizes. The case study of Milk in Finland provides some insights into the effects
of concentration on choice and innovation in one such MS.
189
Supplier concentration at national level

Regarding supplier concentration at national level, Table 26 below presents the average
supplier concentration HHI figures across the 23 product category sample 76. Whilst
supplier concentration is defined at product category level, the averages across product
categories shown in Table 26 is intended to give some indication of the range of
experience across MS in order to help judge the extent to which the sample of MS used
in the econometric analysis reflects the wider experience of the 14 MS for which data
were gathered.
Table 26: Supplier HHI – brand only by sales market share (national level), averaged
across 23 product category sample (source: EY analysis based on © Euromonitor)
Rank Member State Population (m) 2004 2006 2008 2010 2012 CAGR (04-12)
1 Denmark 5.570 2433 2437 2705 2779 2840 2,0%
2 Netherlands 16.693 2575 2635 2926 2889 2838 1,2%
3 Finland 5.388 2792 2729 2767 2741 2594 -0,9%
4 Portugal 10.557 2122 2166 2289 2339 2426 1,7%
5 Belgium 11.047 2096 2239 2324 2396 2337 1,4%
6 Spain 46.174 1776 1914 1957 2018 2179 2,6%
7 France 65.161 1838 1955 1998 2122 2130 1,9%
8 Czech Republic 10.496 1700 1800 2042 2056 2056 2,4%
9 Hungary 9.971 1963 2106 2035 2055 2016 0,3%
10 United Kingdom 62.271 1716 1707 1714 1794 1766 0,4%
11 Romania 21.384 1750 1758 1721 1720 1747 0,0%
12 Poland 38.534 1439 1528 1648 1724 1743 2,4%
13 Italy 60.723 1406 1461 1499 1519 1590 1,5%
14 Germany 81.797 1202 1226 1268 1383 1359 1,5%
As can be seen in the table above, the econometric analysis sample includes the MS with
the 2nd and 3rd lowest supplier concentration levels on average across the 23 product
categories – Italy, with an HHI of 1590 in 2012, and Poland, with an HHI of 1743 in
2012 – the lowest being Germany, with an HHI of 1359 in 2012. Econometric analysis,
on the other hand, does not cover the three MS with the highest average level of
supplier concentration – Denmark (with an HHI of 2840 in 2012), the Netherlands (2838
in 2012) and Finland (2594 in 2012). This said, the effects of supplier concentration in
two of these MS are addressed through case studies – Cheese in the Netherlands and
Milk in Finland. Four MS (Portugal, Belgium, Spain and France) in the econometric
sample feature amongst the top seven MS of the 14 MS sample in terms of the level of
supplier concentration, with an average HHI of between 2130 and 2426 in 2012.
Furthermore a case study on Tomatoes in Belgium studies the effects of high supplier
concentration on this fresh food category.
With regards to the evolution of supplier concentration over time, the econometric
sample covers two of the three MS with the greatest increase in average concentration
from 2004 to 2012 – Spain (with an HHI of 1776 in 2004 and 2179 in 2012) and Poland
(1439 in 2004 to 1743 in 2012). On the other hand, the sample does not consider the
only MS where supplier concentration decreased on average – in Finland (with HHI of
2792 in 2004 and 2594 in 2012), although the case of Milk in Finland is addressed
76
MS ranked in descending order by 2012 HHI figures.
190
through a case study. Of the other MS in the econometric sample, the annual growth
rates were slightly above average compared to the 14 MS sample as a whole: in
Portugal, Belgium, France and Italy, supplier concentration on average increased by
between 1.4% and 1.9%
Figure 133 provides further information on the representativeness of the sample used for
econometric analysis. This figure shows that the sample also includes situations (couple
Member States / product category) of very high and very low supplier concentration
Figure 133: Distribution of supplier concentration (HHI – brand only by sales market
share at national level) for the 23 product categories in each country in 2012 (source:
EY analysis based on © Euromonitor)
10000
Supplier concentration HHI (for each product category)
9000
8000
7000
6000
5000
10th percentile
4000
Highest
3000
Lowest
2000
90th percentile
1000
Sampled MS
0
Countries (ordered by average supplier concentration across product categories
In conclusion, the scope of MS in the econometric analysis tends to represent the range
of situations of supplier concentration in the 14 MS sample; whilst the scope does not
cover the MS with the most and least concentrated suppliers on average in 2012, it does
include the 4th highest (Portugal) and the 2nd lowest (Italy). Furthermore, supplier
concentration in the 2nd and 3rd placed MS is addressed through case studies. The
remaining MS in scope registered mid-range concentration levels when considered
amongst the 14 MS sample. In terms of evolution in this driver, the econometric scope
notably includes the two MS with the highest increase in average supplier concentration
over time, Spain and Poland.
191
Measure of imbalance HHI at national level

Regarding the measure of imbalance (or the log 77 of the ratio of retail concentration
divided by supplier concentration) across all 23 sample product categories over the past
decade, diverse trends are observed in the 14 MS sample. The variety of situations
(across the 14 MS for which supplier concentration data was gathered and 23 product
categories) is represented in the Table 27 below.
(source: EY analysis based on © Planet Retail and © Euromonitor)
Measure of imbalance 2004 2006 2008 2010 2012
In the 14 MS
In the sampled MS
Note: The measure of imbalance is defined as the natural logarithm of the ratio of
national retail concentration HHI to national supplier concentration (in a given product
category) HHI. A value of zero indicates that retailer HHI and supplier HHI are equal.
Values greater than zero indicate higher retailer than supplier concentration; values less
than zero indicate higher supplier than retail concentration.
attest that they are approximately equal numbers of situations in favour of retailers as
they are situations in favour of suppliers. In our sample, due to the absence of MS where
retailers are strongly concentrated, the number of situations where suppliers are in a
dominant position is slightly higher, although both situations are represented.
Figure 133 provides further information on the representativeness of the sample used for
econometric analysis. This figure shows that the sample also includes situations (couple
Member States / product category) of very high and very low measure of imbalance, in
spite of the fact that MS with high retail concentration were not included in the scope.
77
The log transformation is used so that the metric presents a higher retail
concentration HHI and a higher supplier concentration HHI symmetrically. For example,
in the unlogged metric, if retail concentration moves from being at the same level as
supplier concentration to a level that it is twice as high, the imbalance ratio increases
from 1.0 to 2.0, whereas if supplier concentration doubles then the imbalance ratio falls
from 1.0 to 0.5. In the logged metric, the value increases from 0 to 0.693 or falls from
0 to -0.693 in the two examples.
192
Figure 134: Distribution of measure of imbalance for the 23 product categories in each
country in 2012 (source: EY analysis based on © Planet Retail and © Euromonitor)
3,000
Measure of imbalance (for each product category)
2,500
2,000
1,500
1,000
10th percentile
0,500
Highest
0,000 Lowest
90th percentile
-0,500
-1,000 Sampled MS
Countries (ordered by average MoI across product categories
In conclusion, the scope of MS in the econometric analysis tends to represent a broad

range of situations across the 14 MS sample in terms of the level of imbalance and the
trends over time.
193
Private label share at national level

In relation to private label sales share, descriptive statistics show an overall increase in
private label share averaged across the 23 product category sample in all of the 14 MS
sampled, however with significant differences between MS. This is represented in the
Table 28 below.
Table 28: Private label percentage share by sales (national level), averaged across 23
product category sample (source: EY analysis based on © Planet Retail)
Member State Population (m) 2004 2006 2008 2010 2012 CAGR (04-12)
Germany 81.797 30,48 32,18 33,38 33,54 32,93 1,0%
Spain 46.174 20,55 22,36 24,99 28,73 32,11 5,7%
Portugal 10.557 17,26 19,66 23,56 27,37 30,05 7,2%
United Kingdom 62.271 29,15 29,13 29,52 29,52 29,60 0,2%
Belgium 11.047 27,51 28,61 28,98 29,25 29,51 0,9%
Netherlands 16.693 23,75 24,87 25,37 26,76 27,86 2,0%
France 65.161 23,26 24,05 24,93 25,45 24,82 0,8%
Denmark 5.570 17,68 18,69 19,54 20,06 21,60 2,5%
Hungary 9.971 8,65 11,88 15,89 18,51 19,68 10,8%
Finland 5.388 13,61 15,11 16,01 17,60 19,01 4,3%
Italy 60.723 12,41 13,06 13,61 14,75 15,77 3,0%
Czech Republic 10.496 8,22 9,81 11,71 12,44 13,22 6,1%
Poland 38.534 5,48 6,26 6,80 7,98 11,20 9,3%
Romania 21.384 2,59 3,53 3,51 3,81 4,56 7,3%
As is shown in the table above, the econometric analysis sample includes a broad range
of situations both in terms of level of private label share and evolution trends over time.
Private label share averaged across the 23 product category sample in 2012 was highest
in Germany (32.9%), followed by Spain (32.1%) and Portugal (30%) – these latter two
MS forming part of the econometric analysis. At the other end of the spectrum, private
label share was lowest in 2012 on average in Romania (4.6%) followed by Poland
(11.2%), the latter of which is part of the econometric analysis. The remaining MS
featuring in the econometric analysis are distributed evenly amongst the 14 MS sample
shown in Table 25.
In terms of evolution over time, the strongest growth among the 14 MS sample was
observed in Hungary (10.8% compound annual growth rate), followed by Poland (9.3%),
Romania (7.3%) and Portugal (7.2%). All of these MS with the exception of Romania are
included in the econometric analysis sample. On the other hand, growth in private label
share was weakest in the UK (0.2%), France (0.8%) and Belgium (0.9%), the latter two
MS being represented in the econometric analysis.
In conclusion, the scope of MS in the econometric analysis covers a wide range of
situations in terms of level and evolution of private label share.
7.4. Implications of the sample selection process

The most likely kind of bias introduced by the process by which shops are included in
Opus is that we over-represent shops that face more (local) competition, because
according to Nielsen these are the ones that tend to prompt requests for coverage by
competitors.
194
Figure 135 shows the count of shops in the long data set falling into different bands for
the (banner) C5 concentration ratio (averaged over the entire time period). This
suggests that there is reasonable coverage of shops operating in a quite highly
concentrated environment.
Figure 136 uses the broader HHI measure of concentration: about one sixth of the shops
in the sample operate in an area with a HHI that exceeds 2,500, which represents a
reasonably high degree of concentration.
The actual distribution of shops by degree of local competition is unknown, and so the
extent to which these distributions of the sample depart from the distribution of the
population cannot be assessed, but it is clear that the sample includes cases with a
moderate to high degree of concentration in sufficient numbers for these to influence the
econometric results.
Figure 135: Distribution of shops by C5 concentration measure at banner level (long

data set) (source: analysis based on © Nielsen Trade Dimensions sales area data)
C5 ratio
195
Figure 136: Distribution of shops by HHI concentration measure at banner level (long
data set) (source: analysis based on © Nielsen Trade Dimensions sales area data)
HHI (10000 = monopoly)
196
Main features of choice and innovation indicators in the sample data set
8. Main features of choice and innovation indicators in

the sample data set
Below is a summary of the main features of the sample data set with respect to the main
measures of choice and innovation, and note how these features are addressed or
reflected in the econometric analysis.
8.1. Choice
Figure 137 shows the ranking of types of shop according to the number of products
stocked (averaged across shops and product categories). As expected, the order is:
hypermarkets > supermarkets > hard discounters. The figure also shows that the
increase in choice (on this measure) over time was seen in all three types of shop.
Proportionately the increase over time is largest in discounters, next largest in
hypermarkets and smallest in supermarkets; in absolute terms the increase is largest in
hypermarkets, which is the change most easily seen in the figure.
Although the econometric analysis includes a measure for the size of the shop (in floor
space), it also includes a fixed effect for type of shop. The estimated parameters are
statistically significant, which suggests that the amount of choice offered by the shop is
not just a function of size: it is also a matter of format. The estimated parameters reflect
the ranking by type of shop for this choice indicator. The data suggest that the hard
discounter effect may be declining in absolute size over time (the gap is closing between
discounters and other types of shop): the parameters in equations estimated in separate
cross sections for each time period reflect this.
Figure 137: Average number of EAN codes per shop and per product category, presented
by shop type (long data set) (source: analysis based on © Nielsen Opus. Data are for
first period in each year)
250
EAN codes per shop and product
200
150
Hypermarkets
Supermarkets
100
Hard discounters
50
0
2004 2006 2008 2010 2012
Figure 138 shows the average number of EAN codes (across shops and product
categories) by the Member States in our long data set. The figure shows the data only
for hypermarkets, to filter out any effect of a different mix of shop types in our sample in
different Member States.
Again, the increase in choice on this measure is seen in all the Member States.
The average is considerably higher in France than in the other Member States. The
econometric analysis includes a fixed effect for each Member State to capture this.
However, this effect cannot be interpreted as simply adjusting for the difference
observed in Figure 138 because the econometric analysis also includes national product
197
category turnover, an indicator whose scale varies by Member State (simply because of
the different sizes of the economies) but whose impact on choice is not expected to
reflect that difference in scale: national turnover in France for a given product category
might be ten times what it is in Portugal, but that does not mean that one would expect
there to be ten times as many EAN codes in France. Rather, the role of this indicator is
mainly to discriminate between different product categories in the same Member State.
The consequence is that the estimated Member State fixed effect is adjusting for that
difference in scale as well as the difference in levels of EAN codes shown in Figure 138.
Figure 138: Average number of EAN codes per shop and per product category in
hypermarkets in Member States (long data set) (source: analysis based on © Nielsen
Opus. Data are for first period in each year)
300
250
EAN codes per shop and product
200
Italy
Spain
150
France
Portugal
100 Poland
50
0
2004 2006 2008 2010 2012
Figure 139 shows the average number of EAN codes (across all shops in the long data
set) by product category. The figure shows that choice (on this measure) increased in
almost all product categories (the exception is canned vegetables, where the number of
EAN codes decreased after 2008, and is potentially due to incomplete or unreliable data
for this product category in selected MS). The figure also shows that the number of EAN
codes varies across product categories, reflecting the particular features of each type of
product. The econometric analysis includes a fixed effect for each product category to
capture this. For the same reason as discussed above for Member State fixed effects,
the interpretation of the product category fixed effects is complicated by the presence of
the national product category turnover driver, which varies across product categories:
for some product categories, the fact that the number of EAN codes is relatively high or
low may be completely accounted for by the relative size of the product category
turnover, and so the fixed effect for that product category could be close to zero.
198
Figure 139: Average number of EAN codes per shop by product category (source:
analysis based on © Nielsen Opus. Data are for first period in each year)
350 Yoghurt
Biscuits
Ham/delicatessen
300 Starters/pizzas
Chocolate (Bar + Candies)
Milk
Number of EAN codes per shop
250 Butter/margarine
Dessert
Frozen vegetables
200 Ice cream
Fruit juices (ambient)
Cheese
150 Cereals
Savoury snacks
Mineral water
Coffee
100
Canned vegetables
Edible oil
Tea
50
Baby Food
Soft-drinks
Ready-cooked meals
0
2004 2006 2008 2010 2012
Figure 140 shows that the ranking of product categories on this choice indicator is
similar, but not identical, across the Member States in the long data set (again, the data
are for hypermarkets only, to filter out the effect of variations in the shop-type mix
across Member States in our sample). The inclusion of both product category and
Member State fixed effects in the econometric analysis is intended to allow for these
differences (to the extent that they are not explained by other drivers).
199
Figure 140: Average number of EAN codes per shop in hypermarkets in selected Member
States in 2012, presented by product category (source: analysis based on © Nielsen
Opus)
Average number of EAN codes per shop
600
500
400
300
200
100
0
Italy Spain Portugal Poland France
8.2. Innovation
Figure 141 shows the level and trend in the total number of innovative EAN codes by
type of shop. It shows the same difference in levels that was observed for the choice
indicator among the three types of shop. The trends following the recession differ, in
that hard discounters continued to increase the number of innovative products that they
stocked (but from a low level). Again, fixed effects are included for shop types in the
econometric analysis to reflect this difference in levels.
Figure 141: Average number of new EAN codes per shop and per product category,
presented by shop type (long data set) (source: analysis based on © Nielsen Opus. Data
are for first period in each year)
50
45
Innovations per shop and
40
35
30
product
Hypermarkets
25
Supermarkets
20
Hard discounters
15
10
5
0
2006 2008 2010 2012
Figure 142 shows innovations in the sample shops by the Member States in the long
data set. Again, the data are for hypermarkets only, to filter out any difference in the
mix of shops by Member State in our sample. As was the case with the choice indicator,
France ranks highest, but unlike the case for the choice indicator Italy is distinctly
200
lowest. There is a somewhat different trend across Member States, with the number of
innovative products continuing to increase in 2010 in Poland and Portugal, whereas in
the other three Member States the number remains flat or falls. Again, Member State
fixed effects are included in the econometric analysis to reflect the difference in levels:
the difference in trend is left to be explained by other drivers (for example,
macroeconomic drivers to capture the impact of the recession).
Figure 142: Average number of new EAN codes per shop and per product category in
hypermarkets in selected Member States (long data set) (source: analysis based on ©
Nielsen Opus. Data are for first period in each year)
70
60
Innovations per shop and product
50
Italy
40 Spain
France
30
Portugal
20 Poland
10
0
2006 2008 2010 2012
Figure 143 shows the level and trend in innovations in the shops in the long data set by
product category. Differences in trend across product categories are more pronounced
here than they are for choice. For most product categories the number of innovations
peaked in 2008; for some (savoury snacks, canned vegetables, ready-cooked meals and
milk) the peak came in 2010; for desserts and cereals, the number of innovations
increased through to 2012. Product category fixed effects are included to capture the
difference in levels (the part not explained by differences in national product category
turnover); the analysis includes indicators that vary by product category (national
product category turnover and national supplier concentration) to try to account for the
differences in trend.
201
Figure 143: Average number of new EAN codes per shop by product category (long data
set) (source: analysis based on © Nielsen Opus. Data are for first period in each year)
90 Biscuits
Ham/delicatessen
Yoghurt
80
Savoury snacks
70 Ice cream
Average number of innovations per shop
Baby Food
Cheese
60
Soft-drinks
Dessert
50 Coffee
Canned vegetables
40
Starters/pizzas
Cereals
30 Ready-cooked meals
Tea
Frozen vegetables
20 Milk
Butter/margarine
10 Mineral water
Edible oil
0
2006 2008 2010 2012
Figure 144 shows average innovations per shop in hypermarkets in 2012 by Member
State and product category. There is somewhat more variation in the pattern between
Member States than is the case for choice, suggesting that national factors play a
greater role in influencing innovation behaviour in product categories than they do for
choice.
202
Figure 144: Average number of new EAN codes per shop in hypermarkets in selected
Member States in 2012, presented by product category (long data set) (source: analysis
based on © Nielsen Opus)
203
9. Results of the econometric analysis
9.1. Introduction
This section provides the key results from the econometric analysis. Each driver is
discussed in turn with comments on the evidence linking it to choice and innovation.
Graphical analysis is presented to help to illustrate and interpret the econometric results,
recognising the limitation that the graphs are typically limited to showing the relationship
between choice or innovation and one driver, without controlling for the effects of the
other drivers (as the econometric analysis does).
In the reporting, there is a focus on the results obtained from analysis of the long (time
period) data set, and then comments are made on the extent to which those conclusions
change when the equations are estimated over the short (time period) data set. Full
results are reported in Annex F.
Below is a summary of the conclusions. Subsequent sections examine the estimated
impacts of each key driver in turn.
9.2. Summary of results for drivers

This section brings together the results that are explored in more detail in the
subsequent section so as to provide a summary of the findings. For each dependent
variable (the various measures of choice and innovation), a number of different
equations were estimated reflecting
 alternative measures for some of the drivers
 two time periods and associated samples of shops
 alternative econometric methods
For each driver, a parameter estimate with an associated measure of statistical
significance and of economic importance78 is calculated for each equation estimated. In
this summary findings that generally emerged across the alternative equations that were
estimated have been reported; where relevant, cases where the findings changed
markedly between alternative equations have been noted. Fuller details are provided in
the subsequent section and in Annex F.
The estimated scale of the effect of some of the drivers is sensitive to the choice of
whether the long or short time period samples were used. Both the time period and the
selection of countries changes between these two samples, but because a substantial
number of the shops are present in both samples, the difference in results typically 79
reflects the difference between the two time periods: one includes 6 periods prior to the
recession as well as the 4 recession periods, while the other includes 2 periods just prior
to the recession and the 4 recession periods. The fact that some parameter estimates
change depending on the time period used indicates that the impact of the recession on
choice and innovation is not sufficiently captured by the change in the experience of the
drivers when the recession occurred. The relationship between, say, GDP per capita and
choice appears to be different during a period of steady growth than during a period
when recession takes hold. One way of interpreting this is to conclude that our drivers
are not comprehensive and exclude some influences on the behaviour of retailers and
manufacturers that differed greatly in the pre-recession period compared with during the
recession. Clearly various influences on food manufacturers (raw material price
78
See Section 5.3 for the way in which ‘economic importance’ has been defined and
calculated.
79
This was examined by taking the selection of shops used for the long data set but
restricting the period of estimation to that of the short data set.
204
Results of the econometric analysis
volatility, energy cost price volatility, food safety regulations, consumer attitudes to
environmental sustainability) changed during the 2004-12 period but their influence
could be difficult to capture in equations estimated over a data set where the number of
observations mainly comes from the number of shops and product categories rather than
the number of time periods.
Because the econometric estimation is carried out over the dimensions of shops, product
categories and time, the results reflect estimates of the impact of drivers over all three
of these dimensions taken together and not necessarily any one of them. For example,
a result reported for the impact of the unemployment rate is based on the observed
variation over time and geographical areas: there is no separate estimate for the impact
of changes over time versus variation over space. The exception to this is the case of
the Fixed Effects estimator, where each indicator is transformed by subtracting the mean
of each time series for shop and product categories from each time period’s observation,
so that the impact of differences in levels across space (for example, differences in the
level of GDP per capita between one area and another) is removed. Hence, although the
observations are taken from different areas and product categories, the reported results
for the Fixed Effect estimator reflect the different experiences of each shop and product
category with respect to the changes over time rather than the differences in levels
across space.
Not all drivers vary across all the possible dimensions of shops, product categories and
time. Some economic drivers (the unemployment rate and GDP per capita) vary across
local areas and time, but not across shops within the same local area or across product
categories. Some drivers are available at national level only; of these, some vary across
countries, product categories and time; some vary only across countries and time with
no product category dimension. In those cases where a driver does not vary across a
given dimension, it cannot explain variation in choice or innovation that occurs within
that dimension. For example, the regional unemployment rate driver cannot account for
differences in choice or innovation between shops in the same region; instead, the
parameter estimate reflects differences between the experience of the whole set of
shops in a region (and time period) compared with the sets of shops in other regions
(and time periods).
Because the number of countries and time periods is quite small, we regard the
parameter estimates for drivers that vary only across countries and time as having a less
secure basis than those for drivers that vary also across other dimensions: the small
number of observations leaves open the possibility there could be some other
macroeconomic driver omitted from the analysis that is responsible for the differences in
choice or innovation across countries and time periods. For this reason, in the summary
tables (Table 29 and Table 30) we include a column entitled ‘Reduced dimensions’ and
place a flag in it to identify these drivers.
205
9.2.1. Choice
Although statistically significant effects were sometimes found, the impacts of the drivers
that measure indicators that relate directly to retailers and suppliers were mostly small.
The main drivers were found to be the GDP per capita of the region in which the shop is
located, national turnover in the product category, certain shop characteristics (format,
floorspace) and the presence of a new shop opening in the local area: these all had
positive impacts on choice.
206
Table 29: Summary of econometric results for key drivers: choice
Product supplier
Low Product variety Product size variety Product price variety
Driver variety Comments Business explanations
Dim.
Sign Signif. Import. Sign Signif. Import. Sign Signif. Import. Sign Signif. Import.
Retail concentration
Few observations (macro level driver). Result
reflects tendency for price variety to be Too few observations over countries and time to draw conclusions:
Procurement greater in Italy (low retail concentration) than some other trend may have been driving choice in the same
(national) level
 .. .. .. .. .. .. .. .. ..   
in Spain and Portugal (higher concentration), period. Selection of countries does not include cases with the
and the reduction in product price variety highest level of retail concentration.
during the recession.
Negative effect for both product variety and
Shops facing greater competition respond by offering more choice
Local level .. .. ..   .. .. .. .. .. .. .. product size variety , but only significant for
(but scale of effect is small).
product size variety.
Supplier
concentration at Small but significant positive impact for
.. .. ..   .. .. .. .. .. .. ..
procurement product size variety in long data set.
(national) level
Imbalance between
Few observations (macro level driver) for
retailers and
numerator (retail concentration at national Selection of countries does not include cases with the highest level
suppliers at .. .. .. ?  .. .. .. .. ?  
procurement
 level). Direction of impacts not consistent of retail concentration.
between long and short data periods.
(national) level
Private labels
A small positive impact of private label share of assortment on
choice is probably due to the fact that retailers tend to keep
National level .. .. .. .. .. ..   ..   .. branded products in their assortment beside the private labels. In

other words, they don't withdraw as many branded products as
they introduce more private labels.
A small positive impact of private label share of assortment on
Some indication of a negative impact when choice is probably due to the fact that retailers tend to keep
the share of private labels exceeds a certain branded products in their assortment beside the private labels. In
Local level   ..   ..   ..   .. level (which varies depending on the product other words, they don't withdraw as many branded products as
category). they introduce more private labels. However, at local level the high
PL share seems to limit the price scale offered to customers.
Product category
Strong positive impacts for all choice Product categories with high sales turnover are those where there
turnover (sales) at
           .. indicators except product price variety is a greater commercial potential for each SKU. There is also more
procurement
(negative). economic potential for more suppliers in these categories.
(national) level
To face a new competitor, established retailers will seek to retain
New shop opening
        ..   .. Positive impact for all choice indicators. customer loyalty by including additional products to either match
in the local area
competitors or better satisfy existing customers.
General economic drivers
More unemployment tends to change consumer behaviours who
Negligible positive (unexpected) impacts for will probably look for cheaper products and limit their purchase of
Unemployment   ..   ..   ..    all choice indicators except product price more expensive products. As a result, retailers propose more
variety (negative). cheaper products but tend to limit the price scale of their
assortments.
More prosperous areas with higher GDP per capita may tend to
Strong positive impacts for all choice encourage retailers to extend product choice and supplier choice in
GDP per capita          ?  .. indicators except product price variety order to increase the average shopping basket of their customers.
(smaller; negative in short data set) One can also imagine that more expensive products can be
proposed to customers, enlarging the product price variety.
(Population density is a more relevant driver than population size
Population .. .. .. .. .. .. .. .. .. ..   .. Negligible impacts on most choice indicators.
of the region)
It may be that, in densely populated areas, product rotations are
higher than in less densely populated areas. In that case retailers
Population may seek to limit the risk of out of stock products by limiting the
           .. Negative impacts on most choice indicators.
density number of different products on shelves. May also reflect impact
of different competing shop mix (fewer very large shops in cities)
on selection offered in each shop.
As expected (hypermarkets > supermarkets >
not not not not As expected, hypermarkets provide more choice than
Shop type       ?  discounters) except for product price variety in
app. app. app. app. supermarkets, which provide more choice than hard discounters.
long data set.
Larger shops have more shelf space, which enables retailers to
Shop floor space            .. display more different products from a larger variety of suppliers
207
For statistically significant drivers, the ‘Economic importance’ (Import.) column shows the scale of impact of the driver on the dependent variable when the driver is increased by one standard deviation above its mean value (both based on the sample used
for econometric estimation). The symbols used are:
Where a driver is not statistically significant or economically important according to these thresholds, this is denoted by the symbol ‘ ..’
208
9.2.2. Innovation
The economic importance of the drivers was generally larger for innovation than for
choice, although results were not consistent across different innovation measures. In
particular, a different result was often found for the number of new packaging
innovations compared with the other measures of innovation. Results also varied
substantially between the long and short data sets, suggesting that behaviour changed
during the recession in a way that was not fully captured by what happened to the
drivers during that period. Among the indicators that relate directly to retailers and
suppliers, greater concentration among retailers at a local level was associated with less
innovation in the case of new packaging innovations. The econometric results showed a
positive impact on innovation of greater concentration among retailers at the national
level for some innovation indicators (but a negative impact on new packaging), but only
a small number of observations are available for this indicator (it varies only over MS
and years) and so it cannot be regarded as a definitive finding. A negative impact on
some innovation measures was found for greater (national) concentration among
suppliers: there is stronger support for this finding because the indicator varies across
product categories as well as MS and years, providing a much larger number of
observations. When the relative strength of retailer and supplier concentration was
included in the single ‘measure of imbalance’ indicator, a similar result was found
(greater supplier concentration relative to retailer concentration had a negative impact),
but it should be remembered that the selection of countries covered does not include
those with the highest level of national retail concentration. There was less evidence
than was the case with choice that the presence of a new shop opening in the local area
was associated with a positive impact on the offer of existing shops (more innovation).
The impact of the economic drivers included some effects that were unexpected (in the
direction of impact) and these estimates varied substantially between the two data sets.
209
Table 30: Summary of econometric results for key drivers: innovation
New range
Low Opus innovations New products New packaging New formulations
Driver extensions Comments Business explanations
Dim.
Sign Signif. Import. Sign Signif. Import. Sign Signif. Import. Sign Signif. Import. Sign Signif. Import.
Retail concentration
Few observations (macro level driver). Result reflects Too few observations over countries and time to draw conclusions:
Procurement tendency for the number of innovations and the level of some other trend may have been driving innovation in the same
(national) level
          ?     
national retail concentration to rise in some countries period. Selection of countries does not include cases with the
(until recession). highest level of retail concentration.
Some evidence was found of a negative relationship between local
Negative impact on most innovation measures, but not retail concentration and innovation. The main observable impact is
Local level

.. .. .. .. .. ..    .. .. .. .. .. .. usually statistically significant. on new packaging. When they face less competition, retailers seem
to be less prone to introduce innovations on their shelves.
Supplier
concentration at Negative impact on most measures. Positive impact on Suppliers face greater pressure to innovate when competition is

   .. .. .. ?        
procurement new packaging long data set. stronger.
(national) level
Imbalance
Few observations (macro level driver) for numerator
between retailers Consistent with the impacts of retailer and supplier concentration.
(retail concentration at national level). Positive impacts
and suppliers at        ?        
for some measures. Negative impact on new packaging
Selection of countries does not include cases with the highest level
procurement of retail concentration.
in long data set.
(national) level
Private labels
A significant (and sizeable) impact was only found for
National level

?   .. .. .. .. .. .. .. .. .. .. .. .. the Opus innovations measure, but its sign varied
between the long and short data set.
The impact of private labels on innovation is not clear. The share of
Some indication of a negative impact when the share of
private labels does not seem to impact negatively the number of
private labels exceeds a certain level (which varies
Local level   .. .. .. .. .. .. .. .. .. ..   .. innovations, especially new products. However, it seems to have a
 depending on the product category).
small impact on the number of range extensions, which could be less
developed when the share of PL is high.
Product category
turnover (sales) For some indicators, there is a negative impact in the Product categories with high sales turnover offer a greater
.. .. .. ?      ?     
at procurement short data set. commercial potential for investment in innovation.
(national) level
New shop Less evidence that existing retailers respond to new competition by
Only significant (and positive) in random effects for
opening in the .. .. ..    .. .. .. .. .. .. .. .. .. increasing the assortment of new products than by increasing the
some indicators.
local area choice available.
General economic drivers
A higher unemployment rate is generally associated with a smaller
number of innovations, reflecting the underlying macro-economic
situation. Suppliers may be less likely to develop innovations during
difficult economic times, and retailers may also be more hesitant in
Unemployment

         .. .. ..   
offering new innovative products at those times or in places where
the economy is weak. The different result for new packaging
suggests that during the economic crisis there was a shift towards
that form of innovation and away from other forms.
Innovation is probably encouraged by an optimistic attitude from the
Retailer
stakeholders. Therefore there is a positive trend for new product
business     ?            Few observations (macro level driver)
innovation in periods when stakeholders business expectations are
expectations
positive.
Population .. .. .. .. .. .. .. .. ..    .. .. .. Only a significant impact in one case.
In high density areas, we observe less new packagings and new
Population formulations, which may be explained by the need for retailers to
.. .. .. .. .. ..       .. .. ..
density prevent out of stock situations and limit the number of sizes of
products available.
The larger assortment of products available in larger format shops is
As expected (hypermarkets > supermarkets >
not not not not not also reflected in a larger selection of innovative products. The
Shop type 
          discounters). Negative impact of hard discounters is
app. app. app. app. app. smaller range offered by discounters seems to be oriented towards
much larger than for choice.
non-innovative products.
Large significant effects found in random effects
estimator (which compares shops across space), but
As expected, larger shops, for a given format, provide a greater
Shop floor space                typically not in fixed effects estimator (which only
 number of innovative products
detects cases where a shop changes size but not format
over time).
210

For statistically significant drivers, the ‘Economic importance’ (Import.) column shows the scale of impact of the driver on the dependent variable when the driver is increased by one standard deviation above its mean value (both based on the sample used
for econometric estimation). The symbols used are:
Where a driver is not statistically significant or economically important according to these thresholds, this is denoted by the symbol ‘.. ‘.‘
211
9.3. Retail concentration
9.3.1. Retail concentration at the procurement (national) level

Choice
Statistical significance: No, except product price variety
Direction of impact: Negative for product price variety
Economic importance: Large for product price variety
The impact of two alternative measures of retail concentration at the national level was
examined: concentration among retailers in modern retail formats and concentration in
the edible groceries market (both measured using the Herfindahl-Hirschman Index).
Little indication was found of an impact of national retail concentration on choice
(product variety) when the measure used was concentration among retailers in modern
retail formats. Results were generally not statistically significant in fixed effects in the
long period80, 81. For product variety a positive, statistically significant impact was found
in the short period82. For product price variety a negative, statistically significant impact
was found in the short period in both fixed and random effects. 83
When the measure used was concentration among retailers in the edible groceries
market, there was evidence of small, positive, statistically significant impacts on product
variety and product price variety in the long period (in both random and fixed effects) in
the long data period84, and these became larger in the short data period 85. A negative
impact on product price variety was found in the long and short data periods 86.
Both these drivers vary only across Member States and year (not across product
categories or across shops in any given Member State), and so the number of distinct
observations (the number of Member States multiplied by the number of years) is much
smaller than for other drivers. Consequently there is not a strong basis for asserting
that any observed relationship reflects a causal mechanism: most of the variation in
choice in the data set is between shops at local level and between product categories,
across which (in any given Member State) the national retail concentration measure does
not vary. The drivers pick up the association between rising choice and, on some
measures and, in some Member States, rising retail concentration over time and by the
comparison across Member States (having accounted for other indicators that vary
across the same dimensions).
80
This discussion of results draws on an evidence base of over 300 separately-estimated
econometric equations, reflecting differences in the selection of alternative measures of
choice and innovation, drivers, time periods and methods. This set of results is provided
in an accompanying file in which the equations are numbered sequentially [1], [2],… for
ease of reference. Footnotes associated with the findings reported here refer to
particular numbered equations in that file.
81
Equations [6], [42], [78].
82
Equation [24].
83
Equations [113]-[114].
84
Equations [1]-[4], [11-14], [37]-[40] and [47]-[50].
85
Equations [19]-[22], [29]-[32].
86
Equations [109]-[112], [119]-[122], 127]-[130] and [137]-[140].
212
Figure 145 shows the different experiences of the different Member States using the
measure of retail concentration in modern retail formats. The five data points for each
Member State represent the five selected years in the sample: in those Member States
where national retail concentration has been increasing over the past decade, the data
points are ordered by time running from left to right; in those where concentration has
been falling the direction of change over time is from right to left. Each data point
shows, measured on the vertical axis, the average number of EAN codes across shops
and products. The trend towards greater choice over time is reflected in the increase in
the average number of EAN codes in each Member State (an upward movement in the
chart). Italy and France saw a small reduction in national retail concentration over the
period; Spain and Portugal saw quite rapid increases; Poland saw a rapid increase from a
low starting point.
Figure 145: Choice in variety of EAN codes in the sampled shops versus national retail
concentration (source: analysis based on © Nielsen Opus and © Planet Retail. Data are
for first period in each year 2004, 2006, 2008, 2010 and 2012)
Note: Note:  denotes 2004 and  denotes 2012.
The sample does not include Member States that have high concentration levels (the
maximum for the HHI measure shown in Figure 145 is 10,000). The effect that an
increase in concentration has on choice for a Member State where retail concentration is
at lower levels may not be comparable to the effect on choice in a Member State where
retailers are highly concentrated.
213
Innovation
Statistical significance: 1%
Direction of impact: Positive except for new packaging (negative) and new
formulations (ambiguous)
Economic importance: Large (for modern retail measure)
A positive impact of national retail concentration on some measures of innovation was

found, but the number of observations is small four years in each Member State in the
long data set, and three years in the short data set). Again it should be emphasised that
there is not a strong basis for asserting that any observed relationship reflects a causal
mechanism: most of the variation in innovation in the data set is between shops at local
level and between product categories, across which (in any given Member State) the
national retail concentration measure does not vary.
A strong positive impact of national retail concentration in modern retail formats on Opus
innovations was found in the long and short data sets87 and on new products, new
formulations and new range extensions in the long data set88. However, a strong
negative impact on new packaging was found in both long and short data sets 89.
The impact of national retail concentration in the edible groceries market was more
mixed, although broadly consistent with the results for modern retail formats. For the
Opus innovations measure, a positive impact of national retail concentration in the edible
groceries market was found in the long (in Fixed Effects) and especially the short data
sets90. For new products, the estimated impacts were often not statistically significant
for new products in the long data set91 but were positive and statistically significant (in
Fixed Effects) in the short data set92. For new packaging, the impact of national retail
concentration was generally negative (in Fixed Effects) in the long data set 93, and
insignificant in the short data set94. For new formulations, the impacts were generally
positive in the long data set 95 and negative in the short data set 96. For new range
extensions, the impacts were generally insignificant in the long data set 97 but positive in
the short data set98.
Figure 146 shows the variation in experience for Opus innovations in the different
Member States which suggests why a positive impact was found. In three of the Member
States the increase in concentration is generally associated with more innovations over
time except in the last period when innovations fell; in France and Italy the fall in
87
Equations [149]-[150], [169]-[70]
88
Equations [189]-[190], [269]-[270], [309]-[310].
89
Equations [229]-[230], [249]-[250].
90
Equations [146], [156], [158], [165]-[168], [175]-[178], [183]-[184].
91
Equations [185]-[188], [195]-[198], [203]-[204].
92
Equations [206], [216], [218], [224].
93
Equations [226], [230], [235], [244].
94
Equations [245]-[248], [255]-[258], [244].
95
Equations [265]-[270], [275]-[278], 283]-[284].
96
Equations [285]-[288], [290], [295]-[298].
97
Equations [305]-[308], [315]-[318], [323]-[324].
98
Equations [325]-[328], [335]-[338], [343]-[344].
214
innovations in the last period is associated with a reduction in concentration. Although

we have sought to control for the general macroeconomic environment, the small
number of observations available for this national driver means that we cannot be sure
that the positive association generally evident in these charts is not simply the
coincidence of two trends over time: increasing innovations (up until the recession) and
increasing concentration in some Member States.
Figure 146: New EAN codes (innovation) versus national retail concentration (source:
analysis based on © Nielsen Opus and © Planet Retail. Data are for first period in each
year 2006, 2008, 2010 and 2012)
Note:  denotes 2006 and  denotes 2012.
9.3.2. Retail concentration at the local level
Choice
Statistical significance: No (except for product size variety)
Direction of impact: Negative
Economic importance: Small
The impact of two alternative measures of retail concentration faced by each shop at the
local level was examined: concentration by banner and concentration by group (both
measured using the Herfindahl-Hirschman Index applied to shares of floorspace of the
shops located sufficiently close to be competitors to any given shop).
No evidence was found in the long data set of a large statistically significant impact of
greater local retail concentration on any choice indicator. The estimated parameter was
215
generally negative (so an increase in local concentration would be associated with a

reduction in choice), but it was so small as to be economically irrelevant and usually not
statistically significant from zero.
For product variety, the impacts were negative but not statistically significant 99. For
product size variety the impacts were negative, sometimes statistically significant, but
small100. For product supplier variety and product price variety the impacts were not
statistically significant101.
Figure 147 gives an illustration of the lack of relationship. It shows that in each of the
two years, there was no indication that a greater degree of local retail concentration was
associated with either more or less choice, for the sample as a whole or for any of the
three shop types. Between 2004 and 2012 the level of local concentration for the shops
in the sample fell somewhat (there are more shops with higher HHI values in 2004 than
in 2012), and the number of EAN codes rose (across all types of shop), but the
econometric analysis found this negative relationship between local concentration and
choice to be very small in magnitude.
Figure 147: Choice in variety of EAN codes versus local retail concentration by shop type
in 2004 and 2012 (source: Analysis based on © Nielsen Opus and © Nielsen Trade
Dimensions. Data are for first period in each year and cover Italy, Spain, France,
Portugal and Poland.)
Innovation
Statistical significance: No (except for new packaging)
Economic importance: Large for new packaging
Some evidence was found of a negative relationship between local retail concentration
and innovation. For Opus innovations a negative impact of moderate magnitude was
found in the long and short data sets for the Fixed Effects estimator, but the estimate is
99
Equations [7]-[10], [25]-[28].
100
Equations [43]-[46], [61]-[64].
101
Equations [79]-[82], [97]-[100], [115]-[118], [133]-[136]..
216
not statistically significant at the 5% level when standard errors are estimated using the
more cautious method that clusters on CSAs 102. The absence of a strong relationship is
evident in the simple comparison in two years in Figure 148 between the average
number of Opus innovations per shop and product category and the local retail
concentration faced by each shop.
Figure 148: Opus innovations versus local retail concentration by shop type in 2004 and
2012 (source: Analysis based on © Nielsen Opus and © Nielsen Trade Dimensions. Data
are for first period in each year and cover Italy, Spain, France, and Portugal.)
Figure 149 plots innovations per shop and product type for all the years. The left-hand
chart includes the four Member States for which data are available for both indicators
from 2006. In this chart, which combines all the years together, some indication can be
seen of the tendency for the highest number of innovations to be found in locations
where concentration is low, and for more cases where the number of innovations is low
to be found in locations where concentration is higher. However, the high and low
innovation cases seen in Figure 149 are in different Member States (France and Italy, as
the right-hand chart shows): in France alone (for example), the negative relationship is
not evident.
102
Equations [152], [154], [174].
217
Figure 149: New EAN codes (innovation) versus local retail concentration, all shops and
years (source: analysis based on © Nielsen Opus and © Nielsen Trade Dimensions. Data
are for first period in each year of 2006, 2008, 2010 and 2012, and, in the left-hand
chart, cover Italy, Spain, France and Portugal)
Similar results were obtained for new products103 as for Opus innovations, except that a
strong and statistically significant negative effect was found in the short data set in Fixed
Effects104. For new packaging, a strong statistically significant negative impact was found
for the Fixed Effects estimator105. For new formulations and new range extensions the
results were generally negative but statistically insignificant using the cautious estimate
of standard errors106. No statistically significant positive impact was found for any
innovation measure.
9.4. Supplier concentration
Choice
Statistical significance: No (except product size variety)
Direction of impact: Positive for product size variety
There is a focus on supplier concentration at the national level because for most
products this is the relevant level for procurement. Data were not available to measure
supplier concentration at local level adequately107.
103
Equations [[191]-[194], [211]-[214].
104
Equations [212], [214].
105
Equations [232], [234], [252], [254].
106
Equations 271]-[274], [291]-[294], [311]-[314], [331]-[334]/
107
The Opus data set allows us to measure the number of different suppliers of branded
products, but this reflects what retailers have chosen to stock rather than the choices of
suppliers available to retailers in the market. The econometric analysis estimated
218
The impact of two alternative measures of national supplier concentration in product

categories was examined: the share of suppliers in the ‘branded market’ (excluding
private labels altogether), and the share of suppliers in the ‘full market’ (treating private
labels as a single supplier) using the Herfindahl-Hirschman Index. We focus on the
results using the ‘branded market’ measure; the ‘full market’ measure is of interest in
markets where private labels are particularly important.
No general evidence was found of an economically important impact of national supplier
concentration on choice. For product variety, the estimated impact was small and
generally negative in the long and short data sets, but not significant for the brand only
measure108; results were negative and small but statistically significant for the full
market measure. For product size variety the impact was small, positive and statistically
significant in the long data set and insignificant in the short data set 109 (on both
measures). For product supplier variety there was rarely a significant impact110 (on both
measures). For product price variety the impact was mostly small and the sign and
significance varied between the long and short data sets 111 (on both measures).
Figure 150 gives an indication as to why no strong effect was found in the case of
choice. It shows the average number of EAN codes per shop and the level of national
supplier concentration in one period in 2008 for the five MS in the long data set,
distinguishing the product categories. For each product category there are therefore five
data points, one for each MS. Since the variation in national supplier concentration is
much greater in the sample over product categories and MS than over time, the
distribution shown in the chart is not greatly different for any given time period. It can
be seen that there is little indication that the product categories with greater supplier
concentration are those with either more or less choice.
equations including this indicator and generally found negative impacts, but this simply
reflects the expected outcome that few EAN codes (and fewer innovative EAN codes)
108
Equations [1]-[12]], [19]-[30].
109
Equations [[37]-[48], [55]-[66].
110
Equations [73]-[84], [91]-[102].
111
Equations [109]-[120], [127]-[138].
219
Figure 150: Choice in variety of EAN codes versus national supplier concentration by
product category, 2008 (source: analysis based on © Nielsen Opus and © Euromonitor
Portugal and Poland)
Innovation
Statistical significance: 1% level for several innovation indicators
Direction of impact: Mostly negative
Economic importance: Moderate to large
Some evidence was found for a negative impact of national supplier concentration on
innovation. For the Opus innovations measure, a negative statistically significant impact
of national supplier concentration was found in the long and short data sets using the
branded market measure112; using the full market measure there were some negative,
statistically significant results in the long data set but these ceased to be statistically
significant in the short data set113. For new products, a negative impact was found for
the branded market measure in long and short data sets, but it was not statistically
significant114.Using the full market measure, a negative, statistically significant impact
was found in the long data set 115; in the short data set the impact is generally positive
112
Equations [155]-[156], [176].
113
Equations [145]-[154], [165]-[174].
114
Equations [196], [216].
115
220
and sometimes statistically significant for the Fixed Effects estimator116. For new
packaging, a positive impact of national supplier concentration was found in the long
data set117, but it is mostly negative and sometimes statistically significant in Fixed
Effects in the short data set118 (both measures). For new formulations, and new range
extensions the impact is negative and sometimes statistically significant in Fixed Effects
in the long data set119 but not generally statistically significant in Fixed Effects in the
short data set120 (both measures).
The equivalent chart to Figure 150 for Opus innovations is shown in Figure 151. No
obvious relationship is evident in this simple comparison.
Figure 151: Opus innovations versus national supplier concentration by product

category, 2008 (source: analysis based on © Nielsen Opus and © Euromonitor
Portugal.)
116
Equations [206], [208], [210], [212], [214].
117
118
Equations [246], [248], [250], [252], [254], [256].
119
Equations [266], [268], [270], [272], [274], [276], 306], [308], [310], [312], [314],
[316].
120
Equations [286], [288], [290], [292], [294], [296], [326], [328], [330], [332],
[334], [336]..
221
9.5. Measure of imbalance between retailers and suppliers at national

level
Choice
Statistical significance: Various
Direction of impact: Ambiguous for statistically significant cases
Economic importance: Moderate for product price variety
Various combinations of national retailer and national supplier concentration are possible
to construct a measure of imbalance (a ratio of retailer to supplier concentration). We
focus on the ratio of national retail concentration in modern retail formats to national
supplier concentration in the branded market; where relevant we also comment on
results which use national supplier concentration in the full market in the denominator of
the ratio.
Although our sample does not include the Member States with the highest retailer
concentration, the sample still covers a wide range of situations regarding the measure
of imbalance. The two choice indicators for which statistically significant estimates of
impact were found for this measure of imbalance are, unsurprisingly, the ones for which
a similar finding was found for either retail concentration or supplier concentration
(product size variety and product price variety).
There is evidence of a positive relationship between product variety and imbalance in
both the long and short data sets, but its value was small (and it was not statistically
significant for the branded market imbalance measure in the long data set) 121.
For product size variety the evidence is mixed: a small negative statistically significant
relationship in the long data set122 and a small positive statistically significant
relationship in the short data set 123 (both measures). For product supplier variety, no
statistically significant impact was found124. For product price variety a negative
relationship was found in the long data set125 (both measures).
Innovation
Statistical significance: 1% level
Direction of impact: Positive for most indicators of innovation (ambiguous for new
packaging)
Economic importance: Large
The results for the impact of this measure of imbalance on innovation are largely the
mirror image of those for supplier concentration (the denominator in the ratio).
121
Equations [15]-[18], [33]-[36].
122
123
124
Equations [87]-[90], [105]-[108]
125
222
Evidence was found of a statistically significant positive relationship between Opus

innovations and the ratio of retailer to supplier concentration in the long and short data
sets (on both measures of imbalance)126. Similar results were found for new range
extensions127. Similar results were found for new products in the long data set 128, but
the results were not statistically significant (in Fixed Effects) in the short data set 129.
Similar results were found for new formulations in the long data set 130, but the results
were negative (in Fixed Effects) in the short data set 131. A negative relationship was
found in the long data set for new packaging132. This is consistent with the negative
result found for retail concentration and the positive result found for supplier
concentration in the long data set.
In this case it is not easy to see any relationship in the simple graphical comparison of
the two indicators, as Figure 152 illustrates. In those cases where a positive or negative
relationship is found in the econometrics, this only emerges after controlling for other
drivers.
Figure 152: New EAN codes (innovations) versus the ratio of retailer to supplier
concentration (source: analysis based on © Nielsen Opus, © Planet Retail and ©
126
Equations [159]-[162], [179]-[182].
127
Equations [319]-[322], [340], [342].
128
129
Equations [220], [222].
130
131
Equations [300], [302].
132
223
9.6. Private labels
Choice
Direction of impact: Positive (except for impact of local share of private labels on
product price variety, which is negative)
The impact of two alternative measures of the importance of private labels in the market
was examined: the share of private label EAN codes in each local shop’s EAN codes (by
product category), and the share of private labels in the national sales of each product
category. The first of these indicators varies by shop and product category, whereas the
second varies only by product category (and Member State).
Evidence was found of small positive impacts of private labels on most measures of
choice except product price variety where the impact depended on the choice of measure
of private labels.
For product variety, when the indicator used was the share of private label EAN codes in
each local shop’s EAN codes (by product category), a statistically significant positive
impact was found in the long133 and short134 sample data sets, but its value was small.
No statistically significant impact of the national share of private labels was found (in
Fixed Effects estimation) on product variety 135. For product size variety, small positive
statistically significant effects of local private label share were found in the long data
set136, but no statistically significant effects were found in the short data set 137; the
impact of national private label share was insignificant in Fixed Effects 138 (a negative,
statistically significant impact was found in Random Effects 139). For product supplier
variety, small positive statistically significant effects of local private label share were
found in the long and short data sets 140; small positive statistically significant effects
were also found for national private label share in the long data set 141 but the effects
were insignificant in the short data set142. For product price variety, small negative
statistically significant effects of local private label share were found in the long and
short data sets143; small positive statistically significant effects were found for national
133
Equations [1]-[2], [5]-[18].
134
Equations [19]-[20], [23]-[36].
135
Equations [4], [22].
136
Equations [37]-[38], [41]-[54].
137
Equations [55]-[56], [59]-[72].
138
139
140
Equations [[73]-[74], [77]-[92], [95]-[108].
141
142
143
Equations [[109]-[110], [113]-[128], [131]-[144].
224
private label share in the long data set 144 but the effects were mixed/insignificant in the
short data set145.
Figure 153 compares product variety and the share of private labels in the long data set,
distinguishing the three shop types. Each point in the chart is the number of EAN codes
for a given product category and shop. The vast majority of observations (97% for
hypermarkets and 86% for supermarkets) are located in the region where the share of
private label EAN codes in the total number of EAN codes (for a given product category)
is 50% or less, and in that region there is no clear relationship between product variety
and the private label share in this simple comparison. But for the small number of cases
among hypermarkets and supermarkets where the share is high, there is a clear
indication that choice is much reduced. Among discounters (where just over half of the
observations had greater than a 50% private label share), there is no evidence that the
(relatively low) level of choice is reduced as the share of private labels increases. When
the data are examined at the level of separate product categories, the point after which
an increase in share is associated with less choice varies, depending on the product
category.
Figure 153: Choice and the private label share by shop type
An analysis that shows the results for each product category is presented in Annex
11.6.3.
Innovation
Statistical significance: 1% level for a few cases
Direction of impact: No consistent direction found
The results for innovation were mixed, depending on the measure and time period
chosen and no clear impact was found.
144
145
225
For Opus innovations, small positive statistically significant effects of local private labels
share were found in the long data set 146, but these became insignificant (in Fixed Effects)
in the short data set147. The national private labels share was (in Fixed Effects) strongly
positive and significant in the long data set 148 but strongly negative and significant in the
short data set149. For new products and new packaging, local private labels share was
not significant in Fixed Effects in the long data set 150; for new products the impact in the
short data set151 was generally small, positive and borderline statistically significant,
while for new packaging the impact was not generally statistically significant in the short
data set152. The impact of national private labels share was mostly not significant for
these two indicators153, the exception being new packaging in the short data set 154 where
it was borderline significant and negative. For new formulations, the impact of local
private labels share was generally negative but not significant in Fixed Effects in the long
ans short data sets155; the impact of national private label share was negative but not
significant in Fixed Effects in both long and short data sets 156. For new range extensions
the impact of local private labels share was small, negative and statistically significant in
Fixed Effects in the long data set157 but generally not significant in the short data set 158.
Figure 154 shows a similar finding for innovation as for choice: in hypermarkets and
supermarkets, in the minority of cases (a particular product category in a particular
shop) where the private label share is high, there is a fall-off in the number of
innovations. Again, when the data are examined at the level of separate product
categories, the point after which an increase in share is associated with less innovation
varies, depending on the product category.
146
Equations [145]-[146], [149]-[164].
147
Equations [166], [170], [172], [174], [176], [178], [180], [182].
148
Equation [148].
149
Equation [168].
150
Equations [186], [190], [192], [194], [196], [198], [200], [202], [204]; 226], [230],
[232], [234], [236], [238], [240], [242], [244].
151
Equations [205]-[206], [209]-[224];
152
Equations [246], [250], [252], [254], [256], [258], [260], [262], [264].
153
Equations [188], [208], [228].
154
Equation [248].
155
Equations [266], [270], [272], [274], [276], [278], [280], [282], [284], [286],
[290], [292], [294], [296], [298], [300], [302], [304].
156
Equations [308], [328].
157
Equations [306], [310], [312], [314], [316], [3318], [320], [322], [324].
158
Equations [325]-[326], [329]-[344].
226
Figure 154: Innovation and the private label share by shop type
9.7. Product category turnover
Choice
Direction of impact: Positive (except for product price variety where it was
negative)
Economic importance: Large (except for product price variety)
The national level of turnover (sales) in each product category is a statistically significant
and economically important driver of choice. Positive impacts were found for all
measures of choice159 except product price variety160 where it was generally negative but
small. Product categories with high sales turnover are also those where there is a
greater commercial potential, and therefore where suppliers focus on product
development, ultimately accounting for a wide variety of products on offer.
Most of the variation in this driver is between product categories; changes over time are
modest by comparison. The positive relationship between turnover and choice therefore
reflects the fact that products with a larger turnover tend to have a larger number of
EAN codes161.
159
160
161
The econometric specifications include a ‘product-specific’ intercept to control for
differences in the level of choice or innovation associated with each product ‘on average’
across Member States and time. Product-specific and Member-State-specific drivers,
such as product category turnover, account for differences apart from these ‘average’
effects.
227
Figure 155 shows this relationship for one period 162 of the data in a selection of the
Member States in the data set. Each point in the charts represents a single product
category: the horizontal axis shows the national turnover expressed as € per inhabitant,
while the vertical axis shows the average number of EAN codes in the product category
per shop in the data set in that Member State.
Figure 155: Choice in variety of EANs versus national product category sales turnover in
2010 period 1 in four Member States (source: analysis based on © Nielsen Opus and ©
Innovation
Direction of impact: Generally positive
Economic importance: Mostly large
The national level of turnover (sales) in each product category is generally a statistically
significant and economically important driver of innovation, and the reasoning set out
above with respect to choice also applies here for innovation. However, there was more
variation across measures of innovation and across the two data sets than was the case
162
The period chosen is arbitrary: the aim is to show the pattern across product
categories for a given period.
228
for choice. For Opus innovations, the impact was either negative or not statistically
significant (in Fixed Effects)163. Clear positive impacts were mostly found for new
packaging164 and for new range extensions165, but the impacts varied between the long
and short data sets for new products166 and new formulations167
Product categories with high sales turnover may be those where suppliers are likely to
develop innovations. The relationship may be negative in the short period due to the
effect of the crisis, whereby suppliers may invest less in research and development
despite product categorises continuing to grow in size.
Figure 156 shows the relation between turnover and innovation for one period. The
same positive relationship that was seen for choice is observable, although it is less
pronounced.
Figure 156: New EAN codes (innovations) versus national product category sales
turnover in 2010 period 1 in four Member States (source: analysis based on © Nielsen
Opus and © Euromonitor International)
163
Equations [146], [148], [150], [152], [154], [156], [158], [160], [162], [164],
[166], [168], [170], [172], [174], [176], [178], [180], [182], [184].
164
165
166
167
229
9.8. General economic drivers: unemployment
Choice
Statistical significance: Various
Direction of impact: Positive; negative for product price variety
A positive, significant impact was found of the rate of unemployment on most choice
indicators168 except for product price variety, but its scale was small and only sometimes
statistically significant at the 1% level. In the case of product price variety169 the impact
was negative (in both the long and short data sets).
Innovation
Statistical significance: 1% level (in long data set)
Direction of impact: Negative (in long data set)
There was evidence of a statistically significant and reasonably large negative impact on
Opus innovations170 in the long data and (for the Fixed Effects estimator) short data
sets. Similar results were found for new products 171 and new range extensions172. For
new packaging the results were large, positive and statistically significant in the long
data set173 but generally not significant in the short data set 174.For new formulations the
results were generally not significant in the long data set 175 and large, negative and
statistically significant in the short data set176. A higher unemployment rate is generally
associated with a decrease in innovation, due to the underlying macro-economic
situation. Suppliers may be less likely to develop innovations during difficult economic
times, and retailers may also be more hesitant in offering new innovative products.
Figure 157 shows data for the first period in each of the years for which data is available
for the innovation indicator in the long data set. Each point in the graph represents the
average number of innovations for all shops in a location with the same unemployment
rate (either because they are located in the same NUTS 3 region – the level at which the
unemployment rate is recorded – or because the unemployment rate happens to
168
Equations [1]-[108]
169
Equations [[109]-[144].
170
171
172
173
174
175
176
230
coincide with the level in another region). A broadly negative relationship can be
observed when the comparison is made across Member States, as in the chart, and also
for some individual Member States.
Figure 157: New EAN codes (innovations) versus unemployment rate (source: analysis
based on © Nielsen Opus and Eurostat. Innovation data are for first period in each year
2004, 2006, 2008, 2010, 2012)
9.9. General economic drivers: GDP per capita/Retail business

expectations
The GDP per capita driver is intended to capture differences in the level of prosperity
across areas and time. Initial results showed the expected positive relationship between
regional GDP per capita and indicators of choice, but a negative relationship with
indicators of innovation in some specifications. A possible explanation was that variation
in the level of GDP per capita did not capture the impact of the recession adequately for
innovation, which appears to be more sensitive than the indicators of choice to the state
of the macroeconomic environment. An attempt was made to capture this sensitivity by
using the growth in GDP per capita as a driver, but the result was that a negative impact
was found for most innovation indicators in the long data set and a positive impact in the
short data set, apparently because in some cases GDP per capita growth was slowing
down (falling) during the period before the crisis when innovation was still increasing.
An alternative macroeconomic indicator was tried, namely the state of national retail
business expectations with respect to the next three months (chosen as a proxy for
general expectations with regard to household spending rather than as a measure of
retailer attitudes alone) and this was found to have a strong positive relationship with
several of the indicators of innovation. It should be noted that this driver is subject to
the same potential weakness as the national measures of retail concentration, namely
that the number of distinct observations (the number of Member States multiplied by the
number of time periods) is much smaller than for other drivers. But its role in the
analysis is simply to try to control for the broad influence of the state of the
macroeconomy on innovation when examining the impact of other drivers.
231
The final specification for explaining innovation therefore uses national retail business
expectations together with average GDP per capita over the time periods in the data set
(as a measure of the difference in the level of prosperity between regions); in the Fixed
Effects estimator, all indicators that are constant over time are dropped and so average
GDP per capita does not feature in that specification.
Choice: GDP per capita

Statistical significance: 1-5% level
Direction of impact: Positive
Evidence was found of a statistically significant and reasonably large positive (as
expected) impact of GDP per capita on most choice indicators (all except product price
variety)177 in the long and short data sets, although the size of the impact was generally
smaller in the short data set. In the case of product price variety, the impact is mostly
positive, though with varying degrees of statistical significance, in the long data set 178,
and negative in the short data set179.
The relationship is shown in Figure 158. More prosperous areas with higher GDP per
capita may tend to encourage retailers to extend product choice in order to increase the
average shopping basket of their customers.
177
178
179
232
Figure 158: Choice in variety of EAN codes versus GDP per capita (source: analysis
based on © Nielsen Opus and Eurostat. Choice data are for first period in each year
2004, 2006, 2008, 2010, 2012)
Note: GDP per capita uses the Purchasing Power Standard measure.
233
Innovation: Retailer business expectations

A large positive, statistically significant impact was found for retailer business
expectations on Opus innovations in the long and short data sets 180, for new packaging
(in Fixed Effects in the long data set181 and both methods in the short data set 182), and
for new range extensions183. For new products the impact was generally negative but
not always significant in the long data set184, and positive but not significant (in Fixed
Effects) in the short data set185 and for some of the other innovation indicators (not for
new products, and not always in both data sets for the other indicators). Positive retail
business expectations are associated with a favourable macro-economic environment,
therefore encouraging suppliers to develop innovations, and retailers to stock them.
Figure 159 shows Opus innovations and retailer business expectations for the four
Member States for which innovation data are available from 2006, but in this simple
comparison no clear pattern is evident.
Figure 159: Opus innovations versus retailer business expectations (source: analysis
based on © Nielsen Opus and Eurostat), 2006, 2008, 2010, 2012
180
181
Equations [226], [228], [230], [232], [234], [236], [238], [240], [242],
182
183
Equations [305]-[322] and [325]-[342].
184
185
Equations [206], [208], [210], [212], [214], [216], [218], [220], [222].
234
9.10. General economic drivers: population and population density

In Fixed Effects estimation, where indicators are transformed to represent changes from
the average value over time, the change in population is the same as the change in
population density (because area does not change) and so the method cannot identify
the separate contribution of the drivers. Because the variation in these drivers is far
greater across space than time, it was therefore decided to represent them as average
values over time. In the Fixed Effects estimator, all indicators that are constant over
time are dropped and so average population and average population density do not
feature in that specification; the results reported here are therefore only for the Random
Effects estimator.
Choice
Statistical significance: 1% level (for population density)
Economic importance: Moderate
Moderate negative impacts of average population density on several measures of choice
were found for the Random Effects estimator, but it should be remembered that positive
impacts were found for GDP per capita, and areas with a high population density (cities)
tend also to be areas with high GDP per capita. The impact of average population was
generally not statistically significant.
Figure 160 shows the relationship between product variety and population density in the
short data set. In this two-dimensional comparison, the trend is not strong but it can be
seen that there are more cases with greater product variety in the less densely
populated areas.
Figure 160: Choice in variety of EAN codes and population density, 2008-12
Innovation
Statistical significance: 1% (for population density for new packaging and new
formulations)
235
Direction of impact: Negative (in those cases)

Economic importance: Large (in those cases)
Large, statistically significant negative effects on two indicators of innovation (new

packaging and new formulations) were found for average population density using the
Random Effects estimator. The impact of average population was generally not
statistically significant.
Figure 161 shows the relationship between Opus innovations and population density in
the short data set. Again, in this two-dimensional comparison, the trend is not strong
but it can be seen that there are more cases with a greater number of innovations in the
less densely populated areas.
Figure 161: Opus innovations and population density, 2008-12
9.11. Shop characteristics: size, format and the opening of a new shop
in the same local area
Choice: Floorspace
Differences in shop floorspace (for a given format) across space (which is what the
Random Effects estimator tends to capture in a data set in which most of the variation is
over space rather than time) were found to have clear positive statistically significant
impacts on product variety, product size variety and product supplier variety: larger
shops (for a given format) provide more choice on these indicators. For product price
236
variety the Random Effects estimator impacts were generally not statistically significant
at lower levels 186. Differences in changes in floorspace across time (which is what the
Fixed Effects estimator captures) were typically smaller in the short data set, but these
estimates depend on the experience only of those shops that change their floorspace
over time (without changing their format).
Choice: Format
Direction of impact: Positive (larger formats offer more choice)
Fixed effects dummies were included for supermarkets (relative to hypermarkets) and
for hard discounters (relative to supermarkets). For product variety, product size variety
and product supplier variety, the expected ranking was found: hypermarkets provide
more choice than supermarkets, and supermarkets provide more choice than
discounters. Findings were somewhat more mixed for product price variety.
Choice: New shop opening

Economic importance: Small or moderate
Shops that experienced the opening of a new shop in their local area tended themselves
to offer somewhat more choice (on all measures) in the long data set: in the short data
set results were often not statistically significant.
Innovation: Floorspace
As for choice, differences in shop floorspace (for a given format) across space (reflected
in the Random Effects estimator) were found to have clear positive statistically
significant impacts on all measures of innovation: larger shops (for a given format)
provide a greater number of innovative products. Differences in changes in floorspace
across time (reflected in the Fixed Effects estimator) were sometimes smaller than
across space, but this was not the case for new packaging or new formulations.
Innovation: Format
186
Equations [109], [111], [113], [115], [117], [119], [121], [123], [125], [127],
[129], [131], [133], [135], [137], [139], [141], [143].
237
Direction of impact: Positive (larger formats offer a greater number of innovative

products)
For all measures of innovation, the expected ranking was generally found: hypermarkets
provide a greater number of innovative products than supermarkets, and supermarkets
provide a greater number than discounters. The size of the impacts for discounters was
generally larger for the innovation measures than for choice, suggesting that the
difference between what hypermarkets offer and what discounters offer is more
pronounced when the focus is on innovative products than when it is on all products.
Earlier figures in this chapter in which the different shop types have been distinguished
have shown the marked differences in the scale of offering between the three types.
Innovation: New shop opening

Statistical significance: Mostly no
Direction of impact: Positive (in random effects for new products)
Economic importance: Mostly low, except for random effects for new products
The positive impact of the opening of a new shop in the local area on the offer of
innovative products in a given shop (Random Effects estimator, since variation in this
driver is greater across space than across time) was large for new products in the long
data set187, and for new formulations and new range extensions in the short data set188.
It was not generally significant for other measures of innovation. To face a new
competitor, established retailers will seek to retain customer loyalty by including new
and innovative products to either match competitors or better satisfy existing customers,
but there was less evidence for this than for the strategy of providing more choice.
10. Accounting for changes over time in selected shops

The econometric equations that are estimated over the whole data set seek to account
for the variation in choice and innovation by attributing impacts to the various drivers.
Because the data set varies over the dimensions of shops, product categories and time,
variation over all of these dimensions influences the econometric results. Here the
extent to which the equations can account for observed changes in particular places over
time, aggregating across product categories, have been examined. Since the most
important drivers that change over time that were identified in the econometric analysis
are not specific to shops but apply at national or regional level, the equations do not
have strong shop-specific drivers that can discriminate between the performance of
particular shops over time, but the presence of region-specific and national drivers can
produce different outcomes for shops located in different areas and Member States.
187
Equations [185], [187], [189], [191], [193], [195], [197], [199], [201], [203].
188
Equations [285], [287], [289], [291], [293], [295], [297], [299], [301], [303],
[325], [327], [329], [331], [333], [335], [337], [339], [341], [343].
238
Accounting for changes over time in selected shops
One can use an estimated equation to calculate the contribution to the change in product
variety or innovation over time by entering the observed values of the drivers relevant to
any given shop. The sum of these contributions represent the change that is explained
by the equation: the difference between this value and the observed change in product
variety or innovation is the unexplained residual. In order to carry out this analysis one
has to pick just one of the estimated equations and if the impact of a given driver varies
greatly between the alternative equations then its estimated contribution will also vary.
Here results are presented using a random effects equation estimated over the long data
set, measuring the influence of private labels by using the share of private label EANs in
all EANs stocked by each shop and using national retail concentration and national
supplier concentration to represent the concentration drivers.
10.1. Examples of the impacts of the drivers in five shops

The following analysis depicted in
239
Figure 162 to Figure 166 presents examples of shops where the change in choice or
innovation over time predicted by the estimated equation was close to the actual change
in order to illustrate the impacts of the drivers. An example shop is selected from each of
the five member states in the long period data set, where the drivers explained a large
proportion of the change in both choice and innovation over the period 2006 (because
the Opus innovations indicator is first calculated for that year) to 2012. For the purposes
of this illustration, examples of hypermarkets and supermarkets have been chosen.
In the case of Italy, the increase in national retail concentration and national retail sales
can explain most of the growth in choice. Meanwhile, the fall in innovation was driven by
worsening economic conditions of the recession as most of the change was accounted for
by rising unemployment and falling retail business expectations.
In the case of France, the example hypermarket showed strong growth in choice and
innovation that was predominantly driven by the expansion of the floor space in the shop
over the period and growth in national retail sales over the period. As with Italy,
innovation growth was dampen by the recession, offsetting almost all of the predicted
growth from the expansion of the shop.
In the case of Spain, the example supermarket faced the entry of a new competitor shop
over the period and this provided a modest contribution to the growth in product variety
in the supermarket, along with increases in national retail concentration and national
retail sales. Meanwhile, much of the fall in innovation was explained by the large rise in
unemployment in the area.
In the case of Poland, the example hypermarket showed strong growth in choice and
innovation over the period driven by increasing national retail concentration and national
retail sales. Growth was particularly strong for the Polish hypermarket compared to the
shops in other member states, due to a steady rise in GDP per capita along with stable
unemployment and retail business expectations over the period. Innovation was strongly
driven by the entry of a new competitor shop into the area along with the steady rise of
national retail concentration.
In the case of Portugal, the choice equation still over estimates the actual change by
some margin as shown by the negative residual. The estimated impacts show a similar
trend to Spain with national retail concentration and national retail sales accounting for
most of the change over the period. For innovation, the decline over the period is well
explained by the equation with unemployment and retail business expectations
accounting for most of the change.
In all the five shops shown as examples here, supplier concentration has a negligible
impact: this highlights the small coefficients estimated by the equation for both choice
and innovation but equally the modest change in supplier concentration over time.
240
Table 31: Key to the figures showing the contribution of drivers to change in choice and
innovation
Driver code Description

P Labels Local private labels share
Ret Conc National retail concentration HHI (edible grocery) (group)
Sup Conc National supplier concentration HHI (full market)
Shop size Shop floorspace
Unemp Regional unemployment rate
GDP pc Regional GDP per capita
Nat Sales National product category turnover
Ret Bus Exp National retailer business expectations
New Shop New shop opening in the local area
Residual The difference between the observed change in choice/innovation
and the sum of the contributions of the drivers
241
2006-12 in a hypermarket in Italy
242
2006-12 in a hypermarket in France
243
2006-12 in a supermarket in Spain
244
2006-12 in a hypermarket in Poland
245
2006-12 in a supermarket in Portugal
246
10.2. Examples of the impacts of the drivers in five CSAs

Figure 167 shows the average number of EAN codes per shop and product category in
the CSA areas from which hypermarkets 189 in the long data set were sampled. Each
point in the figure represents a single CSA area. The horizontal axis shows the average
number of EAN codes in 2004 (first period), while the vertical axis shows the number in
2012 (first period).
All but one of the points lie above the 45 degree line that is plotted in the figure,
indicating that more choice was available in 2012 than in 2004. The vertical distance
above the line indicates the extent to which the 2012 value exceeded the 2004 value.
The distribution of CSA areas by Member State reflects the findings at national level
shown (for hypermarkets) in Figure 138 above for this indicator of choice. The level in
Italy is the lowest and has increased the least; the level in France is the highest and the
gap compared with other Member States has remained broadly constant; the level in
Poland began low but has increased markedly.
Figure 167: Change in choice (product variety) offered by sample hypermarkets in

consumer shopping areas, 2004-2012 (source: analysis based on © Nielsen Opus)
It should be remembered that the number of sample shops in each CSA is not large
enough for the averages shown in the figure to be regarded as a reliable estimate of the
general level of choice in each CSA. Rather, our purpose is to identify CSAs that include
sampled shops where choice has increased markedly, or by relatively little, so as to
select shops for further investigation of the reasons for the high or low increase over
189
The focus here is on hypermarkets to filter out the impact a different representation
of shop types in different CSAs.
247
time. Extreme cases are likely to reflect special factors such as a change in the size of a
shop, rather than the impact of one of the other drivers of interest. For this reason, we
mainly focus on CSAs where the sample includes more than two hypermarkets.
Five CSAs are circled in Figure 167, and these are the ones that were selected for closer
examination. In each of France and Poland there were sufficient hypermarkets in the
sample in these CSAs to select a high and low case; in Italy only a low case was
selected).
The corresponding data for innovation are shown in Figure 168. Because the number of
innovations rose from 2006 and then fell during the recessions, the CSAs represented by
the points in the figure lie closer to the 45 degree line. The same five CSAs are circled in
the figure.
Figure 168: Change in innovation (total new EAN codes) offered by sample
hypermarkets in consumer shopping areas, 2004-2012 (source: analysis based on ©
Nielsen Opus)
Figure 169 shows the decomposition of the change in product variety for the (average of
the) sample shops in the five CSAs, attributing contributions to the various drivers.
Since we are looking at the difference between 2012 and 2006, all the control variables
that are constant over time (such as Member State fixed effects, or product category
fixed effects) drop out of the comparison (they do not change between the two years).
In the case of the two French CSAs, the fact that Gironde had a much larger increase in
product variety than Hauts-de-Seine is largely unexplained: the driver with the largest
difference between the two CSAs is the ‘Residual’ driver. More precisely, the outcome in
Hauts-de-Seine is largely explained by the contributions of the drivers, but for Gironde
there is a large positive residual. In other words, the outturn for the drivers in Gironde
compared with those in Hauts-de-Seine was not sufficiently different to account for the
difference in outcomes.
In the case of the two Warsaw CSAs, both areas saw fairly similar growth in product
variety. As with the two French CSAs, the difference between the outcomes between
248
Warsaw (High) and Warsaw (low) is unexplained as shown by the large difference in the
residual (relative to the collective impact of the other drivers).
In the case of Bologna, the growth in most of the drivers is lower than in the other four
CSAs, but there is also a substantial negative residual: the outturn for product variety
was even smaller than predicted on the basis of the drivers.
Figure 169: Contribution of drivers accounting for change in product variety 2006-12 in
five CSAs
249
Figure 170 shows the equivalent analysis for the total innovation indicator.
250
Once again, much of the difference in outturn between Gironde and Hauts-de-Seine lies
in the residual factor, but there are also some local differences (the impact of shop
expansion in Gironde) and somewhat different impacts from national influences
(reflecting different product mixes in the two areas, because some of the national effects
are specific to particular products).
One difference between the two Warsaw areas is the contribution that comes from the
stronger growth in the share of private labels in one area.
In Bologna the outturn was very close to as the equation predicted on the basis of the
drivers: positive drivers made small contributions which were largely countered by the
negative impact of the economic crisis.
Figure 170: Contribution of drivers accounting for change in total innovations 2006-12 in
five CSAs
251
252
Annexes
11. Annexes
11.1. Annex A: Illustration of © Mintel GNPD launch types
The Global New Products Database (GNPD) relies on Mintel’s global network of field
associates to identify new and changed FMCG product launches in 50 countries around
the world records are assigned a Launch Type (or Innovation type). Some Launch Types
are dependent on the Brand field190, which is used to document a product range or line
of products. There are five GNPD Launch Types: New Product, New Variety/Range
Extension, New Packaging, New Formulation, Relaunch. Every product in the database is
coded with an innovation Type.
Definitions are as follows:
 New Product: This launch type is dependent on the Brand field. It is assigned
when a new range, line, or family of products is encountered. This launch type
is also used if a brand that already exists on GNPD, in one country, crosses
over to a new sub-category191.
 New Variety/Range Extension: This launch type is dependent on the Brand
field. It is used to document an extension to an existing range of products on
the GNPD.
 New Packaging: This launch type is determined by visually inspecting the
product for changes, and also when terms like New Look, New Packaging, or
New Size are written on pack.
 New Formulation: This launch type is determined when terms such as New
Formula, Even Better, Tastier, Now Lower in Fat, New and Improved, or Great
New Taste are indicated on pack. They do not look at the ingredient list to
determine a new formulation.
 Relaunch: This launch type is determined when specified on pack, via
secondary source information (trade shows, PR, websites, and press) or when a
product has been both significantly repackaged and also reformulated
GNPD products are a representative sampling of the new and/or changed FMCG products
in a country. Each product sample or version thereof is purchased once per country--the
database does not include information on all the regions and store types in which a
product can be found.
Each country’s brand activity is treated independently, so if a range of products exists in
one particular country, any brand activity in another country is treated independently.
190
Brand is a free text field where Mintel GNPD shoppers enter all the brand and range
information off the product packaging. Brand is used in relation to innovation types to
determine whether a new product is a new variety or a new product. A new variety
would be an extension to an existing brand, ie. Danone yogurt in a new flavour. A new
product would be if they haven't seen the brand before in a particular country, ie. Coca-
cola Super Awesome.
191
A new sub-category in this context refers to a sub-category that Mintel GNPD
shoppers have not seen the product in within the same country. For instance,
carbonated soft drinks are launched under the Coca-cola brand regularly but if diapers or
hand cream were launched under Coca-cola that would constitute expanding into a new
sub-category. Each country is treated independently.
253
New Product
This launch type is dependent on the Brand field. It is assigned when a new range, line,
or family of products is encountered. This launch type is also used if a brand that already
exists on GNPD, in one country, crosses over to a new sub-category.
Examples
Strawberry Cereal
Record ID: 1507893
Company: Kellogg
Brand: Kellogg's Special K
Pépites
Category: Breakfast Cereals
Sub-Category: Cold Cereals
Country: France
Store Name: Carrefour
Store Type: Mass
Merchandise/Hype
rmarket
Store Address: Chambourcy
78240
Date Published: Mar 2011
Product source: Shopper
Launch Type: New Product
Price in local currency: €2.92
Price in US Dollars: 3.89
Bar Code: 5050083533365
Product Description
Kellogg's Special K Pépites Fraise (Strawberry) Cereal is made with fruit pieces and
enriched with vitamins B1, B2, PP, B6, B9 and B12 and iron. It contains a maximum of
3% fat and is said to be an innovative product. This cereal retails in a recyclable 375g
pack. A Nature (Natural) variety is also available in this range.
Product Analysis
Package Type: Flexible
Package Material: Plastic unspecified
Pack Size: 375.00 g
Storage: Shelf stable
Alcohol By Volume (%):
Private Label: Branded
Store Type: Mass Merchandise/Hypermarket
254
Annexes
Zucchini with Pasta and Hake Baby Meal

Record ID: 1538772
Company: Numil
Brand: Milupa Las Recetas
De Mamá
Category: Baby Food
Sub-Category: Baby Savoury
Meals & Dishes
Country: Spain
Store Name: AhorraMas
Store Type: Supermarket
Store Address: Cordoba 14007
Date Published: Jun 2011
Bar Code: 3041090830013
Product Description
Milupa Las Recetas De Mamá Calabacín con Pasta de Estrellitas y Merluza (Zucchini with
Pasta and Hake Baby Meal) contains no colouring or preservatives and is said to be low
in salt. The meal can be microwaved or steam cooked and is suitable for babies aged
from eight months. It is made according to a traditional Mediterranean recipe and
provides a portion of vegetables to provide the necessary vitamins for baby growth. The
UHT sterilized product retails in a 2 x 200g pack.
Product Analysis
Package Type: Tub
Package Material: Plastic PP
Pack Size: 200.00 g
255
Butter with Truffle

Record ID: 1566662
Company: Galateo & Friends
Brand: Galateo & Friends
Category: Dairy
Sub-Category: Butter
Country: Italy
Product source: Trade Show
Product Description
Galateo & Friends Burro con Tartufo (Butter with Truffle) is now available. The product is
retailed in a 25g jar and was on display at Tuttofood 2011 trade show in Milan, Italy.
Product Analysis
Package Type:
Package Material:
Pack Size:
Storage: Chilled
Frozen Vegetables for Minestrone

Record ID: 1955964
Company: Agrifood Abruzzo
Brand: Grandi Panieri
Category: Fruit & Vegetables
Sub-Category: Vegetables
Country: Italy
Store Name: Conad superstore
Store Type: Mass
Merchandise/Hype
rmarket
Store Address: Cava dei Tirreni
84013
Date Published: Dec 2012
256
Annexes

Bar Code: 8015536000402
Product Description
Grandi Panieri Il Grande Minestrone Surgelato (Frozen Vegetables for Minestrone) are a
mix of diced vegetables. This product retails in a 1000g pack featuring cooking
instructions.
Product Analysis
Package Material: Plastic LDPE
Pack Size: 1000.00 g
Storage: Frozen
Veal & Poultry Sausages

Record ID: 1901757
Company: Sokolów
Brand: Sokolów Sokoliki
Category: Processed Fish,
Meat & Egg
Products
Sub-Category: Meat Products
Country: Poland
Store Name: Piotr i Pawel
Store Address: Warsaw 02-777
Date Published: Oct 2012
Price in local currency: PLN3.99
Price in Euros: 0.96
Bar Code: 5906712808277
Product Description
257
Sokolów Sokoliki Veal & Poultry Sausages) are low-fat sausages for children. They are
made with 87% meat and natural seasonings. The product is rich in protein and retails in
a 140g pack.
Product Analysis
Package Type: Tray
Pack Size: 140.00 g
Storage: Chilled
258
Annexes
New Variety/Range Extension

This launch type is dependent on the Brand field. It is used to document an extension to
an existing range of products on the GNPD.
Examples
Prepared Noodle Meal
Record ID: 566420
Company: Jean Stalaven
Brand: Rudix
Category: Meals & Meal
Centers
Sub-Category: Prepared Meals
Country: Poland
Date Published: Aug 2006
Launch Type: New
Variety/Range
Extension
Bar Code: 5900961000024
Product Description
Rudix Prepared Noodle Meal is claimed to be free from preservatives and can be heated
up in sauce-pan or microwave. This product is available in a 400g pack.
Product Analysis
Package Type: Tray
Pack Size: 400.00 g
Storage: Chilled
259
Spicy Mince Meat Pizza

Record ID: 1179045
Company: Campofrio Food
Group
Brand: Campofrio Pizza &
Salsa
Category: Meals & Meal
Centers
Sub-Category: Pizzas
Country: Portugal
Launch Type: New
Variety/Range
Extension
Bar Code: 8410320033497
Product Description
Campofrio Pizza & Salsa has launched Parrila Argentina (Spicy Mince Meat Pizza). The
product is available in a 410g pack containing one sachet of Chimichurri sauce, made
with oil, garlic and fine herbs.
Product Analysis
Package Type: Tray
Pack Size: 410.00 g
Storage: Frozen
Cherry Cream Flavoured Alpine Milk

Chocolate
Record ID: 1772680
Company: Kraft Foods
Brand: Milka
Category: Chocolate
Confectionery
Sub-Category: Chocolate Tablets
Country: Czech Republic
260
Annexes
Store Name: Billa

Store Address: Prague 25101
Date Published: Apr 2012
Launch Type: New
Variety/Range
Extension
Price in local currency: CZK24.90
Bar Code: 7622300674595
Product Description
Milka Alpska Mlecna Cokolada (Cherry Cream Flavoured Alpine Milk Chocolate) is now
available. The product retails in a 100g pack.
Product Analysis
Pack Size: 100.00 g
Whole Green Bean Pods

Record ID: 1850716
Company: Bonduelle
Brand: Bonduelle
Category: Fruit & Vegetables
Sub-Category: Vegetables
Country: Czech Republic
Store Name: Kaufland
Store Type: Mass
Merchandise/Hype
rmarket
Store Address: Prague 10100
Launch Type: New
Variety/Range
Extension
261
Price in local currency: CZK37.90

Bar Code: 3083680002295
Product Description
Bonduelle Zelene Fazulove Struky Cele (Whole Green Bean Pods) are now available. The
product retails in a 425ml can.
Product Analysis
Package Type: Can
Package Material: Metal steel
Pack Size: 400.00 g
Black Pepper & Sea Salt Crackers

Record ID: 1185038
Company: Verduijn's
Brand: Verduijn's
Category: Bakery
Sub-Category: Savoury
Biscuits/Crackers
Country: Belgium
Date Published: Sep 2009
Launch Type: New
Variety/Range
Extension
Bar Code: 8713726300539
Product Description
Verduijn's Black Pepper and Sea Salt Crackers are said to be delicious as nibbles with
drinks. This product can be served with a dip or topped with sour cream and salmon. The
product is retailed in a 75g pack. Also available are the following varieties: Sesame and
Sea Salt; and Rosemary and Sea Salt.
262
Annexes
Product Analysis
Package Material: Metallised Film
Pack Size: 2.60 g
New Packaging
This launch type is determined by visually inspecting the product for changes, and also
when terms like New Look, New Packaging, or New Size are written on pack.
Examples
Bolognese penne pasta
Record ID: 1510482
Company: Sodebo
Brand: Pasta Box by Sodeb'O
Category: Meals & Meal Centers
Sub-Category: Instant Pasta
Country: Spain
Store Name: Alcampo
Store Type: Mass
Merchandise/Hypermarket
Store Address: Torrelodones 28240
Date Published: Mar 2011
Launch Type: New Packaging
Bar Code: 3242272252054
Product Description
Pasta Box by Sodeb'O Penne Boloñesa (Bolognese Penne Pasta) has been repackaged
and is now available in a 300g pack complete with a fork. The precooked product can be
prepared in the microwave in two minutes.
Product Analysis
Package Type: Tub
Pack Size: 300.00 g
263
Storage: Chilled
Chili con Carne and Rice Kit

Record ID: 1708460
Company: Carrefour - CMI
Brand: Carrefour
Category: Meals & Meal Centers
Sub-Category: Meal Kits
Country: France
Store Name: Carrefour
Store Type: Mass
Store Address: Montesson 78360
Date Published: Jan 2012
Bar Code: 3270190131038
Product Description
Carrefour Chili con Carne (Chili con Carne and Rice Kit) has been repackaged in a newly
designed 510g box containing a 400g can of chili with beef mixture and a 110g sachet of
long grain rice. The can content can be heated in the microwave once cooked on a pan.
This product serves 2 people.
Product Analysis
Package Type: Can
Package Material: Metal steel
Pack Size: 510.00 g
Private Label: Private Label
264
Annexes
Organic Mini Rice Cakes

Record ID: 1693571
Company: Fiorentini Alimentari
Brand: Fiorentini Bio
Category: Bakery
Sub-Category: Savoury Biscuits/Crackers
Country: Italy
Store Name: Conad superstore
Store Type: Mass
Store Address: Cava dei Tirreni 84013
Bar Code: 8002885000160
Product Description
Fiorentini Bio Mini Gallette di Riso (Organic Mini Rice Cakes) have been repackaged and
now retail in a newly designed 200g pack with a resealable tab.This product from Italian
rice has a low fat content.
Product Analysis
Pack Size: 200.00 g
265
Wholegrain Subs
Record ID: 1918744
Company: Kohberg Brød
Brand: Kohberg
Category: Bakery
Sub-Category: Bread & Bread
Products
Country: Denmark
Store Name: Føtex
Store Address: Aalborg 9000
Date Published: Nov 2012
Price in local currency: DKK15.00
Bar Code: 5701246108325
Product Description
Kohberg Fuldkorns Subs (Wholegrain Subs) have been repackaged in a newly designed
510g pack containing six units. The design features a pink bra to support the fight
against breast cancer, and 1 kr. will be donated to this campaign for each purchased
bag. The packaging bears a Green Keyhole logo for a healthier choice.
Product Analysis
Pack Size: 510.00 g
266
Annexes
Pu-Erh Red Tea with Lemon Aroma

Record ID: 1374368
Company: Foltin Globe
Brand: Vitax
Category: Hot Beverages
Sub-Category: Tea
Country: Hungary
Store Name: Tesco
Store Address: Gödöllö 2100
Date Published: Jul 2010
Price in local currency: HUF286.00
Bar Code: 5902806061054
Product Description
Vitax Pu-Erh Red Tea with Lemon Aroma has been repackaged and is now available in a
30g pack with an updated design. One pack contains 20 x 1.5g tea bags. The tea is said
to be discovered thousand years ago in China and was used in a secret by Chinese
emperors of the country and their families only.
Product Analysis
Package Type: Carton
Package Material: Board white lined
Pack Size: 30.00 g
267
Chocolate Flavour Mini Biscuits

Record ID: 1917236
Company: Kraft Foods
Brand: Lu Prince Mini
Mystery Box
Category: Bakery
Sub-Category: Sweet
Biscuits/Cookies
Country: Belgium
Store Name: Colruyt
Store Address: Nossegem 1930
Bar Code: 5629400759803
Product Description
Lu Prince Mini Mystery Box Biscuits Fourrés Goût Chocolat (Chocolate Flavour Mini
Biscuits) are now available in a limited edition festive treasure box. The product retails in
a pack containing 10 x 42g biscuits and a games booklet.
Product Analysis
Package Material: Metallised Film
Pack Size: 42.00 g
268
Annexes
Coffee Repackaging
Record ID: 385791
Company: Christgau Kaffe
Brand: Christgau
Category: Hot Beverages
Sub-Category: Coffee
Country: Denmark
Product source: Publication
Product Description
The Christgau coffee range has been repackaged in resealable packs with a tiny air hole
to preserve the aroma.
Product Analysis
Package Material:
Pack Size: 250.00 g
Storage:
Patent Number:
Carbonated Drink
Record ID: 1216205
Company: Fashion Drinks
Brand: Moxito by Fashion
Drinks
Category: Carbonated Soft
Drinks
Sub-Category: Carbonated Soft
Drinks
Country: Spain
269

Bar Code: 8435040300117
Product Description
Moxito by Fashion Drinks Carbonated Drink has been repackaged and is now retailed in a
250ml pack featuring a new design. This drink is alcohol-free and is available in an
original citrus and peppermint flavour.
Product Analysis
Package Type: Can
Package Material: Metal aluminium
Pack Size: 250.00 ml
270
Annexes
New Formulation
This launch type is determined by visually looking for key terms on pack like New
Formula, Even Better, Tastier, Now Lower in Fat, New and Improved, Great New Taste,
Now With…, or Better …. We cannot assume that a product is newly reformulated unless
it is clearly stated on pack or we know from secondary sources that this is the case.
Examples
Frying Oil
Record ID: 1455097
Company: Lesieur
Brand: Lesieur Frial
Category: Sauces &
Seasonings
Sub-Category: Oils
Country: France
Store Name: Intermarché
Store Address: Montendre 17240
Launch Type: New Formulation
Bar Code: 3265479327011
Product Description
Lesieur Frial Frying Oil is now available featuring a new recipe containing no palm oil,
which is a source of saturated fat. This product is said to make crunchy and light food
without bad odours and retails in a 1L bottle.
Product Analysis
Package Type: Bottle
Package Material: Plastic PET
Pack Size: 1.00 litre
271
Chocolate and Hazelnut Flavoured Milk

Record ID: 940330
Company: Zott
Brand: Zott Monte Drink
Category: Dairy
Sub-Category: Flavoured Milk
Country: Hungary
Price in local currency: HUF135.00
Bar Code: 4014500024424
Product Description
Zott Monte Drink Chocolate and Hazelnut Flavoured Milk has been reformulated and
made using grape sugar. This product contains calcium and vitamin B12 and is available
in a 200ml bottle.
Product Analysis
Package Material: Plastic PE
Storage: Chilled
Low-Fat Strawberry Yogurt

Record ID: 1813219
Company: Danone
Brand: Danone Vitalinea
SatisfAcción Pro
Category: Dairy
Sub-Category: Spoonable Yogurt
Country: Spain
Store Name: Mercadona
Store Address: Empuriabrava
17487
272
Annexes

Bar Code: 8410500013608
Product Description
Danone Vitalinea SatisfAcción Pro Fresa (Low-Fat Strawberry Yogurt) is now available
with a new formula that is said to have double the protein, providing 10.5g of protein
per portion. This gluten-free yogurt comprises fermented milk with skimmed fresh
cheese and strawberries. This product with a creamy texture is retailed in a 540g pack
with four 135g tubs. Also reformulated in this range are varieties with the following
flavours: Natural; and Peach.
Product Analysis
Package Type: Tub
Package Material: Plastic PS
Pack Size: 135.00 g
Storage: Chilled
Semolina Dessert Reformulation

Record ID: 379401
Company: Emmi
Brand: Emmi Griess Töpfli
Category: Desserts & Ice
Cream
Sub-Category: Chilled Desserts
Country: Portugal
Bar Code: 7610900118380
Product Description
273
Believed to be reformulated is Griess-Töpfli, a semolina dessert with cream, packaged in

a 175g tub.
Product Analysis
Package Type: Tub
Package Material: Plastic PS
Pack Size: 175.00 g
Storage: Chilled
Patent Number:
274
Annexes
Relaunch
This launch type is determined when: there is some wording to the effect that the
product has been relaunched on the packaging or the product does not exist on the
database but there is secondary source information (such as from a press release,
magazine, trade show, website or a shop display) that the product has been relaunched.
Key phrases to look out for include “previously or formerly known as…” and “new name”.
If a product meets the criteria for the new packaging launch type and for the new
formulation launch type, then the relaunch launch type should be selected.
Examples
Multifruit Smoothie
Record ID: 1813068
Company: Marwit
Brand: Marwit Owocudo
Happy
Category: Juice Drinks
Sub-Category: Juice
Country: Poland
Store Name: Real
Store Type: Mass
Merchandise/Hype
rmarket
Store Address: Warsaw 02-801
Launch Type: Relaunch
Bar Code: 5904373000368
Product Description
Marwit Owocudo Happy Sok Wieloowocowy (Multifruit Smoothie) has been relaunched
and comprises pasteurised juice with fruit mousse, partially from concentrate and purée.
It contains no added sugars and only natural sugars. The product retails in a 200ml
bottle.
Product Analysis
Package Material: Glass plain
Storage: Chilled
275

276
Annexes
Sliced Brigante Cheese

Record ID: 1941519
Company: Fratelli Pinna
Azienda Casearia
Brand: F.lli Pinna
Category: Dairy
Sub-Category: Hard Cheese &
Semi-Hard Cheese
Country: Italy
Store Name: Auchan
Store Type: Mass
Merchandise/Hype
rmarket
Store Address: Nola 80035
Bar Code: 8010861005092
Product Description
F.lli Pinna Brigante a Fette (Sliced Brigante Cheese) has been relaunched. The sheep
cheese now retails in a 0.100kg tray pack.
Product Analysis
Package Type: Tray
Pack Size: 100.00 g
Storage: Chilled
277
Natural Mineral Water

Record ID: 1927347
Company: SEAB
Brand: Marque Savoie Aix
Les Bains
Category: Water
Sub-Category: Water
Country: France
Product source: Trade Show
Product Description
Marque Savoie Aix Les Bains Eau Minérale Naturelle (Natural Mineral Water) has been
relaunched with a new brand name and in a newly designed pack. The product retails in
a 0.75L bottle and was on display at the SIAL 2012 trade show, in Paris.
Product Analysis
Package Material: Plastic PET
Pistachio Ice Cream

Record ID: 1957719
Company: Auchan
Brand: Auchan
Category: Desserts & Ice
Cream
Sub-Category: Dairy-Based
Frozen Products
Country: France
Store Name: Auchan
Store Type: Mass
Merchandise/Hype
rmarket
Store Address: Plaisir 78370
278
Annexes

Bar Code: 3254563259420
Product Description
Auchan Glace Pistache (Pistachio Ice Cream) has been reformulated with a new recipe. The
product contains pieces of roasted pistachios and retails in a newly designed 1L tub, which
contains approximately 20 servings.
Product Analysis
Package Type: Tub
Storage: Frozen
Private Label: Private Label
279
11.2. Annex B: Descriptive statistics
11.2.1. Evolution of choice
Evolution of the total number of EAN codes by product category

14000
12000
10000
8000
2004
6000 2006
2008
4000 2010
2012
2000
280
Annexes
11.2.2. Evolution of innovation per product category

Proportion of new product innovations 2004 2006 2008 2010 2012
Baby food (ambient)
Product 54% 31% 31% 41% 31%

Packaging 1% 6% 16% 11% 27%
Formula 9% 6% 12% 16% 8%
Range extension 34% 56% 41% 31% 25%
Relaunch 1% 1% 0% 0% 9%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - baby food
(ambient)
100%
90%
80%
Relaunch
70%
60% Range extension
50%
Formula
40%
30% Packaging
20% Product
10%
0%
2004 2006 2008 2010 2012

Biscuits
Product 51% 38% 44% 44% 33%

Packaging 6% 8% 10% 16% 27%
Formula 2% 4% 6% 5% 4%
Relaunch 1% 1% 0% 0% 5%
TOTAL 100% 100% 100% 100% 100%
Propotion of new product innovations - Biscuits

100%
80% Relaunch
60% Range extension

Formula
40%
Packaging
20%
Product
0%
2004 2006 2008 2010 2012
281

Butter/margarine
Product 54% 39% 46% 46% 35%

Packaging 7% 12% 12% 23% 31%
Formula 5% 1% 5% 6% 4%
Relaunch 0% 3% 0% 0% 4%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations -

Butter/margarine
100%
80% Relaunch
60% Range extension
40% Formula
20% Packaging
0% Product
2004 2006 2008 2010 2012

Canned vegetables
Product 50% 43% 41% 43% 39%

Packaging 21% 10% 13% 22% 19%
Formula 8% 4% 3% 2% 2%
Relaunch 0% 1% 0% 0% 3%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Canned

vegetables
100%
80% Relaunch
60% Range extension
40% Formula
Packaging
20%
Product
0%
2004 2006 2008 2010 2012
282
Annexes

Cereals
Product 59% 43% 44% 37% 29%

Packaging 5% 9% 9% 25% 31%
Formula 8% 4% 11% 10% 7%
Relaunch 3% 3% 0% 0% 6%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Cereals

100%
90%
80%
Relaunch
70%
60% Range extension
50%
Formula
40%
30% Packaging
20%
10%
Product
0%
2004 2006 2008 2010 2012

Cheese
Product 55% 32% 33% 37% 31%

Packaging 7% 13% 18% 23% 29%
Formula 3% 4% 4% 5% 2%
Relaunch 1% 1% 0% 0% 5%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Cheese

100%
90%
80%
70%
Relaunch
60% Range extension
50%
Formula
40%
30% Packaging
20% Product
10%
0%
2004 2006 2008 2010 2012
283

Product 46% 33% 48% 46% 36%

Packaging 3% 8% 11% 18% 26%
Formula 1% 2% 4% 5% 3%
Relaunch 1% 1% 0% 0% 3%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Chocolate

(Bar + Candies)
100%
90%
80% Relaunch
70%
60% Range extension
50%
Formula
40%
30% Packaging
20%
10% Product
0%
2004 2006 2008 2010 2012

Coffee
Product 50% 33% 31% 36% 25%

Packaging 3% 10% 20% 24% 35%
Formula 7% 2% 7% 7% 2%
Relaunch 2% 1% 0% 0% 6%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Coffee

100%
90%
80%
Relaunch
70%
60% Range extension
50%
Formula
40%
30% Packaging
20%
Product
10%
0%
2004 2006 2008 2010 2012
284
Annexes

Dessert
Product 52% 28% 29% 32% 28%

Packaging 3% 10% 10% 13% 19%
Formula 6% 8% 7% 11% 8%
Relaunch 0% 1% 0% 0% 4%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Dessert

100%
80% Relaunch
60% Range extension
Formula
40%
Packaging
20%
Product
0%
2004 2006 2008 2010 2012

Edible oil
Product 67% 55% 53% 52% 45%

Packaging 0% 14% 15% 25% 31%
Formula 0% 3% 2% 1% 1%
Relaunch 0% 1% 1% 1% 2%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Edible oil

100%
80% Relaunch
60% Range extension
40%
Formula
Packaging
20%
Product
0%
2004 2006 2008 2010 2012
285

Product 64% 36% 42% 35% 28%

Packaging 3% 3% 5% 16% 25%
Formula 3% 5% 11% 9% 8%
Relaunch 0% 2% 0% 0% 2%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Fresh pre-

packaged bread
100%
80% Relaunch
60% Range extension
Formula
40%
Packaging
20%
Product
0%
2004 2006 2008 2010 2012

Frozen vegetables
Product 44% 43% 36% 32% 25%

Packaging 4% 4% 8% 17% 21%
Formula 7% 6% 6% 7% 6%
Relaunch 0% 0% 1% 0% 4%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Frozen

vegetables
100%
90%
80%
Relaunch
70%
60% Range extension
50%
Formula
40%
30% Packaging
20%
Product
10%
0%
2004 2006 2008 2010 2012
286
Annexes

Product 45% 36% 44% 39% 31%

Packaging 11% 9% 17% 18% 31%
Formula 1% 1% 3% 5% 2%
Relaunch 1% 0% 0% 1% 5%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Fruit juices

(ambient)
100%
80% Relaunch
60% Range extension
40%
Formula
Packaging
20%
Product
0%
2004 2006 2008 2010 2012

Ham/delicatessen
Product 53% 31% 36% 34% 31%

Packaging 8% 8% 10% 17% 21%
Formula 3% 4% 4% 6% 4%
Relaunch 0% 0% 0% 0% 2%
TOTAL 100% 100% 100% 100% 100%

Ham/delicatessen
100%
80% Relaunch
60% Range extension
40%
Formula
Packaging
20%
Product
0%
2004 2006 2008 2010 2012
287

Ice cream
Product 62% 38% 37% 44% 40%

Packaging 1% 3% 7% 10% 15%
Formula 1% 5% 8% 4% 4%
Relaunch 0% 2% 0% 0% 2%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Ice cream

100%
80% Relaunch
60% Range extension

Formula
40%
Packaging
20%
Product
0%
2004 2006 2008 2010 2012

M ilk
Product 59% 43% 44% 39% 35%

Packaging 8% 12% 22% 31% 35%
Formula 0% 2% 3% 2% 2%
Relaunch 0% 4% 0% 0% 5%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Milk

100%
80% Relaunch
60% Range extension
40%
Formula
Packaging
20%
Product
0%
2004 2006 2008 2010 2012

M ineral water
Product 31% 33% 38% 35% 29%

Packaging 26% 23% 36% 50% 54%
Formula 6% 2% 1% 2% 1%
Relaunch 0% 1% 1% 0% 2%
TOTAL 100% 100% 100% 100% 100%
288
Annexes
Proportion of new product innovations - Mineral water

100%
90%
80%
Relaunch
70%
60% Range extension
50%
Formula
40%
30% Packaging
20% Product
10%
0%
2004 2006 2008 2010 2012

Ready-cooked meals
Product 37% 27% 34% 26% 27%

Packaging 7% 11% 9% 12% 19%
Formula 9% 10% 11% 21% 10%
Relaunch 3% 1% 2% 0% 6%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Ready-cooked

meals
100%
90%
80%
Relaunch
70%
60% Range extension
50%
Formula
40%
30% Packaging
20%
Product
10%
0%
2004 2006 2008 2010 2012

Savoury snacks
Product 53% 44% 41% 47% 38%

Packaging 3% 7% 12% 18% 24%
Formula 1% 2% 8% 5% 4%
Relaunch 1% 1% 0% 1% 4%
TOTAL 100% 100% 100% 100% 100%
289
Proportion of new product innovations - Savoury

snacks
100%
80% Relaunch
60% Range extension
40% Formula
Packaging
20%
Product
0%
2004 2006 2008 2010 2012

Soft-drinks
Product 45% 39% 44% 40% 32%

Packaging 12% 16% 20% 30% 36%
Formula 0% 3% 6% 6% 3%
Relaunch 1% 1% 0% 1% 3%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Soft-

drinks
100%
80% Relaunch
60% Range extension
40% Formula
20% Packaging
0% Product
2004 2006 2008 2010 2012

Starters/pizzas
Product 51% 33% 42% 39% 33%

Packaging 11% 5% 6% 13% 19%
Formula 9% 13% 8% 17% 7%
Relaunch 0% 2% 0% 0% 5%
TOTAL 100% 100% 100% 100% 100%
290
Annexes

Starters/pizzas
100%
80% Relaunch
60% Range extension
40% Formula
20% Packaging
0% Product
2004 2006 2008 2010 2012

Tea
Product 40% 31% 30% 32% 22%

Packaging 8% 8% 11% 17% 21%
Formula 5% 1% 2% 3% 4%
Relaunch 3% 0% 0% 2% 6%
TOTAL 100% 100% 100% 100% 100%
Proportion of new product innovations - Tea

100%
80% Relaunch
60% Range extension
40% Formula
20% Packaging
0%
Product
2004 2006 2008 2010 2012

Yoghurt
Product 35% 33% 36% 36% 32%

Packaging 6% 6% 10% 15% 23%
Formula 1% 5% 5% 7% 6%
Relaunch 1% 1% 0% 0% 3%
TOTAL 100% 100% 100% 100% 100%
291
Proportion of new product innovations - Yoghurt

100%
80% Relaunch
60% Range extension
40% Formula
Packaging
20%
Product
0%
2004 2006 2008 2010 2012
11.2.3. Evolution of private labels per Member State (Euromonitor)

Percentage of private labels per

product category 2004 2006 2008 2010 2012
BELGIUM
Baby food 0,6 0,6 0,7 1,4 1,8
Biscuits 39,6 42,0 43,0 44,5 46,9
Bread 8,1 8,7 9,0 9,2 9,8
Butter/margarine 18,4 18,2 18,1 18,5 18,9
Canned vegetables 46,3 51,5 53,0 54,2 54,4
Cereals 25,1 25,0 26,4 28,7 30,1
Cheese 19,2 19,9 19,3 19,8 20,2
Chocolate 12,0 13,1 12,3 12,8 13,6
Coffee 18,5 18,7 19,0 18,9 18,2
Desserts 26,4 28,6 30,3 31,8 33,0
Edible oil 34,2 40,9 43,2 45,1 47,7
Frozen pizzas/starters 23,4 23,3 21,6 17,5 16,0
Frozen ready cooked meals 34,7 36,5 35,9 35,0 33,5
Frozen vegetables 40,8 45,1 46,6 46,2 44,9
Fruit Juices 42,7 38,0 37,5 39,0 38,3
Ham 59,1 66,4 69,8 68,9 65,9
Ice Cream 23,0 23,0 23,4 23,7 23,9
Milk 64,1 62,7 62,4 62,2 62,2
Mineral water 15,2 16,4 16,0 16,3 16,9
Savoury snacks 23,2 21,9 22,1 22,6 23,1
Soft drinks 16,0 15,7 15,1 14,5 14,0
Tea 21,5 21,5 21,9 21,4 22,6
Yoghurt 20,8 20,2 20,0 20,5 22,9
292
Annexes

CZECH REPUBLIC
Baby food 0,0 0,0 0,7 1,9 2,4
Biscuits 3,1 5,5 6,4 8,2 8,5
Bread 1,9 2,6 2,5 2,6 2,5
Cereals 6,3 9,1 11,3 13,3 13,2
Cheese 3,5 3,2 5,3 5,8 7,3
Chocolate 2,4 5,9 6,9 7,5 8,6
Coffee 3,8 3,8 5,1 5,1 5,1
Desserts 7,9 9,2 9,9 10,1 11,4
Edible oil 20,3 23,9 28,0 30,7 35,0
Fruit Juices 10,8 10,8 9,6 9,0 9,5
Ham 8,7 10,0 11,0 14,0 15,0
Ice Cream 2,8 3,0 4,0 4,1 4,7
Milk 19,6 21,3 24,0 23,6 26,2
Mineral water 4,0 4,4 4,1 3,9 3,2
Savoury snacks 9,4 10,5 11,7 13,7 14,2
Soft drinks 6,6 6,9 7,2 7,6 7,6
Tea 3,8 4,4 4,7 6,1 6,9
Yoghurt 4,7 6,5 7,8 7,3 9,8

DENMARK
Baby food 0,0 0,0 0,0 0,0 0,0
Biscuits 13,7 14,8 16,5 20,5 23,2
Bread 9,9 9,5 10,2 9,0 8,6
Cereals 20,3 22,0 20,6 23,8 22,1
Cheese 10,8 11,6 12,4 13,7 14,5
Chocolate 4,1 4,3 5,0 5,7 6,4
Coffee 12,1 10,5 11,0 11,7 14,8
Desserts 9,4 9,8 9,0 10,4 11,6
Edible oil 26,4 27,3 31,1 32,7 36,7
Fruit Juices 18,2 19,8 21,1 23,7 27,1
Ham 25,9 28,9 28,6 27,7 33,4
Ice Cream 8,3 9,4 11,7 13,2 13,6
Milk 14,5 14,5 15,7 17,8 17,7
Mineral water 12,4 13,4 13,8 12,8 13,6
Savoury snacks 15,4 16,4 16,9 18,7 20,0
Soft drinks 12,5 12,3 14,4 15,2 13,2
Tea 12,9 15,6 16,2 16,5 18,3
Yoghurt 5,1 5,0 6,1 6,4 7,3
293

FINLAND
Baby food 1,6 0,7 0,5 0,0 0,0
Biscuits 8,5 8,8 10,9 12,9 13,8
Bread 2,2 2,2 3,4 5,4 7,5
Cereals 10,2 12,6 15,6 17,3 18,1
Cheese 10,1 11,4 11,6 13,5 15,0
Chocolate 3,4 3,7 4,8 5,5 5,7
Coffee 5,5 5,6 6,3 10,3 21,3
Desserts 7,0 8,0 9,8 11,5 15,3
Edible oil 40,6 41,8 41,9 44,6 45,9
Fruit Juices 15,3 15,9 17,4 18,3 20,3
Ham 11,8 14,4 16,6 19,2 19,6
Ice Cream 6,8 7,1 9,1 9,9 10,8
Milk 1,5 1,8 3,0 4,8 6,7
Mineral water 5,7 8,6 8,9 10,4 11,0
Savoury snacks 18,3 20,5 21,5 23,3 22,8
Soft drinks 5,6 7,2 7,5 9,2 11,2
Tea 12,0 12,1 10,3 11,8 12,4
Yoghurt 6,3 8,1 8,9 9,7 11,1

France
Baby food 1,7 1,9 2,1 3,0 3,1
Biscuits 22,7 24,3 26,1 25,4 21,4
Bread 2,4 2,7 3,2 3,6 3,8
Cereals 16,0 17,4 18,5 15,3 13,5
Cheese 24,5 25,6 27,1 28,3 28,1
Chocolate 8,0 8,1 7,9 7,8 7,2
Coffee 7,0 6,9 6,9 5,8 5,3
Desserts 25,3 27,2 29,5 29,2 29,5
Edible oil 35,4 37,4 39,6 41,6 43,4
Fruit Juices 24,6 25,0 24,0 22,4 21,3
Ham 33,6 34,3 35,1 39,5 38,3
Ice Cream 16,1 16,9 16,7 16,2 14,3
Milk 36,7 38,2 40,9 40,9 42,3
Mineral water 9,8 10,7 11,0 11,7 12,2
Savoury snacks 27,0 28,8 31,0 33,6 32,6
Soft drinks 10,3 10,4 9,6 9,7 9,0
Tea 12,1 16,7 16,1 16,3 15,4
Yoghurt 14,6 15,5 16,6 17,2 17,2
294
Annexes

GERMANY
Baby food 2,7 2,9 2,8 3,1 3,2
Biscuits 35,5 36,5 37,4 36,8 36,9
Bread 20,6 21,1 21,2 19,9 20,1
Cereals 31,9 33,4 34,9 31,4 31,2
Cheese 29,4 30,0 30,9 30,3 30,1
Chocolate 15,9 16,7 17,2 16,1 16,0
Coffee 18,3 21,9 21,9 21,6 21,2
Desserts 36,8 39,4 39,7 39,5 40,0
Edible oil 44,4 45,7 47,9 48,1 47,1
Fruit Juices 29,2 31,0 30,9 31,8 31,5
Ham 69,4 69,7 69,5 72,2 71,4
Ice Cream 21,5 22,0 22,0 21,7 21,9
Milk 57,8 62,7 65,5 65,0 66,8
Mineral water 7,2 10,3 11,4 13,2 13,5
Savoury snacks 24,5 26,5 35,2 37,7 36,0
Soft drinks 19,0 20,6 19,6 18,4 18,0
Tea 19,2 19,8 19,8 19,1 18,5
Yoghurt 22,6 22,7 22,3 22,2 23,2

HUNGARY
Baby food 1,7 1,8 2,1 1,9 2,2
Biscuits 11,1 16,3 21,9 26,4 26,5
Bread 0,6 1,2 1,7 2,2 2,4
Cereals 9,5 17,0 32,9 37,1 38,2
Cheese 3,7 5,6 9,4 10,9 11,5
Chocolate 3,1 6,3 8,5 10,0 10,0
Coffee 4,5 5,4 7,4 7,8 8,0
Desserts 17,1 19,6 23,3 24,7 25,3
Edible oil 25,8 28,0 31,8 33,4 33,3
Fruit Juices 8,4 13,3 17,1 20,9 24,8
Ham 4,1 16,4 26,1 32,3 33,5
Ice Cream 8,7 11,4 13,6 16,2 16,9
Milk 15,0 17,9 21,0 23,3 25,7
Mineral water 3,7 4,4 9,5 14,8 12,3
Savoury snacks 7,8 11,7 17,0 18,2 20,7
Soft drinks 3,7 5,7 8,1 9,8 10,5
Tea 3,6 5,0 6,5 8,3 8,3
Yoghurt 9,7 10,7 13,2 14,6 15,0
295

ITALY
Baby food 1,4 2,0 2,1 2,2 2,4
Biscuits 12,7 14,4 15,4 16,0 16,7
Bread 2,4 2,6 3,0 3,5 3,9
Cereals 5,5 5,5 5,8 5,7 6,1
Cheese 6,0 6,5 6,5 7,1 8,5
Chocolate 4,0 4,1 4,1 4,7 5,1
Coffee 4,5 5,6 5,9 6,6 7,0
Desserts 7,5 7,5 7,4 10,2 11,3
Edible oil 17,6 17,8 17,4 21,1 22,9
Fruit Juices 17,2 17,3 17,2 17,3 17,8
Ham 23,5 24,6 24,8 26,9 28,2
Ice Cream 4,4 4,1 4,0 4,4 4,3
Milk 9,2 11,6 14,5 18,6 20,2
Mineral water 3,6 3,8 3,9 4,1 4,3
Savoury snacks 12,5 12,9 13,6 14,2 14,5
Soft drinks 5,1 5,1 5,7 6,2 7,8
Tea 4,8 6,1 6,5 6,6 7,0
Yoghurt 8,2 8,5 9,0 10,2 10,8

NETHERLANDS
Baby food 0,3 1,2 2,8 4,9 6,0
Biscuits 30,7 31,8 31,5 32,4 33,2
Bread 30,7 29,4 28,6 27,6 28,0
Cereals 12,0 11,9 12,0 11,7 12,3
Cheese 26,9 27,2 27,9 31,1 32,5
Chocolate 11,8 12,6 12,5 12,9 16,9
Coffee 10,1 12,3 14,0 16,4 17,2
Desserts 22,7 25,2 24,2 25,7 29,1
Edible oil 39,7 40,3 41,7 44,8 45,8
Fruit Juices 16,4 18,9 20,1 22,9 23,0
Ham 84,6 86,1 87,0 90,0 88,3
Ice Cream 9,2 10,1 10,6 12,7 13,7
Milk 39,2 39,3 38,2 41,2 43,2
Mineral water 8,8 8,6 8,3 9,8 12,2
Savoury snacks 23,9 26,0 25,0 26,3 25,9
Soft drinks 14,0 13,8 14,2 15,2 15,4
Tea 12,8 13,6 14,2 13,9 16,3
Yoghurt 20,8 20,7 20,7 23,1 26,7
296
Annexes

POLAND
Baby food 0,0 0,0 0,0 0,0 0,0
Biscuits 2,2 4,1 4,3 4,9 12,7
Bread 2,0 2,4 2,6 3,2 8,7
Cereals 6,7 7,8 7,9 8,6 16,1
Cheese 8,4 9,6 11,6 14,6 18,9
Chocolate 3,1 4,6 6,0 7,4 12,0
Coffee 0,7 0,7 0,8 0,7 0,8
Desserts 8,2 9,3 10,2 11,6 15,6
Edible oil 6,6 7,4 7,5 8,7 16,0
Fruit Juices 4,9 5,3 5,5 6,7 6,0
Ham 8,4 7,6 7,5 7,3 7,4
Ice Cream 0,9 1,3 1,4 1,7 4,2
Milk 6,8 7,5 10,3 15,8 22,7
Mineral water 5,6 6,6 6,3 8,1 10,9
Savoury snacks 5,8 7,0 7,6 8,3 9,0
Soft drinks 2,5 3,8 6,2 7,1 8,3
Tea 4,9 5,7 6,1 8,2 12,4
Yoghurt 6,0 6,7 7,2 9,6 15,2

Portugal
Baby food 6,2 6,0 5,8 6,3 6,4
Biscuits 28,1 32,3 36,4 40,2 41,5
Bread 2,7 4,1 4,6 5,1 5,2
Cereals 13,6 16,1 19,6 24,7 25,5
Cheese 9,0 11,7 16,2 20,7 21,8
Chocolate 11,1 10,5 12,4 13,0 13,1
Coffee 9,7 10,7 11,3 10,9 15,6
Desserts 11,6 15,1 20,6 23,6 26,3
Edible oil 22,9 27,4 32,2 37,7 37,9
Fruit Juices 17,2 20,9 25,3 28,3 30,0
Ham 23,8 26,2 33,1 38,2 43,6
Ice Cream 10,3 13,1 14,6 15,9 17,0
Milk 16,8 18,1 20,3 23,0 29,1
Mineral water 6,3 10,8 14,4 22,6 26,8
Savoury snacks 19,2 19,6 23,5 26,3 31,5
Soft drinks 7,2 9,3 17,8 26,6 29,6
Tea 5,1 5,0 6,4 8,5 10,2
Yoghurt 9,9 14,5 17,5 22,8 25,0
297

ROMANIA
Baby food 0,0 0,0 0,0 0,0 0,0
Biscuits 0,0 0,0 0,0 0,0 0,0
Bread 0,0 0,0 0,0 0,0 0,2
Cereals 0,0 0,0 0,0 0,0 0,0
Cheese 0,0 0,0 0,0 0,0 0,0
Chocolate 0,0 0,0 0,0 0,0 0,1
Coffee 0,0 0,0 0,0 0,0 0,0
Desserts 0,0 0,0 0,0 0,0 5,3
Edible oil 2,2 3,0 3,9 8,1 12,6
Fruit Juices 0,0 0,1 0,1 0,2 0,2
Ham 33,6 37,8 34,7 33,7 34,1
Ice Cream 0,0 0,0 0,0 0,0 0,1
Milk 0,0 0,0 0,0 0,0 0,5
Mineral water 1,1 2,1 2,6 3,9 4,8
Savoury snacks 0,0 0,0 0,0 0,0 0,0
Soft drinks 0,0 0,0 0,2 0,7 1,5
Tea 0,0 0,0 0,0 0,0 0,0
Yoghurt 0,0 0,0 0,0 0,0 2,3

SPAIN
Baby food 0,0 0,2 0,6 1,2 3,7
Biscuits 19,8 21,9 24,5 29,1 34,4
Bread 2,9 3,3 3,9 5,7 7,4
Cereals 20,6 23,0 25,5 29,7 37,8
Cheese 9,2 11,0 17,0 20,4 23,2
Chocolate 14,0 14,3 15,6 19,2 20,0
Coffee 17,9 19,0 17,7 21,4 20,3
Desserts 23,2 24,8 27,9 31,0 34,6
Edible oil 40,8 43,3 48,6 49,6 54,4
Fruit Juices 28,9 30,6 32,9 34,4 37,4
Ham 20,5 23,7 27,0 44,7 52,9
Ice Cream 9,0 9,7 14,0 18,8 26,9
Milk 22,9 28,1 32,9 36,9 46,2
Mineral water 10,9 12,3 17,0 21,5 23,4
Savoury snacks 23,9 25,3 31,9 34,9 36,1
Soft drinks 4,5 4,9 5,6 9,2 11,1
Tea 12,0 12,9 12,9 19,8 24,0
Yoghurt 14,9 16,1 19,2 21,3 24,7
298
Annexes

UNITED KINGDOM
Baby food 1,1 1,0 0,6 0,7 0,7
Biscuits 22,1 22,1 21,9 21,1 21,3
Bread 11,8 14,2 14,7 14,0 14,3
Cereals 20,6 20,1 20,7 21,0 21,8
Cheese 37,7 36,6 34,4 32,3 31,5
Chocolate 7,7 7,5 7,7 7,9 8,2
Coffee 12,9 13,8 14,4 16,3 19,8
Desserts 34,8 35,6 36,8 36,7 36,6
Edible oil 55,1 56,8 54,8 51,7 46,6
Fruit Juices 36,5 36,2 36,2 36,7 36,0
Ham 52,6 54,2 58,0 59,2 62,7
Ice Cream 22,1 23,0 22,6 24,9 24,9
Milk 65,5 66,1 66,4 66,3 66,5
Mineral water 28,4 28,2 31,6 30,3 27,8
Savoury snacks 23,5 26,1 28,5 30,4 32,3
Soft drinks 11,6 11,4 10,9 9,6 8,9
Tea 17,7 17,3 14,6 15,6 15,1
Yoghurt 17,3 15,4 14,4 13,8 14,4

Retailer concentration - Retail Group Level - C5 Edible grocery 2004 2006 2008 2010 2012 CAGR
Belgium 51% 55% 57% 58% 59% 2%

Czech Republic 26% 32% 38% 41% 44% 7%
France 56% 57% 58% 59% 60% 1%
Hungary 26% 29% 30% 30% 30% 2%
Italy 19% 19% 20% 21% 21% 1%
Poland 13% 16% 23% 28% 32% 12%
Portugal 37% 38% 44% 49% 54% 5%
Spain 34% 38% 45% 45% 46% 4%
Denmark 51% 50% 54% 57% 58% 2%
Finland 46% 43% 45% 48% 50% 1%
Germany 52% 57% 57% 60% 61% 2%
Netherlands 51% 47% 46% 53% 58% 2%
Romania 4% 7% 10% 15% 20% 22%
United Kingdom 46% 44% 42% 42% 39% -2%
Average 14 MS 37% 38% 41% 43% 45% 4%
Austria 62% 62% 66% 66% 67% 1%
Bulgaria 5% 7% 10% 17% 19% 18%
Croatia 17% 20% 31% 40% 44% 13%
Cyprus 10% 14% 17% 21% 25% 12%
Estonia 48% 59% 73% 71% 76% 6%
Greece 18% 21% 25% 26% 28% 6%
Ireland 36% 37% 37% 37% 39% 1%
Latvia 27% 33% 39% 43% 43% 6%
Lithuania 49% 49% 53% 58% 60% 3%
Luxembourg 54% 55% 55% 56% 54% 0%
Slovakia 19% 31% 39% 41% 42% 10%
Slovenia 34% 43% 48% 51% 51% 5%
Sweden 52% 51% 48% 49% 50% 0%
Source: Planet Retail data
299
Retailer concentration - Retail Group Level - C5 Edible Grocery

2004 - 2012 Evolution
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Ireland
Hungary
Denmark
Finland
Netherlands
Greece
Belgium
Romania
Slovenia
Italy
Spain
Croatia
Cyprus
Latvia
Luxembourg
Sweden
Estonia
Czech Republic
France
Poland
Portugal
Germany
Slovakia
Austria
Lithuania
United Kingdom
Bulgaria
2012
Retailer concentration - Retail Group Level - HHI edible grocery 2004 2006 2008 2010 2012 CAGR
Belgium 622 720 738 776 805 3%

Czech Republic 173 243 9 390 479 14%
France 770 785 808 840 828 1%
Hungary 176 211 240 241 241 4%
Italy 88 91 103 114 113 3%
Poland 50 65 126 198 300 25%
Portugal 314 339 468 633 756 12%
Spain 333 422 587 625 717 10%
Denmark 751 716 846 886 900 2%
Finland 654 699 773 872 968 5%
Germany 621 724 731 852 897 5%
Netherlands 986 848 816 817 1009 0%
Romania 5 13 26 53 121 49%
United Kingdom 495 483 460 438 391 -3%
Average 14 MS 431 454 481 553 609 9%
Austria 1022 1060 1276 1271 1310 3%
Bulgaria 7 13 24 65 92 38%
Croatia 70 111 265 429 545 29%
Cyprus 64 76 102 153 180 14%
Estonia 734 892 1235 1144 1298 7%
Greece 86 106 153 142 169 9%
Ireland 374 380 387 353 365 0%
Latvia 227 393 559 651 669 14%
Lithuania 673 658 681 872 943 4%
Luxembourg 1156 1127 1059 990 936 -3%
Slovakia 93 211 344 378 415 21%
Slovenia 377 547 574 625 598 6%
Sweden 955 896 841 868 885 -1%
300
Annexes
Retailer concentration - Retail Group Level - HHI Edible Grocery

1 400
1 200
1 000
800
600
400
200
2012
301
Retailer concentration - Retail Group Level - C5 Modern Retail 2004 2006 2008 2010 2012 CAGR
Belgium 94% 94% 93% 93% 94% 0%

Czech Republic 69% 76% 85% 85% 85% 3%
France 79% 79% 79% 79% 78% 0%
Hungary 69% 70% 71% 67% 68% 0%
Italy 72% 70% 69% 69% 68% -1%
Poland 53% 59% 72% 72% 74% 4%
Portugal 86% 85% 86% 85% 85% 0%
Spain 69% 70% 76% 75% 72% 1%
Denmark 90% 93% 92% 94% 94% 0%
Finland 97% 100% 100% 100% 100% 0%
Germany 77% 85% 86% 90% 90% 2%
Netherlands 89% 88% 78% 84% 91% 0%
Romania 92% 81% 77% 76% 79% -2%
United Kingdom 86% 83% 83% 85% 85% 0%
Average 14 MS 80% 81% 82% 82% 83% 1%
Austria 92% 91% 95% 95% 95% 0%
Bulgaria 99% 90% 92% 86% 86% -2%
Croatia 85% 78% 78% 87% 87% 0%
Cyprus 100% 100% 100% 100% 100% 0%
Estonia 98% 100% 100% 100% 100% 0%
Greece 82% 82% 84% 86% 88% 1%
Ireland 95% 94% 94% 95% 100% 1%
Latvia 100% 98% 99% 97% 97% 0%
Lithuania 100% 92% 100% 99% 98% 0%
Luxembourg 99% 99% 96% 94% 94% -1%
Slovakia 82% 90% 94% 95% 95% 2%
Slovenia 100% 98% 94% 93% 93% -1%
Sweden 99% 97% 96% 96% 96% 0%
Retailer concentration - Retail Group Level - C5 Modern Retail

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
2012
302
Annexes
Retailer concentration - Retail Group Level - HHI Modern Retail 2004 2006 2008 2010 2012 CAGR
Belgium 2 116 2 062 1 992 1 998 2 023 -1%

Czech Republic 1 199 1 387 1 690 1 701 1 779 5%
France 1 533 1 528 1 492 1 482 1 410 -1%
Hungary 1 251 1 243 1 308 1 198 1 229 0%
Italy 1 299 1 220 1 188 1 192 1 170 -1%
Poland 826 926 1 228 1 353 1 580 8%
Portugal 1 681 1 652 1 830 1 888 1 901 2%
Spain 1 335 1 422 1 686 1 735 1 701 3%
Denmark 2374 2481 2458 2385 2320 0%
Finland 2881 3736 3751 3862 3935 4%
Germany 1059 1266 1307 1604 1648 6%
Netherlands 2972 2893 2279 2043 2478 -2%
Romania 2302 1572 1394 1361 1880 -3%
United Kingdom 1749 1745 1793 1817 1811 0%
Average 14 MS 1 756 1 795 1 814 1 830 1 919 1%
Austria 2262 2263 2615 2598 2617 2%
Bulgaria 2943 2047 1959 1646 1907 -5%
Croatia 1834 1622 1620 1986 2088 2%
Cyprus 6530 4049 3634 3572 2879 -10%
Estonia 2981 2522 2308 2246 2225 -4%
Greece 1708 1648 1681 1603 1682 0%
Ireland 2582 2511 2451 2294 2381 -1%
Latvia 3076 3460 3590 3244 3443 1%
Lithuania 2796 2282 2451 2525 2543 -1%
Luxembourg 3499 3343 2998 2704 2730 -3%
Slovakia 1659 1772 1964 2035 2127 3%
Slovenia 3183 2838 2216 2077 2015 -6%
Sweden 3418 3261 3386 3359 3305 0%
Retailer concentration - Retail Group Level - HHI Modern Retail

7 000
6 000
5 000
4 000
3 000
2 000
1 000
2012
303
11.2.5. Supplier concentration
Belgium 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 2 628 2 626 2 860 2 830 2 867 1,1%

Biscuits 2 189 2 319 1 745 2 256 2 275 0,5%
Bread 12 15 16 16 16 4,0%
Butter/margarine 1 310 1 344 1 228 1 220 1 289 -0,2%
Canned vegetables 1 713 2 190 2 643 3 078 3 109 7,7%
Cereals 5 443 5 465 5 198 5 141 4 895 -1,3%
Cheese 208 249 264 273 238 1,7%
Chocolate 1 981 1 923 2 157 2 108 2 171 1,2%
Coffee 2 574 2 569 2 967 3 248 3 423 3,6%
Desserts 2 254 1 938 1 922 1 841 1 824 -2,6%
Edible oil 2 198 2 075 1 609 1 667 1 668 -3,4%
Frozen pizzas/starters 1 407 1 791 1 489 1 795 1 740 2,7%
Frozen ready cooked meals 3 496 3 975 4 223 4 247 3 577 0,3%
Frozen vegetables 2 472 2 526 2 486 2 477 2 250 -1,2%
Fruit Juices 665 683 688 688 643 -0,4%
Ham/Delicatessen 1 545 1 890 2 218 2 159 1 897 2,6%
Ice Cream 2 341 2 505 2 681 2 908 2 738 2,0%
Milk 2 486 2 345 3 295 3 051 2 976 2,3%
Mineral water 1 643 1 521 1 523 1 680 1 810 1,2%
Savoury snacks 2 544 3 627 4 123 4 349 4 448 7,2%
Soft drinks 4 159 4 076 3 895 3 688 3 675 -1,5%
Tea 1 245 1 367 1 457 1 490 1 738 4,3%
Yoghurt 1 697 2 495 2 781 2 915 2 485 4,9%
Average 2096,1 2239,7 2324,7 2396,8 2337,1
Source: EY analysis based on Euromonitor Passport data
Supplier concentration - Belgium - HHI (brand only)

6 000
5 000
4 000
3 000
2 000
1 000
2012
304
Annexes
Czech Republic 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 2 156 2 179 2 139 2 172 2 152 0,0%

Biscuits 6 563 6 252 6 386 6 188 6 184 -0,7%
Bread 98 165 227 227 376 18,4%
Butter/margarine 929 1 043 1 027 993 991 0,8%
Canned vegetables 1 941 1 492 1 593 1 551 1 457 -3,5%
Cereals 2 026 2 073 2 309 2 354 2 308 1,6%
Cheese 789 792 856 891 920 1,9%
Chocolate 2 857 2 726 2 876 2 874 2 909 0,2%
Coffee 1 783 1 702 1 649 1 752 1 737 -0,3%
Desserts 501 542 546 566 640 3,1%
Edible oil 2 479 2 822 3 030 3 023 3 333 3,8%
Frozen vegetables 1 217 1 371 1 421 1 456 1 453 2,2%
Fruit Juices 1 026 1 018 1 116 1 196 1 203 2,0%
Ice Cream 720 678 775 777 777 1,0%
Milk 839 916 996 951 1 139 3,9%
Mineral water 2 425 3 717 3 514 3 193 2 998 2,7%
Savoury snacks 1 985 1 900 1 776 2 357 2 137 0,9%
Soft drinks 1 080 1 063 951 1 009 1 066 -0,2%
Tea 1 029 895 820 848 877 -2,0%
Yoghurt 1 439 1 121 1 139 1 185 1 064 -3,7%
Average 1700,4 1800,8 2042,2 2056,7 2056,8
Supplier concentration - Czech Republic - HHI (brand only)

9 000
8 000
7 000
6 000
5 000
4 000
3 000
2 000
1 000
2012
305
France 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 2 578 2 507 2 546 2 672 2 690 0,5%

Biscuits 2 682 2 993 3 208 3 479 3 584 3,7%
Bread 24 23 21 21 23 -0,8%
Butter/margarine 1 139 1 154 1 291 1 316 1 329 1,9%
Cereals 3 275 3 305 3 318 3 271 3 182 -0,4%
Cheese 1 216 1 487 1 523 1 569 1 491 2,6%
Chocolate 832 870 896 958 1 002 2,3%
Coffee 1 980 2 016 2 057 2 061 2 098 0,7%
Desserts 1 347 1 167 1 038 930 916 -4,7%
Edible oil 2 832 3 613 3 793 4 343 3 867 4,0%
Frozen ready cooked meals 2 288 1 670 1 452 1 490 1 444 -5,6%
Fruit Juices 754 901 957 1 029 1 124 5,1%
Ham/Delicatessen 698 803 707 1 019 871 2,8%
Ice Cream 1 789 1 715 1 813 1 983 2 074 1,9%
Milk 2 322 2 407 2 318 2 661 2 390 0,4%
Mineral water 1 927 2 120 2 110 2 148 2 415 2,9%
Savoury snacks 978 1 332 1 641 1 741 1 809 8,0%
Soft drinks 3 057 3 294 3 435 3 590 3 603 2,1%
Tea 2 273 2 339 2 294 2 334 2 314 0,2%
Yoghurt 1 975 2 920 3 112 3 107 3 197 6,2%
Average 1838,8 1955,1 1998,8 2122,9 2130,1
Supplier concentration - France - HHI (brand only)

4 500
4 000
3 500
3 000
2 500
2 000
1 500
1 000
500
2012
306
Annexes
Hungary 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 2 261 2 261 2 135 2 171 2 067 -1,1%

Biscuits 5 086 4 900 3 900 4 127 3 976 -3,0%
Bread 2 2 3 3 3 6,4%
Canned vegetables 506 297 177 255 290 -6,7%
Cereals 2 444 2 744 2 660 2 594 2 727 1,4%
Cheese 1 440 1 799 1 681 1 645 1 709 2,2%
Chocolate 1 469 1 446 1 371 1 374 1 349 -1,1%
Coffee 2 131 2 036 2 039 2 043 2 101 -0,2%
Desserts 1 658 1 752 1 467 1 493 1 490 -1,3%
Edible oil 6 326 7 362 6 556 5 802 6 056 -0,5%
Frozen ready cooked meals 1 363 1 342 1 152 1 363 683 -8,3%
Frozen vegetables 838 1 130 976 986 939 1,4%
Fruit Juices 656 843 1 287 1 400 1 474 10,7%
Ham/Delicatessen 818 1 863 2 224 2 373 2 179 13,0%
Ice Cream 3 363 2 991 2 184 2 181 2 259 -4,9%
Milk 1 895 2 696 2 112 1 757 1 784 -0,8%
Mineral water 1 091 1 073 900 1 013 858 -3,0%
Savoury snacks 1 069 1 060 1 249 1 308 1 225 1,7%
Soft drinks 2 142 2 133 1 853 1 800 1 618 -3,4%
Tea 1 683 1 785 1 891 1 656 1 629 -0,4%
Yoghurt 1 410 1 551 1 553 1 572 1 739 2,7%
Average 1964,0 2106,8 2035,1 2055,0 2016,6
Supplier concentration - Hungary - HHI (brand only)

7 000
6 000
5 000
4 000
3 000
2 000
1 000
2012
307
Italy 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 2 956 3 213 3 280 3 339 3 358 1,6%

Biscuits 1 907 2 144 2 171 2 265 2 285 2,3%
Bread 21 21 26 33 39 8,3%
Canned vegetables 319 339 358 385 397 2,8%
Cereals 4 161 4 335 4 457 4 258 4 328 0,5%
Cheese 131 194 185 184 205 5,8%
Chocolate 1 712 1 733 1 842 1 846 1 806 0,7%
Coffee 2 206 2 252 2 261 2 228 1 981 -1,3%
Desserts 840 994 1 311 1 460 1 747 9,6%
Edible oil 534 540 928 844 931 7,2%
Fruit Juices 1 454 1 509 1 531 1 526 1 428 -0,2%
Ham/Delicatessen 239 239 225 263 289 2,4%
Ice Cream 533 428 383 297 245 -9,3%
Milk 1 613 1 704 1 401 1 290 1 240 -3,2%
Mineral water 1 426 1 388 1 370 1 327 1 494 0,6%
Savoury snacks 550 609 648 690 692 2,9%
Soft drinks 1 394 1 537 1 617 1 642 1 725 2,7%
Tea 1 224 1 281 1 290 1 267 1 319 0,9%
Yoghurt 1 916 1 792 1 720 1 779 1 829 -0,6%
Average 1406,6 1461,4 1499,9 1519,1 1590,2
Supplier concentration - Italy - HHI (brand only)

5 000
4 500
4 000
3 500
3 000
2 500
2 000
1 500
1 000
500
2012
308
Annexes
Poland 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 2 688 2 834 3 270 3 276 3 241 2,4%

Biscuits 2 150 2 329 2 293 2 327 2 209 0,3%
Bread 8 4 6 9 17 9,4%
Butter/margarine 303 306 374 443 644 9,9%
Cereals 4 748 4 550 4 253 4 265 4 298 -1,2%
Cheese 977 932 949 1 057 1 090 1,4%
Chocolate 1 055 1 099 1 109 1 113 1 044 -0,1%
Coffee 1 408 1 577 1 615 1 688 1 731 2,6%
Desserts 610 664 811 857 989 6,2%
Edible oil 1 599 1 528 2 825 3 252 3 364 9,7%
Frozen pizzas/starters 2 928 2 760 2 682 2 624 2 553 -1,7%
Frozen ready cooked meals 936 978 954 926 910 -0,3%
Fruit Juices 1 707 2 467 2 482 2 690 2 166 3,0%
Ice Cream 1 350 1 270 1 329 1 278 1 226 -1,2%
Milk 881 1 102 1 618 1 876 1 843 9,7%
Mineral water 617 872 980 1 084 1 257 9,3%
Savoury snacks 914 962 1 062 1 066 1 070 2,0%
Soft drinks 743 738 722 751 821 1,3%
Tea 965 1 353 1 541 1 922 2 228 11,0%
Yoghurt 1 207 1 378 1 543 1 764 1 973 6,3%
Average 1440,0 1528,9 1648,4 1724,6 1743,4
Supplier concentration - Poland - HHI (brand only)

5 000
4 500
4 000
3 500
3 000
2 500
2 000
1 500
1 000
500
2012
309
Portugal 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 2 787 2 770 3 153 3 348 3 341 2,3%

Biscuits 890 1 304 1 432 1 521 1 503 6,8%
Bread 35 38 35 32 30 -1,9%
Cereals 2 804 3 530 3 564 3 894 3 995 4,5%
Cheese 835 936 1 086 1 312 1 321 5,9%
Chocolate 1 207 1 073 1 052 1 163 1 145 -0,7%
Coffee 2 687 2 870 3 002 2 847 3 198 2,2%
Desserts 1 988 1 322 1 336 1 312 1 406 -4,2%
Edible oil 2 025 2 012 2 158 2 563 2 514 2,7%
Fruit Juices 1 462 1 528 1 316 1 976 1 858 3,0%
Ice Cream 3 737 4 108 4 715 5 092 5 066 3,9%
Milk 2 894 2 630 2 437 2 247 3 353 1,9%
Mineral water 670 842 775 736 737 1,2%
Savoury snacks 1 680 1 652 1 592 1 835 1 953 1,9%
Soft drinks 1 610 1 663 1 850 1 971 2 135 3,6%
Tea 1 728 1 784 1 704 1 806 1 998 1,8%
Yoghurt 1 887 1 897 1 841 1 957 2 029 0,9%
Average 2122,8 2166,7 2289,1 2339,4 2426,5
Supplier concentration - Portugal - HHI (brand only)

7 000
6 000
5 000
4 000
3 000
2 000
1 000
2012
310
Annexes
Spain 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 1 651 1 681 1 954 1 832 1 781 1,0%

Biscuits 1 109 1 455 1 540 1 598 1 638 5,0%
Bread 22 29 31 33 50 10,7%
Canned vegetables 814 1 042 1 196 1 186 1 199 5,0%
Cereals 3 683 3 743 3 780 3 431 3 983 1,0%
Cheese 386 446 435 591 606 5,8%
Chocolate 1 298 1 345 1 364 1 495 1 511 1,9%
Coffee 3 376 3 694 4 031 4 546 4 804 4,5%
Desserts 2 930 2 330 2 151 2 030 2 060 -4,3%
Edible oil 1 157 1 307 1 416 1 483 1 739 5,2%
Frozen pizzas/starters 1 011 881 860 760 778 -3,2%
Fruit Juices 1 089 1 110 1 092 1 377 1 510 4,2%
Ice Cream 2 992 3 034 3 066 3 287 3 536 2,1%
Milk 1 448 1 466 1 321 1 749 1 747 2,4%
Mineral water 884 866 900 896 819 -1,0%
Savoury snacks 2 362 2 488 2 639 3 004 3 139 3,6%
Soft drinks 4 311 4 226 4 293 4 381 4 545 0,7%
Tea 1 383 1 520 1 515 1 489 1 448 0,6%
Yoghurt 3 468 5 333 5 712 5 841 6 309 7,8%
Average 1776,1 1914,4 1957,7 2018,3 2179,2
Supplier concentration - Spain - HHI (brand only)

7 000
6 000
5 000
4 000
3 000
2 000
1 000
2012
311
Finland 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 4 064 3 872 4 254 4 315 3 719 -1,1%

Biscuits 5 087 5 301 5 406 5 558 3 150 -5,8%
Bread 1 118 1 126 1 187 1 297 1 315 2,1%
Canned vegetables 2 107 434 450 445 440 -17,8%
Cereals 1 247 1 439 1 360 1 406 1 408 1,5%
Cheese 3 014 3 239 3 604 3 644 3 842 3,1%
Chocolate 2 971 2 978 3 032 3 154 3 021 0,2%
Coffee 4 767 4 827 4 929 4 969 2 871 -6,1%
Desserts 2 843 2 584 2 576 2 325 2 509 -1,6%
Edible oil 1 567 1 500 1 634 1 683 1 759 1,5%
Fruit Juices 1 940 2 087 2 118 2 510 2 426 2,8%
Ice Cream 3 569 3 655 3 628 3 541 4 459 2,8%
Milk 5 331 4 925 4 439 4 303 4 192 -3,0%
Mineral water 2 280 2 255 2 342 2 486 2 445 0,9%
Savoury snacks 3 096 2 920 2 902 2 386 2 500 -2,6%
Soft drinks 1 945 1 781 1 679 1 647 1 591 -2,5%
Tea 2 272 2 462 2 500 2 448 2 471 1,1%
Yoghurt 4 316 4 378 4 191 3 902 3 742 -1,8%
Average 2792,1 2729,2 2767,6 2741,0 2594,2
Supplier concentration - Finland - HHI (brand only)

6 000
5 000
4 000
3 000
2 000
1 000
2012
312
Annexes
Germany 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 2 266 2 257 2 236 2 219 2 249 -0,1%

Biscuits 1 806 2 135 2 059 1 718 1 690 -0,8%
Bread 277 320 310 308 308 1,3%
Butter/margarine 894 954 932 985 972 1,1%
Cereals 1 821 1 757 1 664 1 675 1 677 -1,0%
Cheese 354 433 424 444 447 3,0%
Chocolate 1 108 1 225 1 242 1 264 1 268 1,7%
Coffee 1 338 1 360 1 367 1 327 1 263 -0,7%
Desserts 413 559 631 648 580 4,3%
Edible oil 2 263 1 365 1 009 1 184 1 130 -8,3%
Fruit Juices 354 434 522 597 599 6,8%
Ice Cream 1 527 1 605 1 604 1 576 1 585 0,5%
Milk 441 561 686 745 915 9,6%
Mineral water 390 418 388 397 378 -0,4%
Savoury snacks 1 340 1 309 1 219 1 742 1 747 3,4%
Soft drinks 1 626 1 679 1 611 1 432 1 389 -2,0%
Tea 1 038 1 163 1 260 1 262 1 272 2,6%
Yoghurt 914 930 1 000 1 009 935 0,3%
Average 1202,3 1226,3 1268,4 1383,5 1359,4
Supplier concentration - Germany - HHI (brand only)

3 000
2 500
2 000
1 500
1 000
500
2012
313
Netherlands 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 4 232 4 041 4 565 4 357 4 295 0,2%

Biscuits 1 132 1 020 1 290 1 537 1 565 4,1%
Bread 186 168 173 159 138 -3,7%
Cereals 1 087 1 088 1 100 1 194 1 134 0,5%
Cheese 393 424 683 611 625 6,0%
Chocolate 1 362 1 392 1 418 1 443 1 540 1,5%
Coffee 3 965 3 943 4 569 5 059 4 985 2,9%
Desserts 2 236 1 954 3 407 3 533 3 440 5,5%
Edible oil 1 280 1 534 1 921 2 034 2 266 7,4%
Fruit Juices 1 765 1 617 1 366 1 501 1 415 -2,7%
Ham/Delicatessen 433 438 402 603 396 -1,1%
Ice Cream 4 941 5 152 5 463 6 999 7 305 5,0%
Milk 2 554 2 459 4 593 3 255 3 336 3,4%
Mineral water 4 824 4 191 3 548 2 915 2 266 -9,0%
Savoury snacks 1 913 3 259 3 666 3 696 3 518 7,9%
Soft drinks 1 333 1 204 1 295 1 249 1 207 -1,2%
Tea 4 750 4 017 3 419 3 276 3 390 -4,1%
Yoghurt 2 385 1 854 2 714 1 829 1 698 -4,2%
Average 2575,2 2635,5 2926,2 2889,6 2838,9
Supplier concentration - Netherlands - HHI (brand only)

9 000
8 000
7 000
6 000
5 000
4 000
3 000
2 000
1 000
2012
314
Annexes
Denmark 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 3 391 3 173 2 915 2 793 2 887 -2,0%

Biscuits 2 428 2 553 2 582 2 576 2 214 -1,1%
Bread 1 133 1 167 1 178 1 015 895 -2,9%
Canned vegetables 452 448 540 672 858 8,3%
Cereals 1 861 1 664 2 010 2 134 2 356 3,0%
Cheese 4 136 4 357 4 026 3 987 3 843 -0,9%
Chocolate 1 495 1 505 1 554 1 512 1 608 0,9%
Coffee 2 046 2 162 2 152 2 151 2 087 0,2%
Desserts 5 278 5 327 5 331 5 079 5 462 0,4%
Edible oil 688 770 778 765 826 2,3%
Fruit Juices 3 694 3 416 3 587 3 399 3 237 -1,6%
Ice Cream 1 740 1 739 1 705 2 073 1 994 1,7%
Milk 5 962 5 236 5 959 5 752 5 525 -0,9%
Mineral water 3 009 2 789 2 934 3 102 2 956 -0,2%
Savoury snacks 2 909 2 645 2 419 3 319 3 489 2,3%
Soft drinks 1 769 1 760 1 865 1 870 1 912 1,0%
Tea 2 367 2 399 2 280 2 181 2 161 -1,1%
Yoghurt 7 085 7 051 6 712 6 806 6 664 -0,8%
Average 2661,8 2667,6 2705,6 2779,1 2840,3
Supplier concentration - Denmark - HHI (brand only)

8 000
7 000
6 000
5 000
4 000
3 000
2 000
1 000
2012
315
Romania 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 2 782 2 512 2 530 2 563 2 559 -1,0%

Biscuits 562 544 590 592 662 2,1%
Bread 8 20 42 54 50 25,6%
Butter/margarine 883 1 144 1 013 1 041 1 122 3,0%
Canned vegetables 810 1 009 1 020 1 062 1 134 4,3%
Cereals 1 733 1 557 1 413 1 437 1 487 -1,9%
Cheese 909 1 173 1 230 1 119 1 036 1,6%
Chocolate 1 801 1 654 1 372 1 418 1 673 -0,9%
Coffee 1 801 1 644 2 134 2 328 2 283 3,0%
Desserts 558 906 1 034 918 846 5,3%
Edible oil 1 788 1 551 2 003 1 613 1 644 -1,0%
Frozen pizzas/starters 666 579 549 709 990 5,1%
Frozen vegetables 813 775 546 544 532 -5,2%
Fruit Juices 1 691 1 577 1 328 1 398 1 289 -3,3%
Ham/Delicatessen 1 960 2 106 1 375 1 169 1 124 -6,7%
Ice Cream 1 236 1 447 1 268 1 246 1 241 0,0%
Milk 669 974 983 770 764 1,7%
Mineral water 1 330 1 540 1 638 1 627 1 427 0,9%
Savoury snacks 1 930 1 884 1 583 1 705 1 713 -1,5%
Soft drinks 2 999 2 474 2 388 2 469 2 677 -1,4%
Tea 1 104 1 404 1 493 1 630 1 454 3,5%
Yoghurt 2 230 1 960 2 051 2 161 2 483 1,4%
Average 1750,6 1758,1 1721,0 1720,5 1747,3
Supplier concentration - Romania - HHI (brand only)

12 000
10 000
8 000
6 000
4 000
2 000
2012
316
Annexes
United Kingdom 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food 1 818 1 888 2 028 2 096 2 180 2,3%

Biscuits 881 921 933 997 1 035 2,0%
Bread 555 752 1 001 995 1 028 8,0%
Cereals 1 565 1 430 1 416 1 341 1 274 -2,5%
Cheese 201 206 219 250 244 2,4%
Chocolate 1 801 1 754 1 751 1 951 1 908 0,7%
Coffee 4 070 3 812 3 580 3 227 2 741 -4,8%
Desserts 908 844 802 742 681 -3,5%
Edible oil 1 428 1 295 1 181 993 1 134 -2,8%
Fruit Juices 472 595 785 890 905 8,5%
Ham/Delicatessen 394 356 375 406 435 1,3%
Ice Cream 1 979 1 939 1 879 2 027 2 218 1,4%
Milk 1 828 1 657 1 433 1 276 1 187 -5,3%
Mineral water 2 857 2 910 2 751 2 801 2 581 -1,3%
Savoury snacks 2 361 2 212 2 249 2 132 2 189 -0,9%
Soft drinks 1 987 1 962 2 039 1 964 1 871 -0,7%
Tea 1 733 1 834 1 728 1 674 1 621 -0,8%
Yoghurt 1 977 1 774 1 952 2 116 1 844 -0,9%
Average 1716,6 1707,1 1714,8 1794,6 1766,0
Supplier concentration - United Kingdom - HHI (brand only)

6 000
5 000
4 000
3 000
2 000
1 000
2012
317
CAGR CAGR CAGR

Supplier concentration HHI 2004 2006 2008 2010 2012 2004-2012 2004-2008 2008-2012
Baby food
Belgium 2628 2626 2860 2830 2867 1,1% 2,1% 0,1%
Czech Republic 2156 2179 2139 2172 2152 0,0% -0,2% 0,1%
Denmark 3391 3173 2915 2793 2887 -2,0% -3,7% -0,2%
Finland 4064 3872 4254 4315 3719 -1,1% 1,1% -3,3%
France 2578 2507 2546 2672 2690 0,5% -0,3% 1,4%
Germany 2266 2257 2236 2219 2249 -0,1% -0,3% 0,1%
Hungary 2261 2261 2135 2171 2067 -1,1% -1,4% -0,8%
Italy 2956 3213 3280 3339 3358 1,6% 2,6% 0,6%
Netherlands 4232 4041 4565 4357 4295 0,2% 1,9% -1,5%
Poland 2688 2834 3270 3276 3241 2,4% 5,0% -0,2%
Portugal 2787 2770 3153 3348 3341 2,3% 3,1% 1,5%
Romania 2782 2512 2530 2563 2559 -1,0% -2,3% 0,3%
Spain 1651 1681 1954 1832 1781 1,0% 4,3% -2,3%
United Kingdom 1818 1888 2028 2096 2180 2,3% 2,8% 1,8%
Average 2733 2701 2848 2856 2813 0,4% 1,0% -0,3%
Supplier concentration Baby food - HHI Brand only

5000
4500
4000
3500
3000
2500
2000
1500
1000
500
2012
318
Annexes
CAGR CAGR CAGR

Biscuits
Belgium 2189 2319 1745 2256 2275 0,5% -5,5% 6,9%
Czech Republic 6563 6252 6386 6188 6184 -0,7% -0,7% -0,8%
Denmark 2428 2553 2582 2576 2214 -1,1% 1,5% -3,8%
Finland 5087 5301 5406 5558 3150 -5,8% 1,5% -12,6%
France 2682 2993 3208 3479 3584 3,7% 4,6% 2,8%
Germany 1806 2135 2059 1718 1690 -0,8% 3,3% -4,8%
Hungary 5086 4900 3900 4127 3976 -3,0% -6,4% 0,5%
Italy 1907 2144 2171 2265 2285 2,3% 3,3% 1,3%
Netherlands 1132 1020 1290 1537 1565 4,1% 3,3% 5,0%
Poland 2150 2329 2293 2327 2209 0,3% 1,6% -0,9%
Portugal 890 1304 1432 1521 1503 6,8% 12,6% 1,2%
Romania 562 544 590 592 662 2,1% 1,2% 2,9%
Spain 1109 1455 1540 1598 1638 5,0% 8,6% 1,5%
United Kingdom 881 921 933 997 1035 2,0% 1,4% 2,6%
Average 2462 2584 2538 2624 2426 -0,2% 0,8% -1,1%
Supplier concentration Biscuits - HHI Brand only

7000
6000
5000
4000
3000
2000
1000
2012
319
CAGR CAGR CAGR

Bread
Belgium 12 15 16 16 16 4,0% 7,5% 0,6%
Czech Republic 98 165 227 227 376 18,4% 23,5% 13,5%
Denmark 1133 1167 1178 1015 895 -2,9% 1,0% -6,6%
Finland 1118 1126 1187 1297 1315 2,1% 1,5% 2,6%
France 24 23 21 21 23 -0,8% -3,4% 1,9%
Germany 277 320 310 308 308 1,3% 2,9% -0,2%
Hungary 2 2 3 3 3 6,4% 17,4% -3,5%
Italy 21 21 26 33 39 8,3% 6,2% 10,4%
Netherlands 186 168 173 159 138 -3,7% -1,8% -5,5%
Poland 8 4 6 9 17 9,4% -9,7% 32,6%
Portugal 35 38 35 32 30 -1,9% -0,2% -3,6%
Romania 8 20 42 54 50 25,6% 50,6% 4,7%
Spain 22 29 31 33 50 10,7% 9,3% 12,1%
United Kingdom 555 752 1001 995 1028 8,0% 15,9% 0,7%
Average 250 275 304 300 306 2,6% 5,0% 0,2%
Supplier concentration Bread - HHI Brand only

1400
1200
1000
800
600
400
200
2012
320
Annexes
CAGR CAGR CAGR

Butter/margarine
Belgium 1310 1344 1228 1220 1289 -0,2% -1,6% 1,2%
Czech Republic 929 1043 1027 993 991 0,8% 2,5% -0,9%
Denmark 2962 2978 2992 3067 3263 1,2% 0,3% 2,2%
Finland 1598 1661 1832 1882 2580 6,2% 3,5% 8,9%
France 1139 1154 1291 1316 1329 1,9% 3,2% 0,7%
Germany 894 954 932 985 972 1,1% 1,1% 1,1%
Hungary 2607 2475 2305 2409 2395 -1,1% -3,0% 1,0%
Italy 1065 1149 1118 1155 1428 3,7% 1,2% 6,3%
Netherlands 2249 2660 2783 3073 3392 5,3% 5,5% 5,1%
Poland 303 306 374 443 644 9,9% 5,4% 14,6%
Portugal 2745 2832 3043 3144 3044 1,3% 2,6% 0,0%
Romania 883 1144 1013 1041 1122 3,0% 3,5% 2,6%
Spain 1144 1207 1324 1216 1056 -1,0% 3,7% -5,5%
United Kingdom 2684 2724 3098 4960 5025 8,2% 3,7% 12,8%
Average 1608 1688 1740 1922 2038 3,0% 2,0% 4,0%
Supplier concentration Butter/margarine - HHI Brand only

6000
5000
4000
3000
2000
1000
2012
321
CAGR CAGR CAGR

Canned vegetables 1577 1531 1610 1603 1642
Belgium 1713 2190 2643 3078 3109 7,7% 11,4% 4,1%
Czech Republic 1941 1492 1593 1551 1457 -3,5% -4,8% -2,2%
Denmark 452 448 540 672 858 8,3% 4,5% 12,3%
Finland 2107 434 450 445 440 -17,8% -32,0% -0,5%
France 2149 2226 2296 2366 2512 2,0% 1,7% 2,3%
Germany 1298 1436 1503 2051 2294 7,4% 3,7% 11,1%
Hungary 506 297 177 255 290 -6,7% -23,1% 13,1%
Italy 319 339 358 385 397 2,8% 3,0% 2,6%
Netherlands 3276 3379 3064 3186 3126 -0,6% -1,7% 0,5%
Poland 2246 2256 2284 2275 2248 0,0% 0,4% -0,4%
Portugal 2232 2436 2721 1349 1388 -5,8% 5,1% -15,5%
Romania 810 1009 1020 1062 1134 4,3% 5,9% 2,7%
Spain 814 1042 1196 1186 1199 5,0% 10,1% 0,1%
United Kingdom 2217 2451 2695 2586 2530 1,7% 5,0% -1,6%
Average 1577 1531 1610 1603 1642 0,5% 0,5% 0,5%
Supplier concentration Canned vegetables - HHI Brand only

3500
3000
2500
2000
1500
1000
500
2012
322
Annexes
CAGR CAGR CAGR

Cereals
Belgium 5443 5465 5198 5141 4895 -1,3% -1,1% -1,5%
Czech Republic 2026 2073 2309 2354 2308 1,6% 3,3% 0,0%
Denmark 1861 1664 2010 2134 2356 3,0% 1,9% 4,1%
Finland 1247 1439 1360 1406 1408 1,5% 2,2% 0,9%
France 3275 3305 3318 3271 3182 -0,4% 0,3% -1,0%
Germany 1821 1757 1664 1675 1677 -1,0% -2,2% 0,2%
Hungary 2444 2744 2660 2594 2727 1,4% 2,1% 0,6%
Italy 4161 4335 4457 4258 4328 0,5% 1,7% -0,7%
Netherlands 1087 1088 1100 1194 1134 0,5% 0,3% 0,8%
Poland 4748 4550 4253 4265 4298 -1,2% -2,7% 0,3%
Portugal 2804 3530 3564 3894 3995 4,5% 6,2% 2,9%
Romania 1733 1557 1413 1437 1487 -1,9% -5,0% 1,3%
Spain 3683 3743 3780 3431 3983 1,0% 0,7% 1,3%
United Kingdom 1565 1430 1416 1341 1274 -2,5% -2,5% -2,6%
Average 2707 2763 2750 2743 2789 0,4% 0,4% 0,4%
Supplier concentration Cereals - HHI Brand only

6000
5000
4000
3000
2000
1000
2012
323
CAGR CAGR CAGR

Cheese
Belgium 208 249 264 273 238 1,7% 6,1% -2,6%
Czech Republic 789 792 856 891 920 1,9% 2,1% 1,8%
Denmark 4136 4357 4026 3987 3843 -0,9% -0,7% -1,2%
Finland 3014 3239 3604 3644 3842 3,1% 4,6% 1,6%
France 1216 1487 1523 1569 1491 2,6% 5,8% -0,5%
Germany 354 433 424 444 447 3,0% 4,6% 1,4%
Hungary 1440 1799 1681 1645 1709 2,2% 3,9% 0,4%
Italy 131 194 185 184 205 5,8% 8,9% 2,7%
Netherlands 393 424 683 611 625 6,0% 14,8% -2,2%
Poland 977 932 949 1057 1090 1,4% -0,7% 3,5%
Portugal 835 936 1086 1312 1321 5,9% 6,8% 5,0%
Romania 909 1173 1230 1119 1036 1,6% 7,8% -4,2%
Spain 386 446 435 591 606 5,8% 3,1% 8,6%
United Kingdom 201 206 219 250 244 2,4% 2,1% 2,7%
Average 1071 1191 1226 1256 1258 2,0% 3,4% 0,7%
Supplier concentration Cheese - HHI Brand only

4500
4000
3500
3000
2500
2000
1500
1000
500
2012
324
Annexes
CAGR CAGR CAGR

Chocolate
Belgium 1981 1923 2157 2108 2171 1,2% 2,2% 0,2%
Czech Republic 2857 2726 2876 2874 2909 0,2% 0,2% 0,3%
Denmark 1495 1505 1554 1512 1608 0,9% 1,0% 0,9%
Finland 2971 2978 3032 3154 3021 0,2% 0,5% -0,1%
France 832 870 896 958 1002 2,3% 1,9% 2,8%
Germany 1108 1225 1242 1264 1268 1,7% 2,9% 0,5%
Hungary 1469 1446 1371 1374 1349 -1,1% -1,7% -0,4%
Italy 1712 1733 1842 1846 1806 0,7% 1,8% -0,5%
Netherlands 1362 1392 1418 1443 1540 1,5% 1,0% 2,1%
Poland 1055 1099 1109 1113 1044 -0,1% 1,3% -1,5%
Portugal 1207 1073 1052 1163 1145 -0,7% -3,4% 2,1%
Romania 1801 1654 1372 1418 1673 -0,9% -6,6% 5,1%
Spain 1298 1345 1364 1495 1511 1,9% 1,2% 2,6%
United Kingdom 1801 1754 1751 1951 1908 0,7% -0,7% 2,2%
Average 1639 1623 1645 1691 1711 0,5% 0,1% 1,0%
Supplier concentration Chocolate - HHI Brand only

3500
3000
2500
2000
1500
1000
500
2012
325
CAGR CAGR CAGR

Coffee
Belgium 2574 2569 2967 3248 3423 3,6% 3,6% 3,6%
Czech Republic 1783 1702 1649 1752 1737 -0,3% -1,9% 1,3%
Denmark 2046 2162 2152 2151 2087 0,2% 1,3% -0,8%
Finland 4767 4827 4929 4969 2871 -6,1% 0,8% -12,6%
France 1980 2016 2057 2061 2098 0,7% 1,0% 0,5%
Germany 1338 1360 1367 1327 1263 -0,7% 0,5% -2,0%
Hungary 2131 2036 2039 2043 2101 -0,2% -1,1% 0,8%
Italy 2206 2252 2261 2228 1981 -1,3% 0,6% -3,2%
Netherlands 3965 3943 4569 5059 4985 2,9% 3,6% 2,2%
Poland 1408 1577 1615 1688 1731 2,6% 3,5% 1,7%
Portugal 2687 2870 3002 2847 3198 2,2% 2,8% 1,6%
Romania 1801 1644 2134 2328 2283 3,0% 4,3% 1,7%
Spain 3376 3694 4031 4546 4804 4,5% 4,5% 4,5%
United Kingdom 4070 3812 3580 3227 2741 -4,8% -3,2% -6,5%
Average 2581 2605 2739 2820 2664 0,4% 1,5% -0,7%
Supplier concentration Coffee - HHI Brand only

6000
5000
4000
3000
2000
1000
2012
326
Annexes
CAGR CAGR CAGR

Desserts
Belgium 2254 1938 1922 1841 1824 -2,6% -3,9% -1,3%
Czech Republic 501 542 546 566 640 3,1% 2,2% 4,1%
Denmark 27 42 5331 5079 5462 94,6% 276,2% 0,6%
Finland 2843 2584 2576 2325 2509 -1,6% -2,4% -0,7%
France 1347 1167 1038 930 916 -4,7% -6,3% -3,1%
Germany 413 559 631 648 580 4,3% 11,2% -2,1%
Hungary 1658 1752 1467 1493 1490 -1,3% -3,0% 0,4%
Italy 840 994 1311 1460 1747 9,6% 11,8% 7,4%
Netherlands 2236 1954 3407 3533 3440 5,5% 11,1% 0,2%
Poland 610 664 811 857 989 6,2% 7,4% 5,1%
Portugal 1988 1322 1336 1312 1406 -4,2% -9,5% 1,3%
Romania 558 906 1034 918 846 5,3% 16,7% -4,9%
Spain 2930 2330 2151 2030 2060 -4,3% -7,4% -1,1%
United Kingdom 908 844 802 742 681 -3,5% -3,1% -4,0%
Average 1365 1257 1740 1695 1756 3,2% 6,3% 0,2%
Supplier concentration Desserts - HHI Brand only

6000
5000
4000
3000
2000
1000
2012
327
CAGR CAGR CAGR

Edible oil
Belgium 2198 2075 1609 1667 1668 -3,4% -7,5% 0,9%
Czech Republic 2479 2822 3030 3023 3333 3,8% 5,1% 2,4%
Denmark 688 770 778 765 826 2,3% 3,1% 1,5%
Finland 1567 1500 1634 1683 1759 1,5% 1,1% 1,9%
France 2832 3613 3793 4343 3867 4,0% 7,6% 0,5%
Germany 2263 1365 1009 1184 1130 -8,3% -18,3% 2,9%
Hungary 6326 7362 6556 5802 6056 -0,5% 0,9% -2,0%
Italy 534 540 928 844 931 7,2% 14,8% 0,1%
Netherlands 1280 1534 1921 2034 2266 7,4% 10,7% 4,2%
Poland 1599 1528 2825 3252 3364 9,7% 15,3% 4,5%
Portugal 2025 2012 2158 2563 2514 2,7% 1,6% 3,9%
Romania 1788 1551 2003 1613 1644 -1,0% 2,9% -4,8%
Spain 1157 1307 1416 1483 1739 5,2% 5,2% 5,3%
United Kingdom 1428 1295 1181 993 1134 -2,8% -4,6% -1,0%
Average 2012 2091 2203 2232 2302 1,7% 2,3% 1,1%
Supplier concentration Edible oil - HHI Brand only

7000
6000
5000
4000
3000
2000
1000
2012
328
Annexes
CAGR CAGR CAGR

Belgium 1407 1791 1489 1795 1740 2,7% 1,4% 4,0%
Czech Republic 2590 3960 8743 8502 7904 15,0% 35,5% -2,5%
Denmark 1104 1338 1486 1696 1849 6,7% 7,7% 5,6%
Finland 1459 1512 1550 1607 1560 0,8% 1,5% 0,2%
France 2662 2564 2620 2972 3222 2,4% -0,4% 5,3%
Germany 2340 2276 2276 2472 2524 0,9% -0,7% 2,6%
Hungary 2913 2917 5133 5941 5830 9,1% 15,2% 3,2%
Italy 1686 1795 1839 1867 1797 0,8% 2,2% -0,6%
Netherlands 3811 4739 5396 5113 4851 3,1% 9,1% -2,6%
Poland 2928 2760 2682 2624 2553 -1,7% -2,2% -1,2%
Portugal 2130 1948 2999 3021 2979 4,3% 8,9% -0,2%
Romania 666 579 549 709 990 5,1% -4,7% 15,9%
Spain 1011 881 860 760 778 -3,2% -4,0% -2,5%
United Kingdom 1686 1769 1667 2005 2080 2,7% -0,3% 5,7%
Average 2028 2202 2806 2935 2904 4,6% 8,5% 0,9%
Supplier concentration Frozen pizzas/starters - HHI Brand only

9000
8000
7000
6000
5000
4000
3000
2000
1000
2012
329
CAGR CAGR CAGR

Frozen ready cooked meals
Belgium 3496 3975 4223 4247 3577 0,3% 4,8% -4,1%
Czech Republic 1604 1825 1760 1963 2460 5,5% 2,3% 8,7%
Denmark 2300 2325 2512 3192 3350 4,8% 2,2% 7,5%
Finland 2865 2874 2929 2998 3043 0,8% 0,6% 1,0%
France 2288 1670 1452 1490 1444 -5,6% -10,7% -0,1%
Germany 1522 1626 2073 2833 1958 3,2% 8,0% -1,4%
Hungary 1363 1342 1152 1363 683 -8,3% -4,1% -12,3%
Italy 2551 2508 2532 2658 3485 4,0% -0,2% 8,3%
Netherlands 7392 8403 8789 8251 7674 0,5% 4,4% -3,3%
Poland 936 978 954 926 910 -0,3% 0,5% -1,2%
Portugal 5799 5194 5188 5121 5367 -1,0% -2,7% 0,9%
Romania 10000 10000 10000 10000 10000 0,0% 0,0% 0,0%
Spain 1519 1680 1515 1452 1498 -0,2% -0,1% -0,3%
United Kingdom 1223 1207 1048 1171 1196 -0,3% -3,8% 3,4%
Average 3204 3258 3295 3405 3332 0,5% 0,7% 0,3%
Supplier concentration Frozen ready cooked meals - HHI Brand only

12000
10000
8000
6000
4000
2000
2012
330
Annexes
CAGR CAGR CAGR

Frozen vegetables
Belgium 2472 2526 2486 2477 2250 -1,2% 0,1% -2,5%
Czech Republic 1217 1371 1421 1456 1453 2,2% 3,9% 0,6%
Denmark 2036 2281 2324 2356 3112 5,4% 3,4% 7,6%
Finland 3078 3247 3410 2757 2772 -1,3% 2,6% -5,1%
France 1516 1542 1526 1764 1835 2,4% 0,2% 4,7%
Germany 1014 1111 1806 2139 2101 9,5% 15,5% 3,9%
Hungary 838 1130 976 986 939 1,4% 3,9% -1,0%
Italy 1914 1906 2005 2338 2527 3,5% 1,2% 6,0%
Netherlands 1731 1681 1680 1588 1733 0,0% -0,7% 0,8%
Poland 1463 1596 1579 1518 1549 0,7% 1,9% -0,5%
Portugal 2619 2811 2668 2831 2743 0,6% 0,5% 0,7%
Romania 813 775 546 544 532 -5,2% -9,5% -0,6%
Spain 1595 1697 1791 1728 2851 7,5% 2,9% 12,3%
United Kingdom 2856 2974 2831 2676 2712 -0,6% -0,2% -1,1%
Average 1797 1903 1932 1940 2079 1,8% 1,8% 1,9%
Supplier concentration Frozen vegetables - HHI Brand only

3500
3000
2500
2000
1500
1000
500
2012
331
CAGR CAGR CAGR

Fruit Juices
Belgium 665 683 688 688 643 -0,4% 0,8% -1,7%
Czech Republic 1026 1018 1116 1196 1203 2,0% 2,1% 1,9%
Denmark 3694 3416 3587 3399 3237 -1,6% -0,7% -2,5%
Finland 1940 2087 2118 2510 2426 2,8% 2,2% 3,5%
France 754 901 957 1029 1124 5,1% 6,1% 4,1%
Germany 354 434 522 597 599 6,8% 10,2% 3,5%
Hungary 656 843 1287 1400 1474 10,7% 18,4% 3,5%
Italy 1454 1509 1531 1526 1428 -0,2% 1,3% -1,7%
Netherlands 1765 1617 1366 1501 1415 -2,7% -6,2% 0,9%
Poland 1707 2467 2482 2690 2166 3,0% 9,8% -3,4%
Portugal 1462 1528 1316 1976 1858 3,0% -2,6% 9,0%
Romania 1691 1577 1328 1398 1289 -3,3% -5,9% -0,7%
Spain 1089 1110 1092 1377 1510 4,2% 0,1% 8,4%
United Kingdom 472 595 785 890 905 8,5% 13,6% 3,6%
Average 1338 1413 1441 1584 1520 1,6% 1,9% 1,3%
Supplier concentration Fruit juices - HHI Brand only

4000
3500
3000
2500
2000
1500
1000
500
2012
332
Annexes
CAGR CAGR CAGR

Ham
Belgium 1545 1890 2218 2159 1897 2,6% 9,5% -3,8%
Czech Republic 1034 1165 1323 1276 1222 2,1% 6,4% -2,0%
Denmark 1375 2272 2390 2422 2779 9,2% 14,8% 3,8%
Finland 1682 1716 1703 1782 1849 1,2% 0,3% 2,1%
France 698 803 707 1019 871 2,8% 0,3% 5,4%
Germany 1311 1291 1351 1794 1988 5,3% 0,7% 10,1%
Hungary 818 1863 2224 2373 2179 13,0% 28,4% -0,5%
Italy 239 239 225 263 289 2,4% -1,5% 6,4%
Netherlands 433 438 402 603 396 -1,1% -1,8% -0,4%
Poland 1617 1608 1632 1605 1626 0,1% 0,2% -0,1%
Portugal 2375 2654 2983 2729 2707 1,6% 5,9% -2,4%
Romania 1960 2106 1375 1169 1124 -6,7% -8,5% -4,9%
Spain 1215 1451 1100 1015 1518 2,8% -2,5% 8,4%
United Kingdom 394 356 375 406 435 1,3% -1,2% 3,8%
Average 1193 1418 1429 1472 1491 2,8% 4,6% 1,1%
Supplier concentration Ham - HHI Brand only

3000
2500
2000
1500
1000
500
2012
333
CAGR CAGR CAGR

Ice Cream
Belgium 2341 2505 2681 2908 2738 2,0% 3,4% 0,5%
Czech Republic 720 678 775 777 777 1,0% 1,8% 0,1%
Denmark 1740 1739 1705 2073 1994 1,7% -0,5% 4,0%
Finland 3569 3655 3628 3541 4459 2,8% 0,4% 5,3%
France 1789 1715 1813 1983 2074 1,9% 0,3% 3,4%
Germany 1527 1605 1604 1576 1585 0,5% 1,2% -0,3%
Hungary 3363 2991 2184 2181 2259 -4,9% -10,2% 0,8%
Italy 533 428 383 297 245 -9,3% -7,9% -10,5%
Netherlands 4941 5152 5463 6999 7305 5,0% 2,5% 7,5%
Poland 1350 1270 1329 1278 1226 -1,2% -0,4% -2,0%
Portugal 3737 4108 4715 5092 5066 3,9% 6,0% 1,8%
Romania 1236 1447 1268 1246 1241 0,0% 0,6% -0,5%
Spain 2992 3034 3066 3287 3536 2,1% 0,6% 3,6%
United Kingdom 1979 1939 1879 2027 2218 1,4% -1,3% 4,2%
Average 2273 2305 2321 2519 2623 1,8% 0,5% 3,1%
Supplier concentration Ice Cream - HHI Brand only

8000
7000
6000
5000
4000
3000
2000
1000
2012
334
Annexes
CAGR CAGR CAGR

Milk
Belgium 2486 2345 3295 3051 2976 2,3% 7,3% -2,5%
Czech Republic 839 916 996 951 1139 3,9% 4,4% 3,4%
Denmark 5962 5236 5959 5752 5525 -0,9% 0,0% -1,9%
Finland 5331 4925 4439 4303 4192 -3,0% -4,5% -1,4%
France 2322 2407 2318 2661 2390 0,4% 0,0% 0,8%
Germany 441 561 686 745 915 9,6% 11,7% 7,5%
Hungary 1895 2696 2112 1757 1784 -0,8% 2,7% -4,1%
Italy 1613 1704 1401 1290 1240 -3,2% -3,5% -3,0%
Netherlands 2554 2459 4593 3255 3336 3,4% 15,8% -7,7%
Poland 881 1102 1618 1876 1843 9,7% 16,4% 3,3%
Portugal 2894 2630 2437 2247 3353 1,9% -4,2% 8,3%
Romania 669 974 983 770 764 1,7% 10,1% -6,1%
Spain 1448 1466 1321 1749 1747 2,4% -2,3% 7,2%
United Kingdom 1828 1657 1433 1276 1187 -5,3% -5,9% -4,6%
Average 2226 2220 2399 2263 2314 0,5% 1,9% -0,9%
Supplier concentration Milk - HHI Brand only

7000
6000
5000
4000
3000
2000
1000
2012
335
CAGR CAGR CAGR

Mineral water
Belgium 1643 1521 1523 1680 1810 1,2% -1,9% 4,4%
Czech Republic 2425 3717 3514 3193 2998 2,7% 9,7% -3,9%
Denmark 3009 2789 2934 3102 2956 -0,2% -0,6% 0,2%
Finland 2280 2255 2342 2486 2445 0,9% 0,7% 1,1%
France 1927 2120 2110 2148 2415 2,9% 2,3% 3,4%
Germany 390 418 388 397 378 -0,4% -0,1% -0,7%
Hungary 1091 1073 900 1013 858 -3,0% -4,7% -1,2%
Italy 1426 1388 1370 1327 1494 0,6% -1,0% 2,2%
Netherlands 4824 4191 3548 2915 2266 -9,0% -7,4% -10,6%
Poland 617 872 980 1084 1257 9,3% 12,3% 6,4%
Portugal 670 842 775 736 737 1,2% 3,7% -1,2%
Romania 1330 1540 1638 1627 1427 0,9% 5,3% -3,4%
Spain 884 866 900 896 819 -1,0% 0,4% -2,3%
United Kingdom 2857 2910 2751 2801 2581 -1,3% -0,9% -1,6%
Average 1812 1893 1834 1815 1746 -0,5% 0,3% -1,2%
Supplier concentration Mineral water - HHI Brand only

6000
5000
4000
3000
2000
1000
2012
336
Annexes
CAGR CAGR CAGR

Savoury snacks
Belgium 2544 3627 4123 4349 4448 7,2% 12,8% 1,9%
Czech Republic 1985 1900 1776 2357 2137 0,9% -2,7% 4,7%
Denmark 2909 2645 2419 3319 3489 2,3% -4,5% 9,6%
Finland 3096 2920 2902 2386 2500 -2,6% -1,6% -3,7%
France 978 1332 1641 1741 1809 8,0% 13,8% 2,5%
Germany 1340 1309 1219 1742 1747 3,4% -2,3% 9,4%
Hungary 1069 1060 1249 1308 1225 1,7% 4,0% -0,5%
Italy 550 609 648 690 692 2,9% 4,2% 1,7%
Netherlands 1913 3259 3666 3696 3518 7,9% 17,7% -1,0%
Poland 914 962 1062 1066 1070 2,0% 3,8% 0,2%
Portugal 1680 1652 1592 1835 1953 1,9% -1,3% 5,2%
Romania 1930 1884 1583 1705 1713 -1,5% -4,8% 2,0%
Spain 2362 2488 2639 3004 3139 3,6% 2,8% 4,4%
United Kingdom 2361 2212 2249 2132 2189 -0,9% -1,2% -0,7%
Average 1831 1990 2055 2238 2259 2,7% 2,9% 2,4%
Supplier concentration Savoury snacks - HHI Brand only

5000
4500
4000
3500
3000
2500
2000
1500
1000
500
2012
337
CAGR CAGR CAGR

Soft drinks
Belgium 4159 4076 3895 3688 3675 -1,5% -1,6% -1,4%
Czech Republic 1080 1063 951 1009 1066 -0,2% -3,1% 2,9%
Denmark 1769 1760 1865 1870 1912 1,0% 1,3% 0,6%
Finland 1945 1781 1679 1647 1591 -2,5% -3,6% -1,3%
France 3057 3294 3435 3590 3603 2,1% 3,0% 1,2%
Germany 1626 1679 1611 1432 1389 -2,0% -0,2% -3,6%
Hungary 2142 2133 1853 1800 1618 -3,4% -3,6% -3,3%
Italy 1394 1537 1617 1642 1725 2,7% 3,8% 1,6%
Netherlands 1333 1204 1295 1249 1207 -1,2% -0,7% -1,7%
Poland 743 738 722 751 821 1,3% -0,7% 3,3%
Portugal 1610 1663 1850 1971 2135 3,6% 3,5% 3,6%
Romania 2999 2474 2388 2469 2677 -1,4% -5,5% 2,9%
Spain 4311 4226 4293 4381 4545 0,7% -0,1% 1,4%
United Kingdom 1987 1962 2039 1964 1871 -0,7% 0,7% -2,1%
Average 2154 2113 2107 2105 2131 -0,1% -0,6% 0,3%
Supplier concentration Soft drinks - HHI Brand only

5000
4500
4000
3500
3000
2500
2000
1500
1000
500
2012
338
Annexes
CAGR CAGR CAGR

Tea
Belgium 1245 1367 1457 1490 1738 4,3% 4,0% 4,5%
Czech Republic 1029 895 820 848 877 -2,0% -5,5% 1,7%
Denmark 2367 2399 2280 2181 2161 -1,1% -0,9% -1,3%
Finland 2272 2462 2500 2448 2471 1,1% 2,4% -0,3%
France 2273 2339 2294 2334 2314 0,2% 0,2% 0,2%
Germany 1038 1163 1260 1262 1272 2,6% 5,0% 0,2%
Hungary 1683 1785 1891 1656 1629 -0,4% 2,9% -3,7%
Italy 1224 1281 1290 1267 1319 0,9% 1,3% 0,6%
Netherlands 4750 4017 3419 3276 3390 -4,1% -7,9% -0,2%
Poland 965 1353 1541 1922 2228 11,0% 12,4% 9,7%
Portugal 1728 1784 1704 1806 1998 1,8% -0,3% 4,1%
Romania 1104 1404 1493 1630 1454 3,5% 7,9% -0,7%
Spain 1383 1520 1515 1489 1448 0,6% 2,3% -1,1%
United Kingdom 1733 1834 1728 1674 1621 -0,8% -0,1% -1,6%
Average 1771 1829 1799 1806 1851 0,6% 0,4% 0,7%
Supplier concentration Tea - HHI Brand only

5000
4500
4000
3500
3000
2500
2000
1500
1000
500
2012
339
CAGR CAGR CAGR

Yoghurt
Belgium 1697 2495 2781 2915 2485 4,9% 13,1% -2,8%
Czech Republic 1439 1121 1139 1185 1064 -3,7% -5,7% -1,7%
Denmark 7085 7051 6712 6806 6664 -0,8% -1,3% -0,2%
Finland 4316 4378 4191 3902 3742 -1,8% -0,7% -2,8%
France 1975 2920 3112 3107 3197 6,2% 12,0% 0,7%
Germany 914 930 1000 1009 935 0,3% 2,3% -1,6%
Hungary 1410 1551 1553 1572 1739 2,7% 2,4% 2,9%
Italy 1916 1792 1720 1779 1829 -0,6% -2,7% 1,6%
Netherlands 2385 1854 2714 1829 1698 -4,2% 3,3% -11,1%
Poland 1207 1378 1543 1764 1973 6,3% 6,3% 6,3%
Portugal 1887 1897 1841 1957 2029 0,9% -0,6% 2,5%
Romania 2230 1960 2051 2161 2483 1,4% -2,1% 4,9%
Spain 3468 5333 5712 5841 6309 7,8% 13,3% 2,5%
United Kingdom 1977 1774 1952 2116 1844 -0,9% -0,3% -1,4%
Average 2422 2602 2716 2710 2714 1,4% 2,9% 0,0%
Supplier concentration Yoghurt - HHI Brand only

8000
7000
6000
5000
4000
3000
2000
1000
2012
11.2.6. Measure of imbalance
340
Annexes
Belgium 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,094 -0,105 -0,157 -0,151 -0,151 6,1%

Biscuits -0,015 -0,051 0,057 -0,053 -0,051 16,8%
Bread 2,251 2,135 2,099 2,084 2,095 -0,9%
Butter/margarine 0,208 0,186 0,210 0,214 0,196 -0,8%
Canned vegetables 0,092 -0,026 -0,123 -0,188 -0,187 n.a
Cereals -0,410 -0,423 -0,417 -0,410 -0,384 -0,8%
Cheese 1,007 0,919 0,877 0,865 0,930 -1,0%
Chocolate 0,029 0,030 -0,035 -0,023 -0,031 n.a
Coffee -0,085 -0,095 -0,173 -0,211 -0,228 13,1%
Desserts -0,028 0,027 0,015 0,035 0,045 n.a
Edible oil -0,017 -0,003 0,093 0,079 0,084 n.a
Frozen pizzas/starters 0,177 0,061 0,126 0,046 0,065 -11,7%
Frozen ready cooked meals -0,218 -0,285 -0,326 -0,328 -0,247 1,6%
Frozen vegetables -0,068 -0,088 -0,096 -0,093 -0,046 -4,7%
Fruit Juices 0,503 0,480 0,462 0,463 0,498 -0,1%
Ham/Delicatessen 0,137 0,038 -0,047 -0,034 0,028 -18,0%
Ice Cream -0,044 -0,085 -0,129 -0,163 -0,132 14,7%
Milk -0,070 -0,056 -0,219 -0,184 -0,168 11,5%
Mineral water 0,110 0,132 0,116 0,075 0,048 -9,8%
Savoury snacks -0,080 -0,245 -0,316 -0,338 -0,342 19,9%
Soft drinks -0,294 -0,296 -0,291 -0,266 -0,259 -1,5%
Tea 0,230 0,178 0,136 0,127 0,066 -14,5%
Yoghurt 0,096 -0,083 -0,145 -0,164 -0,089 n.a
Average 0,149 0,102 0,075 0,060 0,076
Source: EY analysis based on Euromonitor Passport data and Source: Planet Retail data
Measure of imbalance brand only - Belgium

2,0
1,5
1,0
0,5
0,0
-0,5
2012
341
Czech Republic 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,255 -0,196 -0,102 -0,106 -0,082 -13,1%

Biscuits -0,738 -0,654 -0,577 -0,561 -0,541 -3,8%
Bread 1,090 0,925 0,872 0,874 0,675 -5,8%
Butter/margarine 0,111 0,124 0,217 0,234 0,254 10,9%
Canned vegetables -0,209 -0,032 0,026 0,040 0,087 n.a
Cereals -0,228 -0,174 -0,135 -0,141 -0,113 -8,4%
Cheese 0,182 0,243 0,295 0,281 0,287 5,8%
Chocolate -0,377 -0,293 -0,231 -0,228 -0,214 -6,9%
Coffee -0,172 -0,089 0,011 -0,013 0,010 n.a
Desserts 0,379 0,408 0,491 0,478 0,444 2,0%
Edible oil -0,315 -0,308 -0,254 -0,250 -0,273 -1,8%
Frozen pizzas/starters -0,334 -0,456 -0,714 -0,699 -0,648 8,6%
Frozen vegetables -0,006 0,005 0,075 0,068 0,088 n.a
Fruit Juices 0,068 0,134 0,180 0,153 0,170 12,2%
Ham/Delicatessen 0,064 0,076 0,106 0,125 0,163 12,3%
Ice Cream 0,222 0,311 0,339 0,340 0,360 6,2%
Milk 0,155 0,180 0,230 0,252 0,194 2,8%
Mineral water -0,306 -0,428 -0,318 -0,274 -0,227 -3,7%
Savoury snacks -0,219 -0,137 -0,022 -0,142 -0,080 -11,9%
Soft drinks 0,045 0,116 0,250 0,227 0,222 22,0%
Tea 0,066 0,190 0,314 0,302 0,307 21,1%
Yoghurt -0,079 0,092 0,171 0,157 0,223 n.a
Average -0,043 -0,004 0,053 0,046 0,051
Measure of imbalance brand only - Czech Republic2004 - 2012 Evolution
1,0
0,8
0,6
0,4
0,2
0,0
-0,2
-0,4
-0,6
-0,8
2012
342
Annexes
France 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,226 -0,215 -0,232 -0,256 -0,281 2,8%

Biscuits -0,243 -0,292 -0,332 -0,371 -0,405 6,6%
Bread 1,802 1,828 1,851 1,847 1,793 -0,1%
Canned vegetables -0,147 -0,163 -0,187 -0,203 -0,251 6,9%
Cereals -0,329 -0,335 -0,347 -0,344 -0,354 0,9%
Cheese 0,101 0,012 -0,009 -0,025 -0,024 n.a
Chocolate 0,265 0,244 0,222 0,190 0,148 -7,0%
Coffee -0,111 -0,120 -0,139 -0,143 -0,173 5,7%
Desserts 0,056 0,117 0,158 0,202 0,187 16,2%
Edible oil -0,266 -0,374 -0,405 -0,467 -0,438 6,4%
Frozen ready cooked meals -0,174 -0,039 0,012 -0,002 -0,010 -29,6%
Frozen vegetables 0,005 -0,004 -0,010 -0,076 -0,114 n.a
Fruit Juices 0,308 0,229 0,193 0,159 0,099 -13,3%
Ham/Delicatessen 0,342 0,279 0,324 0,163 0,209 -6,0%
Ice Cream -0,067 -0,050 -0,084 -0,126 -0,168 12,2%
Milk -0,180 -0,197 -0,191 -0,254 -0,229 3,1%
Mineral water -0,099 -0,142 -0,150 -0,161 -0,234 11,3%
Savoury snacks 0,195 0,060 -0,041 -0,070 -0,108 n.a
Soft drinks -0,300 -0,334 -0,362 -0,384 -0,408 3,9%
Tea -0,171 -0,185 -0,187 -0,197 -0,215 2,9%
Yoghurt -0,110 -0,281 -0,319 -0,321 -0,356 15,8%
Average 0,024 -0,003 -0,018 -0,047 -0,072
Measure of imbalance brand only - France

1,5
1,0
0,5
0,0
-0,5
2012
343
Hungary 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,257 -0,260 -0,213 -0,258 -0,226 -1,6%

Biscuits -0,609 -0,596 -0,475 -0,537 -0,510 -2,2%
Bread 2,890 2,887 2,631 2,564 2,666 -1,0%
Butter/margarine -0,319 -0,299 -0,246 -0,303 -0,290 -1,2%
Canned vegetables 0,393 0,622 0,869 0,672 0,628 6,0%
Cereals -0,291 -0,344 -0,308 -0,335 -0,346 2,2%
Cheese -0,061 -0,161 -0,109 -0,138 -0,143 11,2%
Chocolate -0,070 -0,066 -0,021 -0,059 -0,040 -6,6%
Coffee -0,231 -0,214 -0,193 -0,232 -0,233 0,1%
Desserts -0,123 -0,149 -0,050 -0,096 -0,084 -4,7%
Edible oil -0,704 -0,773 -0,700 -0,685 -0,693 -0,2%
Frozen ready cooked meals -0,038 -0,033 0,055 -0,056 0,255 n.a
Frozen vegetables 0,174 0,041 0,127 0,085 0,117 -4,9%
Fruit Juices 0,280 0,169 0,007 -0,068 -0,079 n.a
Ham/Delicatessen 0,184 -0,176 -0,231 -0,297 -0,249 n.a
Ice Cream -0,430 -0,381 -0,223 -0,260 -0,264 -5,9%
Milk -0,180 -0,336 -0,208 -0,166 -0,162 -1,4%
Mineral water 0,059 0,064 0,162 0,073 0,156 12,9%
Savoury snacks 0,068 0,069 0,020 -0,038 0,001 -38,1%
Soft drinks -0,234 -0,234 -0,151 -0,177 -0,119 -8,0%
Tea -0,129 -0,157 -0,160 -0,140 -0,122 -0,7%
Yoghurt -0,052 -0,096 -0,075 -0,118 -0,151 14,2%
Average -0,002 -0,034 -0,004 -0,055 -0,025
Measure of imbalance brand only - Hungary

2,8
2,3
1,8
1,3
0,8
0,3
-0,2
-0,7
2012
344
Annexes
Italy 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,357 -0,421 -0,441 -0,447 -0,458 3,2%

Biscuits -0,167 -0,245 -0,262 -0,279 -0,291 7,2%
Bread 1,801 1,761 1,657 1,555 1,478 -2,4%
Butter/margarine 0,087 0,026 0,026 0,014 -0,087 n.a
Canned vegetables 0,611 0,556 0,521 0,491 0,470 -3,2%
Cereals -0,505 -0,551 -0,574 -0,553 -0,568 1,5%
Cheese 0,996 0,798 0,809 0,810 0,755 -3,4%
Chocolate -0,120 -0,152 -0,191 -0,190 -0,189 5,8%
Coffee -0,230 -0,266 -0,279 -0,272 -0,229 -0,1%
Desserts 0,190 0,089 -0,043 -0,088 -0,174 n.a
Edible oil 0,386 0,354 0,107 0,150 0,099 -15,6%
Frozen vegetables -0,168 -0,194 -0,227 -0,293 -0,334 9,0%
Fruit Juices -0,049 -0,092 -0,110 -0,107 -0,087 7,4%
Ice Cream 0,387 0,454 0,492 0,604 0,679 7,3%
Milk -0,094 -0,145 -0,072 -0,034 -0,025 -15,1%
Mineral water -0,040 -0,056 -0,062 -0,047 -0,106 12,9%
Savoury snacks 0,373 0,302 0,263 0,238 0,228 -6,0%
Soft drinks -0,030 -0,100 -0,134 -0,139 -0,169 23,8%
Tea 0,026 -0,021 -0,036 -0,027 -0,052 n.a
Yoghurt -0,169 -0,167 -0,161 -0,174 -0,194 1,8%
Average 0,142 0,094 0,065 0,058 0,030
Measure of imbalance brand only - Italy

1,8
1,3
0,8
0,3
-0,2
-0,7
2012
345
Poland 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,513 -0,486 -0,425 -0,384 -0,312 -6,0%

Biscuits -0,416 -0,400 -0,271 -0,235 -0,146 -12,3%
Bread 1,989 2,422 2,339 2,178 1,958 -0,2%
Canned vegetables -0,435 -0,386 -0,269 -0,226 -0,153 -12,2%
Cereals -0,760 -0,691 -0,539 -0,499 -0,435 -6,7%
Cheese -0,073 -0,003 0,112 0,107 0,161 n.a
Chocolate -0,106 -0,074 0,044 0,085 0,180 n.a
Coffee -0,232 -0,231 -0,119 -0,096 -0,040 -19,8%
Desserts 0,131 0,144 0,180 0,198 0,203 5,6%
Edible oil -0,287 -0,217 -0,362 -0,381 -0,328 1,7%
Frozen pizzas/starters -0,550 -0,474 -0,339 -0,288 -0,209 -11,4%
Frozen ready cooked meals -0,054 -0,024 0,109 0,165 0,239 n.a
Frozen vegetables -0,248 -0,236 -0,109 -0,050 0,009 n.a
Fruit Juices -0,316 -0,425 -0,306 -0,299 -0,137 -9,9%
Ham/Delicatessen -0,292 -0,239 -0,123 -0,074 -0,012 -32,6%
Ice Cream -0,214 -0,137 -0,034 0,025 0,110 n.a
Milk -0,028 -0,075 -0,120 -0,142 -0,067 11,5%
Mineral water 0,127 0,026 0,098 0,096 0,099 -3,0%
Savoury snacks -0,044 -0,016 0,063 0,104 0,169 n.a
Soft drinks 0,046 0,099 0,231 0,256 0,284 25,6%
Tea -0,068 -0,164 -0,099 -0,153 -0,149 10,4%
Yoghurt -0,165 -0,173 -0,099 -0,115 -0,096 -6,5%
Average -0,090 -0,056 0,021 0,033 0,075
Measure of imbalance brand only - Poland

1,7
1,2
0,7
0,2
-0,3
-0,8
2012
346
Annexes
Portugal 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,219 -0,224 -0,236 -0,249 -0,245 1,4%

Biscuits 0,276 0,103 0,107 0,094 0,102 -11,7%
Bread 1,681 1,640 1,721 1,777 1,801 0,9%
Butter/margarine -0,213 -0,234 -0,221 -0,222 -0,205 -0,5%
Canned vegetables -0,123 -0,169 -0,172 0,146 0,136 n.a
Cereals -0,222 -0,330 -0,289 -0,314 -0,323 4,8%
Cheese 0,304 0,247 0,227 0,158 0,158 -7,9%
Chocolate 0,144 0,187 0,241 0,210 0,220 5,4%
Coffee -0,204 -0,240 -0,215 -0,178 -0,226 1,3%
Desserts -0,073 0,097 0,137 0,158 0,131 n.a
Edible oil -0,081 -0,086 -0,072 -0,133 -0,121 5,2%
Frozen ready cooked meals -0,538 -0,498 -0,453 -0,433 -0,451 -2,2%
Fruit Juices 0,061 0,034 0,143 -0,020 0,010 -20,3%
Ham/Delicatessen -0,150 -0,206 -0,212 -0,160 -0,154 0,3%
Ice Cream -0,347 -0,396 -0,411 -0,431 -0,426 2,6%
Milk -0,236 -0,202 -0,124 -0,076 -0,247 0,6%
Mineral water 0,400 0,292 0,373 0,409 0,411 0,4%
Savoury snacks 0,000 0,000 0,061 0,012 -0,012 n.a
Soft drinks 0,019 -0,003 -0,005 -0,019 -0,051 n.a
Tea -0,012 -0,033 0,031 0,019 -0,022 7,9%
Yoghurt -0,050 -0,060 -0,003 -0,016 -0,028 -6,8%
Average 0,005 -0,017 0,011 0,015 0,005
Measure of imbalance brand only - Portugal

1,9
1,4
0,9
0,4
-0,1
-0,6
2012
347
Spain 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,092 -0,073 -0,064 -0,024 -0,020 -17,5%

Biscuits 0,080 -0,010 0,039 0,036 0,017 -17,9%
Bread 1,783 1,693 1,729 1,715 1,535 -1,9%
Cereals -0,441 -0,420 -0,351 -0,296 -0,369 -2,2%
Cheese 0,539 0,503 0,588 0,468 0,448 -2,3%
Chocolate 0,012 0,024 0,092 0,065 0,052 20,1%
Coffee -0,403 -0,415 -0,379 -0,418 -0,451 1,4%
Desserts -0,342 -0,215 -0,106 -0,068 -0,083 -16,2%
Edible oil 0,062 0,036 0,076 0,068 -0,009 n.a
Frozen pizzas/starters 0,120 0,208 0,292 0,358 0,340 13,8%
Frozen ready cooked meals -0,056 -0,073 0,047 0,077 0,055 n.a
Frozen vegetables -0,077 -0,077 -0,026 0,002 -0,224 14,2%
Fruit Juices 0,088 0,108 0,188 0,100 0,052 -6,4%
Ham/Delicatessen 0,041 -0,009 0,186 0,233 0,050 2,5%
Ice Cream -0,351 -0,329 -0,260 -0,278 -0,318 -1,2%
Milk -0,035 -0,013 0,106 -0,003 -0,012 -13,1%
Mineral water 0,179 0,215 0,273 0,287 0,317 7,4%
Savoury snacks -0,248 -0,243 -0,195 -0,238 -0,266 0,9%
Soft drinks -0,509 -0,473 -0,406 -0,402 -0,427 -2,2%
Tea -0,016 -0,029 0,046 0,066 0,070 n.a
Yoghurt -0,415 -0,574 -0,530 -0,527 -0,569 4,0%
Average 0,009 0,002 0,070 0,067 0,024
Measure of imbalance brand only - Spain

1,3
0,8
0,3
-0,2
-0,7
2012
348
Annexes
Finland 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,071 0,071 0,109 0,141 0,134 n.a

Biscuits 0,074 0,165 0,162 0,176 0,250 16,4%
Bread 0,405 0,505 0,503 0,580 0,643 5,9%
Butter/margarine -0,012 0,098 0,098 0,100 0,081 n.a
Cereals 0,190 0,351 0,271 0,258 0,223 2,0%
Cheese -0,157 -0,067 -0,031 -0,014 0,010 n.a
Chocolate 0,285 0,395 0,383 0,407 0,389 4,0%
Coffee 0,149 0,237 0,241 0,254 0,276 8,0%
Desserts -0,263 -0,154 -0,153 -0,119 -0,142 -7,4%
Edible oil 0,622 0,686 0,683 0,703 0,678 1,1%
Frozen ready cooked meals 0,098 0,206 0,174 0,083 0,070 -4,1%
Fruit Juices -0,108 0,039 0,020 0,055 0,085 n.a
Ice Cream 0,219 0,332 0,343 0,270 0,295 3,8%
Milk -0,316 -0,147 -0,201 -0,173 -0,147 -9,1%
Mineral water -0,019 0,127 0,107 0,095 0,124 n.a
Savoury snacks -0,004 0,150 0,190 0,066 0,052 n.a
Soft drinks 0,212 0,327 0,304 0,315 0,313 5,0%
Tea 0,085 0,192 0,216 0,248 0,260 14,9%
Yoghurt -0,391 -0,276 -0,253 -0,246 -0,229 -6,5%
Average 0,117 0,219 0,209 0,206 0,200
Measure of imbalance brand only - Finland

0,9
0,7
0,5
0,3
0,1
-0,1
-0,3
-0,5
2012
349
Germany 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,330 -0,251 -0,233 -0,141 -0,135 -10,6%

Biscuits -0,232 -0,227 -0,197 -0,030 -0,011 -31,7%
Bread 0,582 0,597 0,625 0,717 0,728 2,8%
Cereals -0,235 -0,142 -0,105 -0,019 -0,008 -34,8%
Cheese 0,476 0,466 0,489 0,558 0,566 2,2%
Chocolate -0,020 0,014 0,022 0,104 0,114 n.a
Coffee -0,102 -0,031 -0,019 0,082 0,115 n.a
Desserts 0,409 0,355 0,316 0,394 0,454 1,3%
Edible oil -0,330 -0,033 0,112 0,132 0,164 n.a
Frozen pizzas/starters -0,344 -0,255 -0,241 -0,188 -0,185 -7,5%
Frozen ready cooked meals -0,157 -0,109 -0,200 -0,247 -0,075 -8,9%
Frozen vegetables 0,019 0,057 -0,140 -0,125 -0,106 n.a
Fruit Juices 0,476 0,465 0,399 0,429 0,440 -1,0%
Ham/Delicatessen -0,093 -0,008 -0,014 -0,049 -0,081 -1,6%
Ice Cream -0,159 -0,103 -0,089 0,008 0,017 n.a
Milk 0,380 0,353 0,280 0,333 0,255 -4,9%
Mineral water 0,434 0,481 0,528 0,606 0,640 5,0%
Savoury snacks -0,102 -0,014 0,031 -0,036 -0,025 -16,0%
Soft drinks -0,186 -0,122 -0,091 0,049 0,074 n.a
Tea 0,009 0,037 0,016 0,104 0,113 37,7%
Yoghurt 0,064 0,134 0,117 0,201 0,246 18,4%
Average 0,024 0,075 0,073 0,130 0,147
Measure of imbalance brand only - Germany

0,8
0,6
0,4
0,2
0,0
-0,2
-0,4
2012
350
Annexes
Netherlands 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,153 -0,145 -0,302 -0,329 -0,239 5,7%

Biscuits 0,419 0,453 0,247 0,124 0,200 -8,9%
Bread 1,204 1,236 1,120 1,108 1,254 0,5%
Butter/margarine 0,121 0,036 -0,087 -0,177 -0,136 n.a
Cereals 0,437 0,425 0,316 0,233 0,340 -3,1%
Cheese 0,878 0,834 0,524 0,524 0,598 -4,7%
Chocolate 0,339 0,318 0,206 0,151 0,207 -6,0%
Coffee -0,125 -0,135 -0,302 -0,394 -0,304 11,7%
Desserts 0,124 0,170 -0,175 -0,238 -0,142 n.a
Edible oil 0,366 0,275 0,074 0,002 0,039 -24,5%
Fruit Juices 0,226 0,253 0,222 0,134 0,243 0,9%
Ice Cream -0,221 -0,251 -0,380 -0,535 -0,470 9,9%
Milk 0,066 0,070 -0,304 -0,202 -0,129 n.a
Mineral water -0,210 -0,161 -0,192 -0,154 0,039 n.a
Savoury snacks 0,191 -0,052 -0,206 -0,257 -0,152 n.a
Soft drinks 0,348 0,381 0,246 0,214 0,312 -1,3%
Tea -0,204 -0,143 -0,176 -0,205 -0,136 -4,9%
Yoghurt 0,096 0,193 -0,076 0,048 0,164 7,0%
Average 0,192 0,177 0,024 -0,022 0,076
Measure of imbalance brand only - Netherlands

1,4
1,2
1,0
0,8
0,6
0,4
0,2
0,0
-0,2
-0,4
-0,6
2012
351
Denmark 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,155 -0,107 -0,074 -0,069 -0,095 -5,9%

Biscuits -0,010 -0,012 -0,021 -0,034 0,020 n.a
Bread 0,321 0,327 0,319 0,371 0,413 3,2%
Butter/margarine -0,096 -0,079 -0,085 -0,109 -0,148 5,6%
Cereals 0,106 0,173 0,087 0,048 -0,007 n.a
Cheese -0,241 -0,245 -0,214 -0,223 -0,219 -1,2%
Chocolate 0,201 0,217 0,199 0,198 0,159 -2,9%
Coffee 0,064 0,060 0,058 0,045 0,046 -4,1%
Desserts -0,347 -0,332 -0,336 -0,328 -0,372 0,9%
Edible oil 0,538 0,508 0,500 0,494 0,448 -2,2%
Frozen ready cooked meals 0,014 0,028 -0,009 -0,127 -0,159 n.a
Frozen vegetables 0,067 0,036 0,024 0,005 -0,127 n.a
Fruit Juices -0,192 -0,139 -0,164 -0,154 -0,145 -3,5%
Ham/Delicatessen 0,237 0,038 0,012 -0,007 -0,078 n.a
Ice Cream 0,135 0,154 0,159 0,061 0,066 -8,6%
Milk -0,400 -0,324 -0,385 -0,382 -0,377 -0,7%
Mineral water -0,103 -0,051 -0,077 -0,114 -0,105 0,3%
Savoury snacks -0,088 -0,028 0,007 -0,144 -0,177 9,1%
Soft drinks 0,128 0,149 0,120 0,106 0,084 -5,1%
Tea 0,001 0,015 0,033 0,039 0,031 48,7%
Yoghurt -0,475 -0,454 -0,436 -0,455 -0,458 -0,4%
Average 0,033 0,041 0,026 -0,004 -0,029
Measure of imbalance brand only - Denmark

0,8
0,6
0,4
0,2
0,0
-0,2
-0,4
-0,6
2012
352
Annexes
Romania 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,082 -0,204 -0,259 -0,275 -0,134 6,3%

Biscuits 0,612 0,461 0,373 0,361 0,454 -3,7%
Bread 2,451 1,899 1,522 1,398 1,572 -5,4%
Cereals 0,123 0,004 -0,006 -0,024 0,102 -2,4%
Cheese 0,404 0,127 0,055 0,085 0,259 -5,4%
Chocolate 0,107 -0,022 0,007 -0,018 0,051 -8,9%
Coffee 0,107 -0,020 -0,185 -0,233 -0,084 n.a
Desserts 0,615 0,239 0,130 0,171 0,347 -6,9%
Edible oil 0,110 0,006 -0,157 -0,074 0,058 -7,6%
Frozen vegetables 0,452 0,307 0,407 0,398 0,548 2,4%
Fruit Juices 0,134 -0,001 0,021 -0,012 0,164 2,6%
Ham/Delicatessen 0,070 -0,127 0,006 0,066 0,223 15,6%
Ice Cream 0,270 0,036 0,041 0,038 0,180 -4,9%
Milk 0,537 0,208 0,152 0,247 0,391 -3,9%
Mineral water 0,238 0,009 -0,070 -0,078 0,120 -8,2%
Savoury snacks 0,077 -0,079 -0,055 -0,098 0,041 -7,6%
Soft drinks -0,115 -0,197 -0,234 -0,259 -0,153 3,7%
Tea 0,319 0,049 -0,030 -0,078 0,111 -12,3%
Yoghurt 0,014 -0,096 -0,168 -0,201 -0,121 n.a
Average 0,314 0,111 0,060 0,046 0,179
Measure of imbalance brand only - Romania

2,5
2,0
1,5
1,0
0,5
0,0
-0,5
-1,0
2012
353
United Kingdom 2004 2006 2008 2010 2012 CAGR 2004-2012
Baby food -0,017 -0,034 -0,053 -0,062 -0,081 21,7%

Biscuits 0,298 0,278 0,284 0,261 0,243 -2,5%
Bread 0,498 0,366 0,253 0,262 0,246 -8,5%
Butter/margarine -0,186 -0,194 -0,237 -0,436 -0,443 11,5%
Cereals 0,048 0,086 0,103 0,132 0,153 15,5%
Cheese 0,939 0,929 0,913 0,861 0,871 -0,9%
Chocolate -0,013 -0,002 0,010 -0,031 -0,023 7,5%
Coffee -0,367 -0,339 -0,300 -0,249 -0,180 -8,5%
Desserts 0,285 0,315 0,350 0,389 0,425 5,1%
Edible oil 0,088 0,129 0,181 0,262 0,203 11,1%
Frozen pizzas/starters 0,016 -0,006 0,032 -0,043 -0,060 n.a
Frozen ready cooked meals 0,155 0,160 0,233 0,191 0,180 1,9%
Fruit Juices 0,569 0,467 0,359 0,310 0,301 -7,6%
Ice Cream -0,054 -0,046 -0,020 -0,048 -0,088 6,4%
Milk -0,019 0,023 0,098 0,154 0,183 n.a
Mineral water -0,213 -0,222 -0,186 -0,188 -0,154 -4,0%
Savoury snacks -0,130 -0,103 -0,098 -0,069 -0,082 -5,6%
Soft drinks -0,055 -0,051 -0,056 -0,034 -0,014 -15,6%
Tea 0,004 -0,022 0,016 0,036 0,048 36,6%
Yoghurt -0,053 -0,007 -0,037 -0,066 -0,008 -21,4%
Average 0,092 0,089 0,093 0,085 0,088
Measure of imbalance brand only - United Kingdom

1,0
0,8
0,6
0,4
0,2
0,0
-0,2
-0,4
-0,6
2012
354
Annexes
CAGR CAGR CAGR

Measure of imbalance 2004 2006 2008 2010 2012 2004-2012 2004-2008 2008-2012
Baby food
Belgium -0,094 -0,105 -0,157 -0,151 -0,151 6,1% 13,7% -0,9%
Czech Republic -0,255 -0,196 -0,102 -0,106 -0,082 -13,1% -20,4% -5,2%
Denmark -0,155 -0,107 -0,074 -0,069 -0,095 -5,9% -16,8% 6,4%
Finland -0,149 -0,016 -0,055 -0,048 0,025 #NUM! -22,2% #NUM!
France -0,226 -0,215 -0,232 -0,256 -0,281 2,8% 0,7% 4,9%
Germany -0,330 -0,251 -0,233 -0,141 -0,135 -10,6% -8,3% -12,8%
Hungary -0,257 -0,260 -0,213 -0,258 -0,226 -1,6% -4,6% 1,5%
Italy -0,357 -0,421 -0,441 -0,447 -0,458 3,2% 5,4% 0,9%
Netherlands -0,153 -0,145 -0,302 -0,329 -0,239 5,7% 18,4% -5,7%
Poland -0,513 -0,486 -0,425 -0,384 -0,312 -6,0% -4,6% -7,4%
Portugal -0,219 -0,224 -0,236 -0,249 -0,245 1,4% 1,9% 0,9%
Romania -0,082 -0,204 -0,259 -0,275 -0,134 6,3% 33,2% -15,2%
Spain -0,092 -0,073 -0,064 -0,024 -0,020 -17,5% -8,7% -25,4%
United Kingdom -0,017 -0,034 -0,053 -0,062 -0,081 21,7% 33,5% 10,9%
Average -0,192 -0,177 -0,196 -0,193 -0,166 -1,8% 0,5% -4,0%
Measure of Imbalance Baby food

0,100
0,000
-0,100
-0,200
-0,300
-0,400
-0,500
-0,600
2012
355
CAGR CAGR CAGR

Biscuits
Belgium -0,015 -0,051 0,057 -0,053 -0,051 16,8% #NUM! #NUM!
Czech Republic -0,738 -0,654 -0,577 -0,561 -0,541 -3,8% -6,0% -1,6%
Denmark -0,010 -0,012 -0,021 -0,034 0,020 #NUM! 21,1% #NUM!
Finland -0,247 -0,152 -0,159 -0,158 0,097 #NUM! -10,5% #NUM!
France -0,243 -0,292 -0,332 -0,371 -0,405 6,6% 8,2% 5,1%
Germany -0,232 -0,227 -0,197 -0,030 -0,011 -31,7% -4,0% -51,5%
Hungary -0,609 -0,596 -0,475 -0,537 -0,510 -2,2% -6,1% 1,8%
Italy -0,167 -0,245 -0,262 -0,279 -0,291 7,2% 12,0% 2,7%
Netherlands 0,419 0,453 0,247 0,124 0,200 -8,9% -12,4% -5,2%
Poland -0,416 -0,400 -0,271 -0,235 -0,146 -12,3% -10,1% -14,4%
Portugal 0,276 0,103 0,107 0,094 0,102 -11,7% -21,2% -1,1%
Romania 0,612 0,461 0,373 0,361 0,454 -3,7% -11,6% 5,0%
Spain 0,080 -0,010 0,039 0,036 0,017 -17,9% -16,4% -19,3%
United Kingdom 0,298 0,278 0,284 0,261 0,243 -2,5% -1,2% -3,8%
Average -0,147 -0,158 -0,146 -0,157 -0,102 -4,5% -0,2% -8,6%
Measure of Imbalance Biscuits
0,800
0,600
0,400
0,200
0,000
-0,200
-0,400
-0,600
-0,800
-1,000
2012
356
Annexes
CAGR CAGR CAGR

Bread
Belgium 2,251 2,135 2,099 2,084 2,095 -0,9% -1,7% 0,0%
Czech Republic 1,090 0,925 0,872 0,874 0,675 -5,8% -5,4% -6,2%
Denmark 0,321 0,327 0,319 0,371 0,413 3,2% -0,1% 6,7%
Finland 0,411 0,521 0,500 0,474 0,476 1,8% 5,0% -1,2%
France 1,802 1,828 1,851 1,847 1,793 -0,1% 0,7% -0,8%
Germany 0,582 0,597 0,625 0,717 0,728 2,8% 1,8% 3,9%
Hungary 2,890 2,887 2,631 2,564 2,666 -1,0% -2,3% 0,3%
Italy 1,801 1,761 1,657 1,555 1,478 -2,4% -2,1% -2,8%
Netherlands 1,204 1,236 1,120 1,108 1,254 0,5% -1,8% 2,9%
Poland 1,989 2,422 2,339 2,178 1,958 -0,2% 4,1% -4,3%
Portugal 1,681 1,640 1,721 1,777 1,801 0,9% 0,6% 1,1%
Romania 2,451 1,899 1,522 1,398 1,572 -5,4% -11,2% 0,8%
Spain 1,783 1,693 1,729 1,715 1,535 -1,9% -0,8% -2,9%
United Kingdom 0,498 0,366 0,253 0,262 0,246 -8,5% -15,6% -0,7%
Average 0,847 0,815 0,776 0,785 0,797 -0,8% -2,2% 0,7%
Measure of Imbalance Bread
3,500
3,000
2,500
2,000
1,500
1,000
0,500
0,000
2012
357
CAGR CAGR CAGR

Butter/margarine
Belgium 0,208 0,186 0,210 0,214 0,196 -0,8% 0,2% -1,8%
Czech Republic 0,111 0,124 0,217 0,234 0,254 10,9% 18,2% 4,1%
Denmark -0,096 -0,079 -0,085 -0,109 -0,148 5,6% -2,9% 14,8%
Finland 0,256 0,352 0,311 0,312 0,183 -4,1% 5,0% -12,4%
France 0,129 0,122 0,063 0,052 0,026 -18,3% -16,5% -20,0%
Germany 0,074 0,123 0,147 0,212 0,229 15,2% 18,8% 11,8%
Hungary -0,319 -0,299 -0,246 -0,303 -0,290 -1,2% -6,3% 4,2%
Italy 0,087 0,026 0,026 0,014 -0,087 #NUM! -25,8% #NUM!
Netherlands 0,121 0,036 -0,087 -0,177 -0,136 #NUM! #NUM! 12,0%
Poland 0,436 0,481 0,517 0,484 0,390 -1,4% 4,3% -6,8%
Portugal -0,213 -0,234 -0,221 -0,222 -0,205 -0,5% 0,9% -1,9%
Romania 0,416 0,138 0,139 0,116 0,224 -7,4% -24,0% 12,8%
Spain 0,067 0,071 0,105 0,154 0,207 15,2% 12,0% 18,5%
United Kingdom -0,186 -0,194 -0,237 -0,436 -0,443 11,5% 6,3% 16,9%
Average 0,038 0,027 0,018 -0,021 -0,026 #NUM! -17,0% #NUM!
Measure of Imbalance Butter/margarine
0,500
0,400
0,300
0,200
0,100
0,000
-0,100
-0,200
-0,300
-0,400
-0,500
2012
358
Annexes
CAGR CAGR CAGR

Canned vegetables
Belgium 0,092 -0,026 -0,123 -0,188 -0,187 #NUM! #NUM! 11,0%
Czech Republic -0,209 -0,032 0,026 0,040 0,087 #NUM! #NUM! 35,6%
Denmark 0,720 0,743 0,658 0,550 0,432 -6,2% -2,2% -10,0%
Finland 0,136 0,935 0,921 0,938 0,951 27,5% 61,4% 0,8%
France -0,147 -0,163 -0,187 -0,203 -0,251 6,9% 6,3% 7,6%
Germany -0,089 -0,055 -0,061 -0,107 -0,144 6,2% -9,0% 24,1%
Hungary 0,393 0,622 0,869 0,672 0,628 6,0% 21,9% -7,8%
Italy 0,611 0,556 0,521 0,491 0,470 -3,2% -3,9% -2,5%
Netherlands -0,042 -0,067 -0,128 -0,193 -0,101 11,5% 32,1% -5,9%
Poland -0,435 -0,386 -0,269 -0,226 -0,153 -12,2% -11,3% -13,2%
Portugal -0,123 -0,169 -0,172 0,146 0,136 #NUM! 8,8% #NUM!
Romania 0,454 0,192 0,136 0,108 0,220 -8,7% -26,1% 12,8%
Spain 0,215 0,135 0,149 0,165 0,152 -4,2% -8,7% 0,5%
United Kingdom -0,103 -0,148 -0,177 -0,153 -0,145 4,4% 14,5% -4,8%
Average 0,047 0,069 0,052 0,057 0,068 4,8% 2,7% 6,9%
Measure of Imbalance Canned vegetables
1,200
1,000
0,800
0,600
0,400
0,200
0,000
-0,200
-0,400
-0,600
2012
359
CAGR CAGR CAGR

Cereals
Belgium -0,410 -0,423 -0,417 -0,410 -0,384 -0,8% 0,4% -2,0%
Czech Republic -0,228 -0,174 -0,135 -0,141 -0,113 -8,4% -12,2% -4,4%
Denmark 0,106 0,173 0,087 0,048 -0,007 #NUM! -4,6% #NUM!
Finland 0,364 0,414 0,441 0,439 0,446 2,6% 4,9% 0,3%
France -0,329 -0,335 -0,347 -0,344 -0,354 0,9% 1,3% 0,5%
Germany -0,235 -0,142 -0,105 -0,019 -0,008 -34,8% -18,3% -48,0%
Hungary -0,291 -0,344 -0,308 -0,335 -0,346 2,2% 1,5% 2,9%
Italy -0,505 -0,551 -0,574 -0,553 -0,568 1,5% 3,2% -0,3%
Netherlands 0,437 0,425 0,316 0,233 0,340 -3,1% -7,7% 1,8%
Poland -0,760 -0,691 -0,539 -0,499 -0,435 -6,7% -8,2% -5,3%
Portugal -0,222 -0,330 -0,289 -0,314 -0,323 4,8% 6,8% 2,8%
Romania 0,123 0,004 -0,006 -0,024 0,102 -2,4% #NUM! #NUM!
Spain -0,441 -0,420 -0,351 -0,296 -0,369 -2,2% -5,6% 1,3%
United Kingdom 0,048 0,086 0,103 0,132 0,153 15,5% 20,8% 10,5%
Average -0,188 -0,187 -0,181 -0,176 -0,162 -1,8% -1,0% -2,6%
Measure of Imbalance Cereals
0,600
0,400
0,200
0,000
-0,200
-0,400
-0,600
-0,800
-1,000
2012
360
Annexes
CAGR CAGR CAGR

Cheese
Belgium 1,007 0,919 0,877 0,865 0,930 -1,0% -3,4% 1,5%
Czech Republic 0,182 0,243 0,295 0,281 0,287 5,8% 12,9% -0,7%
Denmark -0,241 -0,245 -0,214 -0,223 -0,219 -1,2% -2,9% 0,6%
Finland -0,020 0,062 0,017 0,025 0,010 #NUM! #NUM! -12,1%
France 0,101 0,012 -0,009 -0,025 -0,024 #NUM! #NUM! 28,4%
Germany 0,476 0,466 0,489 0,558 0,566 2,2% 0,7% 3,7%
Hungary -0,061 -0,161 -0,109 -0,138 -0,143 11,2% 15,5% 7,0%
Italy 0,996 0,798 0,809 0,810 0,755 -3,4% -5,1% -1,7%
Netherlands 0,878 0,834 0,524 0,524 0,598 -4,7% -12,1% 3,4%
Poland -0,073 -0,003 0,112 0,107 0,161 #NUM! #NUM! 9,6%
Portugal 0,304 0,247 0,227 0,158 0,158 -7,9% -7,1% -8,6%
Romania 0,404 0,127 0,055 0,085 0,259 -5,4% -39,3% 47,5%
Spain 0,539 0,503 0,588 0,468 0,448 -2,3% 2,2% -6,6%
United Kingdom 0,939 0,929 0,913 0,861 0,871 -0,9% -0,7% -1,2%
Average 0,215 0,178 0,170 0,164 0,183 -2,0% -5,7% 1,9%
Measure of Imbalance Cheese
1,200
1,000
0,800
0,600
0,400
0,200
0,000
-0,200
-0,400
2012
361
CAGR CAGR CAGR

Chocolate
Belgium 0,029 0,030 -0,035 -0,023 -0,031 #NUM! #NUM! -2,8%
Czech Republic -0,377 -0,293 -0,231 -0,228 -0,214 -6,9% -11,5% -1,9%
Denmark 0,201 0,217 0,199 0,198 0,159 -2,9% -0,2% -5,4%
Finland -0,013 0,098 0,092 0,088 0,115 #NUM! #NUM! 5,6%
France 0,265 0,244 0,222 0,190 0,148 -7,0% -4,4% -9,6%
Germany -0,020 0,014 0,022 0,104 0,114 #NUM! #NUM! 50,4%
Hungary -0,070 -0,066 -0,021 -0,059 -0,040 -6,6% -26,2% 18,3%
Italy -0,120 -0,152 -0,191 -0,190 -0,189 5,8% 12,3% -0,3%
Netherlands 0,339 0,318 0,206 0,151 0,207 -6,0% -11,7% 0,0%
Poland -0,106 -0,074 0,044 0,085 0,180 #NUM! #NUM! 42,1%
Portugal 0,144 0,187 0,241 0,210 0,220 5,4% 13,7% -2,2%
Romania 0,107 -0,022 0,007 -0,018 0,051 -8,9% -49,4% 64,1%
Spain 0,012 0,024 0,092 0,065 0,052 20,1% 66,6% -13,4%
United Kingdom -0,013 -0,002 0,010 -0,031 -0,023 7,5% #NUM! #NUM!
Average 0,030 0,044 0,042 0,034 0,050 6,6% 9,2% 4,1%
Measure of Imbalance Chocolate
0,400
0,300
0,200
0,100
0,000
-0,100
-0,200
-0,300
-0,400
-0,500
2012
362
Annexes
CAGR CAGR CAGR

Coffee
Belgium -0,085 -0,095 -0,173 -0,211 -0,228 13,1% 19,4% 7,2%
Czech Republic -0,172 -0,089 0,011 -0,013 0,010 #NUM! #NUM! -0,8%
Denmark 0,064 0,060 0,058 0,045 0,046 -4,1% -2,7% -5,5%
Finland -0,219 -0,111 -0,119 -0,109 0,137 #NUM! -14,2% #NUM!
France -0,111 -0,120 -0,139 -0,143 -0,173 5,7% 5,9% 5,5%
Germany -0,102 -0,031 -0,019 0,082 0,115 #NUM! -33,9% #NUM!
Hungary -0,231 -0,214 -0,193 -0,232 -0,233 0,1% -4,4% 4,8%
Italy -0,230 -0,266 -0,279 -0,272 -0,229 -0,1% 5,0% -4,9%
Netherlands -0,125 -0,135 -0,302 -0,394 -0,304 11,7% 24,6% 0,1%
Poland -0,232 -0,231 -0,119 -0,096 -0,040 -19,8% -15,3% -24,0%
Portugal -0,204 -0,240 -0,215 -0,178 -0,226 1,3% 1,4% 1,3%
Romania 0,107 -0,020 -0,185 -0,233 -0,084 #NUM! #NUM! -17,8%
Spain -0,403 -0,415 -0,379 -0,418 -0,451 1,4% -1,6% 4,5%
United Kingdom -0,367 -0,339 -0,300 -0,249 -0,180 -8,5% -4,9% -12,0%
Average -0,167 -0,162 -0,179 -0,188 -0,143 -2,0% 1,7% -5,5%
Measure of Imbalance Coffee
0,200
0,100
0,000
-0,100
-0,200
-0,300
-0,400
-0,500
2012
363
CAGR CAGR CAGR

Desserts
Belgium -0,028 0,027 0,015 0,035 0,045 #NUM! #NUM! 30,5%
Czech Republic 0,379 0,408 0,491 0,478 0,444 2,0% 6,7% -2,5%
Denmark 1,950 1,768 -0,336 -0,328 -0,372 #NUM! #NUM! 2,5%
Finland 0,006 0,160 0,163 0,220 0,196 55,4% 130,8% 4,6%
France 0,056 0,117 0,158 0,202 0,187 16,2% 29,4% 4,3%
Germany 0,409 0,355 0,316 0,394 0,454 1,3% -6,2% 9,5%
Hungary -0,123 -0,149 -0,050 -0,096 -0,084 -4,7% -20,1% 13,7%
Italy 0,190 0,089 -0,043 -0,088 -0,174 #NUM! #NUM! 42,0%
Netherlands 0,124 0,170 -0,175 -0,238 -0,142 #NUM! #NUM! -5,0%
Poland 0,131 0,144 0,180 0,198 0,203 5,6% 8,2% 3,1%
Portugal -0,073 0,097 0,137 0,158 0,131 #NUM! #NUM! -1,1%
Romania 0,615 0,239 0,130 0,171 0,347 -6,9% -32,2% 27,8%
Spain -0,342 -0,215 -0,106 -0,068 -0,083 -16,2% -25,4% -5,9%
United Kingdom 0,285 0,315 0,350 0,389 0,425 5,1% 5,3% 5,0%
Average 0,109 0,155 0,018 0,033 0,038 -12,2% -36,2% 20,8%
Measure of Imbalance Desserts
2,500
2,000
1,500
1,000
0,500
0,000
-0,500
2012
364
Annexes
CAGR CAGR CAGR

Edible oil
Belgium -0,017 -0,003 0,093 0,079 0,084 #NUM! #NUM! -2,5%
Czech Republic -0,315 -0,308 -0,254 -0,250 -0,273 -1,8% -5,3% 1,8%
Denmark 0,538 0,508 0,500 0,494 0,448 -2,2% -1,8% -2,7%
Finland 0,265 0,396 0,361 0,361 0,350 3,5% 8,1% -0,8%
France -0,266 -0,374 -0,405 -0,467 -0,438 6,4% 11,0% 2,0%
Germany -0,330 -0,033 0,112 0,132 0,164 #NUM! #NUM! 9,8%
Hungary -0,704 -0,773 -0,700 -0,685 -0,693 -0,2% -0,1% -0,3%
Italy 0,386 0,354 0,107 0,150 0,099 -15,6% -27,4% -1,9%
Netherlands 0,366 0,275 0,074 0,002 0,039 -24,5% -32,9% -15,0%
Poland -0,287 -0,217 -0,362 -0,381 -0,328 1,7% 6,0% -2,4%
Portugal -0,081 -0,086 -0,072 -0,133 -0,121 5,2% -3,0% 14,2%
Romania 0,110 0,006 -0,157 -0,074 0,058 -7,6% #NUM! #NUM!
Spain 0,062 0,036 0,076 0,068 -0,009 #NUM! 5,2% #NUM!
United Kingdom 0,088 0,129 0,181 0,262 0,203 11,1% 19,9% 2,9%
Average -0,059 -0,066 -0,084 -0,086 -0,079 3,7% 9,2% -1,6%
Measure of Imbalance Edible oil
0,600
0,400
0,200
0,000
-0,200
-0,400
-0,600
-0,800
2012
365
CAGR CAGR CAGR

Belgium 0,177 0,061 0,126 0,046 0,065 -11,7% -8,1% -15,2%
Czech Republic -0,334 -0,456 -0,714 -0,699 -0,648 8,6% 20,9% -2,4%
Denmark 0,333 0,268 0,219 0,148 0,099 -14,1% -10,0% -18,0%
Finland 0,296 0,393 0,384 0,381 0,402 3,9% 6,7% 1,2%
France -0,240 -0,225 -0,244 -0,302 -0,359 5,2% 0,5% 10,1%
Germany -0,344 -0,255 -0,241 -0,188 -0,185 -7,5% -8,6% -6,4%
Hungary -0,367 -0,370 -0,594 -0,695 -0,676 7,9% 12,8% 3,3%
Italy -0,113 -0,168 -0,190 -0,195 -0,186 6,5% 13,8% -0,4%
Netherlands -0,108 -0,214 -0,374 -0,398 -0,292 13,2% 36,5% -6,0%
Poland -0,550 -0,474 -0,339 -0,288 -0,209 -11,4% -11,4% -11,5%
Portugal -0,103 -0,072 -0,215 -0,204 -0,195 8,4% 20,2% -2,3%
Romania 0,539 0,433 0,405 0,283 0,279 -7,9% -6,9% -8,9%
Spain 0,120 0,208 0,292 0,358 0,340 13,8% 24,8% 3,9%
United Kingdom 0,016 -0,006 0,032 -0,043 -0,060 #NUM! 18,8% #NUM!
Average -0,063 -0,089 -0,189 -0,205 -0,180 14,1% 31,9% -1,3%
Measure of Imbalance Frozen pizzas/starters
0,600
0,400
0,200
0,000
-0,200
-0,400
-0,600
-0,800
2012
366
Annexes
CAGR CAGR CAGR

Frozen ready cooked meals
Belgium -0,218 -0,285 -0,326 -0,328 -0,247 1,6% 10,6% -6,7%
Czech Republic -0,126 -0,119 -0,017 -0,062 -0,141 1,3% -39,1% 68,5%
Denmark 0,014 0,028 -0,009 -0,127 -0,159 #NUM! #NUM! 102,7%
Finland 0,002 0,114 0,107 0,110 0,112 60,9% 156,5% 0,9%
France -0,174 -0,039 0,012 -0,002 -0,010 -29,6% #NUM! #NUM!
Germany -0,157 -0,109 -0,200 -0,247 -0,075 -8,9% 6,2% -21,8%
Hungary -0,038 -0,033 0,055 -0,056 0,255 #NUM! #NUM! 46,8%
Italy -0,293 -0,313 -0,329 -0,348 -0,474 6,2% 2,9% 9,6%
Netherlands -0,396 -0,463 -0,586 -0,606 -0,491 2,7% 10,3% -4,3%
Poland -0,054 -0,024 0,109 0,165 0,239 #NUM! #NUM! 21,6%
Portugal -0,538 -0,498 -0,453 -0,433 -0,451 -2,2% -4,2% -0,1%
Romania -0,638 -0,804 -0,856 -0,866 -0,726 1,6% 7,6% -4,0%
Spain -0,056 -0,073 0,047 0,077 0,055 #NUM! #NUM! 4,4%
United Kingdom 0,155 0,160 0,233 0,191 0,180 1,9% 10,7% -6,3%
Average -0,261 -0,259 -0,259 -0,270 -0,240 -1,1% -0,2% -1,9%
Measure of Imbalance Frozen ready cooked meals
0,400
0,200
0,000
-0,200
-0,400
-0,600
-0,800
2012
367
CAGR CAGR CAGR

Frozen vegetables
Belgium -0,068 -0,088 -0,096 -0,093 -0,046 -4,7% 9,3% -16,8%
Czech Republic -0,006 0,005 0,075 0,068 0,088 #NUM! #NUM! 4,0%
Denmark 0,067 0,036 0,024 0,005 -0,127 #NUM! -22,2% #NUM!
Finland -0,029 0,061 0,041 0,146 0,152 #NUM! #NUM! 38,5%
France 0,005 -0,004 -0,010 -0,076 -0,114 #NUM! #NUM! 85,1%
Germany 0,019 0,057 -0,140 -0,125 -0,106 #NUM! #NUM! -6,9%
Hungary 0,174 0,041 0,127 0,085 0,117 -4,9% -7,6% -2,1%
Italy -0,168 -0,194 -0,227 -0,293 -0,334 9,0% 7,8% 10,2%
Netherlands 0,235 0,236 0,133 0,109 0,155 -5,0% -13,3% 4,0%
Poland -0,248 -0,236 -0,109 -0,050 0,009 #NUM! -18,6% #NUM!
Portugal -0,193 -0,231 -0,164 -0,176 -0,159 -2,3% -4,0% -0,7%
Romania 0,452 0,307 0,407 0,398 0,548 2,4% -2,6% 7,7%
Spain -0,077 -0,077 -0,026 0,002 -0,224 14,2% -23,7% 71,0%
United Kingdom -0,213 -0,232 -0,198 -0,168 -0,175 -2,4% -1,8% -3,0%
Average -0,010 -0,025 -0,027 -0,025 -0,035 16,6% 27,9% 6,2%
Measure of Imbalance Frozen vegetables
0,600
0,500
0,400
0,300
0,200
0,100
0,000
-0,100
-0,200
-0,300
-0,400
2012
368
Annexes
CAGR CAGR CAGR

Fruit Juices
Belgium 0,503 0,480 0,462 0,463 0,498 -0,1% -2,1% 1,9%
Czech Republic 0,068 0,134 0,180 0,153 0,170 12,2% 27,7% -1,5%
Denmark -0,192 -0,139 -0,164 -0,154 -0,145 -3,5% -3,9% -3,1%
Finland 0,172 0,253 0,248 0,187 0,210 2,6% 9,7% -4,1%
France 0,308 0,229 0,193 0,159 0,099 -13,3% -11,1% -15,4%
Germany 0,476 0,465 0,399 0,429 0,440 -1,0% -4,4% 2,5%
Hungary 0,280 0,169 0,007 -0,068 -0,079 #NUM! -60,3% #NUM!
Italy -0,049 -0,092 -0,110 -0,107 -0,087 7,4% 22,6% -5,9%
Netherlands 0,226 0,253 0,222 0,134 0,243 0,9% -0,4% 2,3%
Poland -0,316 -0,425 -0,306 -0,299 -0,137 -9,9% -0,8% -18,2%
Portugal 0,061 0,034 0,143 -0,020 0,010 -20,3% 23,9% -48,7%
Romania 0,134 -0,001 0,021 -0,012 0,164 2,6% -36,9% 66,8%
Spain 0,088 0,108 0,188 0,100 0,052 -6,4% 20,9% -27,6%
United Kingdom 0,569 0,467 0,359 0,310 0,301 -7,6% -10,9% -4,3%
Average 0,118 0,104 0,100 0,063 0,101 -1,9% -4,1% 0,3%
Measure of Imbalance Fruit Juices
0,700
0,600
0,500
0,400
0,300
0,200
0,100
0,000
-0,100
-0,200
-0,300
-0,400
2012
369
CAGR CAGR CAGR

Ham
Belgium 0,137 0,038 -0,047 -0,034 0,028 -18,0% #NUM! #NUM!
Czech Republic 0,064 0,076 0,106 0,125 0,163 12,3% 13,4% 11,3%
Denmark 0,237 0,038 0,012 -0,007 -0,078 #NUM! -52,4% #NUM!
Finland 0,234 0,338 0,343 0,336 0,328 4,3% 10,1% -1,1%
France 0,342 0,279 0,324 0,163 0,209 -6,0% -1,3% -10,4%
Germany -0,093 -0,008 -0,014 -0,049 -0,081 -1,6% -37,4% 54,5%
Hungary 0,184 -0,176 -0,231 -0,297 -0,249 #NUM! #NUM! 1,9%
Italy 0,736 0,708 0,723 0,656 0,608 -2,4% -0,4% -4,2%
Netherlands 0,837 0,820 0,754 0,530 0,796 -0,6% -2,6% 1,4%
Poland -0,292 -0,239 -0,123 -0,074 -0,012 -32,6% -19,3% -43,6%
Portugal -0,150 -0,206 -0,212 -0,160 -0,154 0,3% 9,1% -7,8%
Romania 0,070 -0,127 0,006 0,066 0,223 15,6% -45,9% 147,2%
Spain 0,041 -0,009 0,186 0,233 0,050 2,5% 46,1% -28,1%
United Kingdom 0,648 0,690 0,680 0,651 0,619 -0,6% 1,2% -2,3%
Average 0,168 0,102 0,104 0,094 0,109 -5,2% -11,4% 1,4%
Measure of Imbalance Ham
1,000
0,800
0,600
0,400
0,200
0,000
-0,200
-0,400
2012
370
Annexes
CAGR CAGR CAGR

Ice Cream
Belgium -0,044 -0,085 -0,129 -0,163 -0,132 14,7% 30,9% 0,5%
Czech Republic 0,222 0,311 0,339 0,340 0,360 6,2% 11,2% 1,5%
Denmark 0,135 0,154 0,159 0,061 0,066 -8,6% 4,2% -19,8%
Finland -0,093 0,009 0,015 0,038 -0,054 -6,5% #NUM! #NUM!
France -0,067 -0,050 -0,084 -0,126 -0,168 12,2% 6,0% 18,7%
Germany -0,159 -0,103 -0,089 0,008 0,017 #NUM! -13,5% #NUM!
Hungary -0,430 -0,381 -0,223 -0,260 -0,264 -5,9% -15,1% 4,4%
Italy 0,387 0,454 0,492 0,604 0,679 7,3% 6,2% 8,4%
Netherlands -0,221 -0,251 -0,380 -0,535 -0,470 9,9% 14,5% 5,5%
Poland -0,214 -0,137 -0,034 0,025 0,110 #NUM! -36,7% #NUM!
Portugal -0,347 -0,396 -0,411 -0,431 -0,426 2,6% 4,3% 0,9%
Romania 0,270 0,036 0,041 0,038 0,180 -4,9% -37,5% 44,6%
Spain -0,351 -0,329 -0,260 -0,278 -0,318 -1,2% -7,2% 5,2%
United Kingdom -0,054 -0,046 -0,020 -0,048 -0,088 6,4% -21,6% 44,3%
Average -0,112 -0,109 -0,107 -0,139 -0,136 2,4% -1,2% 6,1%
Measure of Imbalance Ice Cream
0,800
0,600
0,400
0,200
0,000
-0,200
-0,400
-0,600
2012
371
CAGR CAGR CAGR

Milk
Belgium -0,070 -0,056 -0,219 -0,184 -0,168 11,5% 32,9% -6,4%
Czech Republic 0,155 0,180 0,230 0,252 0,194 2,8% 10,3% -4,2%
Denmark -0,400 -0,324 -0,385 -0,382 -0,377 -0,7% -1,0% -0,5%
Finland -0,267 -0,120 -0,073 -0,047 -0,027 -24,7% -27,7% -21,7%
France -0,180 -0,197 -0,191 -0,254 -0,229 3,1% 1,5% 4,6%
Germany 0,380 0,353 0,280 0,333 0,255 -4,9% -7,4% -2,3%
Hungary -0,180 -0,336 -0,208 -0,166 -0,162 -1,4% 3,6% -6,1%
Italy -0,094 -0,145 -0,072 -0,034 -0,025 -15,1% -6,6% -22,9%
Netherlands 0,066 0,070 -0,304 -0,202 -0,129 #NUM! #NUM! -19,3%
Poland -0,028 -0,075 -0,120 -0,142 -0,067 11,5% 43,8% -13,5%
Portugal -0,236 -0,202 -0,124 -0,076 -0,247 0,6% -14,8% 18,7%
Romania 0,537 0,208 0,152 0,247 0,391 -3,9% -27,1% 26,7%
Spain -0,035 -0,013 0,106 -0,003 -0,012 -13,1% #NUM! #NUM!
United Kingdom -0,019 0,023 0,098 0,154 0,183 #NUM! #NUM! 17,1%
Average -0,103 -0,092 -0,121 -0,092 -0,081 -2,9% 4,2% -9,6%
Measure of Imbalance Milk
0,600
0,400
0,200
0,000
-0,200
-0,400
-0,600
2012
372
Annexes
CAGR CAGR CAGR

Mineral water
Belgium 0,110 0,132 0,116 0,075 0,048 -9,8% 1,5% -19,8%
Czech Republic -0,306 -0,428 -0,318 -0,274 -0,227 -3,7% 1,0% -8,1%
Denmark -0,103 -0,051 -0,077 -0,114 -0,105 0,3% -7,1% 8,1%
Finland 0,102 0,219 0,205 0,191 0,207 9,3% 19,1% 0,3%
France -0,099 -0,142 -0,150 -0,161 -0,234 11,3% 11,0% 11,6%
Germany 0,434 0,481 0,528 0,606 0,640 5,0% 5,0% 4,9%
Hungary 0,059 0,064 0,162 0,073 0,156 12,9% 28,6% -0,9%
Italy -0,040 -0,056 -0,062 -0,047 -0,106 12,9% 11,3% 14,4%
Netherlands -0,210 -0,161 -0,192 -0,154 0,039 #NUM! -2,2% #NUM!
Poland 0,127 0,026 0,098 0,096 0,099 -3,0% -6,2% 0,4%
Portugal 0,400 0,292 0,373 0,409 0,411 0,4% -1,7% 2,4%
Romania 0,238 0,009 -0,070 -0,078 0,120 -8,2% #NUM! #NUM!
Spain 0,179 0,215 0,273 0,287 0,317 7,4% 11,1% 3,9%
United Kingdom -0,213 -0,222 -0,186 -0,188 -0,154 -4,0% -3,4% -4,6%
Average -0,014 -0,023 -0,005 0,004 0,041 #NUM! -23,7% #NUM!
Measure of Imbalance Mineral water
0,800
0,600
0,400
0,200
0,000
-0,200
-0,400
2012
373
CAGR CAGR CAGR

Savoury snacks
Belgium -0,080 -0,245 -0,316 -0,338 -0,342 19,9% 40,9% 2,0%
Czech Republic -0,219 -0,137 -0,022 -0,142 -0,080 -11,9% -44,0% 38,7%
Denmark -0,088 -0,028 0,007 -0,144 -0,177 9,1% #NUM! #NUM!
Finland -0,031 0,107 0,112 0,209 0,197 #NUM! #NUM! 15,3%
France 0,195 0,060 -0,041 -0,070 -0,108 #NUM! #NUM! 27,4%
Germany -0,102 -0,014 0,031 -0,036 -0,025 -16,0% #NUM! #NUM!
Hungary 0,068 0,069 0,020 -0,038 0,001 -38,1% -26,5% -47,8%
Italy 0,373 0,302 0,263 0,238 0,228 -6,0% -8,4% -3,5%
Netherlands 0,191 -0,052 -0,206 -0,257 -0,152 #NUM! #NUM! -7,3%
Poland -0,044 -0,016 0,063 0,104 0,169 #NUM! #NUM! 28,1%
Portugal 0,000 0,000 0,061 0,012 -0,012 #NUM! 245,6% #NUM!
Romania 0,077 -0,079 -0,055 -0,098 0,041 -7,6% #NUM! #NUM!
Spain -0,248 -0,243 -0,195 -0,238 -0,266 0,9% -5,9% 8,1%
United Kingdom -0,130 -0,103 -0,098 -0,069 -0,082 -5,6% -6,8% -4,4%
Average -0,018 -0,045 -0,054 -0,087 -0,071 18,5% 31,2% 7,0%
Measure of Imbalance Savoury snacks
0,500
0,400
0,300
0,200
0,100
0,000
-0,100
-0,200
-0,300
-0,400
2012
374
Annexes
CAGR CAGR CAGR

Soft drinks
Belgium -0,294 -0,296 -0,291 -0,266 -0,259 -1,5% -0,2% -2,9%
Czech Republic 0,045 0,116 0,250 0,227 0,222 22,0% 53,2% -2,9%
Denmark 0,128 0,149 0,120 0,106 0,084 -5,1% -1,5% -8,5%
Finland 0,171 0,322 0,349 0,370 0,393 11,0% 19,6% 3,0%
France -0,300 -0,334 -0,362 -0,384 -0,408 3,9% 4,8% 3,0%
Germany -0,186 -0,122 -0,091 0,049 0,074 #NUM! -16,5% #NUM!
Hungary -0,234 -0,234 -0,151 -0,177 -0,119 -8,0% -10,3% -5,8%
Italy -0,030 -0,100 -0,134 -0,139 -0,169 23,8% 44,8% 6,0%
Netherlands 0,348 0,381 0,246 0,214 0,312 -1,3% -8,4% 6,2%
Poland 0,046 0,099 0,231 0,256 0,284 25,6% 49,6% 5,4%
Portugal 0,019 -0,003 -0,005 -0,019 -0,051 #NUM! #NUM! 80,8%
Romania -0,115 -0,197 -0,234 -0,259 -0,153 3,7% 19,4% -10,0%
Spain -0,509 -0,473 -0,406 -0,402 -0,427 -2,2% -5,5% 1,3%
United Kingdom -0,055 -0,051 -0,056 -0,034 -0,014 -15,6% 0,2% -28,9%
Average -0,089 -0,071 -0,065 -0,061 -0,046 -8,0% -7,5% -8,5%
Measure of Imbalance Soft drinks
0,500
0,400
0,300
0,200
0,100
0,000
-0,100
-0,200
-0,300
-0,400
-0,500
-0,600
2012
375
CAGR CAGR CAGR

Tea
Belgium 0,230 0,178 0,136 0,127 0,066 -14,5% -12,4% -16,5%
Czech Republic 0,066 0,190 0,314 0,302 0,307 21,1% 47,5% -0,5%
Denmark 0,001 0,015 0,033 0,039 0,031 48,7% 124,1% -1,4%
Finland 0,103 0,181 0,176 0,198 0,202 8,8% 14,4% 3,5%
France -0,171 -0,185 -0,187 -0,197 -0,215 2,9% 2,2% 3,6%
Germany 0,009 0,037 0,016 0,104 0,113 37,7% 16,2% 63,2%
Hungary -0,129 -0,157 -0,160 -0,140 -0,122 -0,7% 5,6% -6,5%
Italy 0,026 -0,021 -0,036 -0,027 -0,052 #NUM! #NUM! 10,0%
Netherlands -0,204 -0,143 -0,176 -0,205 -0,136 -4,9% -3,6% -6,2%
Poland -0,068 -0,164 -0,099 -0,153 -0,149 10,4% 9,9% 10,9%
Portugal -0,012 -0,033 0,031 0,019 -0,022 7,9% #NUM! #NUM!
Romania 0,319 0,049 -0,030 -0,078 0,111 -12,3% #NUM! #NUM!
Spain -0,016 -0,029 0,046 0,066 0,070 #NUM! #NUM! 10,8%
United Kingdom 0,004 -0,022 0,016 0,036 0,048 36,6% 41,8% 31,5%
Average -0,004 -0,008 0,004 0,006 0,016 #NUM! #NUM! 44,8%
Measure of Imbalance Tea
0,400
0,300
0,200
0,100
0,000
-0,100
-0,200
-0,300
2012
376
Annexes
CAGR CAGR CAGR

Yoghurt
Belgium 0,096 -0,083 -0,145 -0,164 -0,089 #NUM! #NUM! -11,4%
Czech Republic -0,079 0,092 0,171 0,157 0,223 #NUM! #NUM! 6,9%
Denmark -0,475 -0,454 -0,436 -0,455 -0,458 -0,4% -2,1% 1,2%
Finland -0,175 -0,069 -0,048 -0,004 0,022 #NUM! -27,6% #NUM!
France -0,110 -0,281 -0,319 -0,321 -0,356 15,8% 30,5% 2,7%
Germany 0,064 0,134 0,117 0,201 0,246 18,4% 16,3% 20,5%
Hungary -0,052 -0,096 -0,075 -0,118 -0,151 14,2% 9,4% 19,3%
Italy -0,169 -0,167 -0,161 -0,174 -0,194 1,8% -1,2% 4,8%
Netherlands 0,096 0,193 -0,076 0,048 0,164 7,0% #NUM! #NUM!
Poland -0,165 -0,173 -0,099 -0,115 -0,096 -6,5% -11,9% -0,7%
Portugal -0,050 -0,060 -0,003 -0,016 -0,028 -6,8% -52,3% 82,1%
Romania 0,014 -0,096 -0,168 -0,201 -0,121 #NUM! #NUM! -7,8%
Spain -0,415 -0,574 -0,530 -0,527 -0,569 4,0% 6,3% 1,8%
United Kingdom -0,053 -0,007 -0,037 -0,066 -0,008 -21,4% -8,8% -32,3%
Average -0,140 -0,161 -0,175 -0,171 -0,150 0,9% 5,8% -3,7%
Measure of Imbalance Yoghurt
0,300
0,200
0,100
0,000
-0,100
-0,200
-0,300
-0,400
-0,500
-0,600
-0,700
2012
377
11.3. Annex C: Design of the econometric analysis
11.3.1. The general specification and choice of indicators

The objective of the econometric analysis is to analyse the historical evidence for the
impact of priori drivers on each of choice and innovation. The analysis models the
behaviour of each shop and the selection of products that it offers, and seeks to explain
this with reference to various national and local drivers. It is important to note that this
differs from modelling the total assortment available to consumers from the shops to
which they have access, which would include the impact of a change in the number and
mix of types of shops in the local area. The number and mix of shops is examined and
reported in the descriptive analysis of this study.
The relationships of interest are expressed below:
[choice or innovation]s,p,t = f {
shop types,t
shop sizes,t
private labels sharen/s,p,t
retailers' concentrationn/s,t
suppliers' concentrationn/s,p,t
[or imbalance (retailer vs supplier concentration)n/s,p,t]
socio-demographic indicatorc,t
rural/urban categoryc
product category turnovern,p,t
economic prosperityc/n,t
Member Staten
product categoryp
yeary
seasonm
new competitor shop openings,t
}
where the indices used are:

c consumer shopping area
m month in the year (2nd quarter or 4th quarter),
n Member State
p product category
s shop
t time period (two per year, every second year)
y year
378
Annexes
Variables and alternative indicators

The following tables note the alternative empirical indicators used to represent the
conceptual variables in the broad specification outlined above. Alternative measures of
choice and innovation are generally shown to be moderately or strongly correlated (see
Table 32 and Table 33). The stronger the correlation, the more we expect the estimation
results for the different measures to be broadly similar. However, correlation between
national and local retail concentration is low (see Table 34, Table 35 and Table 36)
because national measures do not vary across shops in the same country. This is also
the case for supplier concentration.
379
Table 32: Correlations between choice variables (long data set)
Choice Product Variety Product Price Product Size

Variety Variety
Product Price Variety -0.23
Product Size Variety 0.76 -0.18
Product Supplier Variety 0.64 0.1 0.55
Table 33: Correlations between innovation variables (long data set)
Innovation Opus New Product New New Formula

Innovations Packaging
New Product 0.74
New Packaging 0.43 0.52
New Formula 0.62 0.52 0.43
New Range extension 0.78 0.74 0.52 0.60
Table 34: Correlations between national and local supplier concentrations (long data
set)
Supplier National National

National C5 HHI full National C5 HHI brand
full market market brand only only Local C5
National HHI full
market 0.87
National C5 brand
only 0.75 0.68
National HHI
brand only 0.63 0.83 0.81
Local C5 0.21 0.29 0.4 0.41
Local HHI 0.04 0.12 0.24 0.26 0.73
380
Annexes
Table 35: Correlations between national and local retail concentrations (long data set)
Retail Local C5 Local HHI Local C5 Local HHI

Concentration Floorespace Floorspace Local C5 Local HHI Shop Share Shop share Local C5 Local HHI
(Banner) (Banner) Floorspace Floorspace (Banner) (Banner) Shop share Shop share
National group
C5 Edible
Grocery 0.16 0.07 0.21 0.03 0.17 0.1 0.21 0.13
National group
HHI Edible
Grocery 0.15 0.06 0.19 0.02 0.17 0.1 0.2 0.12
National group
C5 Modern
Retail 0.2 0.1 0.29 0.09 0.18 0.11 0.27 0.14
National group
HHI Modern
Retail 0.12 0.01 0.02 -0.08 0.23 0.14 0.14 0.06
National
banner C5
Edible Grocery 0.16 0.06 0.19 0.02 0.18 0.11 0.2 0.12
National
banner HHI
Edible Grocery 0.13 0.03 0.1 -0.04 0.18 0.11 0.16 0.09
National
banner C5
Modern Retail 0.17 0.06 0.17 0.02 0.21 0.13 0.21 0.1
National
banner HHI
Modern Retail 0.03 -0.05 -0.17 -0.18 0.2 0.12 0.01 -0.02
381
Table 36: Correlations between selected measures of national and local retail concentrations (long data set)
Retail Concentration National National

group HHI banner HHI National Local HHI Local HHI
Edible National HHI Edible banner HHI Floorspace Local HHI Shop share
Grocery Modern Retail Grocery Modern Retail (Banner) Floorspace (Banner)
National group HHI Modern
Retail 0.67
National banner HHI Edible

Grocery 0.97 0.77
National banner HHI Modern

Retail 0.44 0.89 0.63
Local HHI Floorspace

(Banner) 0.06 0.01 0.03 -0.05
Local HHI Floorspace 0.02 -0.08 -0.04 -0.18 0.87
Local HHI Shop share

(Banner) 0.1 0.14 0.11 0.12 0.84 0.69
Local HHI Shop share 0.12 0.06 0.09 -0.02 0.68 0.81 0.75
382
Annexes
Table 37: Variables and alternative indicators
Conceptual variable Empirical indicator Units

Choice
a Product variety (no. of unique products in given shop) #
b Product size variety (no. of unique product sizes in a given product #
category in a given shop)
c Product supplier variety (No. of unique brand owners) #
d Product price variety (Average of coefficient of prices across a #
category, shop over time)
Innovation a Innovation (new EAN codes) observed in shop sample using Nielson #
Opus data
b New products (see Mintel GNDP definition) %
c New packaging (see Mintel GNDP definition) %
d New formulation/packaging (see Mintel GNDP definition) %
e New range extensions (see Mintel GNDP definition) %
Shop type Shop type dummy (base: HM) #
Shop size Shop size m2
Private labels share a Private labels SKU share in shops %
b Private labels national market sales share %
Retailers' concentration a Retail concentration (HHI) at local level - % of shops and % of Value
floorspace
b Retail concentration (c(k)) at local level - % of shops and % of %
floorspace
c Retail concentration (HHI) at national level - % market share at Value
banner level and retail group level
d Retail concentration (c(k)) at national level - % market share at %
banner level and retail group level
Suppliers' concentration a Supplier concentration (HHI) at local level - % of SKUs Value
383
Conceptual variable Empirical indicator Units

b Supplier concentration (c(k)) at local level - % of SKUs %
c Supplier concentration (HHI) at national level - % market share Value
d Supplier concentration (c(k)) at national level - % market share %
Measure of imbalance (ratio of concentration) (included Ratio of retail concentration HHI to supplier concentration HHI Value
as an alternative to showing retailers’ and suppliers’
concentration separately)
Socio-demographic indicator a Population size #
b GDP per capita #
Rural/urban type a Population density #
b Rural/intermediate/urban dummy (base: PU) #
Product category turnover Product category turnover at national level € million
Economic prosperity a Unemployment rate (by region and time period) %
b Retail business expectations for the next 3 months, converted to an Index
index where 100 corresponds to ‘no change’
c Unemployment rate %
Country Dummy (base: Italy) #
Product category Dummy (base: first product category – baby food) #
Year Dummy (base: first year – 2004 or 2008) #
nd
Season Dummy (base: season 1 – 2 quarter) #
New shop opening Dummy (= 1 if a new shop opened in the catchment area) Dummy
For dummy variables, ‘base’ indicates the category for which a dummy variable will not be included (to avoid multicollinearity), and so
the ‘base’ equation (prior to the addition of dummy effects) will represent this category.
In all cases the variables appear in log linear transformation, except for dummies which are just linear
384
Annexes
11.4. Annex D: The data sets

The observations in the data set span three dimensions by shop, product and time
period although some drivers do not vary over all of these dimensions (for example,
some national drivers vary only between Member States and over time). In addition,
the sample was limited by the need for Trade Dimensions data for all time periods for
the calculation of local retail concentration.
The final data sets are balanced panel data sets including all indicators and drivers.
Due to variation in the availability of data, two data sets were used; a long data set
covering the period 2004H1 to 2012H2 and a short data set covering the period
2008H1 to 2012H2 but with more Member States. The econometric analysis was
performed on both of these data sets. The Table 38 below illustrates the difference in
coverage between the two data sets:
Table 38: Country and shop coverage in short and long data sets
Long Data set No. of shops Short Data set No. of shops
(2004H1 - 2012H2) (2008H1 - 2012H2)
Italy 80 Italy 83
Spain 42 Spain 42
France 131 Belgium 9
Portugal 19 France 131
Poland* 24 Portugal 19
Poland 29
Hungary 24
Total 296 Total 337
Single Member State estimates

For countries where there are sufficient observations (at least 10 shops which
excludes Belgium), it was possible to estimate an equation for that country alone,
which allows the parameter estimates for all drivers to change (whereas when the
data are pooled across countries the only country-specific parameter is the country
dummy). However, this excludes from the analysis a comparison across Member
States and this is important for national drivers that only vary over time and Member
States.
11.5. Annex E: Econometric estimation issues

Our choice of econometric estimation methods needs to take account of certain issues
that may be present in the process that we are modelling.
Unobserved heterogeneity among shops

This is the standard issue that arises with data where the unit of observation is an
individual (a shop, in this case). It considers the possibility that there is some
difference between the observed outcome for choice/innovation that is due to
385
something specific about the shop that is not already captured in the drivers. In a
pure cross section there is no way of identifying such effects, but in panel data (where
indicators are measured for the same shops over different time periods) it is
conventional to seek to use the information available for shops over time to detect
such (time-invariant) effects and thereby improve the estimates of the effects of the
observed drivers. Since the shops are a sample drawn from a wider population, we
prefer to use a random effects specification if the data support this (Hausman test),
but we also calculate the fixed-effects (within) estimator.
Spatial dependence
The literature on spatial econometrics identifies different kinds of spatial dependence
which call for different methods. By spatial dependence we mean the possibility that
outcomes in a shop are affected not just by the characteristics of that shop (including
the area/country in which it lies) but by the behaviour of nearby shops and/or the
characteristics of nearby areas.
Depending on the nature of the spatial dependence that is present, if we do not apply
a method that takes account of such dependence then the result may be that the
standard errors for parameter estimates are incorrectly estimated (so that we are
misled in our assessment of the statistical significance of our parameter estimates for
the drivers) or that the parameter estimates themselves are incorrectly estimated (so
that we incorrectly attribute an influence to a given driver).
A spatial weight matrix, W, is given by assumption, which measures any given shop’s
spatial dependence on every other shop. Conventionally this is constructed as a
declining function of distance (often the reciprocal of the square of distance is used),
so that nearby shops are assumed to have a large influence and distant shops to have
negligible influence. This spatial weight matrix is then used both to test for spatial
dependence and in methods that seek to account for that dependence.
We use Moran’s I to test for spatial dependence in estimated residuals. This provides a
diagnostic suggesting misspecification in an equation that does not adequately account
for spatial dependence. Moran’s I is calculated for cross sections and is used to for
every estimated equation. The spatial econometrics literature developed methods to
address various kinds of spatial dependence in cross sections:
 spatial lag of exogenous variables
 spatial correlation of residuals
 spatial lag of endogenous variables
A specification of spatial lag of exogenous variables can then be estimated using
ordinary least squares by including additional regressors. If X is the (N x k) matrix of
regressors then WX is the matrix of spatially lagged regressors (where N is the
number of shops and k is the number of regressors).
Estimation of models that assume spatial lag of endogenous variables and spatial
correlation of residuals (the so-called SARAR specification) requires a more
sophisticated estimation technique (maximum likelihood, two-stage least squares or
general method of moments).
With the growth in popularity of panel data approaches, the spatial econometrics
methods that were originally designed for cross-sections have been extended to panel
data applications. The software to implement such methods has been developed and
made available by some academics as an extension of existing software (such as
386
Annexes
Stata, R or MATLAB), but not all such libraries are sufficiently general to cope with the
dimensions that are present in our data set (disaggregation over time, space and also
by product type).
In practice, when we undertook the estimation work the Hausman test was rejected in
most specifications indicating that the random effects model is inconsistent and may
not approach the true value even as sample size increases. However, the fixed-effects
estimator proved to be more vulnerable to spatial dependence, and so we have
reported and drawn on both types of estimator in summarising conclusions about the
impacts of the drivers.
A particular form of spatial dependence arises when it is believed that the residuals
(which capture all the reasons for variation in the dependent variable that are not
accounted for by the drivers that have been included) could be ‘clustered’, that is
related to one another by geographical area. The shops in this study are located in
common consumer shopping areas and the possibility arises that there are unobserved
(i.e. not taken into account in the indicators that are included in the analysis)
influences at the local level that affect all shops in the same area. In that case the
estimated standard errors associated with each parameter estimate, which are used to
assess whether it is statistically significantly different from zero, would be
underestimated if no allowance were made for clustering. The results reported here
use standard errors estimated on the assumption of clustering at the CSA level so as
to take a cautious approach to reporting statistical significance of results. In many
cases the parameter estimates that are treated as statistically insignificant as a result
of taking this approach are those that are in any case so small as to be economically
irrelevant.
11.6. Annex F: Results of the econometric analysis
The following discussion on the results of the econometric analysis is organised around
key testable hypothesis. The hypotheses are based on expectations that emerged
from the descriptive analysis.
11.6.1. Choice
Hypothesis: Retail concentration at procurement level is a driver of the evolution of
choice in all its components
Most models with the exception of those exploring the effect on product price variety
indicate evidence of a positive effect of national retail concentration: as concentration
among retailers at national level has increased, so has choice. On the other hand, the
evidence indicates a negative effect for product price variety. Also, the effect for
product supplier variety is not statistically significant for the fixed effects model.
Hypothesis: The growing emergence of private labels, in part due to the increased
presence of discount stores appears to have played a role in the evolution of choice
A statistically significant positive effect of private labels was estimated for product
variety, product size variety (not significant in the short period) and product supplier
variety, and a negative effect for product price variety, but in all cases the size was
very small.
387
Hypothesis: The economic crisis has negatively impacted the evolution of choice in all
its components
The estimated impact of the unemployment rate on choice was positive, rather than
negative, but in any case small.This hypothesis is largely not supported in the results.
The exception to this is the price variety measure where a small negative effect was
found. However, the equations also include a generally positive impact of GDP per
capita as a measure of prosperity, and so the expected negative impact of the
economic crisis comes through this measure.
Hypothesis: Shop type has strongly impacted the level and evolution of choice in all its
components.
The evidence supports this hypothesis for product variety, product size variety and
product supplier variety but the evidence is less clear for product price variety. The
‘base’ for shop type is hypermarket: the estimated impacts indicate that for all but
product price variety, supermarkets and hard discounters broadly offer less choice
than hypermarkets, and hard discounters offer less choice than supermarkets for
product variety, product size variety and product supplier. In contrast, in the case of
product price variety the (negative) hard discounter effect was not generally larger
than the (negative) supermarket effect (both compared to hypermarkets).
Hypothesis: National product category turnover appears to have an impact on the

evolution of choice in all its components
National product category turnover, which can be conceptualised as market size, is

shown to have a large positive effect on product variety, product size variety and
product supplier variety although the effect on product price variety is negative, small
and in the case of one model insignificant. The effect on the measures other than
product price variety is generally smaller in the short period and in the case of product
size variety no longer significant in the random effects model. However, the results
generally suggest that much more choice is provided in product categories with larger
turnover, but the choice of prices available to consumers is somewhat smaller.
Hypothesis: Supplier concentration at procurement level is a driver of the evolution of

The results do not support this hypothesis. The evidence is mixed: the impacts are
small, not always statistically significant and vary in sign.
Hypothesis: Measure of imbalance at procurement level is a driver of the evolution of

A very small positive impact from the imbalance between retailer and supplier
concentrations on product variety was found. In contrast, evidence is found to indicate
a negative effect of imbalance on product size variety and product price variety
388
Annexes
suggesting as the concentration of retailers relative to suppliers increases, the variety

of product sizes decreases and retailers reduce the variety of product prices. But the
size of the effects is small.
Hypothesis: Average population size, average population density, GDP per capita and
new shop opening are drivers of the evolution of choice in all its components
The effect of average (over time) population size is found to be have no statistically
significant effect on all but the product price variety measure where it is found to
have a positive effect. However, average (over time) population density is found to
have a negative effect in all models with the exception of the product price variety
models. This would suggest less choice in more densely populated CSAs but should be
taken in the context of the findings for GDP per capita. The impact of GDP per capita is
found to be broadly positive in all but the product price variety model where the
evidence is mixed. This suggests two offsetting effects, since the more densely
populated areas (cities) tend also to have higher GDP per capita. The more affluent
the local economy in the CSA the more choice but more densely populated areas will
have less choice.
The opening of a new shop has a positive effect on all choice indicators in the existing
shops although the effect is often insignificant in the short period. Generally, the
results suggest that existing shops increase the choice available to consumers when
faced with the competition provided by the opening of a new shop in the same area.
To face a new competitor, established retailers seek to retain customer loyalty; they
modify the product assortment and potentially extend their product offer by including
products the competitors are offering that they do not currently stock and/or offering
new products to better satisfy existing customers.
389
Table 39: Results - Product Variety
Product Variety
Long Short Long Long Long Long Long Long Short Long Long Long Long Long
period RE Period RE period RE period RE period period RE period RE period FE Period FE period FE period FE period period FE period FE
(Separate (Separate (Imbalanc (Separate RE (Separate (Separate (Separate (Separate (Imbalanc (Separate FE (Separate (Separate
concentrat concentrat e) †† concentrat (Separa concentrat concentrat concentrat concentrat e) †† concentrat (Separa concentrat concentrat
ion ion ion te ion ion ion ion ion te ion ion
measures) measures) measures) concent measures) measures) measures) measures) measures) concent measures) measures)
† † † ration † † † † † ration † †
measur measur
es) † es) †
Local Private
labels share 0.014*** 0.006*** 0.015*** 0.015*** 0.014*** - 0.016*** 0.013*** 0.005** 0.013*** 0.014*** 0.013*** - 0.014***
(0.002) (0.002) (0.002) (0.002) (0.002) - (0.003) (0.003) (0.002) (0.003) (0.003) (0.003) - (0.003)
National
Private labels
share - - - -0.006 - - - - -0.001 -
- - - (0.007) - - - - (0.007) -
National
retail
concentratio
n HHI
(group,
edible
groceries) 0.144*** 0.216*** - - 0.143*** 0.161*** - 0.074*** 0.186*** - - 0.078*** 0.087*** -
(0.02) (0.041) - - (0.02) (0.02) - (0.021) (0.048) - - (0.021) (0.021) -
National
retail
concentratio
n HHI
(group,
modern
retail) - 0.143** - - - 0.013 - -
- (0.058) - - - (0.058) - -
Local retail - - - - -0.034 - - - - -0.051

concentratio
390
Annexes
Product Variety
† † † ration † † † † † ration † †
measur measur
es) † es) †
n HHI
(group,
floorspace)
- - - - (0.03) - - - - (0.04)
National
supplier
concentratio
n HHI (full
market) -0.003 -0.018 -0.002 0.003 -0.012 -0.035*** -0.057** -0.044*** -0.024** -0.049***
(0.009) (0.014) (0.01) (0.008) (0.01) (0.013) (0.022) (0.012) (0.012) (0.013)
National
supplier
concentratio
n HHI
(brands
only) - 0.013 - -0.004
- (0.013) - (0.019)
Imbalance - - 0.004 - - - - - - 0.005 - - - -
- - (0.017) - - - - - - (0.023) - - - -
Average
Population
density -0.086*** -0.045*** -0.117*** -0.108*** -0.085*** -0.088*** -0.092*** - - - - - - -
(0.011) (0.013) (0.018) (0.014) (0.011) (0.012) (0.013) - - - - - - -
Shop floor
space 0.253*** 0.219*** 0.249*** 0.252*** 0.254*** 0.256*** 0.253*** 0.246*** 0.153*** 0.231*** 0.236*** 0.244*** 0.242*** 0.243***
391
Product Variety
† † † ration † † † † † ration † †
measur measur
es) † es) †
(0.017) (0.02) (0.018) (0.018) (0.017) (0.018) (0.055) (0.056) (0.055) (0.057) (0.055)
Average
Population 0.016 0.036 0.016 0.016 0.016 0.016 0.021 - - - - - - -
(0.023) (0.023) (0.026) (0.025) (0.023) (0.024) (0.024) - - - - - - -
Unemployme
nt 0.024 0.062*** 0.056** 0.056** 0.024 0.032 0.07*** 0.049** 0.075*** 0.071*** 0.066*** 0.052** 0.055*** 0.068**
(0.022) (0.02) (0.025) (0.024) (0.022) (0.022) (0.025) (0.021) (0.018) (0.024) (0.023) (0.021) (0.02) (0.026)
Regional
GDP per
Capita 0.389*** 0.164*** 0.571*** 0.519*** 0.386*** 0.411*** 0.477*** 0.684*** 0.331** 0.822*** 0.819*** 0.675*** 0.702*** 0.771***
(0.064) (0.05) (0.05) (0.053) (0.064) (0.064) (0.068) (0.094) (0.162) (0.073) (0.074) (0.095) (0.095) (0.09)
National
Product
Category
Turnover 0.36*** 0.228*** 0.42*** 0.405*** 0.362*** 0.368*** 0.429*** 0.432*** 0.268*** 0.476*** 0.476*** 0.43*** 0.441*** 0.468***
(0.022) (0.024) (0.024) (0.024) (0.023) (0.021) (0.023) (0.043) (0.044) (0.046) (0.047) (0.043) (0.043) (0.049)
Supermarket
Dummy -0.179*** -1.393*** -0.179*** -0.182*** -0.179*** -0.177*** -0.179*** -0.138** -0.097** -0.132** -0.134** -0.137** -0.134** -0.138**
(0.035) (0.085) (0.036) (0.035) (0.035) (0.035) (0.034) (0.052) (0.04) (0.053) (0.053) (0.052) (0.053) (0.052)
Hard
Discounter
Dummy -1.277*** -0.364*** -1.293*** -1.286*** -1.277*** -1.267*** -1.256*** -1.103*** -1.098*** -1.122*** -1.116*** -1.106*** -1.104*** -1.112***
(0.11) (0.092) (0.11) (0.111) (0.11) (0.109) (0.116) (0.047) (0.049) (0.047) (0.048) (0.047) (0.048) (0.045)
392
Annexes
Product Variety
† † † ration † † † † † ration † †
measur measur
es) † es) †
New shop
opening 0.085*** 0.012 0.097*** 0.099*** 0.084*** 0.087*** 0.091*** 0.065*** 0.006 0.065*** 0.065*** 0.065*** 0.067*** 0.062***
(0.011) (0.016) (0.011) (0.012) (0.011) (0.011) (0.012) (0.011) (0.017) (0.011) (0.011) (0.011) (0.012) (0.011)
Seasonal
Dummy 0.026*** 0.022*** 0.026*** 0.026*** 0.026*** 0.026*** 0.021*** 0.026*** 0.022*** 0.026*** 0.026*** 0.026*** 0.026*** 0.021***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
BIC - - - - - - - - -20502.9 - - - - -
2
Within R 0.261 0.073 0.259 0.259 0.261 0.255 0.216 0.265 0.075 0.263 0.263 0.265 0.258 0.22
Between R2 0.826 0.811 0.816 0.819 0.826 0.823 0.849 0.45 0.469 0.415 0.411 0.457 0.452 0.43
Overall R2 0.777 0.777 0.768 0.771 0.777 0.774 0.797 0.433 0.451 0.4 0.396 0.439 0.434 0.412
Hausman 625.98**
Test 431.54*** 252.92*** 1245.71*** 1110.99*** * 640.58*** 772.79*** - - - - - - -
Moran’s I (0.172 - (0.18 - (0.171 - (0.181 - (0.115 - (0.413 -

(0.477 - (0.185 - (0.233 - (0.411 - (0.416 - (0.41 - (0.412 - (0.372 -
(Range) 0.357) 0.348) 0.344) 0.313)
0.348) 0.537) 0.36) 0.513) 0.297) 0.503) 0.505) 0.512) 0.512) 0.469)
Note: All specifications use standard errors derived by clustering on consumer shopping areas and include product and country fixed effects (not reported). Standard
errors are presented in parentheses in the row below each coefficient. *** indicates significant at the 1% level of significance, ** at the 5% and * at the 10%.
Moran’s I is calculated for each time period; the table shows the range of test statistics over the time periods, and the level of significance indicated is the average p-
value across the time periods. † Separate concentration measures to refers to models including both retailer and supplier concentration measures rather than just the
imbalance between the two. †† Imbalance refers to models which include only the measure of imbalance between retailer and supplier concentration measures
rather than both measures
393
Table 40: Results - Product Size Variety
Product Size Variety
(Separate (Separate (Imbalanc (Separate RE (Separate (Separate (Separate (Separate (Imbalanc (Separate FE (Separate (Separat
concentra concentra e) †† concentra (Separa concentra concentra concentra concentra e) †† concentra (Separ concentrati e
tion tion tion te tion tion tion tion tion ate on concentra
measures measures measures concent measures measures measures measures measures concen measures) tion
)† )† )† ration )† )† )† )† )† tration † measures
measur measu )†
es) † res) †
Local Private
labels share 0.01*** 0.002 0.011*** 0.011*** 0.01*** - 0.008*** 0.008** 0 0.008** 0.008** 0.008** - 0.006**
(0.003) (0.002) (0.003) (0.003) (0.003) - (0.002) (0.004) (0.002) (0.004) (0.004) (0.004) - (0.002)
National
Private
labels share - - - -0.015* - - - 0.005 -
- - - (0.008) - - - (0.006) -
National
retail
concentratio
n HHI
(group,
edible 0.048**
groceries ) 0.138*** 0.151*** - - 0.128*** 0.154*** - 0.062*** 0.15*** - - * 0.067*** -
(0.016) (0.048) - - (0.016) (0.017) - (0.019) (0.05) - - (0.017) (0.019) -
National
retail
concentratio
n HHI
(group,
modern
retail) - 0.131** - - - 0.014 - -
- (0.056) - - - (0.051) - -
Local retail - - - - -0.043** - - - -0.054*

concentratio
394
Annexes
measur measu )†
es) † res) †
n HHI
(group,
floorspace)
- - - - (0.021) - - - (0.031)
National
supplier
concentratio
n HHI (full
market) 0.085*** -0.02 0.085*** 0.089*** 0.03*** 0.09*** -0.012 - 0.082*** - 0.097*** 0.034**
(0.017) (0.022) (0.018) (0.017) (0.008) (0.023) (0.033) - (0.023) - (0.024) (0.014)
National
supplier
concentratio
n HHI
(brands 0.095**
only) - 0.067*** - *
- (0.021) - (0.032)
Imbalance - - -0.042** - - - - - - -0.054*** - - - -
- - (0.02) - - - - - - (0.018) - - - -
Average
Population
density -0.064*** -0.029*** -0.099*** -0.088*** -0.065*** -0.066*** -0.053*** - - - - - - -
(0.017) (0.011) (0.027) (0.021) (0.018) (0.018) (0.013) - - - - - - -
Shop floor 0.163**

space 0.181*** 0.167*** 0.176*** 0.179*** 0.182*** 0.182*** 0.167*** 0.164*** 0.069** 0.15*** 0.147*** * 0.162*** 0.162***
395
measur measu )†
es) † res) †
(0.018) (0.017) (0.019) (0.02) (0.019) (0.019) (0.015) (0.046) (0.033) (0.047) (0.046) (0.045) (0.046) (0.046)
Average
Population 0.012 0.011 0.012 0.011 0.012 0.011 0.013 - - - - - - -
(0.019) (0.016) (0.023) (0.022) (0.019) (0.019) (0.018) - - - - - - -
Unemploym
ent 0.056*** 0.092*** 0.076*** 0.086*** 0.046** 0.063*** 0.089*** 0.065*** 0.084*** 0.077*** 0.066** 0.052** 0.068*** 0.074***
(0.021) (0.018) (0.025) (0.022) (0.022) (0.02) (0.025) (0.022) (0.016) (0.023) (0.026) (0.023) (0.022) (0.026)
Regional
GDP per 0.383**
Capita 0.335*** 0.151*** 0.535*** 0.471*** 0.336*** 0.35*** 0.304*** 0.39*** 0.211 0.544*** 0.527*** * 0.4*** 0.47***
(0.076) (0.052) (0.09) (0.072) (0.077) (0.077) (0.063) (0.139) (0.136) (0.126) (0.125) (0.141) (0.139) (0.131)
National
Product
Category 0.565**
Turnover 0.283*** 0.051 0.343*** 0.33*** 0.282*** 0.289*** 0.408*** 0.559*** 0.381*** 0.605*** 0.606*** * 0.565*** 0.612***
(0.041) (0.064) (0.039) (0.04) (0.043) (0.04) (0.018) (0.046) (0.039) (0.048) (0.049) (0.046) (0.046) (0.053)
Supermarket
Dummy -0.126*** -0.795*** -0.125*** -0.128*** -0.126*** -0.124*** -0.138*** -0.101*** -0.052** -0.094** -0.092** -0.1*** -0.098** -0.1***
(0.026) (0.072) (0.027) (0.027) (0.026) (0.026) (0.023) (0.037) (0.026) (0.037) (0.037) (0.036) (0.036) (0.036)
Hard -
Discounter 0.635**
Dummy -0.776*** -0.289*** -0.791*** -0.785*** -0.773*** -0.77*** -0.851*** -0.641*** -0.638*** -0.656*** -0.652*** * -0.641*** -0.644***
(0.113) (0.091) (0.113) (0.114) (0.113) (0.112) (0.13) (0.035) (0.04) (0.036) (0.036) (0.035) (0.036) (0.034)
396
Annexes
measur measu )†
es) † res) †
New shop 0.073**

opening 0.089*** -0.003 0.097*** 0.101*** 0.088*** 0.091*** 0.106*** 0.075*** -0.002 0.073*** 0.07*** * 0.076*** 0.074***
(0.021) (0.017) (0.02) (0.021) (0.021) (0.022) (0.02) (0.025) (0.018) (0.024) (0.024) (0.024) (0.025) (0.025)
Seasonal 0.026**
Dummy 0.026*** 0.025*** 0.026*** 0.026*** 0.026*** 0.027*** 0.022*** 0.026*** 0.025*** 0.026*** 0.026*** * 0.027*** 0.022***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.002) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.002)
BIC - - - - - - - - 28679.7 56165.7 - - - -

2
Within R 0.097 0.016 0.096 0.259 0.097 0.096 0.143 0.1 0.019 0.099 0.099 0.1 0.099 0.148
Between R2 0.547 0.572 0.534 0.819 0.545 0.544 0.858 0.154 0.141 0.14 0.139 0.15 0.151 0.198
Overall R2 0.484 0.511 0.474 0.771 0.483 0.482 0.772 0.144 0.127 0.131 0.13 0.141 0.141 0.187
Hausman 610.02**
Test 628.07*** 196.80*** 1137.77*** 705.59*** * 549.36*** 483.5*** - - - - - - -
Moran’s I (0.313 - (0.413 -

(0.416 - (0.318 - (0.316 - (0.314 - (0.313 - (0.105 - (0.294 - (0.426 - (0.424 - (0.424 - (0.422 - (0.348 -
(Range)
0.491) 0.472) 0.495) 0.493) 0.492) 0.493) 0.363) 0.513) 0.354) 0.509) 0.51) 0.511) 0.512) 0.482)
397
Table 41: Results - Product Supplier Variety
Product Supplier Variety
period Period period period period period period period Period period period period period period
RE RE RE RE RE RE RE FE FE FE FE FE FE FE
(Separa (Separa (Imbala (Separa (Separa (Separa (Separa (Separa (Separa (Imbala (Separa (Separa (Separa (Separa
te te nce) †† te te te te te te nce) †† te te te te
concent concent concent concent concent concent concent concent concent concent concent concent
ration ration ration ration ration ration ration ration ration ration ration ration
measur measur measur measur measur measur measur measur measur measur measur measur
es) † es) † es) † es) † es) † es) † es) † es) † es) † es) † es) † es) †
Local
Private
labels
share 0.013*** 0.006*** 0.014*** 0.014*** 0.013*** - 0.016*** 0.014*** 0.007*** 0.014*** 0.014*** 0.014*** - 0.016***
(0.002) (0.002) (0.002) (0.002) (0.002) - (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) - (0.003)
National
Private
labels
share - - - 0.015*** - - - - 0.025*** -
- - - (0.005) - - - - (0.006) -
National
retail
concentrati
on HHI
(group,
edible
groceries ) 0.079*** 0.062** - - 0.078*** 0.086*** - 0.028 0.032 - - 0.027 0.03* -
(0.018) (0.032) - - (0.017) (0.017) - (0.018) (0.035) - - (0.018) (0.018) -
National
retail
concentrati
on HHI
(group,
modern
retail) - 0.071 - - - - -0.022 - - -
398
Annexes
- (0.047) - - - - (0.049) - - -
Local retail
concentrati
on HHI
(group,
floorspace) - - - - 0.012 - - - - 0.011
- - - - (0.027) - - - - (0.035)
National
supplier
concentrati
on HHI
(full
market) -0.006 -0.009 -0.006 0.002 -0.021** -0.004 -0.012 -0.01 0.011 -0.029*
(0.01) (0.011) (0.011) (0.01) (0.01) (0.016) (0.027) (0.017) (0.017) (0.016)
National
supplier
concentrati
on HHI
(brands
only) - 0.006 - 0.025
- (0.011) - (0.018)
Imbalance - - 0.002 - - - - - - -0.025 - - - -
- - (0.011) - - - - - - (0.017) - - - -
Average
Population
density -0.059*** -0.039** -0.076*** -0.072*** -0.059*** -0.061*** -0.066*** - - - - - - -
399
(0.013) (0.015) (0.019) (0.017) (0.012) (0.013) (0.015) - - - - - - -
Shop floor
space 0.18*** 0.123*** 0.177*** 0.179*** 0.18*** 0.183*** 0.185*** 0.133*** 0.017 0.123*** 0.126*** 0.131*** 0.129*** 0.129***
(0.016) (0.024) (0.016) (0.016) (0.016) (0.016) (0.016) (0.038) (0.039) (0.037) (0.037) (0.037) (0.04) (0.036)
Average
Population 0.016 0.04 0.016 0.016 0.017 0.016 0.018 - - - - - - -
(0.023) (0.025) (0.025) (0.025) (0.023) (0.024) (0.025) - - - - - - -
Unemploy
ment 0.012 0.052** 0.03 0.029 0.012 0.014 0.039** 0.039** 0.063*** 0.045** 0.046** 0.038** 0.041** 0.044**
(0.017) (0.021) (0.019) (0.018) (0.017) (0.016) (0.019) (0.016) (0.019) (0.018) (0.018) (0.017) (0.016) (0.02)
Regional
GDP per
Capita 0.27*** 0.184*** 0.371*** 0.346*** 0.268*** 0.286*** 0.36*** 0.567*** 0.436*** 0.626*** 0.636*** 0.559*** 0.581*** 0.601***
(0.045) (0.057) (0.037) (0.039) (0.045) (0.045) (0.048) (0.061) (0.136) (0.052) (0.054) (0.061) (0.062) (0.066)
National
Product
Category
Turnover 0.178*** 0.161*** 0.213*** 0.204*** 0.18*** 0.188*** 0.214*** 0.165*** 0.198*** 0.185*** 0.185*** 0.165*** 0.18*** 0.213***
(0.018) (0.02) (0.02) (0.019) (0.019) (0.017) (0.016) (0.032) (0.037) (0.032) (0.033) (0.031) (0.031) (0.029)
Supermark
et Dummy -0.117*** -0.227*** -0.117*** -0.118*** -0.117*** -0.115*** -0.113*** -0.052* -0.048*** -0.047* -0.048* -0.051* -0.047* -0.05*
(0.02) (0.041) (0.021) (0.021) (0.02) (0.02) (0.02) (0.026) (0.011) (0.026) (0.026) (0.026) (0.027) (0.026)
Hard -0.943*** -1.167*** -0.952*** -0.949*** -0.943*** -0.933*** -0.913*** -0.601*** -0.617*** -0.61*** -0.609*** -0.602*** -0.599*** -0.606***
Discounter
400
Annexes
Dummy
(0.135) (0.119) (0.135) (0.135) (0.135) (0.132) (0.135) (0.029) (0.032) (0.028) (0.029) (0.028) (0.029) (0.028)
New shop
opening 0.034*** 0.011 0.04*** 0.041*** 0.034*** 0.035*** 0.034*** 0.017** -0.003 0.016** 0.017** 0.017** 0.018** 0.015*
(0.007) (0.015) (0.008) (0.008) (0.007) (0.007) (0.009) (0.007) (0.015) (0.007) (0.007) (0.007) (0.007) (0.007)
Seasonal
Dummy 0.016*** 0.021*** 0.016*** 0.016*** 0.016*** 0.017*** 0.014*** 0.016*** 0.02*** 0.016*** 0.016*** 0.016*** 0.017*** 0.014***
(0.002) (0.003) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.003) (0.002) (0.002) (0.002) (0.002) (0.002)
BIC - - - - - - - - -29445.7 - - - - -
Within R2 0.127 0.029 0.126 0.259 0.127 0.115 0.104 0.132 0.036 0.132 0.263 0.132 0.121 0.109
Between
R2 0.77 0.758 0.764 0.819 0.77 0.77 0.787 0.293 0.178 0.285 0.411 0.283 0.287 0.364
Overall R2 0.712 0.72 0.706 0.771 0.711 0.71 0.727 0.279 0.17 0.27 0.396 0.27 0.272 0.342
Hausman 557.61** 634.08** 1365.23* 1065.95* 821.51** 847.26** 773.75**

Test * * ** ** * * * - - - - - - -
Moran’s I (0.124 - (0.438 - (0.128 - (0.127 - (0.123 - (0.131 - (0.098 - (0.388 - (0.122 - (0.388 - (0.384 - (0.391 - (0.396 - (0.353 -
(Range) 0.186) 0.501) 0.196) 0.193) 0.186) 0.192) 0.179) 0.448) 0.17) 0.45) 0.447) 0.45) 0.458) 0.417)
401
Table 42: Results - Product Price Variety
Product Price Variety
concentra concentra e) †† concentra (Separa concentra concentra concentra concentra e) †† concentra (Separa concentra e
tion tion tion te tion tion tion tion tion te tion concentr
measures measures measures concent measures measures measures measures measures concent measures ation
)† )† )† ration )† )† )† )† )† ration )† measure
measur measur s) †
es) † es) †
Local Private
labels share -0.002*** -0.004*** -0.003*** -0.002*** -0.002*** - -0.003*** -0.001 -0.002** -0.002*** -0.002** -0.001 - -0.001*
(0.001) (0.001) (0.001) (0.001) (0.001) - (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) - (0.001)
National
Private labels
share - - - 0.021*** - - - - 0.034*** -
- - - (0.003) - - - - (0.003) -
National
retail
concentratio
n HHI
(group,
edible
groceries ) -0.141*** -0.099*** - - -0.145*** -0.153*** - -0.172*** -0.075*** - - -0.174*** -0.189*** -
(0.009) (0.02) - - (0.009) (0.009) - (0.008) (0.024) - - (0.009) (0.009) -
National
retail
concentratio
n HHI
(group,
modern
retail) - -0.31*** - - - - -0.326*** - - -
- (0.02) - - - - (0.022) - - -
Local retail - - - - 0.001 - - - - 0.01

concentratio
402
Annexes
es) † es) †
n HHI
(group,
floorspace)
- - - - (0.013) - - - - (0.02)
National
supplier
concentratio
n HHI (full
market) 0.024*** 0.016*** 0.017*** 0.025*** 0.038*** 0.019* -0.004 0.016* 0.022** 0.054***
(0.004) (0.003) (0.003) (0.004) (0.004) (0.009) (0.012) (0.008) (0.009) (0.01)
National
supplier
concentratio
n HHI
(brands
only) - 0.024*** - 0.005
- (0.004) - (0.006)
Imbalance - - -0.065*** - - - - - - -0.092*** - - - -
- - (0.006) - - - - - - (0.009) - - - -
Average
Population
density -0.023*** -0.004 0.003 -0.013** -0.023*** -0.023*** -0.016** - - - - - - -
(0.005) (0.006) (0.01) (0.005) (0.005) (0.005) (0.006) - - - - - - -
Shop floor
space 0.007 -0.015 0.01 0.006 0.007 0.006 0.013* 0.047** -0.05** 0.061*** 0.036* 0.048** 0.048** 0.056***
403
es) † es) †
(0.007) (0.009) (0.007) (0.007) (0.007) (0.007) (0.007) (0.019) (0.025) (0.019) (0.02) (0.019) (0.019) (0.019)
Average
Population 0.025*** 0.037*** 0.025** 0.025*** 0.025*** 0.025*** 0.025*** - - - - - - -
(0.009) (0.011) (0.01) (0.009) (0.009) (0.009) (0.01) - - - - - - -
Unemployme
nt -0.096*** -0.105*** -0.133*** -0.127*** -0.1*** -0.101*** -0.15*** -0.09*** -0.121*** -0.14*** -0.129*** -0.092*** -0.098*** -0.158***
(0.006) (0.008) (0.009) (0.006) (0.006) (0.006) (0.008) (0.006) (0.01) (0.008) (0.006) (0.006) (0.006) (0.008)
Regional
GDP per
Capita 0.057** -0.042** -0.094** 0 0.056** 0.054** 0.017 0.202*** -0.139** -0.113 0.064 0.205*** 0.194*** 0.101*
(0.026) (0.02) (0.043) (0.024) (0.026) (0.027) (0.037) (0.041) (0.069) (0.087) (0.045) (0.041) (0.041) (0.059)
National
Product
Category
Turnover -0.016** -0.049*** -0.056*** -0.039*** -0.016** -0.016** -0.044*** 0.007 -0.081*** -0.085*** -0.047** 0.009 0.014 -0.03
(0.007) (0.007) (0.008) (0.007) (0.006) (0.006) (0.007) (0.018) (0.026) (0.029) (0.019) (0.018) (0.018) (0.022)
Supermarket
Dummy -0.031*** -0.033** -0.027*** -0.025*** -0.03*** -0.031*** -0.028*** -0.043*** -0.007 -0.046*** -0.034** -0.043*** -0.043*** -0.044***
(0.009) (0.013) (0.008) (0.009) (0.009) (0.009) (0.009) (0.014) (0.01) (0.013) (0.014) (0.014) (0.014) (0.014)
Hard
Discounter
Dummy -0.031 -0.133*** -0.02 -0.028 -0.031 -0.032 -0.005 0.032** -0.024 0.061*** 0.037** 0.034** 0.036** 0.071***
(0.043) (0.038) (0.043) (0.043) (0.043) (0.044) (0.046) (0.014) (0.021) (0.014) (0.015) (0.014) (0.014) (0.015)
404
Annexes
es) † es) †
New shop
opening 0.064*** 0.033*** 0.045*** 0.048*** 0.064*** 0.063*** 0.052*** 0.052*** 0.033*** 0.047*** 0.04*** 0.052*** 0.049*** 0.041***
(0.006) (0.006) (0.006) (0.006) (0.006) (0.006) (0.007) (0.006) (0.006) (0.007) (0.006) (0.006) (0.006) (0.007)
Seasonal
Dummy -0.001 -0.01*** -0.001 -0.001 -0.001 -0.001 -0.001* -0.001 -0.01*** -0.001 -0.001 -0.001 -0.001 -0.001*
(0.001) (0.002) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.002) (0.001) (0.001) (0.001) (0.001) (0.001)
BIC - - - - - - - - -37656.9 - - - - -
2
Within R 0.103 0.044 0.078 0.259 0.102 0.105 0.099 0.106 0.046 0.081 0.263 0.106 0.109 0.101
Between R2 0.598 0.479 0.599 0.819 0.598 0.592 0.618 0.341 0.099 0.011 0.411 0.347 0.354 0.044
Overall R2 0.491 0.375 0.487 0.771 0.492 0.486 0.507 0.29 0.084 0.016 0.396 0.295 0.3 0.053
Hausman 619.47**
Test 274.54*** 153.41*** 305.56*** 259.25*** * 1607.44*** 1977.38*** - - - - - - -
Moran’s I (0.11 - (0.163 - (0.128 - (0.11 - (0.109 - (0.108 - (0.107 - (0.213 - (0.059 - (0.306 - (0.263 - (0.207 - (0.199 - (0.285 -
(Range) 0.252) 0.311) 0.255) 0.248) 0.249) 0.249) 0.248) 0.364) 0.15) 0.431) 0.393) 0.357) 0.353) 0.428)
405
11.6.2. Innovation
Hypothesis: Innovation has increased due to consolidation of suppliers and producer

organisations
The relationship between supplier concentration and innovation is more often negative
than positive for many of the measures of innovation, which contradicts this
hypothesis, although evidence of large positive effects is found for new packaging.
Hypothesis: Retail concentration at procurement level does not appear to have had a
noticeable effect on the innovation evolution (number and type)
The results for this hypothesis are mixed and vary by model and measure of
innovation and no consistent result emerges to indicate that retail concentration
influences innovation evolution in any conclusive way.
Hypothesis: Shop type has strongly impacted the level and evolution of innovation.
The evidence supports this hypothesis for all measures of innovation. As found for the
choice indicators above, supermarkets and hard discounters are found to have fewer
innovative products in comparison to the base category hypermarkets and hard
discounters have fewer than supermarkets. The negative hard discounter impact is
much larger for innovation than for choice.
Hypothesis: The economic crisis has negatively impacted the evolution of innovation in
terms of new EANs (Opus innovations)
Some evidence in support of our hypothesis is found for most of the measures of
innovation using unemployment as a proxy for the economic crisis. Nevertheless, the
effect on new formulation is only significant in the short period and new packaging
finds a positive effect. The exception to this is new packaging which reports a positive
effect for all specifications (except the short period) suggesting that retailers may
favour incremental innovations in packaging as opposed to introducing entirely new
products during periods of economic crisis.
Another measure of the economic crisis is the measure of Retailer expectations which
reflects the forward-looking business sentiment among retailers. For Opus innovations,
new formulation, new packaging and new range extensions, this is found to be broadly
positive, so that stronger expectations are associated with more innovation, although
for some measures it is insignificant.
Hypothesis: National product category turnover appears to have an impact on the

evolution of choice in all its components
406
Annexes
There is evidence to support this hypothesis although the results for fixed effects
models varies in statistical significance and sign. In contrast, the random effects
models provide evidence of statistically significant positive effect on all measures of
innovation. This suggests that national product category turnover, which can be
conceptualised as market size, allows greater opportunities for innovation although
when unobservable fixed effects are controlled for, robust evidence is only found for
new packaging, new products and new formulation. This relationship turns negative
for new products and new formulation in the short period.
Hypothesis: Measure of imbalance at procurement level is a driver of the evolution of

the number and type of innovations
The imbalance between retailer and supplier concentrations is found to increase

innovation for Opus innovations and new products. In contrast the imbalance is found
to decrease innovation for new packaging, new formulation and new range extensions.
Hypothesis: Average population size, average population density, average GDP per
capita and new shop opening are drivers of the evolution of the number and type of
innovations
The evidence for the effect of average (over time) population is weak and mostly
insignificant for the measures of innovation. Average (over time) population density is
also mostly insignificant except for some evidence of a negative effect for new
formulations and new packaging. Average GDP per capita is also mostly insignificant
but is positive for the models where average population density is negative in new
packaging.
The evidence for an effect of new shop opening is weak and mostly insignificant but is
positive where it is statistically significant and the strongest evidence of an effect is
found for new products or new range extensions.
Hypothesis: The growing emergence of private labels, in part due to the increased
presence of discount stores appears to have played a role in the evolution of
innovation
Some evidence is found in the random effects models to suggest a small positive
relationship between measures of innovation and the local share of private labels, but
the evidence is less strong in the fixed effects models.
407
Table 43: Results - Opus Innovations
Opus Innovations
)† )† )† ration )† )† )† )† )† ration )† measures
measur measur )†
es) † es) †
Local Private
labels share 0.061*** 0.054*** 0.06*** 0.057*** 0.06*** - 0.072*** 0.055** 0.007 0.059** 0.051** 0.055** - 0.066**
(0.018) (0.021) (0.018) (0.018) (0.018) - (0.025) (0.024) (0.019) (0.023) (0.024) (0.024) - (0.03)
National
Private
labels share - - - -0.07 - - - - 0.632*** -
- - - (0.054) - - - - (0.224) -
National
retail
concentratio
n HHI
(group,
edible
groceries) -0.009 1.005** - - 0.019 0.087 - 0.3* 1.673*** - - 0.382** 0.073 -
(0.152) (0.416) - - (0.152) (0.148) - (0.169) (0.4) - - (0.177) (0.17) -
National
retail
concentratio
n HHI
(group,
modern
retail) - 2.136*** - - - - 2.602*** - - -
- (0.627) - - - - (0.618) - - -
408
Annexes
Opus Innovations
measur measur )†
es) † es) †
Local retail
concentratio
n HHI
(group,
floorspace) - - - - -0.197 - - - - -0.524
- - - - (0.149) - - - - (0.551)
National
supplier
concentratio
n HHI (full
market) -0.096** -0.039 - -0.088* -0.109** 0.007 -0.484** -0.074 - -0.3 -0.303 -0.641**
(0.044) (0.058) - (0.048) (0.046) (0.045) (0.23) (0.436) - (0.222) (0.207) (0.246)
National
supplier
concentratio
n HHI
(brands
only) - -0.167*** - -0.48**
- (0.051) - (0.221)
Average
Population
density -0.029 -0.034 -0.029 -0.02 -0.03 -0.04 -0.012 - - - - - - -
(0.047) (0.075) (0.047) (0.048) (0.047) (0.045) (0.047) - - - - - - -
Imbalance - - 0.245*** - - - - - - 1.213*** - - - -
- - (0.047) - - - - - - (0.278) - - - -
409
Opus Innovations
measur measur )†
es) † es) †
Shop floor
space 0.853*** 0.744*** 0.853*** 0.863*** 0.853*** 0.865*** 0.893*** 0.247 0.857* 0.325 0.384 0.245 0.203 0.268
(0.102) (0.096) (0.101) (0.102) (0.102) (0.103) (0.107) (0.423) (0.49) (0.416) (0.405) (0.423) (0.433) (0.443)
Average
Population -0.044 0.156 -0.043 -0.038 -0.044 -0.042 -0.039 - - - - - - -
(0.093) (0.123) (0.093) (0.094) (0.092) (0.092) (0.093) - - - - - - -
Unemployme
nt -0.619*** -0.39 -0.599*** -0.637*** -0.606*** -0.607*** -0.719*** -0.755*** -1.903*** -0.578** -0.693*** -0.693*** -0.847*** -0.874***
(0.195) (0.343) (0.206) (0.193) (0.195) (0.195) (0.225) (0.23) (0.578) (0.258) (0.224) (0.249) (0.202) (0.289)
Average
regional GDP
per capita 0.007 -0.063 0.014 -0.014 0.012 0.088 -0.031 - - - - - - -
(0.156) (0.273) (0.155) (0.156) (0.155) (0.144) (0.221) - - - - - - -
National
Product
Category
Turnover 0.666*** 0.705*** 0.613*** 0.649*** 0.642*** 0.682*** 0.921*** -0.614* 0.927 -0.451 -0.822** -0.674* -0.391 0.06
(0.062) (0.075) (0.065) (0.066) (0.064) (0.062) (0.047) (0.339) (0.744) (0.3) (0.345) (0.345) (0.351) (0.336)
Supermarket
Dummy -0.613*** -0.734*** -0.616*** -0.612*** -0.615*** -0.617*** -0.575*** -0.2 -0.819* -0.273 -0.305 -0.205 -0.192 -0.19
(0.134) (0.153) (0.132) (0.132) (0.134) (0.136) (0.135) (0.401) (0.431) (0.41) (0.404) (0.403) (0.394) (0.388)
Hard
Discounter
Dummy -2.193*** -2.261*** -2.191*** -2.163*** -2.192*** -2.153*** -2.27*** -1.502*** -0.786 -1.509*** -1.302*** -1.533*** -1.476*** -1.412***
410
Annexes
Opus Innovations
measur measur )†
es) † es) †
(0.329) (0.362) (0.327) (0.327) (0.329) (0.33) (0.361) (0.438) (0.722) (0.433) (0.439) (0.445) (0.435) (0.451)
New shop
opening 0.191 0.213 0.174 -0.003 0.188 0.182 0.14 0.064 -0.245 0.059 -0.127 0.051 0.067 0.007
(0.13) (0.162) (0.118) (0.123) (0.129) (0.129) (0.123) (0.173) (0.313) (0.162) (0.174) (0.172) (0.176) (0.191)
Retailer
Expectations 1.335*** -1.141*** 1.321*** 1.385*** 1.328*** 1.306*** 1.43*** 1.138*** 2.544*** 1.063*** 1.097*** 1.12*** 1.169*** 1.222***
(0.272) (0.326) (0.284) (0.279) (0.272) (0.272) (0.291) (0.272) (0.456) (0.277) (0.275) (0.273) (0.275) (0.294)
Seasonal
Dummy -3.209*** -3.424*** -3.21*** -3.205*** -3.209*** -3.208*** -3.252*** -3.222*** -3.431*** -3.227*** -3.224*** -3.223*** -3.217*** -3.267***
(0.084) (0.098) (0.084) (0.084) (0.084) (0.084) (0.095) (0.083) (0.098) (0.083) (0.083) (0.083) (0.083) (0.096)
BIC - - - - - - - 257823.3 142908.6 257757.6 16281.7 257826.7 257828.5 237031.2

2
Within R 0.23 0.285 0.23 0.189 0.23 0.229 0.235 0.231 0.287 0.232 0.191 0.231 0.231 0.236
Between R2 0.604 0.486 0.604 0.823 0.604 0.601 0.636 0.037 0.143 0.044 0.128 0.045 0.014 0.076
Overall R2 0.352 0.369 0.352 0.45 0.352 0.35 0.364 0.161 0.198 0.161 0.035 0.166 0.14 0.183
Hausman
Test 127.19*** 121.03*** 122.1*** 112.46*** 87.23*** 120.3*** 104.8*** - - - - - - -
Moran’s I (0.051 - (0.154 -

(0.056 - (0.05 - (0.052 - (0.05 - (0.05 - (0.048 - (0.167 - (0.151 - (0.171 - (0.155 - (0.162 - (0.134 -
(Range)
0.171) 0.146) 0.169) 0.161) 0.17) 0.171) 0.166) 0.364) 0.378) 0.362) 0.365) 0.366) 0.374) 0.325)
411
412
Annexes
Table 44: Results - New Products
New Products
es) † es) †
Local Private
labels share 0.03* 0.043*** 0.03* 0.028* 0.031* - 0.056*** 0.006 0.03* 0.006 0.003 0.004 - 0.024
(0.017) (0.016) (0.017) (0.017) (0.017) - (0.02) (0.017) (0.016) (0.018) (0.017) (0.017) - (0.018)
National
Private labels
share - - - -0.036 - - - - -0.204 -
- - - (0.048) - - - - (0.29) -
National
retail
concentratio
n HHI
(group,
edible
groceries ) -0.002 0.329 - - 0.053 0.047 - 0.08 1.286** - - 0.162 0.177 -
(0.205) (0.565) - - (0.211) (0.213) - (0.232) (0.585) - - (0.236) (0.267) -
National
retail
concentratio
n HHI
(group,
modern
retail) - 1.658** - - - - 1.693** - - -
- (0.725) - - - - (0.759) - - -
Local retail - - - - -0.301 - - - - -0.531

concentratio
413
New Products
es) † es) †
n HHI
(group,
edible
groceries
- - - - (0.189) - - - - (0.574)
National
supplier
concentratio
n HHI (full
market) -0.38*** -0.435*** - -0.374*** -0.387*** -0.259*** -0.811*** 0.313 - -0.677** -0.845*** -0.511**
(0.057) (0.066) - (0.059) (0.057) (0.059) (0.239) (0.604) - (0.251) (0.219) (0.227)
National
supplier
concentratio
n HHI
(brands only) - -0.358*** - -0.346
- (0.054) - (0.348)
Average
Population
density -0.054 -0.069 -0.054 -0.047 -0.054 -0.059 -0.027 - - - - - - -
(0.068) (0.079) (0.068) (0.069) (0.068) (0.068) (0.069) - - - - - - -
Imbalance - - 0.405*** - - - - - - 0.877*** - - - -
- - (0.053) - - - - - - (0.303) - - - -
Shop floor
space 1.339*** 1.237*** 1.338*** 1.346*** 1.337*** 1.345*** 1.422*** 0.291 0.62 0.331 0.39 0.262 0.288 0.289
414
Annexes
New Products
es) † es) †
(0.114) (0.126) (0.112) (0.113) (0.113) (0.114) (0.117) (0.37) (0.552) (0.367) (0.357) (0.373) (0.368) (0.395)
Average
Population 0.053 0.195 0.054 0.057 0.053 0.054 0.053 - - - - - - -
(0.124) (0.136) (0.123) (0.124) (0.123) (0.124) (0.127) - - - - - - -
Unemployme
nt -1.054*** -1.049*** -1.011*** -1.066*** -1.02*** -1.05*** -1.196*** -1.385*** -3.038*** -1.24*** -1.363*** -1.303*** -1.343*** -1.794***
(0.172) (0.329) (0.173) (0.164) (0.173) (0.171) (0.191) (0.229) (0.655) (0.251) (0.222) (0.251) (0.206) (0.3)
Average
regional GDP
per capita -0.017 -0.07 -0.003 -0.033 -0.006 0.023 -0.085 - - - - - - -
(0.232) (0.269) (0.232) (0.233) (0.23) (0.226) (0.307) - - - - - - -
National
Product
Category
Turnover 0.676*** 0.654*** 0.673*** 0.665*** 0.691*** 0.684*** 0.852*** 1.397*** -1.82** 1.351** 1.148** 1.301** 1.329*** 3.001***
(0.072) (0.101) (0.07) (0.076) (0.068) (0.073) (0.071) (0.501) (0.775) (0.564) (0.559) (0.506) (0.476) (0.512)
Supermarket
Dummy -1.046*** -0.919*** -1.051*** -1.045*** -1.049*** -1.047*** -0.944*** -0.581 0.305 -0.626 -0.653 -0.573 -0.583 -0.521
(0.146) (0.192) (0.145) (0.147) (0.146) (0.148) (0.145) (0.412) (0.887) (0.423) (0.414) (0.416) (0.412) (0.393)
Hard
Discounter
Dummy -4.029*** -4.413*** -4.028*** -4.008*** -4.03*** -4.009*** -3.733*** -5.92*** -3.969*** -5.933*** -5.757*** -5.984*** -5.942*** -5.717***
(0.332) (0.33) (0.328) (0.332) (0.332) (0.334) (0.339) (0.425) (0.956) (0.41) (0.43) (0.431) (0.422) (0.411)
415
New Products
es) † es) †
New shop
opening 0.383*** 0.298* 0.353*** 0.238** 0.374*** 0.378*** 0.28** 0.265* -0.147 0.216* 0.105 0.245 0.261 0.143
(0.13) (0.18) (0.102) (0.121) (0.13) (0.129) (0.129) (0.157) (0.26) (0.124) (0.155) (0.157) (0.157) (0.149)
Retailer
Expectations -0.47* -2.851*** -0.49* -0.432 -0.481* -0.486* -0.353 -0.54* 0.164 -0.594** -0.558** -0.562** -0.559** -0.3
(0.277) (0.315) (0.289) (0.283) (0.278) (0.277) (0.288) (0.271) (0.54) (0.277) (0.272) (0.276) (0.267) (0.28)
Seasonal
Dummy -6.164*** -5.974*** -6.165*** -6.161*** -6.164*** -6.163*** -6.237*** -6.167*** -5.989*** -6.171*** -6.168*** -6.169*** -6.168*** -6.232***
(0.089) (0.147) (0.09) (0.09) (0.089) (0.089) (0.095) (0.089) (0.145) (0.09) (0.09) (0.089) (0.09) (0.096)
BIC - - - - - - - 294419.4 164024.6 294405.8 16281.7 294434 294418.3 269778.6

2
Within R 0.332 0.355 0.332 0.189 0.332 0.332 0.341 0.333 0.357 0.333 0.191 0.332 0.333 0.342
2
Between R 0.652 0.542 0.652 0.823 0.652 0.651 0.666 0.22 0.023 0.24 0.128 0.247 0.23 0.173
Overall R2 0.424 0.426 0.424 0.45 0.424 0.424 0.431 0.282 0.174 0.291 0.035 0.296 0.287 0.229
Hausman
Test 3492.47*** 188.76*** 76.62*** 78.04*** 72.49*** 71.07*** 104.8*** - - - - - - -
Moran’s I (0.083 - (0.083 - (0.08 - (0.084 - (0.084 - (0.062 - (0.133- (0.124 - (0.121 - (0.104 - (0.126 - (0.187 -
(Range) 0.189) (0.07 - 0.11) 0.188) 0.181) 0.189) 0.19) 0.191) 0.246) (0.231- 0.32) 0.24) 0.238) 0.222) 0.242) 0.273)
416
Annexes
Table 45: Results - New Packaging
New Packaging
† † † ration † † † † † ration † †
measur measur
es) † es) †
Local Private
labels share 0.046** 0.027* 0.05*** 0.056*** 0.045** - 0.069*** 0.001 -0.011 -0.003 0.011 0.004 - 0.041
(0.018) (0.016) (0.018) (0.018) (0.018) - (0.021) (0.029) (0.02) (0.028) (0.027) (0.03) - (0.039)
National
Private labels
share - - - -0.226*** - - - - -0.034 -
- - - (0.04) - - - - (0.243) -
National retail
concentration
HHI (group,
edible
groceries ) 0.7** -0.276 - - 0.644** 0.845*** - -0.506* -0.407 - - -1.023*** -0.49 -
(0.28) (0.434) - - (0.27) (0.28) - (0.292) (0.451) - - (0.273) (0.33) -
National retail
concentration
HHI (group,
modern retail) - -4.55*** - - - - -5.755*** - - -
- (1.004) - - - - (0.87) - - -
Local retail
concentration
HHI (group,
floorspace) - - - - -0.537** - - - - -2.521***
National retail - - - - (0.263) - - - - (0.631)

concentration
417
New Packaging
† † † ration † † † † † ration † †
measur measur
es) † es) †
HHI (group,
edible
groceries )
National
supplier
concentration
HHI (full
market) 0.514*** 0.455*** - 0.506*** 0.496*** 0.601*** 2.866*** -0.248 - 2.44*** 2.86*** 3.161***
(0.068) (0.088) - (0.071) (0.069) (0.082) (0.255) (0.73) - (0.208) (0.247) (0.242)
National
supplier
concentration
HHI (brands
only) - 0.382*** - 3.141***
- (0.06) - (0.381)
Average
Population
density -0.178*** -0.113 -0.191*** -0.21*** -0.177*** -0.187*** -0.188*** - - - - - - -
(0.069) (0.088) (0.068) (0.066) (0.068) (0.069) (0.072) - - - - - - -
Imbalance - - -0.564*** - - - - - - -4.312*** - - - -
- - (0.073) - - - - - - (0.457) - - - -
Shop floor
space 1.055*** 0.822*** 1.043*** 1.022*** 1.058*** 1.063*** 1.131*** 2.978*** 1.865*** 2.721*** 2.66*** 2.973*** 2.977*** 3.03***
(0.112) (0.104) (0.11) (0.109) (0.112) (0.113) (0.12) (0.646) (0.6) (0.603) (0.612) (0.635) (0.654) (0.614)
Average 0.063 0.24* 0.057 0.047 0.063 0.065 0.061 - - - - - - -
418
Annexes
New Packaging
† † † ration † † † † † ration † †
measur measur
es) † es) †
Population
(0.125) (0.128) (0.124) (0.125) (0.124) (0.126) (0.125) - - - - - - -
Unemploymen
t 1.803*** -0.156 1.864*** 1.957*** 1.763*** 1.834*** 1.882*** 2.609*** -0.008 1.974*** 2.495*** 2.223*** 2.615*** 2.036***
(0.25) (0.295) (0.23) (0.238) (0.249) (0.258) (0.243) (0.225) (0.735) (0.227) (0.199) (0.242) (0.223) (0.225)
Average
regional GDP
per capita 1.001** 0.193 1.043*** 1.105*** 0.988** 1.07*** 1.236** - - - - - - -
(0.396) (0.259) (0.399) (0.407) (0.392) (0.406) (0.616) - - - - - - -
National
Product
Category
Turnover 1.964*** 1.61*** 2.007*** 2.029*** 1.908*** 1.982*** 2.594*** 8.231*** 3.322** 8.274*** 8.848*** 8.587*** 8.22*** 9.529***
(0.131) (0.148) (0.14) (0.125) (0.136) (0.129) (0.14) (0.686) (1.367) (0.632) (0.665) (0.72) (0.705) (0.769)
Supermarket
Dummy -1.325*** -1.137*** -1.337*** -1.347*** -1.322*** -1.33*** -1.36*** -1.927*** -0.862* -1.682*** -1.691*** -1.887*** -1.928*** -1.854***
(0.168) (0.167) (0.166) (0.165) (0.168) (0.173) (0.164) (0.431) (0.467) (0.394) (0.398) (0.421) (0.431) (0.385)
Hard
Discounter
Dummy -3.326*** -3.385*** -3.371*** -3.429*** -3.325*** -3.3*** -3.466*** -4.144*** -2.475*** -4.165*** -4.63*** -3.968*** -4.148*** -3.835***
(0.49) (0.437) (0.489) (0.492) (0.489) (0.488) (0.554) (0.662) (0.844) (0.551) (0.576) (0.643) (0.671) (0.588)
New shop
opening -0.363 0.198 -0.149 0.208 -0.352 -0.373 0.007 -0.225 -0.12 -0.029 0.246 -0.147 -0.226 0.025
419
New Packaging
† † † ration † † † † † ration † †
measur measur
es) † es) †
(0.23) (0.239) (0.174) (0.199) (0.23) (0.229) (0.237) (0.244) (0.384) (0.173) (0.209) (0.242) (0.242) (0.24)
Retailer
Expectations -0.081 1.197*** -0.224 -0.358 -0.073 -0.114 -0.314 0.844*** 1.49*** 1.089*** 0.922*** 0.952*** 0.841*** 1.014***
(0.28) (0.328) (0.277) (0.277) (0.277) (0.281) (0.274) (0.28) (0.523) (0.262) (0.275) (0.265) (0.284) (0.26)
Seasonal
Dummy -4.343*** -4.333*** -4.352*** -4.362*** -4.342*** -4.343*** -4.482*** -4.277*** -4.329*** -4.261*** -4.273*** -4.27*** -4.278*** -4.384***
(0.063) (0.084) (0.061) (0.06) (0.063) (0.063) (0.069) (0.062) (0.085) (0.06) (0.06) (0.061) (0.062) (0.069)
BIC - - - - - - - 297005.9 161126.5 296747.9 16281.7 297029.3 297005.8 271944.2
Within R2 0.21 0.242 0.211 0.189 0.21 0.211 0.225 0.222 0.243 0.225 0.191 0.221 0.222 0.239
Between R2 0.701 0.699 0.699 0.823 0.701 0.701 0.676 0.5 0.451 0.491 0.128 0.49 0.5 0.477
2
Overall R 0.391 0.478 0.39 0.45 0.391 0.391 0.381 0.258 0.338 0.257 0.035 0.253 0.258 0.237
Hausman Test 950.04**

6447.77*** 285.50*** 1187.67*** 1050.96*** * 951.66*** 1101.45*** - - - - - - -
Moran’s I (0.125 - (0.126 - (0.124 - (0.128 - (0.123 - (0.103 - (0.123 - (0.195 - (0.381 - (0.371 - (0.361 - (0.373 -
(Range) 0.263) (0.12 - 0.192) 0.266) 0.274) 0.265) 0.263) 0.255) 0.266) 0.357) (0.366 - 0.5) 0.499) 0.499) 0.493) 0.491)
Note: All specifications use standard errors derived by clustering on consumer shopping areas and include time, product and country fixed effects (not reported).
Standard errors are presented in parentheses in the row below each coefficient. *** indicates significant at the 1% level of significance, ** at the 5% and * at the
10%. Moran’s I is calculated for each time period; the table shows the range of test statistics over the time periods, and the level of significance indicated is the
average p-value across the time periods. † Separate concentration measures to refers to models including both retailer and supplier concentration measures rather
than just the imbalance between the two. †† Imbalance refers to models which include only the measure of imbalance between retailer and supplier concentration
measures rather than both measures
420
Annexes
Table 46: Results - New Formulation
New Formulation
period RE Period RE period RE period RE period period RE period RE period FE Period FE period FE period FE period period FE period FFE
† † † ration † † † † † ration † †
measur measur
es) † es) †
Local Private
labels share 0.037** 0.051*** 0.043*** 0.041*** 0.036** - 0.07*** -0.021 -0.028 -0.012 -0.016 -0.02 - -0.013
(0.015) (0.014) (0.016) (0.016) (0.015) - (0.017) (0.02) (0.021) (0.019) (0.019) (0.02) - (0.027)
National
Private labels
share - - - -0.08 - - - - -0.006 -
- - - (0.052) - - - - (0.289) -
National retail
concentration
HHI (group,
edible
groceries ) 1.023*** -2.023*** - - 1.082*** 1.098*** - 0.746*** -1.668*** - - 0.856*** 0.727*** -
(0.164) (0.396) - - (0.17) (0.165) - (0.227) (0.34) - - (0.229) (0.229) -
National retail
concentration
HHI (group,
modern retail) - 1.666*** - - - - 1.183* - - -
- (0.473) - - - - (0.615) - - -
Local retail
concentration
HHI (group,
floorspace) - - - - -0.142 - - - - -0.352
- - - - (0.159) - - - - (0.267)
421
New Formulation
† † † ration † † † † † ration † †
measur measur
es) † es) †
National
supplier
concentration
HHI (full
market) -0.212*** -0.349*** - -0.201*** -0.222*** -0.333*** -0.231 -0.247 - -0.222 -0.254 -0.465**
(0.063) (0.07) - (0.064) (0.061) (0.071) (0.175) (0.44) - (0.186) (0.191) (0.188)
National
supplier
concentration
HHI (brands
only) - -0.348*** - -0.836***
- (0.053) - (0.177)
Average
Population
density -0.148*** -0.13* -0.166*** -0.159*** -0.148*** -0.154*** -0.148*** - - - - - - -
(0.045) (0.074) (0.044) (0.045) (0.045) (0.045) (0.044) - - - - - - -
Imbalance - - 0.346*** - - - - - - 0.826*** - - - -
- - (0.061) - - - - - - (0.221) - - - -
Shop floor
space 0.624*** 0.62*** 0.602*** 0.61*** 0.623*** 0.631*** 0.635*** 0.836* 0.159 0.846* 0.847** 0.868** 0.851** 0.79*
(0.073) (0.081) (0.073) (0.073) (0.073) (0.074) (0.078) (0.414) (0.396) (0.42) (0.417) (0.415) (0.411) (0.425)
Average
Population 0.155** 0.215** 0.149* 0.152** 0.156** 0.157** 0.142* - - - - - - -
(0.077) (0.098) (0.077) (0.076) (0.077) (0.077) (0.076) - - - - - - -
422
Annexes
New Formulation
† † † ration † † † † † ration † †
measur measur
es) † es) †
Unemploymen
t -0.009 -0.517* 0.187 0.145 0.016 0 0.159 0.012 -1.712*** 0.216 0.138 0.072 0.002 0.09
(0.125) (0.264) (0.117) (0.109) (0.127) (0.127) (0.123) (0.16) (0.501) (0.15) (0.144) (0.161) (0.139) (0.19)
Average
regional GDP
per capita 0.226 0.017 0.318** 0.293* 0.235 0.276* 0.262 - - - - - - -
(0.147) (0.201) (0.156) (0.153) (0.148) (0.153) (0.183) - - - - - - -
National
Product
Category
Turnover 1.431*** 1.553*** 1.423*** 1.462*** 1.385*** 1.442*** 1.635*** 3.491*** -16.998*** 4.137*** 3.996*** 3.457*** 3.485*** 4.125***
(0.066) (0.125) (0.064) (0.069) (0.061) (0.067) (0.089) (0.444) (1.071) (0.363) (0.439) (0.435) (0.384) (0.546)
Supermarket
Dummy -0.672*** -0.441*** -0.703*** -0.697*** -0.675*** -0.674*** -0.72*** -1.039** 0.272 -1.075*** -1.071*** -1.064*** -1.04** -1.017***
(0.115) (0.113) (0.113) (0.114) (0.115) (0.118) (0.119) (0.387) (0.18) (0.387) (0.393) (0.391) (0.387) (0.366)
Hard
Discounter
Dummy -2.509*** -2.104*** -2.567*** -2.548*** -2.508*** -2.485*** -2.73*** -6.097*** -4.633*** -6.252*** -6.179*** -6.096*** -6.085*** -6.256***
(0.437) (0.375) (0.44) (0.44) (0.437) (0.433) (0.483) (0.434) (0.443) (0.431) (0.427) (0.435) (0.427) (0.428)
New shop
opening -0.052 0.222** 0.17 0.052 -0.059 -0.058 0.137 -0.041 -0.003 0.121 0.058 -0.05 -0.038 0.183
(0.103) (0.111) (0.108) (0.111) (0.103) (0.103) (0.132) (0.149) (0.15) (0.125) (0.149) (0.148) (0.148) (0.164)
Retailer
Expectations 0.523*** -3.586*** 0.24 0.302 0.507*** 0.502*** 0.128 0.738*** -0.378 0.643*** 0.669*** 0.72*** 0.745*** 0.482**
423
New Formulation
† † † ration † † † † † ration † †
measur measur
es) † es) †
(0.19) (0.449) (0.213) (0.207) (0.191) (0.187) (0.193) (0.218) (0.345) (0.235) (0.236) (0.213) (0.209) (0.223)
Seasonal
Dummy -3.992*** -4.038*** -4.011*** -4.007*** -3.993*** -3.992*** -4.182*** -3.975*** -4.072*** -3.982*** -3.98*** -3.976*** -3.975*** -4.152***
(0.116) (0.191) (0.116) (0.117) (0.116) (0.116) (0.12) (0.115) (0.188) (0.116) (0.116) (0.115) (0.116) (0.12)
BIC - - - - - - - 295733.3 162010.2 295740.8 16281.7 295721.8 295735.6 272267.7
Within R2 0.175 0.212 0.174 0.189 0.175 0.175 0.183 0.176 0.233 0.176 0.191 0.176 0.176 0.185
2
Between R 0.755 0.693 0.756 0.823 0.756 0.754 0.755 0.326 0.194 0.256 0.128 0.315 0.325 0.231
2
Overall R 0.394 0.438 0.394 0.45 0.394 0.394 0.395 0.206 0.069 0.167 0.035 0.201 0.206 0.161
Hausman Test 1544.97*** 193.95*** 134.83*** 114.27*** 90.01*** 62.59*** 112.37*** - - - - - - -
Moran’s I (0.186 - (0.189 - (0.189 - (0.19 - (0.191 - (0.186 - (0.251 - (0.287 - (0.274 - (0.256 - (0.251 - (0.275 -
(Range) (0.19 - 0.336) 0.266) 0.337) 0.336) 0.335) 0.336) 0.315) 0.406) (0.613 - 0.62) 0.426) 0.417) 0.412) 0.407) 0.411)
424
Annexes
Table 47: Results - New Range extensions
New Range extensions
Long Short Long Long Long Long Long Long Short Long Long Long Long Long period
period RE Period RE period RE period RE period period RE period RE period FE Period FE period FE period FE period period FE FE
concentrat concentrati e) †† concentrati (Separat concentrati concentrati concentrati concentrati e) †† concentrati (Separat concentrati concentrati
ion on on e on on on on on e on on
measures) measures) measures) concentr measures) measures) measures) measures) measures) concentr measures) measures)
† † † ation † † † † † ation † †
measure measure
s) † s) †
Local Private
labels share -0.008 0.007 -0.008 -0.011 -0.007 - 0.014 -0.051** -0.001 -0.051** -0.056** -0.053** - -0.063**
(0.019) (0.021) (0.019) (0.019) (0.019) - (0.026) (0.023) (0.019) (0.023) (0.024) (0.023) - (0.03)
National
Private labels
share - - - -0.347*** - - - - -0.359 -
- - - (0.056) - - - - (0.272) -
National
retail
concentration
HHI (group,
edible
groceries ) -0.124 5.846*** - - -0.121 0.011 - 0.083 6.444*** - - 0.203 0.193 -
(0.177) (0.643) - - (0.178) (0.175) - (0.177) (0.603) - - (0.187) (0.2) -
National
retail
concentration
HHI (group,
modern
retail) - 1.932*** - - - - 2.178*** - - -
- (0.692) - - - - (0.576) - - -
Local retail
concentration
HHI (group,
floorspace) - - - - -0.134 - - - - 0.16
425
† † † ation † † † † † ation † †
measure measure
s) † s) †
- - - - (0.196) - - - - (0.448)
National
supplier
concentration
HHI (full
market) -0.124 0.207*** - -0.118 -0.138* 0.054 -0.846*** -0.422 - -0.672*** -0.969*** -0.742***
(0.081) (0.066) - (0.084) (0.082) (0.08) (0.2) (0.573) - (0.207) (0.213) (0.212)
National
supplier
concentration
HHI (brands
only) - -0.037 - -0.652**
- (0.077) - (0.286)
Average
Population
density -0.066 -0.08 -0.064 -0.056 -0.067 -0.067 -0.062 - - - - - - -
(0.063) (0.079) (0.063) (0.064) (0.063) (0.063) (0.065) - - - - - - -
Imbalance - - 0.114 - - - - - - 1.266*** - - - -
- - (0.073) - - - - - - (0.28) - - - -
Shop floor
space 1.219*** 1.037*** 1.221*** 1.23*** 1.218*** 1.214*** 1.276*** 0.491 0.69 0.572* 0.62* 0.477 0.531 0.382
(0.107) (0.102) (0.107) (0.107) (0.107) (0.107) (0.117) (0.336) (0.577) (0.327) (0.326) (0.334) (0.336) (0.369)
Average
Population 0.043 0.186 0.044 0.049 0.043 0.043 0.047 - - - - - - -
426
Annexes
† † † ation † † † † † ation † †
measure measure
s) † s) †
(0.103) (0.121) (0.103) (0.104) (0.103) (0.102) (0.107) - - - - - - -
Unemployme
nt -0.562*** -0.948*** -0.571*** -0.598*** -0.555*** -0.513*** -0.809*** -0.571*** -2.537*** -0.396* -0.546** -0.471** -0.529*** -0.982***
(0.169) (0.342) (0.176) (0.171) (0.169) (0.173) (0.195) (0.2) (0.618) (0.215) (0.203) (0.21) (0.192) (0.249)
Average
regional GDP
per capita 0.12 -0.084 0.113 0.09 0.122 0.128 0.13 - - - - - - -
(0.246) (0.278) (0.245) (0.245) (0.246) (0.241) (0.301) - - - - - - -
National
Product
Category
Turnover 1.239*** 1.416*** 1.239*** 1.219*** 1.273*** 1.249*** 1.565*** 0.256 4.025*** 0.179 -0.084 0.153 0.126 2.409***
(0.105) (0.119) (0.1) (0.111) (0.099) (0.098) (0.072) (0.563) (0.877) (0.571) (0.599) (0.566) (0.604) (0.398)
Supermarket
Dummy -0.903*** -1.012*** -0.902*** -0.899*** -0.904*** -0.909*** -0.832*** -0.049 -0.703* -0.123 -0.141 -0.051 -0.054 0.038
(0.152) (0.16) (0.153) (0.152) (0.152) (0.153) (0.15) (0.359) (0.359) (0.368) (0.36) (0.364) (0.365) (0.356)
Hard
Discounter
Dummy -4.625*** -4.825*** -4.617*** -4.591*** -4.625*** -4.638*** -4.417*** -8.923*** -8.604*** -8.898*** -8.708*** -8.982*** -8.927*** -8.888***
(0.432) (0.329) (0.43) (0.431) (0.432) (0.436) (0.479) (0.365) (0.677) (0.363) (0.37) (0.367) (0.372) (0.367)
New shop
opening 0.195* 0.308** 0.155 -0.008 0.194* 0.191* 0.215* 0.144 0.148 0.067 -0.067 0.122 0.145 0.211
(0.105) (0.139) (0.101) (0.094) (0.105) (0.107) (0.121) (0.132) (0.28) (0.124) (0.126) (0.131) (0.134) (0.159)
427
† † † ation † † † † † ation † †
measure measure
s) † s) †
Retailer
Expectations 2.807*** 1.382*** 2.832*** 2.88*** 2.808*** 2.79*** 2.985*** 2.635*** 3.994*** 2.568*** 2.614*** 2.607*** 2.624*** 2.93***
(0.228) (0.452) (0.226) (0.227) (0.229) (0.229) (0.241) (0.227) (0.442) (0.227) (0.228) (0.229) (0.227) (0.254)
Seasonal
Dummy -5.052*** -5.192*** -5.05*** -5.047*** -5.052*** -5.053*** -5.062*** -5.062*** -5.198*** -5.066*** -5.063*** -5.065*** -5.062***
(0.08) (0.113) (0.08) (0.08) (0.08) (0.08) (0.085) (0.08) (0.112) (0.081) (0.081) (0.081) (0.086)
BIC - - - - - - - 289762.1 160817 289725.6 16281.7 289772.2 264816.6

2
Within R 0.288 0.335 0.288 0.189 0.288 0.288 0.298 0.29 0.337 0.29 0.191 0.289 0.3
Between R2 0.695 0.592 0.695 0.823 0.695 0.698 0.719 0.247 0.262 0.245 0.128 0.251 0.27
Overall R2 0.434 0.448 0.434 0.45 0.434 0.435 0.444 0.261 0.21 0.26 0.035 0.263 0.248
Hausman
Test 5204.52*** 171.07*** 142.2*** 123.02*** 111.65*** 105.98*** 150.48*** - - - - - - -
Moran’s I (0.101 - (0.073 - (0.101 - (0.1 - (0.099 - (0.067 - (0.151 - (0.374 - (0.154 - (0.159 - (0.15 - (0.152 -
(Range) 0.194) 0.125) 0.193) (0.1 - 0.186) 0.194) 0.195) 0.178) 0.275) 0.452) 0.283) 0.277) 0.265) 0.275) (0.144 - 0.285)
428
11.6.3. Focus on private labels results
The figures below show the impacts of private label penetration on choice for each
product category.
429
431
433
435
437
439
441
443
445
447
449
European Commission
Study on the economic impact of modern retail on choice and innovation in the EU
food sector
Final report
Luxembourg: Publications Office of the European Union
2014 – 450 pp. – 21.0 x 29.7 cm
ISBN 978-92-79-40324-8
doi: 10.2763/77405
KD-02-14-955-EN-N
doi 10.2763/77405

Retail EU Study

Uploaded by

Copyright:

Available Formats

Retail EU Study

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Retail EU Study

Uploaded by

Copyright:

Available Formats

The economic impact of

The economic impact of modern retail on choice and

Cambridge Econometrics Ltd.

More information on the European Union is available on the Internet: http://europa.eu

Cataloguing data can be found at the end of this publication.

Luxembourg: Publications Office of the European Union, 2014

© European Union, 2014

1.1. Objectives of the study

Measure of imbalance (national level only) - 2004-2012 ©Planet retail, ©Euromonitor n n n n n n n n n n n n n n 14

Evolution of other a priori drivers

Six case studies bringing additional qualitative information

1.3. Background: a strong development of modern retail across Europe,

1.4. Evolution of choice: choice offered to consumers has notably

Choice in alternative products on the shelves of shops increased in all 9 MS of the

1.5. Evolution of innovation: a steady stream of innovation was made

40% Range extension

2004-2012 sample: Proportion of innovations by MS (local level) (source: EY analysis based on ©

Product Packaging Formula Range extension Relaunch

2004-2012 sample: Proportion of innovations by product categories (local level) (source: EY

The trend towards an increasing proportion of packaging innovations was observed

1.6. Evolution of concentration: concentration of retailers and suppliers

3501 < < 4000 1501 < < 2000

1001 < < 1500

2001 < < 2500 N/A

2004-2012 HHI Modern Retail across EU MS sample at a national level

At local level in terms of share of floorspace, retail concentration decreased by -1.1%

1.7. Conclusions regarding factors driving choice

The impact of the measure of imbalance between modern retailers and

Summary of econometric results for key drivers: choice

Product supplier Product price

1.8. Conclusions regarding factors driving innovation

Greater concentration among suppliers at procurement level was

Summary of econometric results for key drivers: innovation

2.1. Objectives of the study

2.2. Motivations behind study

 On one hand, modern retail is acknowledged as being a very competitive

Turbulence in commodities markets beginning in 2007 and subsequent consumer price

Whilst quantitative evidence to date is insufficient to robustly support these claims, a

2.3. Structure of the final report

Prior to responding to each of the study questions addressed in the descriptive

2.4. Limitations of this report

 Innovation: the definition of innovation from an operational perspective for this

2.5. Different tasks of the study

2.5.5. Tasks 5&6

3. Background of the study

3.1. Europe retail sector in brief

3.1.1. Definition of modern grocery retail

28 EY analysis based on © Planet Retail

29 EY analysis based on © Planet Retail

independent and often family-owned businesses with non-organised distribution

30 EY analysis based on © Planet Retail

3.1.2. Prevalence of modern grocery retail in the EU

Source: EY analysis based on © Planet Retail, 2011

This figure presents an opportunity to draw the reader’s attention to a number of

31 Estimates based on © Planet Retail

32 Estimates based on © Planet Retail

and vegetable market, non-branded neighbourhood stores, butchers and bakers)33.

3.2. Recent evolutions in the grocery retail sector in the EU

3.2.1. Increase in number of shops and commercial sales areas