Local Government Studies, 2013

Vol. 39, No. 2, 182–207, http://dx.doi.org/10.1080/03003930.2012.693075

Budget Transparency in Local

Governments: An Empirical Analysis
JOSÉ CAAMAÑO-ALEGRE*, SANTIAGO LAGO-PEÑAS**,
FRANCISCO REYES-SANTIAS** &
AURORA SANTIAGO-BOUBETA***
*University of Santiago de Compostela, Spain, **University of Vigo, REDE, and IEB, Spain,
***Municipality of Pontevedra, Spain

ABSTRACT The aim of this paper is to shed additional light on the determinants of
budget transparency in local governments. Our work is based on a Likert-type survey
questionnaire specifically designed to measure budget transparency in small municipa-
lities. The questionnaire is based on the IMF’s revised Code of Good Practices on Fiscal
Transparency (2007). Results from 33 Galician municipalities are used to assess its
internal consistency and to test a battery of hypotheses on the determinants of budget
transparency. While several previous findings of the literature are confirmed, some new
results are also obtained.

KEY WORDS: Government accounting, local public finance, municipal budgeting,

budget transparency

1. Introduction
Despite a recent upsurge in popularity, concerns regarding budget transparency
have a long tradition. The establishment of representative democracies through-
out the Western world brought with it a resistance to the secrecy that had
previously been the norm in governmental accounting. At the time, budget
transparency was closely linked to notions of executive accountability and
principles of clarity in budgeting.
However, in recent decades the term has been widely adopted by policy
makers and popular media, and its meaning has evolved. Budget transparency
is not just an element of governmental accountability to Parliament, but rather a
tool for facilitating a relationship between public budgeting and market require-
ments, civil society demands and citizen participation.

Correspondence Address: Professor Santiago Lago-Peñas, Department of Applied Economics,

University of Vigo, Campus de Ourense, Spain. E-mail: [email protected]

© 2013 Taylor & Francis

 Budget Transparency in Local Governments 183

Budget transparency is particularly important in local government.

Globalization has given rise to a greater recognition of the role of local
government, demonstrated by a widespread resort to the subsidiary principle
and a growing municipal participation in public policies. The strengthening of
local governments along with substantial changes in the way those govern-
ments operate has kindled the interest of stakeholders in knowing what
governments do, how and at what price. Fairly often, international organisa-
tions involved in promoting budget transparency consider local governments
to be ideal for testing new systems or arrangements before they are imple-
mented at higher levels.
However, available empirical research on budget transparency tends to focus
on national or regional governments. Moreover, studies generally rely on a few
crude indicators or survey data from dichotomous-item questionnaires, often
extracted from instruments designed for other purposes. Aimed at filling this
gap, this paper presents a Likert-type survey questionnaire to measure budget
transparency in municipalities. The use of a Likert scale1 promotes content
validity2 and discreteness with a limited number of survey items. Rather than
focusing on the existence of formal safeguards as evidence of budget transpar-
ency, Likert scales make it possible to analyze the degree to which municipal
practices and/or behaviors satisfy certain transparency-related criteria. Based
on the answers to this survey by financial comptrollers from 33 Galician
municipalities, we perform an empirical analysis on the determinants of
transparency.
The paper is organised as follows: section 2 presents the definition, scope,
and international standards for budget transparency. Relevant empirical
studies on the determinants of budget transparency are also reviewed.
Section 3 presents our survey and evaluates its consistency. Both the
descriptive analysis and econometrics are contained in section 3. Section 4
concludes.

2. Approach to budget transparency: definition, scope, and determinants

The notion of transparency is far from clear. According to Hood (2006, p. 3),
“like many other notions of a quasi-religious nature, transparency is more often
preached than practised, more often invoked than defined, and indeed might
ironically be said to be mystic in essence, at least to some extent”. As a general
governmental requirement, transparency goes beyond mere access to informa-
tion, demanding that information prove understandable to external stakeholders.
However, the editors’ note of the OECD Best Practices for Budget Transparency
does not state this requirement specifically when it defines transparency as
“openness about policy intentions, formulation and implementation”, and budget
transparency as “the full disclosure of all relevant fiscal information in a timely
and systematic manner” (OECD 2001, p. 7). The OECD requires preparation of
184 J. Caamaño-Alegre et al.

budget reports, disclosure of specific data, and the establishment of procedures to

ensure integrity (Table 1).
Although exposing the detailed development of the three pillars would be
cumbersome, there are certain key elements which deserve to be highlighted.
Regarding the budget reports, the OECD Best Practices outlines both minimum
content and specific times when budget reports must be issued. Some prescrip-
tions are based on classical principles such as universality or gross budgeting.
Others, instead, are built upon more modern requirements that are still
neglected in many countries, particularly in local governments3. A similar
disparity is observed in the specific disclosures that form the second pillar
and standard governmental practices. Few local governments in the world
disclose the sensitivity analyses as required in standards 2.1 and 2.3 in fine,
to show the effects of changes in macroeconomic variables like interest or
exchange rates on budgeting. Concerning the integrity, control and account-
ability standards of the third pillar, a year-end report’s audit by the Supreme
Audit Institution is more common. Conversely, other arrangements like adjust-
ing information on previous periods to provide temporal comparability, or
making all reports available on line, seem to remain challenging for many
local governments.
Recently, revised versions of the Code and the Manual on fiscal transparency
of the International Monetary Fund were issued (IMF 2007a, 2007b). Built upon
Kopits and Craig’s (1998) definition of fiscal transparency, the code sets stan-
dards relative to the openness to the public about the structure and functions of
government, fiscal policy intentions, public sector accounts, and fiscal projec-
tions. The IMF Code is structured around four pillars: (i) clarity of roles and
responsibilities, (ii) open budget processes, (iii) public availability of informa-
tion, and (iv) assurance of integrity. The first of these pillars can be considered
beyond the scope of the OECD Best Practices, since that pillar focuses on the
establishment of a clear distinction between public corporations, the private
sector and the government. This is not the case with the last three pillars,
which maintain multiple similarities with the OECD Best Practices, but go further
in several respects4. We will thus limit the scope of budget transparency to the

Table 1. The “three pillars” of the OECD Best Practices for Budget Transparency

Budget Reports Specific Disclosures Ensuring Integrity

• The Budget • Economic Assumptions • Accounting Policies

• Pre-Budget Report • Tax Expenditures • Systems and Responsibilities
• Monthly Report • Financial Liabilities & Assets • Audit
• Mid-Year Report • Non-Financial Assets • Parliamentary and Public Scrutiny
• Year-End Report • Employee Pension Obligations
• Pre-Election Report • Contingent Liabilities
• Long-Term Report

Source: Blöndal (2006, p. 4–7).

 Budget Transparency in Local Governments 185

standards outlined in the last three pillars of the IMF Code, viewed as an updated
and less detailed version of the OECD Best Practices.
Transparency in public financial management is also receiving increased
attention in public administration, political science and economics research.
Empirical research focuses primarily on testing the benefits of fiscal transpar-
ency. This is the case of the work by Alt and Lassen (2006) focused on the
effects of fiscal transparency on the accumulation of public debt in OECD
countries. When dealing with endogeneity problems of transparency as
regressor, they also find that measures of political competition, presidential
system and common law variables do well in explaining variation in trans-
parency, whereas debt level has no statistically significant effect5.
A more direct analysis of transparency determinants is presented in Alt,
Lassen and Rose (2006). Based on work by Hanssen (2004), the authors
hypothesise that budgeting practices will be more transparent in systems
marked by high political competition. In situations of high political turnover,
incumbents will try to tie the hands of their potential successors (partisan
adversaries) by reinforcing arrangements for transparency. In a similar vein,
political polarisation increases transparency; however, a more polarised polity
could impede a cohesive reform policy and thus become a hindrance to
transparency. To these political but nonpartisan explanations, a partisan
hypothesis is added that Democratic incumbents are more prone than
Republicans to increase transparency, because of the greater transparency
required for directing additional resources toward expand the public sector6.
Along with political variables, the authors also note that financial outcomes
may positively or negatively impact transparency. Incumbents may restrict
access to information in order to avoid blame for poor fiscal performance, or
facilitate information access to get credit for a favourable fiscal record. To test
these theories, the authors construct an annual transparency score for the
years between 1972–2002 using questionnaire data from budget officers in
all 50 U.S. states. The authors then construct the following equation for
estimating fiscal transparency:

yst ¼ αst þ γt þ χpst þ β0 xst þ vst ½1

where yst represents the level of fiscal transparency in state s in year t, αs is a

state fixed effect, γt is a year indicator variable, pst is the measure of political
competition, xst represents other political and economic variables, and vst is
the error term. The xst term includes key fiscal variables such as deficit, debt,
and revenue (all measured in real per capita terms). The term also includes
socioeconomic controls such as real per capita income, income squared,
population size, population squared, percentage elderly, and percentage
school-aged. To allow for the fact that institutional changes occur, a series
of lags for the independent variables are included and the lagged dependent
variable is included as regressor.
186 J. Caamaño-Alegre et al.

The authors find that political competition tends to increase the level of
fiscal transparency, whereas polarisation appears as positively or negatively
related to transparency depending on the regression specifications. A
Democratic legislature is positively associated with transparency when con-
trolling for the effects of debt, fiscal imbalance, and polarisation; the same
occurs when using an Arellano-Bond first-difference GMM estimation of the
main equations rather than fixed effects. Higher levels of debt are associated
with lower transparency, while fiscal imbalance (in the form of higher
surpluses or deficits) appears to contribute to greater transparency.
According the authors, this “conjunction of debt and deficit results suggest
that a deficit motivates reform more where the stock of debt is lower, that
is, where the deficit more resembles a surprise or ‘crisis’” (Alt, Lassen and
Rose 2006, p. 47).
Esteller-Moré and Polo-Otero (2008) combine political competition vari-
ables with tax pressure (proxied by the property tax rate) to explain fiscal
transparency using data from Catalonian municipalities. The study also
includes electoral participation, population size, and percent elderly as control
variables. The resultant fiscal transparency indicator is a measure of municipal
accountability, specifically, the timely submission of the following nine docu-
ments to the region’s Supreme Auditing Institution: (i) opening budget, (ii)
budget liquidation, (iii) budgetary results, (iv) closed years’ budget liquida-
tion, (v) cash flow statement, (vi) cash flow remainder, (vii) balance sheet,
(viii) result statement, and (ix) debt statement. Unfortunately, this indicator
does not account for the accuracy of the documents, and tends to behave
almost as a dichotomous variable, given that the most of municipalities
confront the submission of such documents as an all-or-nothing decision7.
Nevertheless, the indicator is objective and available for a series of consecu-
tive years and for a substantial number of municipalities. Logistic model
estimation results confirm that electoral competition has a positive impact
on fiscal transparency, whereas per capita debt has a negative impact. Among
the control variables, electoral participation promotes fiscal transparency,
while population negatively affects transparency, but only for the large muni-
cipalities (defined as those with more than 5,000 inhabitants).

3. Empirical analysis
3.1. Questionnaire design and municipal governments surveyed
To measure budget transparency in Galician municipalities, we designed a
survey questionnaire composed of 15 items based on the second, third and
fourth pillars of the IMF’s revised Code of Good Practices on Fiscal
Transparency (2007). Items one through five correspond to the second
pillar (open budget processes), items six through ten to the third (public
availability of information), and items 11 through 15 to the fourth pillar
 Budget Transparency in Local Governments 187

(assurances of integrity). As indicated above, adherence to the standards

outlined in the first pillar of the IMF Code was not measured given that its
content was not covered by the OECD Best Practices for Budget
Transparency, used here as a reference to define the scope of budget transpar-
ency. Moreover, the most of the contents of the first pillar are beyond the
municipalities’ control. Definition of roles and competences along with a
legal framework for public management are determined by basic national
legislation, and are not applied at municipalities’ discretion8.
The questionnaire items are not only based on the IMF Code but on our
own practical knowledge about budgetary processes at the local level of
government. In particular, we tried to construct items to go beyond nominal
fiscal transparency to measure effective transparency, as well (Heald 2003,
2006). The item selection was aimed at covering all the areas of budget
content and budgetary process, but avoiding redundancy. To start, we con-
structed a list of 50 items which touched on the three pillars of interest within
the IMF Code. In a second step, the number of items was reduced to 15 in
order that the questionnaire not be perceived as too time-consuming. The
questions were based on a Likert (1932) scale, i.e., a summative scale
consisting of Likert-type items. These items are designed as statements with
which the respondents express their agreement or disagreement. We use
numbers from 1 to 5 instead of the common wording: strongly disagree,
disagree, no opinion, agree, and strongly agree.
The survey questionnaire was sent to 40 Galician municipalities. Responses
from 35 were received, but two had to be discarded due to various short-
comings. The 33 municipalities included in the final sample are listed by
alphabetical order in Table 3. The population is over 50,000 in three munici-
palities (13, 24, and 28), between 20,001 and 50,000 in eight municipalities
(1, 3, 10, 14, 23, 25, 27, and 33), between 10,001 and 20,000 in twelve (2, 5, 6,
9, 15, 17, 18, 22, 29, 30, 31, and 32), between 5,001 and 10,000 in seven
(7, 12, 16, 19, 20, 21, and 26) and below 5,000 in the remaining three
municipalities (4, 8, and 11). In all cases, the survey was answered by the
public officials responsible for pre-auditing transactions and/or accounting.
Those officials were selected by means of a national competitive examination,
and they have a deep knowledge of financial and fiscal laws as well as the
budgetary reality of their municipalities.

3.2. Dimensionality and internal consistency issues

The most common approach to the analysis of internal consistency of a
social research instrument is an item analysis which considers both item-
total correlation and inter-correlation among the items (Colton and Covert
2007, p. 72, 265–267). Item analyses and Cronbach’s (1951) alpha coeffi-
cients are optimal tools for the assessment of internal consistency in a
unidimensional scale with items measuring different substantive areas within
a single concept.
188 J. Caamaño-Alegre et al.

Table 2. Survey questionnaire on municipal budget transparency

QUESTIONNAIRE FOR DATA GATHERING

(SURVEY ON MUNICIPAL BUDGET TRANSPARENCY)

MUNICIPAL GOVERNMENT:..............................................................................................................
GOVERNMENT OFFICIAL
(position of the surveyed):.......................................................................................................................
INSTRUCTIONS:
Please indicate your level of agreement with the following statements as they pertain to your
municipality: 5 indicates strong agreement, 4 agreement, 3 neither agree nor disagree, 2 disagreement
and 1 strong disagreement.
Fully disagree Fully agree

1) The budget is made in accordance with the

established timeline (with no or rare
continuing appropriations) and
documentation is submitted to the opposition
parties early enough for their analysis (i.e.
earlier than scheduled budget meetings) ............................... 1 2 3 4 5
2) Budget estimates and underlying economic
assumptions are accurate and realistic................................... 1 2 3 4 5
3) Over the fiscal year, periodic information on
budget execution is submitted on time to the
entire Assembly (art. 207 TRLHL) ....................................... 1 2 3 4 5
4) Budget liquidation and general account are
presented on time and incorporate the
elements necessary to critically evaluate
management (justification of budgetary
changes, interannual comparison,
explanation of the primary deviations ...) ................................. 1 2 3 4 5
5) Budget liquidation is subject to public debate
in the Assembly...................................................................... 1 2 3 4 5
6) Financial accounting accurately reflects the
patrimonial situation (amortisations,
consistency principle, inventory, non-
accounted invoices) ................................................................ 1 2 3 4 5
7) Arrears and non-accounted invoices do not
significantly distort budget liquidation. ................................. 1 2 3 4 5
8) Spending programs are submitted to an
efficiency analysis (cost-benefit, cost-
utility…) and, in the case of investments, to a
study of their effects on running costs .................................. 1 2 3 4 5
9) Citizens participate, to some extent, in
creating a spending budget (it would
comprise of the public consultation procedure
and other ways in which a transparent
participation is assured) ......................................................... 1 2 3 4 5
10) Budgetary information is available and easily
accessible on- and off-line for consultation by
citizens.................................................................................... 1 2 3 4 5
11) In local public employment, conditions for
access and promotion are established and
fulfilled in such a way as to ensure
transparency and equal opportunity ....................................... 1 2 3 4 5
12) Actual observance of the public procurement
regulations guarantees the integrity, 1 2 3 4 5
(continued)
 Budget Transparency in Local Governments 189

Table 2. (Continued ).

QUESTIONNAIRE FOR DATA GATHERING

(SURVEY ON MUNICIPAL BUDGET TRANSPARENCY)

MUNICIPAL GOVERNMENT:..............................................................................................................
GOVERNMENT OFFICIAL
(position of the surveyed):.......................................................................................................................
INSTRUCTIONS:
Please indicate your level of agreement with the following statements as they pertain to your
municipality: 5 indicates strong agreement, 4 agreement, 3 neither agree nor disagree, 2 disagreement
and 1 strong disagreement.
Fully disagree Fully agree

transparency, impartiality and competition in

this municipal management area............................................
13) Budgeted grants are almost always awarded
by competitive merit based review........................................ 1 2 3 4 5
14) At least a 70% of the budgetary expenditures
are submitted for
pre-audit.................................................................................. 1 2 3 4 5
15) External auditing bodies receive timely
documentation of the municipality’s financial
management, and issue reports, from which
the municipality takes corrective action (in
this case, “documentation” refers to budget
liquidation, general account, lists of contracts
and inter-governmental agreements, list of
agreements to regularise irregular
expenditures, etc.) .................................................................. 1 2 3 4 5

Thus, in order to determine whether ITA is optimal, we should begin by using one
of the many factor-analytic techniques currently available to make sure that there are
no large departures from unidimensionality (Cortina 1993, p. 103). Taking into
account the low number of observations, several principal component analysis
tests confirmed that the instrument effectively was mainly unidimensional9.
In a second step, we evaluate internal consistency by examining inter-item
correlations, item-total correlations and Cronbach’s alpha. This classical
approach was used for the following reasons: (i) it is easily computable in
standard statistical packages; (ii) it compensates for the limited sample size by
providing a synthetic measure based on all correlation matrix elements; and (iii)
as multidimensionality would increase the underestimating bias of Cronbach’s α,
a high α is a reliable indicator of internal consistency10.
Prior to inter-item correlation analysis, the Kolgomorov-Smirnov test on normal-
ity of data was performed. According to results, the hypothesis of normal distribu-
tion should not be rejected (p-value < 0.05). Thus, the relationships among items can
be measured through a Pearson correlation coefficient11, as shown in Table 4.
190 J. Caamaño-Alegre et al.

Table 3. Municipalities included in the sample

1 A Estrada
2 A Pobra do Caramiñal
3 Ames
4 Baltar
5 Betanzos
6 Bueu
7 Caldas de Reis
8 Calvos de Randín
9 Cambados
10 Cangas
11 Castro Caldelas
12 Cedeira
13 Lugo
14 Marín
15 Moaña
16 Mondariz
17 Monforte de Lemos
18 Mos
19 Oroso
20 Ortigueira
21 Padrón
22 Poio
23 Ponteareas
24 Pontevedra
25 Redondela
26 Ribadavia
27 Riveira
28 Santiago de Compostela
29 Sarria
30 Teo
31 Tui
32 Verín
33 Vilagarcía de Arousa

In order to obtain a global appraisal of internal consistency, mean inter-item

correlation and an alpha coefficient for the total score were computed, with
resulting values of 0.27 and 0.85, respectively12. This alpha clearly satisfies the
widely held rule of thumb proposed by Nunnaly (1978), that an alpha of 0.7 or
higher indicates internal consistency. Indeed, such α is within the ideal range
between the consistency threshold (0.8) and the redundancy threshold (0.9)13.
The next step is to examine the consistency of each item with the scale as
a whole, so we can identify any items which do not represent the latent
variable being measured (budget transparency). The key results are in the last
two columns of Table 5. The penultimate column indicates the correlation
between each item and the average of the other items, while the last column
reveals what the consistency of our scale would be if we would delete the
given item (DeCoster 2004, p. 47). If an item has a low correlation with total
and its deletion causes a considerable increase in the alpha score, the item
is suspected of inconsistency. In our case, items 8 and 5 are the least
Table 4. Correlation analysis by Pearson coefficient

ITEM 1 ITEM 2 ITEM 3 ITEM 4 ITEM 5 ITEM 6 ITEM 7 ITEM 8 ITEM 9 ITEM 10 ITEM 11 ITEM 12 ITEM 13 ITEM 14

ITEM 2 Pearson Correlation 0.39

Sig. (2–tailed) 0.03 .
ITEM 3 Pearson Correlation 0.28 0.31
Sig. (2–tailed) 0.11 0.08 .
ITEM 4 Pearson Correlation 0.59 0.33 0.19
Sig. (2–tailed) 0.00 0.06 0.29 .
ITEM 5 Pearson Correlation 0.31 0.06 –0.01 0.31
Sig. (2–tailed) 0.08 0.75 0.97 0.08 .
ITEM 6 Pearson Correlation 0.35 0.09 0.04 0.37 0.39
Sig. (2–tailed) 0.05 0.62 0.82 0.03 0.03 .
ITEM 7 Pearson Correlation 0.36 0.46 0.35 0.47 0.27 0.44
Sig. (2–tailed) 0.04 0.01 0.05 0.01 0.13 0.01 .
ITEM 8 Pearson Correlation 0.01 0.09 0.37 0.15 –0.08 0.12 –0.03
Sig. (2–tailed) 0.97 0.61 0.03 0.41 0.65 0.50 0.89 .
ITEM 9 Pearson Correlation 0.09 0.41 0.45 –0.10 0.21 0.25 0.45 0.16
Sig. (2–tailed) 0.61 0.02 0.01 0.59 0.23 0.16 0.01 0.39 .
ITEM 10 Pearson Correlation 0.02 0.27 0.37 0.19 0.06 0.21 0.59 0.19 0.32
Sig. (2–tailed) 0.92 0.13 0.03 0.29 0.72 0.24 0.00 0.29 0.07 .
ITEM 11 Pearson Correlation 0.32 0.40 0.26 0.17 0.14 0.27 0.46 –0.19 0.12 0.33
Sig. (2–tailed) 0.07 0.02 0.14 0.34 0.42 0.13 0.01 0.30 0.51 0.06 .
ITEM 12 Pearson Correlation 0.26 0.72 0.31 0.31 0.07 0.24 0.52 0.02 0.32 0.30 0.64
Sig. (2–tailed) 0.15 0.00 0.08 0.07 0.70 0.18 0.00 0.93 0.07 0.09 0.00 .
ITEM 13 Pearson Correlation 0.04 0.58 0.14 0.23 0.05 0.20 0.29 –0.03 0.32 0.04 0.31 0.62
Sig. (2–tailed) 0.84 0.00 0.42 0.21 0.80 0.26 0.11 0.86 0.07 0.82 0.08 0.00 .
ITEM 14 Pearson Correlation 0.23 0.20 –0.02 0.28 0.18 0.51 0.58 –0.08 0.34 0.20 0.28 0.39 0.42
Sig. (2–tailed) 0.20 0.26 0.90 0.12 0.30 0.00 0.00 0.66 0.05 0.27 0.11 0.03 0.02 .
ITEM 15 Pearson Correlation 0.73 0.41 0.44 0.68 0.27 0.45 0.49 0.28 0.17 0.15 0.32 0.41 0.14 0.26
Sig. (2–tailed) 0.00 0.02 0.01 0.00 0.12 0.01 0.00 0.12 0.35 0.40 0.07 0.02 0.44 0.14
Budget Transparency in Local Governments

Source: Authors’ calculations.

191
192 J. Caamaño-Alegre et al.

Table 5. Item-Total Statistics

Scale Mean if Scale Variance if Corrected Item-Total Cronbach’s Alpha if

Item Deleted Item Deleted Correlation Item Deleted

ITEM 1 36.45 83.506 0.510 0.840

ITEM 2 35.70 84.405 0.591 0.836
ITEM 3 36.79 86.610 0.414 0.845
ITEM 4 35.70 83.593 0.535 0.838
ITEM 5 35.76 87.439 0.285 0.856
ITEM 6 36.52 86.445 0.505 0.840
ITEM 7 35.76 77.939 0.745 0.824
ITEM 8 37.27 95.080 0.109 0.856
ITEM 9 36.82 87.466 0.434 0.844
ITEM 10 36.21 87.610 0.399 0.846
ITEM 11 35.73 85.267 0.490 0.841
ITEM 12 35.55 83.568 0.652 0.833
ITEM 13 35.91 87.710 0.408 0.845
ITEM 14 35.18 84.903 0.478 0.842
ITEM 15 36.00 82.500 0.670 0.831

correlated with total and the only ones whose deletion causes an increase in
α. Although these increases are not considerable, the correlation of item 8
with the average of the other items is lower than the normal rule of thumb
cited by Steiner and Norman (1995), of at least 0.20. It suggests that item 8 is
not reflecting only budget transparency but also financial management sophis-
tication. Indeed, we expected a peculiar behavior in this item because most
Galician municipalities have not yet developed a performance management
approach, but other practical considerations led us to maintain the item.
Finally, internal consistency of our three subscales was also evaluated by
the alpha coefficient, obtaining the following values: Open Budget Processes
(items 1 to 5), 0.649; Public Availability of Information (items 6 to 10),
0.674; and Assurances of Integrity (items 11 to 15), 0.748. While these
values are lower than the total alpha for all 15 items together, they can be
considered adequate for subscales with a limited number of items. Corrected
item-total correlation and alpha if item deleted were also computed for each
subscale, in order to examine the consistency of each item with the subscale
as a whole. Again, item 8 is the only item whose correlation with the average
of the remaining items of its subscale is lower than 0.2. This result is
consistent with those of an additional PCA/FA on our subscales, according
to which all of the second subscale’s items besides item 8 loaded onto a
single factor.

3.3. Descriptive analysis of the survey results

Respondents’ mean scores by item show the distribution reflected in figure 1,
with most of the municipalities clustered around the sample mean (2.5778).
 Budget Transparency in Local Governments 193

12

10

8
Frequ en cy

Mean = 2.5778
Std. Dev. = 0.65737
N = 33
0
1.00 2.00 3.00 4.00
MEAN
Figure 1. Survey results distribution.

Insofar as this score is below 3, the average perception resulting from the survey
is like a fail grade in budget transparency. Descriptive statistics by item are
shown in Table 6.
The elements perceived as more transparent are those covered by items 12 and
14. Both items are proxies for the observance of public procurement regulations
and the pre-audit coverage. At the opposite extreme, the less transparent areas are
those represented by items 8, 9, and 3: efficiency analysis prior to approval of
spending programs, citizen participation in budget making, and periodic submis-
sion of spending information. This last result is easily understandable: while the
Spanish Local Finances Act (TRLHL) requires access to information on budget
execution, the definition is left up to each municipal government, leading to a
nearly total negligence of this obligation.
Item 8 has the lowest standard deviation of any item (0.747), suggesting that
municipalities rarely conduct efficiency analysis prior to the approval of
194 J. Caamaño-Alegre et al.

Table 6. Descriptive Statistics

N Minimum Maximum Sum Mean Std. Deviation

ITEM 1 33 1 5 73 2.21 1.293

ITEM 2 33 1 5 98 2.97 1.075
ITEM 3 33 1 5 62 1.88 1.193
ITEM 4 33 1 5 98 2.97 1.237
ITEM 5 33 1 5 96 2.91 1.444
ITEM 6 33 1 4 71 2.15 1.034
ITEM 7 33 1 5 96 2.91 1.331
ITEM 8 33 1 4 46 1.39 0.747
ITEM 9 33 1 4 61 1.85 1.064
ITEM 10 33 1 5 81 2.45 1.121
ITEM 11 33 1 5 97 2.94 1.171
ITEM 12 33 1 5 103 3.12 1.053
ITEM 13 33 1 5 91 2.76 1.091
ITEM 14 33 1 5 115 3.48 1.228
ITEM 15 33 1 5 88 2.67 1.109

Source: Authors’ survey.

spending programs. The mean item score is 1.39, with a minimum possible score
of 1. The highest standard deviations are those for items 5 (1.444), 7 (1.331), and
1 (1.293), which suggests a certain heterogeneity among municipalities with
regard to the budget liquidation, arrears and non-accounted invoices, timeliness
of budget making and openness to the opposition.
Based on the survey results, a fiscal transparency index is constructed in order
to rank the 33 municipalities with regards to transparency. The index is calcu-
lated by summing the scores recorded for all survey items. So,
X
General Transparency Index ¼ Xi ½2

where X are the item scores, and i varies from 1 to 15.

From this straightforward calculation, we calculate an index score for each
municipality, resulting in the following ranking table.
Differences in scores among municipalities are significant, with the top-ranked
municipality (Lugo) scoring 10% higher than the second-ranked municipality
(Monforte de Lemos) which, in turn, scores three times higher than the lowest
ranked municipality (Ponteareas). It is equally remarkable that the municipalities
ranked lowest for budget transparency are those in the worst economic situation
(Santiago, Cambados, Cangas, Ponteareas, Betanzos), as determined by the
Galician Supreme Audit Institution.

3.4. Econometric analysis of the determinants of budget transparency

3.4.1. Econometric specification and data. In order to disentangle the
determinants of budget transparency in local governments, we have
 Budget Transparency in Local Governments 195

Table 7. Municipal ranking table according to our Fiscal Transparency Index

Position Municipality Index Score Position Municipality Index Score

1 Lugo 62 18 Tui 39
2 Monforte de Lemos 56 19 Oroso 38
3 Rivadavia 52 20 Santiago de Compostela 37
4 Ortigueira 50 21 Moaña 36
5 A Estrada 47 21 Calvos de Randín 36
5 Padrón 47 23 Redondela 35
5 Riveira 47 24 Marín 32
8 Pontevedra 44 25 Baltar 31
8 Cedeira 44 25 Cambados 31
8 Caldas de Reis 44 25 Cangas 31
11 A Pobra do Caramiñal 43 28 Bueu 30
11 Poio 43 29 Mos 27
13 Castro Caldelas 42 30 Betanzos 24
13 Sarria 42 30 Verín 24
13 Teo 42 32 Mondariz 21
16 Ames 40 33 Ponteareas 19
16 Vilagarcía de Arousa 40

clustered explicative variables into three categories: socioeconomic factors,

fiscal variables, and political factors. Preliminary analyses were run in order
to test potential multicollinearity14. All variables and data sources are defined
in table 8. Two are socioeconomic variables (population size and unemploy-
ment rate), four are fiscal variables (per capita fiscal balance, expenditure over
municipal GDP, per capita debt, and tax pressure), and four represent political
features (incumbent’s ideology, effective number of political parties, coalition
governments, and political participation). Table 9 reports descriptive statistics
for both endogenous and exogenous variables.
The expected sign of the independent variables is the following:

● POP: The expected sign is positive. Larger administrative staffs in large

municipalities make it easier to meet transparency requirements (Esteller-
Moré and Polo-Otero 2008)15.
● U: The expected sign is negative: the worse the economic situation, the
stronger the temptation to conceal fiscal stress16.
● BAL: The effect of this variable is ambiguous. As explained in section 2, the
incumbents would hover between their desires to disengage from adverse
fiscal results and to take credit for positive fiscal outcomes (Alt, Lassen and
Rose 2006)17.
● E: Ambiguous. While a larger governmental budget may increase the demand
for fiscal transparency and provide the administrative support necessary
to meet transparency requirements (La Porte, Demchak and Jong 2002), a
larger budget may also increase opportunities for corruption and rent-seeking
(Alesina and Angeletos 2005)18.
● D: Ambiguous according to the same logic as BAL.
196 J. Caamaño-Alegre et al.

Table 8. Definition of explicative variables and data sources

Category Name Definition Data Source

Socio-economic POP Number of inhabitants expressed www.ige.eu

in thousands (2008)
Socio-economic U Unemployment rate expressed in www.ine.es
percentage (annual average
2008)
Fiscal BAL Municipal budget balance www.ccontasgalicia.es and
expressed in per capita euros www.ine.es
(2006)
Fiscal E Municipal public expenditure over www.ccontasgalicia.es
GDP expressed in percentage www.ine.es and Caixanova
(2006). Data on municipal (2008)
GDP is provided by Caixanova
(2008).
Fiscal D Municipal public debt expressed www.ccontasgalicia.es and
in per capita euros (2006) www.ine.es
Fiscal T Municipal taxes expressed in per www.ccontasgalicia.es and
capita euros (2006) www.ine.es
Political PART Political participation in the last http://www.mir.es/DGPI/
local election (2007) expressed
in percentage
Political COAL Dummy variable coded 1 in the Own authors from several
case of coalition incumbents sources
and 0 otherwise (2008)
Political LEFT Dummy variable coded 1 in the Own authors from several
case of leftist incumbents and 0 sources
otherwise (2008)
Political ENPP Effective number of political Authors calculations with data
parties. According to Laakso from http://www.mir.es/
and Taagapera (1979) it is DGPI/
computed using tbe following
formula: N ¼ Pn1 2 . Where n
i¼1
pi
is the effective number of
parties and p2i the square of
each party’s proportion of all
votes in 2007 municipal
elections.

● T: Positive. When the tax burden is higher, the taxpayers’ demand for budget
transparency is stronger19.
● PART: Positive. A higher voter turnout would indicate that citizens have a
stronger interest in government activities (La Porte, Demchak and Jong 2002:
428; Esteller-Moré and Polo-Otero 2008)20.
● COAL: Ambiguous. While Alt, Lassen and Rose (2006) argue that coalitions
lead to greater transparency, Lago-Peñas and Lago-Peñas (2010) show that
more political parties corresponds to less electoral control over incumbents.
● LEFT: Positive. Leftist incumbents are more likely to increase revenues to
expand public services, and therefore face the greater transparency require-
ments from the public21.
 Budget Transparency in Local Governments 197

Table 9. Descriptive statistics of variables

Variable Mean Standard Deviation

GTI 38.7 9.86

PTI1 12.9 4.05
PTI2 10.8 3.54
PTI3 15.0 3.99
POP 21.9 23.6
U 6.66 1.61
BAL 23.7 79.8
E 4.70 2.10
D 482.5 215.3
T 84.23 26.8
PART 66.26 5.69
COAL 0.51 DUMMY VARIABLE (0/1)
LEFT 0.66 DUMMY VARIABLE (0/1)
ENPP 3.25 0.55

● ENPP: Positive. According to Alt, Lassen and Rose (2006), incumbents will
be more likely to increase transparency when leaving office, in order to tie the
hands of their successor.

Two basic econometric specifications were estimated. While specification [3] is

used to analyse the determinants of the global transparency index (GTI), speci-
fication [4] was constructed using information on the three partial indexes (PTI)
corresponding to the subscales defined above. The list of explanatory variables is
the same in both cases:
X
n X
m X
p
GTIi ¼ α þ βh  Xhi þ δh  Whi þ γh  Zhi þ εi ½3
h¼1 h¼1 h¼1

X
n X
m X
p
PTIji ¼ αj þ βhj  Xhji þ δhj  Whji þ γhj  Zhji þ εji ðwith j ¼ 1; 2; 3Þ
h¼1 h¼1 h¼1
½4

In both expressions X is the set of n socioeconomic variables, W is the set of m

fiscal variables, and Z is the set of p political factors. Subindex i refers to
municipality (i = 1,…33) and subindex j to the partial transparency index (j = 1,
…3). The decomposition of the general transparency index into three partial indexes
P
(GTIi ¼ 3j¼1 PTIji ) allows us to triple the number of observations (from 33 to 99)
and to check for the existence of different effects of regressors on the several
areas of budget transparency. In the case of equation [3], we present results
obtained with both the full list of regressors and after dropping irrelevant
variables22, in order to gain precision and degrees of freedom in estimates
(Table 10). In the case of equation [4], only results obtained from this second
step are reported (Table 11).
198 J. Caamaño-Alegre et al.

Table 10. Econometric estimates of equation [3]

(4)
Independent variable (1) (2) (3) (dy/dx)

Intercept 28.11 37.09*** 0.004

(1.03) (3.62) (0.01)
POP 0.07
(0.73)
U –2.31*** –2.11*** –0.12*** –2.91
(3.04) (2.75) (2.95)
BAL 0.04*** 0.05*** 0.003*** 0.06
(2.65) (3.87) (4.19)
E –0.98* –0.89* –0.05* –1.20
(1.73) (1.78) (1.89)
D 0.004
(0.49)
T –0.025
(0.29)
PART 0.13
(0.54)
COAL –5.59** –6.45** –0.35** –8.75
(2.15) (2.02) (2.15)
LEFT 9.00** 9.35*** 0.50*** 12.46
(2.59) (3.31) (3.46)
ENPP 5.11 4.83** 0.260*** 6.51
(1.48) (2.34) (2.51)
R2 0.590 0.549
Log pseudo-likelihood 33.62
RESET (p-value) 0.166 0.075
Number of observations 33 33 33
Econometric Method Robust OLS Robust OLS Robust QMLE

Notes: Robust t-statistics (columns 1 and 2) or z-statistics (column 3) are in parentheses, below the
coefficients. In column (4) dy/dx is the marginal effect computed with all variables held at their mean
values (QMLE coefficients). R2 is the coefficient of determination.***p < 0.01, **p < 0.05, *p < 0.1.

3.4.2. Econometric methods and results. Equations [3] and [4] are estimated
using Ordinary Least Squares. Given that several tests (White, Breusch-Pagan-
Godfrey, Cook-Weisberg) detected the presence of heteroskedasticity, standard
errors are replaced by White robust errors in the case of specification [3] and
clustered errors by municipality in the case of specification [4]. Multicollinearity
is not a serious concern according to the so-called “Klein’s rule”.
In both cases, the Ramsey’s RESET test is computed23. Specification errors
may be discarded at usual significance levels in all cases. Goodness of fit
significantly increases when moving form specification [3] (R2 = 0.549) to
specification [4] (R2 = 0.959). Results are not reported. Normality of residuals
was also confirmed performing the Jarque-Bera test.
We also tested the endogeneity of some of the right-hand variables using a
graphical display of the cumulative sum (CUSUM) of the recursive residuals
associated with a specific ordering of cross-sectional data (de Luna and
 Budget Transparency in Local Governments 199

Table 11. Econometric estimates of equation [4]

(1) (2) (3) dy/dx

Explained Variable PTI1 PTI1 PTI1

–3.70 –0.88**
Intercept (0.46) (0.034)
U –0.86** –0.05*** –1.21
(2.46) (2.70)
BAL 0.018*** 0.001*** 0.02
(2.92) (3.15)
E –0.46 –0.25* –0.64
(1.60) (1.71)
D 0.007** 0.0004*** 0.010
(2.56) (2.79)
PART 0.20* 0.01** 0.27
(1.96) (2.08)
COAL –2.59* –0.14** –3.58
(1.82) (1.98)
LEFT 2.48* 0.13** 3.31
(1.97) (2.03)
ENPP 2.14** 0.12** 3.02
(2.05) (2.29)
Explained variable PTI2 PTI2 PTI2
Intercept 8.24*** –0.24***
(7.72) (4.20)
POP 0.05* 0.003* 0.07
(1.75) (1.86)
COAL –1.79 –0.10* –2.49
(1.51) (1.64)
LEFT 3.43*** 0.19*** 4.70
(2.81) (3.00)
Explained Variable PTI3 PTI3 PTI3
Intercept 20.85*** 0.48**
(5.51) (2.46)
U –0.88** –0.05** –1.24
(2.23) (2.43)
BAL 0.016** 0.001** 0.022
(2.09) (2.34)
E –0.45 –0.26 –0.47
(1.36) (1.56)
D 0.0003 0.00002 0.0003
(0.08) (0.11)
COAL –1.97* –0.12** –2.94
(1.98) (2.19)
LEFT 3.81*** 0.22*** 5.24
(3.08) (3.48)
R2 0.959
Log pseudo-likelihood –44.06
RESET (p-value) 0.569
Number of observations 99 99
Econometric Method Clustered OLS Clustered QMLE

Notes: Robust t-statistics (column 1) or z-statistics (column 2) are in parentheses, below the
coefficients. In column (3) dy/dx is the marginal effect computed with all variables held at their
mean values (QMLE coefficients). R2 is the coefficient of determination.***p < 0.01, **p < 0.05,
*p < 0.1.
200 J. Caamaño-Alegre et al.

Johansson 2008). In particular we tested the endogeneity of variables BAL, E and

D (T is not relevant to the explanation of GTI or PTIj) by sorting the data with
respect to those variables and looking at recursive residuals obtained from the
estimates in column 1 of Table 10. Figures 2 to 4 display the CUSUM plots for
BAL, E and D and show no evidence of endogeneity.
Finally, we checked the robustness of the results using the quasi-maximum
likelihood estimator (QMLE) suggested by Papke and Woolridge (1996) to deal
with constrained fractional response variables. By definition, variables GTI and
PTI are constrained on the interval [15–75] and [5–25] respectively. However,
OLS does not guarantee that the predicted values of the dependent variable lie
on the unit interval. The QMLE do not suffer from this shortcoming. To
facilitate the comparison between the OLS and the QMLE estimates, in the
last column of Tables 10 and 11 the marginal effects of regressors in the sixth
model are shown. Since the value of the marginal effect depends on the value
of all variables in the model, they have been computed with all variables held
constant at their mean values, except in the case of dummy variables, computed
for discrete changes from 0 to 1. Results from OLS and QMLE are the same,
with coefficients and z-statistics slightly higher in absolute value in the case of
the latter.
In the case of the global transparency index, five regressors are significant
at the 5% level (U, BAL, COAL, LEFT, ENPP) and one is significant at
10% (E). Variables ENPP, LEFT, and BAL have a positive coefficient and

15

10

–5

–10

–15
12 14 16 18 20 22 24 26 28 30 32

CUSUM 5% Significance

Figure 2. CUSUM plot. Data sorted with respect to variable E.

 Budget Transparency in Local Governments 201

15

10

–5

–10

–15
12 14 16 18 20 22 24 26 28 30 32

CUSUM 5% Significance

Figure 3. CUSUM plot. Data sorted with respect to variable BAL.

U, E, and COAL show negative coefficients. Hence transparency is higher in

the case of leftist incumbents and increases with electoral competition and
better budget balance. Conversely, unemployment, coalition incumbents, and
budget size in terms of municipal GDP reduce global transparency. These
results deserve some comments. Our previous hypotheses regarding the
unemployment impact on transparency are consistent with estimates obtained
for both U and BAL. The most surprising result is the negative sign on COAL.
It suggests a trade-off between the desires of tying the partner’s hands
(transparency) and escaping the vigilance of the partner (opacity), and is
resolved in favor of the latter. This is understandable given that settling new
control or transparency arrangements is more conflictive when controllers and
controlees are in different parties; “turning a blind eye” is one way to reduce
conflict to tolerable levels24.
Results for partial indexes are also very interesting. The sign of all
coefficients is the same as than in Table 10. Variable POP is now significant
at the 10% level under PTI2 with a positive sign. Variable PART has also
become significant to explain PTI1: this is the only partial index that increases
with political participation. Variables LEFT and COAL are the only variables
relevant as a predictor for all the three partial indexes. Fiscal variables are
important predictors of PTI1 but not the rest of partial indexes, except BAL in
the case of PTI3. Unemployment is a significant predictor of both the first
202 J. Caamaño-Alegre et al.

15

10

–5

–10

–15
12 14 16 18 20 22 24 26 28 30 32

CUSUM 5% Significance

Figure 4. CUSUM plot. Data sorted with respect to variable D.

and the third partial indexes. Finally, ENPP is only relevant in the case
of PTI1.
Using the marginal effects computed with QMLE for the global index, we can
compare the relative quantitative relevance of the most significant variables. The
effect of LEFT is strong: leftist governments tend to enjoy transparency levels
12.46 points higher than more conservative governments. Coalition incumbents
reduce its transparency by 8.75 points. The unemployment rate reduces global
transparency by 2.91 points. An unitary increase in the effective number of
political parties increases the global index by 6.51, but the standard deviation
of this variable is low (0.55). Finally, the marginal effect of BAL is very small but
the standard deviation is large (more than three times the mean). Hence, its effect
is also important in some cases.

4. Concluding remarks
From a methodological standpoint, our first conclusion is that relying on
subjectivity-laden data from internal sources does not necessarily suppose
an overoptimistic bias in transparency scores. Second, results made clear
the necessity of a more in-depth examination of the survey questionnaires’
dimensionality and internal consistency. Certain features of the budgetary
 Budget Transparency in Local Governments 203

process may be attributes of more than one latent variable, rather than
reflecting only budget transparency.
Finally, while some of our empirical results on the determinants of budget
transparency confirm previous findings, others point to new relationships. This is
the case for the statistical significance of unemployment, the negative relation-
ship between coalitions and transparency, and the positive impact of debt on the
first partial transparency index. This last result stands in contrast to the negative
effect reported by Alt, Lassen and Rose (2006) and Esteller-Moré and Polo-Otero
(2008), and the negative influence of deficit on transparency. This contrast could,
however, have a suitable explanation. While debt may have been accumulated
during past fiscal stresses and not to be attributable to the current or previous
government, deficit supposes a recent and even ongoing fiscal stress. Thus, our
results may be suggesting that governments are more prone to enhance transpar-
ency when inheriting a heavy fiscal burden (high debt) and enacting sound
spending policies (low deficit).

Acknowledgements
We are grateful to Philip Whiteman and one anonymous referee for their
insightful comments. The usual disclaimer applies. The research was financially
supported by the Spanish Ministry of Science and Innovation (ECO2010-
15553).

Notes
1. In this article, the word “scale” usually refers to “an instrument made up of multiple
items that have a relationship to each other as well as to the concept of interest
[budget transparency, in our case]” (Colton and Covert 2007, p. 249). A Likert scale
is a scale composed of Likert-type items, as we explain in subsection 3.1. Although
there are precedents for using Likert-type items in measuring budget transparency at
the national level (Lavielle, Pérez and Hofbauer 2003, Pérez 2005), leading studies
on determinants of budget transparency have tended to rely on the above-mentioned
kinds of data.
2. According to Colton and Covert (2007, p. 68), content validity is the degree to which
an instrument is representative of the topic and process being investigated.
3. For instance: (i) inclusion of comparative and non-financial information, (ii) explana-
tion of deviations, (iii) consistency in format between the budget and the final report,
(iv) issuing of monthly reports within four weeks, and (v) audit and issuing of the
year-end report within of six months.
4. The second pillar also requires “clear mechanisms for the coordination and manage-
ment of budgetary and extra-budgetary activities within the overall fiscal policy
framework”. The third and fourth pillars incorporate innovative provisions such as:
i) publication of a periodic report on long-term public finances;
ii) openness in purchase and sale of public assets;
iii) mechanisms to monitor follow-up actions recommended by the national auditor;
iv) assessment of the fiscal forecasts, the macroeconomic forecasts, and their under-
lying assumptions by independent experts; and
v) institutional independence of the national statistical body.
204 J. Caamaño-Alegre et al.

5. See also Jarmuzek (2006). Although he does not detail the effects of his instrumental
variables on fiscal transparency, the variables used in the model include a political
competition index (i.e. past turnover), the rule of law, and the media freedom index
compiled by the Freedom House.
6. The authors base this hypothesis on model implications from Ferejohn’s (1999).
7. As Islam (2006, p. 154) acknowledges, transparency indicators “could be strength-
ened by considering not just the frequency and availability of data but also the quality
of the data produced by governments.”
8. Although we are well aware of the threats that public corporations and other entities
depending on a municipality may pose to fiscal transparency, these issues have been
considered beyond the scope of the more restricted notion of budget transparency
adopted in our empirical research. All in all, revenues and expenditures of the
autonomous bodies or agencies depending on a municipality are included in the
consolidated budget of local governments. Regarding public corporations, only three
of the municipalities under study have any of them.
9. Plotting the magnitude of the successive eigenvalues and applying the Cattell’s
(1966) Scree test, a sharp drop in eigenvalues from component one is observed.
10. Within a hierarchical factor model framework, Zinbarg et al. (2005) demonstrate that α
underestimates reliability in the first three of four theoretical scenarios: (1) unidimen-
sional scales with unequal general factor loadings, (2) multidimensional scales with
equal general factor loadings, and (3) multidimensional scales with unequal general
factor loadings. Only in the fourth case, unidimensional scales with equal general factor
loadings – i.e., essential tau equivalence – is α as appropriate as ω to measure
reliability. Vehkalahti, Puntanen and Tarkkonen (2006, p. 2) also use alpha’s under-
estimating bias to justify the search for and proposal of a new estimator suggesting that
this estimator, called Tarkkonen’s rho, provides a better alternative for Cronbach’s α.
11. According Revelle (2010, p. 216), the correlations associated with an ordinal scale are
not Pearson’s but Spearman’s. However, Garson (2008) explains that ordinality in
Likert scales refers only to an ordinal relationship of values within a single item:
Likert response values are ordinal within any given item but sets of Likert items are not
necessarily ordinal with respect to each other, and they can be used to form indexes.
12. When calculated from the correlation matrix (Table 4), a standardised alpha of 0.849 is
obtained – the same could be calculated from the covariance matrix by standardizing and
summing all items in the scale. In our case, the raw alpha provided by SPSS (0.850) is
practically equal to the standardised alpha. In Table 5, raw alphas from SPSS are reported.
13. Although the widely-accepted social science criterion is that alpha should be 0.70 or higher
for a set of items to be considered a scale, and even a lenient cut-off of 0.60 is common in
exploratory research, many researchers require a score of 0.80 for a “good scale” (Garson
2008, 2010). On the other hand, a high alpha may also suggest a high level of item
redundancy, wherein essentially the same item is rephrased in several different ways (Boyle
1991). Thus, for example, Fitzpatrick et al. (1998, p. 23) consider that alpha values should
not be higher than 0.9 for scales which are used as research tools to compare groups. A very
high α suggests that there is some redundancy among items, and the possibility that the
items together are addressing a rather narrow aspect of an attribute.
14. In particular, we detected a high correlation between aging, population size, munici-
pal public expenditure over municipal GDP, and per capita GDP. Youth migrate from
backward municipalities in terms of GDP, bringing about both the ageing and
reduction of total population. Insofar as local expenditure is partially financed by
equalisation grants, the public expenditure/GDP ratio tends to be higher in the
backward municipalities. Dropping per capita GDP and ageing from the specification
substantially reduced multicollinearity among the independent variables.
15. Alt, Lassen and Rose (2006) find a non-significant effect. Esteller-Moré and Polo-Otero
(2008) obtain a negative effect, but only for municipalities with over 5,000 inhabitants.
 Budget Transparency in Local Governments 205

16. LaFaive (2009) detects a negative relationship between unemployment and the
transparency of a Michigan’s program for economic development. However,
Peixoto (2010) finds no correlation between the US states’ levels of unemployment
and the transparency of their recovery websites. Moreover, Andersen and Nielsen
(2010, p. 28) suggest that the extremely damaging nature of procyclical fiscal policies
during recessions may trigger reforms that increase the degree of fiscal transparency.
Other studies on transparency also consider unemployment (Alt, Lassen and Skilling
2001, Rosendorff and Vreeland 2009), but not as a determinant of transparency.
17. In Alt, Lassen and Rose’s (2006) empirical analysis, the resulting sign is positive for
imbalance (both surplus and deficit), implying that a greater deficit leads to greater
fiscal transparency.
18. La Porte, Demchak and Jong (2002) fail to obtain empirical support for their hypothe-
sised positive effect of government size on openness, although in some of their tests
the “government size” variable does not remain after removing accounting for obvious
collinearity. Bastida and Benito (2007, p. 690–691) observe that the relative size of
central government presents a low significant (p-value = 0.079) negative relationship
(–0.260) with budget transparency. According to the authors, although the study
shows that larger central governments are linked to lower levels of transparency, the
low significance prevents using the evidence to draw a strong conclusion.
19. Alt, Lassen and Rose (2006) find that per capita general revenues are not significant
in most estimates. In Esteller-Moré and Polo-Otero (2008) taxes are only significant
at 10% in a cluster model, with the expected positive sign.
20. Although La Porte, Demchak and Jong (2002) find no empirical support for the hypothe-
sized impact of democracy on web site openness, they appear not include the “voting”
variable in testing. Esteller-Moré and Polo-Otero (2008), however, confirm a positive
relationship between electoral participation and fiscal transparency, and Rosendorff and
Vreeland (2009) do the same for democracy and transparency. Finally, Bastida and
Benito (2007, p. 690) find no significant correlations between their democratic-level
variables and budget transparency, although they suggest that the direction of the
relationships supports the notion that greater levels of political and civil liberty corre-
spond to higher fulfilment of the OECD Best Practices for Budget Transparency.
21. As pointed out in the above mentioned section, Alt, Lassen and Rose’s (2006)
econometric analysis corroborates this positive relationship when controlling for the
effects of debt, fiscal imbalance, and polarisation, as well as when using an Arellano-
Bond GMM first-difference estimation of their main equations. By contrast, Esteller-
Moré and Polo-Otero (2008) find the variable is not significant, and Bastida and Benito
(2007, p. 692) do not find any relationship between ideology and budget transparency.
22. Variables defined as those with a t-statistic below unity (in absolute value) in the first
step: 4 variables in the case of equation [3] and up to 13 variables in the case of [4].
A Wald test on the joint significance of those variables clearly failed to reject the null
hypothesis in both cases (p-values = 0.73 and 0.42, respectively)
23. RESET is a general test for the following types of specification errors: Omitted variables,
incorrect functional form, and correlation between the exogenous variables and the random
term which may be caused by measurement error or simultaneity, among other things.
24. Following a referee’s suggestion, we also included the rate of female aldermen in the
council as explanatory variable in specification [3], but it was not statistically
significant (p-value > 0.50). In this sense, it is interesting to note that Piotrowski
and Van Ryzin (2007) find that this variable is relevant to explain the demand for
safety transparency, but not to explain differences in fiscal transparency.

References
Alesina, A. and Angeletos, G.M., 2005. Corruption, inequality and fairness. Journal of
Monetary Economics, 52 (7), 1227–1244.
206 J. Caamaño-Alegre et al.

Alt, J.E. and Lassen, D.D., 2006. Fiscal transparency, political parties, and debt in OECD
countries. European Economic Review, 50, 1403–1439.
Alt, J.E., Lassen, D.D., and Rose, S., 2006. The causes of fiscal transparency: evidence
from the U.S. states. IMF Staff Papers, 53, 30–57.
Alt, J.E., Lassen, D.D., and Skilling, D., 2001. Fiscal transparency, gubernatorial popu-
larity, and the scale of government: evidence from the States. EPRU Working Paper
Series, 16. Copenhagen: Economic Policy Research Unit, University of Copenhagen.
Andersen, A.L. and Nielsen, L.H.W., 2010. Fiscal transparency and procyclical fiscal
policy. EPRU Working Paper Series, 1. Copenhagen: Economic Policy Research Unit,
University of Copenhagen.
Bastida, F. and Benito, B., 2007. Central government budget practices and transparency:
an international comparison. Public Administration, 85 (3), 667–716.
Blöndal, J.R., 2006. Budget transparency: experiences from OECD countries. Meeting of
Senior Budget Officials from Latin American Countries, Santiago de Chile, 25 January.
Boyle, G.J., 1991. Does item homogeneity indicate internal consistency or item redundancy
in psychometric scales? Personality and Individual Differences, 12 (3), 291–294.
Caixanova, 2008. Atlas socioeconómico de Galicia 2008. Vigo: Caixanova.
Cattell, R.B., 1966. The scree test for the number of factors. Multivariate Behavioral
Research, 1 (2), 245–276.
Colton, D. and Covert, R.W., 2007. Designing and constructing instruments for social
research and evaluation. San Francisco: Jossey-Bass.
Cortina, J.M., 1993. What is coefficient alpha? An examination of theory and applications.
Journal of Applied Psychology, 78 (1), 98–104.
Cronbach, L., 1951. Coefficient alpha and the internal structure of tests. Psychometrika,
16 (3), 297–333.
DeCoster, J., 2004. Data Analysis in SPSS. Available from http://www.stat-help.com/spss.pdf
De Luna, X. and Johansson, P., 2008. Graphical diagnostics of endogeneity. In: D.
Millimet, J.A. Smith, and E. Vytlacil eds. Advances in econometrics 21 (modelling
and evaluating treatment effects in econometrics. Amsterdam: Elsevier, 147–166.
Esteller-Moré, A. and Polo-Otero, J., 2008. Analysis of the causes of fiscal transparency:
Evidence from Catalonian municipalities. Unpublished paper.
Ferejohn, J., 1999. Accountability and authority: toward a theory of political account-
ability. In: A. Przeworski, S.C. Stokes and B. Manin, eds. Democracy, accountability
and representation, Cambridge: Cambridge University Press, 131–153.
Fitzpatrick, R., et al., 1998. Evaluating patient-based outcome measures for use in clinical
trials. Health Technology Assessment, 2 (14), i–74.
Garson, G.D., 2008. Scales and standard measures [online]. Available from: http://faculty.
chass.ncsu.edu/garson/PA765/standard.htm [Accessed 22 April 2010].
Garson, G.D., 2010. Reliability analysis. Available from: http://faculty.chass.ncsu.edu/
garson/PA765/reliab.htm
Hanssen, F.A., 2004. Is there a politically optimal degree of judicial independence?
American Economic Review, 94, 712–729.
Heald, D., 2003. Fiscal transparency: concepts, measurement and UK practice. Public
administration, 81 (4), 723–759.
Heald, D., 2006. Varieties of transparency. In: C. Hood and D. Heald, eds. Transparency:
the key to better governance? Oxford: Oxford University Press, 25–43.
Hood, C., 2006. Transparency in Historical perspective. In: C. Hood and D. Heald, eds.
Transparency: the key to better governance? Oxford: Oxford University Press, 3–23.
IMF, 2007a. Code of good practices on fiscal transparency (2007) [online]. Washington,
DC: International Monetary Fund. Available from: http://www.imf.org/external/np/pp/
2007/eng/051507c.pdf [Accessed 1 October 2009].
IMF, 2007b. Manual on fiscal transparency (2007) [online]. Washington DC: International
Monetary Fund. Available from: http://www.imf.org/external/np/pp/2007/eng/101907m.pdf
[Accessed 1 October 2009].
 Budget Transparency in Local Governments 207

Islam, R., 2006. Does more transparency go along with better governance? Economics
and Politics, 18 (2): 121–167.
Jarmuzek, M., 2006. Does fiscal transparency matter? The evidence from transition
economies [online]. Available from: http://www.cerge.cuni.cz/pdf/gdn/RRCV_77_
paper_03.pdf [Accessed 31 July 2009].
Kopits, G. and Craig, J., 1998. Transparency in government operations. IMF Occasional
Papers, 158. Washington, DC: International Monetary Fund.
Laakso, M. and Taagepera, R., 1979. Efective number of parties: a measure with applica-
tion to Western Europe. Comparative Political Studies, 13, 3–27.
LaFaive, M.D., 2009. MEGA, the MEDC and the loss of sunshine. Policy Brief 5,
Midland, MI: Mackinac Center for Public Policy.
Lago-Peñas, I. and Lago-Peñas, S., 2010. Decentralization and electoral accountability.
Environment and Planning C: Government and Policy, 28, 318–334.
La Porte, T.M., Demchak, C.C. and Jong, M. de, 2002. Democracy and bureaucracy in the
age of the web: empirical findings and theoretical speculations. Administration &
Society, 34 (4), 411–446.
Lavielle, B., Pérez, M., and Hofbauer, H. (coords), 2003. The Latin American index of
budget transparency 2003: a comparison of 10 countries [online]. Lima, Peru: Fundar.
Available from: http://www.internationalbudget.org/themes/BudTrans/English.pdf
[Accessed 30 July 2010].
Likert, R., 1932. A technique for the measurement of attitudes. Archives of Psychology,
140, 1–55.
Nunnaly, J., 1978. Psychometric theory. New York: McGraw-Hill.
OECD (Organisation for Economic Cooperation and Development), 2001. OECD best
practices for budget transparency [online]. Paris: OECD. Available from: http://www.
oecd.org/dataoecd/33/13/1905258.pdf [Accessed 30 July 2010].
Papke, L.E. and Wooldridge, J. M., 1996. Econometric methods for fractional response
variables with an application to 401(k) plan participation rates. Journal of Applied
Econometrics, 11, 619–632.
Peixoto, T., 2010. Transparency on the web: a democrat(ic) virtue? [online]. Personal
Democracy Forum. TechPresident. Available from: http://techpresident.com/blog-entry/
transparency-web-democrat-virtue [Accessed 30 July 2010].
Pérez, M. (coord.), 2005. Latin American index of budget transparency 2005: a compar-
ison of 8 countries [online]. Managua, Nicaragua: CISAS. Available from: http://www.
fundar.org.mx/indice2005/docs/Regional%20Transparency%20Report%202005.pdf.
[Accessed 30 July 2010].
Piotrowski, S.J. and Van Ryzin, G.G., 2007. Citizen attitudes toward transparency in local
government. The American Review of Public Administration, 37 (3), 306–323.
Revelle, W., 2010. An introduction to psychometric theory with applications in R [online].
Available from: http://personality-project.org/r/book/ [Accessed 22 May 2010].
Rosendorff, B.P. and Vreeland, J.R., 2009. Electoral accountability and data dissemina-
tion: the effect of democracy on transparency [online]. Available from: https://files.nyu.
edu/bpr1/public/papers/BPRJRV.pdf [Accessed 30 July 2010].
Streiner, D.L. and Norman, G.R., 1995. Health measurement scales: a practical guide to
their development and use, 2nd ed. Oxford: Oxford University Press.
Vehkalahti, K., Puntanen, S., and Tarkkonen, L., 2006. Estimation of reliability: a better
alternative for Cronbach’s alpha [online]. Reports on Mathematics 430, Department of
Mathematics and Statistics, University of Helsinki, Helsinki, Finland. Available from:
http://mathstat.helsinki.fi/reports/Preprint430.pdf [Accessed 23 April 2010].
Zinbarg, R., et al., 2005. Cronbach’s α, Revelle’s β, and McDonald’s ωh: their relations with each
other and two alternative conceptualizations of reliability. Psychometrika, 70 (1), 123–133.
Copyright of Local Government Studies is the property of Routledge and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.