BRUEGEL

POLICY
CONTRIBUTION
ISSUE 2015/09
JUNE 2015

THE EFFECTS OF
ULTRA-LOOSE
MONETARY POLICIES
ON INEQUALITY
GRGORY CLAEYS, ZSOLT DARVAS, LVARO LEANDRO AND
THOMAS WALSH

Highlights
Low interest rates, asset purchases and other accommodative monetary policy
measures tend to increase asset prices and thereby benefit the wealthier segments of society, at least in the short-term, given that asset holdings are mainly
concentrated among richest households.
Such policies also support employment, economic activity, incomes and inflation,
which can benefit the poor and middle-class, which have incomes more dependent on employment and which tend to spend a large share of their income on debt
service.
Monetary policy should focus on its mandate, while fiscal and social policies
should address widening inequalities by revising the national social redistribution
systems for improved efficiency, intergenerational equity and fair burden sharing
between the wealthy and poor.
Telephone
+32 2 227 4210
[email protected]
www.bruegel.org

Grgory Claeys ([email protected]) is a Research Fellow at Bruegel. Zsolt

Darvas ([email protected]) is a Senior Fellow at Bruegel. lvaro Leandro
([email protected]) and Thomas Walsh ([email protected]) are
Research Assistants at Bruegel. This Policy Contribution was prepared for the
European Parliament Committee on Economic and Monetary Affairs ahead of the
European Parliaments Monetary Dialogue with European Central Bank President Mario
Draghi on 15 June 2015. Copyright remains with the European Parliament at all times.
We thank our Bruegel colleagues for their helpful comments and suggestions and
Sbastien Prez-Duarte for his valuable support in relation to the European Central
Bank's Household Finance and Consumption Survey.

BRU EGE L

POLICY

CONTRIBUTION

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

02

THE EFFECTS OF ULTRA-LOOSE MONETARY

POLICIES ON INEQUALITY
GRGORY CLAEYS, ZSOLT DARVAS, LVARO LEANDRO AND THOMAS WALSH, JUNE 2015
1 INTRODUCTION
Since 2008, all major central banks have engaged
in monetary easing through conventional interest
rate cuts, and through unconventional measures,
such as asset purchases, long-maturity lending
and forward guidance about intended future monetary policy actions. We call these unconventional
measures ultra-loose monetary policies (ULMP).
Such measures have increased significantly the
size, and changed the composition of, the central
banks balance sheets1. The main reason for these
various unconventional policies and low interest
rates is that central banks try to set interest rates
at, or around the so-called 'natural rate' of interest,
a level consistent with low and stable inflation and
with an economy near its potential. In the last few
years, the too-low and below-target inflation, low
inflation expectations, the low level of capital utilisation and the high level of unemployment suggest that the natural rate of interest has been well
below the policy rate, which has been constrained
by the zero lower bound.
While these various monetary easing measures are
justified from a macroeconomic perspective, and in
fact the European Central Bank should have adopted
expansionary measures much earlier (Claeys et al,
2014), they might have various side effects.

1. See Claeys (2014) for an

overview of such policies.

One possible concern is the impact on financial

stability. By analysing various theoretical considerations and the current situation of the euro area,
we (Claeys and Darvas, 2015) concluded that the
risks to financial stability of ultra-loose monetary
policy in the euro area could be low. We argued
that monetary policy should focus on its primary
mandate of area-wide price stability, and other
policies should be deployed whenever the financial cycle deviates from the economic cycle or
when heterogeneous financial developments in
the euro area require financial tightening in some
but not all countries. These policies include micro-

prudential supervision, macro-prudential oversight, fiscal policy and regulation of sectors that
pose financial stability risks, such as construction.
Another potential concern is the impact of ultraloose monetary policy on income and wealth distribution. Several observers, such as Cohen
(2014), Stiglitz (2015) and Acemoglu and Johnson (2012), have accused central banks of favouring the rich and fuelling the increase in income
and wealth inequality. Inequality is a concern from
both social and economic perspectives (Piketty,
2014). The long-held view of economists that
there exists an inherent trade-off between efficiency and equality (Okun, 1975) has recently
come into question, with inequality itself being put
forward as the potential cause of the crisis. High
levels of inequality might urge households to rely
on debt financing to maintain living standards,
which might have been an important driver of the
housing boom in the pre-crisis period in the US,
and thereby the consequent bust (Rajan, 2010;
Van Treeck, 2014). Ostry et al (2014) claim that
greater inequality could reduce the level and duration of periods of growth, while greater inequality
can also be linked with greater financial instability (Skott, 2013; Vandemoortele, 2009). For the
euro area, Darvas and Wolff (2014) showed that
countries with greater inequality tended to have
higher household borrowing prior to the crisis,
resulting in more subdued consumption growth
during the crisis. The resulting high private debt,
high unemployment, poverty and more limited
access to education undermine long-term growth
and social and political stability.
The rise of inequality is mainly seen as a long-term
trend resulting from deep structural changes that
could be attributed to skill-biased technological
change, globalisation, demography, institutional
and political changes and in particular changes in
fiscal, educational and labour institutions (Piketty,
2014). Using the Gini coefficient and the share of

BRU EGE L

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

POLICY

CONTRIBUTION

03
income going to the top one percent, Figure 1
shows that income inequality in major advanced
countries declined somewhat after the second
world war until about the 1970s, when it started
to increase in most countries. Figure 1 also shows
that there are major differences between countries. For example, Germany is more equal than
the United States or the United Kingdom. Figure A1
in the Annex reveals significant differences
between euro-area countries.

the central bank might purchase other securities

and thereby the prices of all kinds of assets can
increase (portfolio rebalancing effect). Fourth,
asset purchases by central banks can also
improve market functioning and liquidity, thereby
reducing liquidity premia, which can further raise
asset prices. And finally, ultra-loose monetary
policies can convince investors that interest rates
will remain low for a long period, which can affect
future corporate earnings and raise asset prices.

This Policy Contribution assesses the impact of

ultra-loose monetary policies on income and
wealth distribution in the euro area. Section 2
assesses the potential impacts through financial
markets, while section 3 considers the impacts
through changes in the macroeconomic situation.
Section 4 concludes.

Empirical estimates for the United Kingdom and

United States by Joyce et al (2011), Meier (2009),
Gagnon et al (2011) and Baumeister and Benati
(2010) found significant effects of asset purchases on the prices of the assets purchased, and
also on other securities not included in the purchase programmes, including equity prices. However, as argued by Dobbs, Koller and Lund (2014),
the effect of asset purchases on equity prices
might not be as strong as is often reported, for
both theoretical and empirical reasons. First, a
rational investor should regard the current ultralow interest rate environment as temporary, and
thus should not reduce the discount rate to value
future cash flows. As Figure 5 of Claeys and Darvas
(2015) shows, P/E ratios have remained close to
their long-term average in the US, UK and euro
area, suggesting that share prices might not have
been boosted extraordinarily, but might have primarily rebounded from extremely low levels.
Second, according to calculations in Dobbs et al
(2014), the implied real cost of equity, which represents the compensation investors require for

2 THE IMPACT OF ULTRA-LOOSE MONETARY

POLICIES ON INEQUALITY THROUGH
FINANCIAL MARKETS
2.1 The impact through asset prices
One of the main channels through which ultraloose monetary policies affect income and wealth
distribution is changes in asset prices. First, lower
central bank interest rates reduce the interest
rates on securities (such as government and corporate bonds) and increase their prices. Second,
asset purchases result in increases in the prices
of the assets purchased, and a further fall in their
yields. Third, sellers of the assets purchased by
Figure 1: Measures of inequality
A: Gini coefficient of income inequality
(after taxes and transfers), 1960-2013

40

B: Share of income going to the 1%

(before taxes and transfers), 1946-2012

21
19

35

17
15

30

13
25

11
9

20

United States

United Kingdom

Spain

France

Germany

Italy

2010

2006

2002

1998

1994

1990

1986

1982

1978

1974

1970

1966

1962

1958

1954

1950

1946

2011

2008

2005

2002

1999

1996

1993

1990

1987

1984

1981

1978

1975

1972

1969

1966

1963

15

1960

Japan

Source (Panel A): Standardised World Income Inequality Database. Note: the Gini coefficient ranges from 0 to 100, with 100
indicating complete inequality. It is a function of the surface between the Lorenz curve (which is the cumulative distribution
function of the probability distribution of income) and the line of equality. Source (Panel B): Top World Incomes Database
(http://topincomes.parisschoolofeconomics.eu/).

BRU EGE L

POLICY

CONTRIBUTION

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

04
investing in equities instead of risk-free securities
such as Treasuries, has not fallen to a level that
would be expected in the context of a big boost in
equity prices. Finally, in order for the portfoliorebalancing channel to work, equity must be seen
by investors as a close substitute for fixed-income
assets. The authors give some reasons why this
might not be the case: high volatility in the equity
market, which should deter investment in equity,
or the retreat by US retail investors from equity
mutual funds and exchange-traded funds. Overall,
Dobbs et al estimate that, if interest rates rise to
their long-term historical average levels in five
years, low rates will have resulted in an increase in
equity prices of only about one percent.
More generally, the effects of monetary policy on
asset prices should average out over the long
term. First, the exit from quantitative easing and
the tightening of monetary policy through interest
rate rises should have the opposite downward
effect on asset prices. Second, equity prices are

ultimately a function of the profitability of firms

and even though they can diverge from their fundamental values in the short-term, they should
not diverge permanently. While monetary policy
should boost economic activity and thereby corporate profits in the short-term, the so-called longrun neutrality hypothesis suggests that it does
not have such an effect in the long-term.
While the above literature review suggests some
ambiguity about the extent and duration of asset
price increases after asset purchases, asset price
increases at least in the short-term can have significant distributional consequences given that
asset holdings are very much concentrated
among the richest households. The Household
Finance and Consumption Survey (HFCS) of the
European Central Bank2, shows that differences in
net wealth between the wealthy and the poor are
huge (Figure 2 and Figure A2 of the Annex)3. Figure
3 also shows that poorer households hold generally fewer financial assets except deposits.

Figure 2: Net wealth by wealth percentiles in the euro area and its four largest countries
1400

Wealth percentiles
1200

2. The Household Finance

and Consumption Statistics
(HCFS) survey by the European Central Bank collected
household-level data on
households finances and
consumption in 15 European countries (Belgium,
Germany, Greece, Spain,
France, Italy, Cyprus, Luxembourg, Malta, Netherlands, Austria, Portugal,
Slovenia, Slovakia and Finland). Data was collected in
2010 and 2011 in most
countries. Henceforth,
when we refer to the euro
area, we refer to these 15
countries only.
3. Figures 2 and 3 show
data according the net
wealth percentiles: the
figures according to income
percentiles are very similar.

Mean net wealth ( 000s)

Bottom 20%
1000

20-40%
800

40-60%
600

60-80%
400

80-90%
200

90-100%
0
-200

Euro area

France

Germany

Italy

Spain

Source: ECB HFCS (2013) Note: Net Wealth is the difference between total household assets and total household liabilities.
Total assets include real and financial assets. Euro area refers to the aggregate of the 15 countries included in the HFCS (see
footnote 2).

Figure 3: Share of euro-area households with holdings of financial assets by wealth percentiles (%)
100

Wealth percentiles
80

Bottom 20%
20-40%

60

40-60%
40

60-80%
80-90%

20

90-100%
0

Deposits

Mutual Funds

Bonds

Shares

Money owed Pension and Other financial

to household life insurance
assets

Source: ECB HFCS (2013). Note: Euro area refers to the aggregate of the 15 countries included in the HFCS (see footnote 2).

BRU EGE L

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

POLICY

CONTRIBUTION

05
However, while current asset price increases
benefit those that have large holdings of assets
today, they also make future buyers of these
assets worse off, as they will have to purchase
them at higher prices. In general, it is older
households that tend to hold these assets and
plan to sell them in the future in order to maintain
their consumption, while younger households will
buy these assets in the future in order to save for
retirement. This will have distributional effects
across generations.

between countries and whether we consider

income or wealth distributions. In most southern
euro area countries and in Slovakia even among
low-income households there is a high rate of
home ownership, whereas in Austria, France and
Germany home ownership is much more dependent on income (Panel A of Figure 4 and Figure A4 of
the Annex). Since housing wealth constitutes a
significant fraction of total net wealth, especially
for low-wealth households, unsurprisingly lowwealth households tend not to be home owners
(Panel B of Figure 4).

Another important aspect is housing. By reducing

long-term yields, ULMP can also have an impact on
long-term mortgage interest rates. For example, for
the United States, Bivens (2015) reports that a
100 basis points decline in mortgage interest rates
boosts home prices by 7 percent. Similarly, the
portfolio rebalancing channel could increase the
demand for housing further4. As the cost of mortgages goes down, it should put some upward pressure on housing prices. As can be seen in Figure 6
of Claeys and Darvas (2015), house prices have
been falling throughout the euro area since the
bursting of the bubble in 2007. There was a minor
increase in real house prices in Germany from
2010, but the level of real house prices in 2014
was still below the 2000 level. Earlier ECB monetary policy measures might have prevented a
deeper fall in prices, while the more recent asset
purchases might lead to house price increases.

ULMP, by raising housing prices, will benefit all

homeowners. For households with lower incomes,
however, real estate assets represent a much
larger share of their total assets than for richer
households. Therefore it is possible that ULMP will
reduce inequality through the housing channel we
have just described.
As with other assets, rising housing prices will
benefit current homeowners at the expense of
future buyers, who will tend to be young people.
As Figure 5 on the next page and Figure A5 of the
Annex shows, home ownership tends to be
dependent on the age of the head of the household, though in most southern euro-area countries
and in Slovakia home ownership is relatively high
even among the 16-34 age group.
These findings are confirmed by a recent working
paper by Adam and Tzamourani (2015). Using
data from the HFCS, they show that the median
household strongly benefits from housing price
increases, while capital gains from bond-price and

As we can see from the HFCS, home ownership is

prevalent even among intermediate income and
wealth groups (see Figure 4 and Figure A4 of the
Annex). There are however some differences

Figure 4: Home ownership by income and wealth percentiles in the euro area and its four largest
countries (%)
A: By income distribution

B: By wealth distribution

100

100

90

90

80

80

70

70

60

60

50

50

40

40

30

30

20

20

10

10
0

0
Euro area

France

Germany

Bottom 20%

Italy

20-40%

Spain

40-60%

Euro area

60-80%

France

80-90%

Germany

Italy

Spain

90-100%

Source: ECB HFCS. Note: the bars indicate the % of households in each income/wealth group that own their main residence.

4. In principle, supply can

respond to increase in
demand and leave housing
prices unchanged. Yet
experience suggests that
sizeable expansions of the
construction sector used to
coincide with house price
increases, suggesting that
increased demand for
housing used to have an
impact on housing prices.

BRU EGE L

POLICY

CONTRIBUTION

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

06
Figure 5: Home ownership by age of the head of
the household in the euro area and its four
largest countries (% households)

equity-price increases are shared among relatively few households.

2.2 The impact through interest rates

100
90
80
70
60
50
40
30
20
10
0
Euro area

16-34

France

Germany

35-44

45-54

Italy

Spain

55-64

65-74

75+

Source: ECB HFCS.

While ULMP has a positive impact on asset prices,

which benefits those who are holding them when
the measures are implemented, it also reduces
the expected returns on these assets for those
who are buying the assets at a high price. These
two effects might affect different age groups
within income and wealth groups differently. For
example, the young generation of the rich, who are
acquiring financial assets, might suffer relatively
more from the reduced income than older rich generations, who will largely benefit from the stock
effect.

Figure 6: Debt and debt service in the euro area

A) Share of households with debt (percent)
A1: By income distribution

50
45
40
35
30
25
20
15
10
5
0

A2: By wealth distribution

45
40
35
30
25
20
15
10
5
0

Mortgage debt

Other debt
Bottom 20%

20-40%

Mortgage debt
40-60%

60-80%

80-90%

Other debt
90-100%

B) Median value of debt among those who have debt ( thousands)

B1: By income distribution

120

B2: By wealth distribution

160
140

100

120
80

100
80

60

60

40

40
20

20
0

Mortgage debt

Other debt
Bottom 20%

20-40%

Mortgage debt
40-60%

60-80%

80-90%

Other debt
90-100%

C) Debt servicing burden among those who have debt (as % of income) in the euro area
C1: By income distribution

25

20
18
16
14
12
10
8
6
4
2
0

20
15
10
5
0

Bottom 20% 20%-40%

Source: ECB HFCS.

40%-60%

60%-80%

80%-90%

Top10%

C2: By wealth distribution

Bottom 20% 20%-40%

40%-60%

60%-80%

80%-90%

Top10%

BRU EGE L

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

POLICY

CONTRIBUTION

07
More generally, lower interest rates are likely to
reduce the financial revenues of savers, who tend
to be rich, and benefit debtors, which tend to be
households from the middle-class to the rich.
Panels A and B of Figure 6 show that very few lowincome and poor households have mortgage debt
and while many have other debts (such as overdrafts or credit cards), the median value of nonmortgage debt is very small relative to mortgage
debt. However, the debt service to income ratio is
the highest for low-income households (Panel C
of Figure 6), implying that they would benefit the
most from a reduced mortgage interest rate. Country-specific data reported in Figure A6 of the Annex
underlines that this finding applies generally
across the euro area.
Another important element, emphasised by
Beraja et al (2015), is that ULMP can widen
inequality not only between income quintiles but
also between regions (or between countries in the
case of the euro area). Beraja et al (2015) show
that in the US, while in the aggregate asset purchases resulted in more mortgage originations,
refinancing, cash-outs, and consequently consumer spending, these effects were much
stronger in regions with lower mortgage loan-tovalue ratios (LTVs). Regions with numerous homeowners whose house market price is below the
value of their mortgage (ie in negative home
equity), however, do not benefit as much from
these stimulative effects because it is more difficult and expensive for them to refinance their
mortgages. This effect, which could lead to the
Figure 7: Median loan to value ratios of main
residences in euro-area countries
60
50

amplification of regional inequality, could be

important in the euro area, where there is significant disparity in median LTVs in different countries
(Figure 7) and where there are differences in the
evolution of house prices too. In countries in which
LTVs are higher and house prices have fallen more
(Figure 8 and Figure A8 of the Annex), the share of
debtors facing difficulties in refinancing their
loans should be higher and they should benefit
less from the monetary policy accommodation.
The size of the effect of ULMP could therefore
depend on whether homeowners have a fixed or
variable rate mortgage, and how easy or costly it is
to remortgage. In some countries most mortgages
are fixed rate, which means that homeowners will
have to refinance in order to benefit from lower
interest rates, and refinancing can in some cases
be very costly. According to the Bank of Spain
(2009), the Research Institute for Housing America (2010), and the European Mortgage Federation (2012), Austria, Belgium, Denmark, France,
Germany and the Netherlands mostly have fixedrate mortgages, while Greece, Hungary, Ireland,
Portugal, Spain, Sweden and the United Kingdom
mostly have variable-rate mortgages. In Italy there
is a mix of both. In the countries with dominantly
variable-rate mortgages, households with a mortgage would benefit automatically from lower rates,
while in countries with fixed-rate mortgages, only
households that are able to refinance would benefit from the lower interest rates.
Again, it is important to distinguish between shortterm and medium-term effects. In the short term,
low rates and ULMP can have negative effects on
net savers, but not in the medium term when interest rates normalise.
3 THE IMPACT OF ULTRA-LOOSE MONETARY
POLICIES ON INEQUALITY THROUGH THE
MACROECONOMY

40
30
20

Slovenia

Malta

Austria

Luxembourg

Italy

Belgium

Spain

Greece

France

Cyprus

Slovakia

Portugal

Source: ECB HFCS.

Euro area

Finland

Germany

Netherlands

10

Those claiming that ULMP is worsening inequality

mainly focus on the fact that unconventional monetary policy works primarily by raising asset
prices, as documented in the previous section,
resulting in distributional effects in favour of those
holding assets. However, one of the most important effects that unconventional monetary policy
might have on inequality is its potential impact on

BRU EGE L

POLICY

CONTRIBUTION

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

08
the general macroeconomic environment boosting GDP, raising inflation back to target and supporting employment. Economic gains from these
positive developments could again be unequally
distributed, but possibly in a different direction to
the benefits accruing from asset-price increases.
Households and firms make spending and saving
decisions based on their expectations of future
income. ULMP can affect the decisions of households and firms in several ways: through a virtuous cycle of higher revenues and incomes, higher
asset prices and wealth effects, higher collateral
values, and through higher expected inflation.
Higher expected inflation will induce households
and firms to bring consumption spending forward
to protect their purchasing power. Higher household and firm spending, in a more benign borrowing environment, should boost inflation and GDP
and reduce unemployment. Higher asset prices
will increase household and firm wealth, increasing spending, and will increase the value of assets
that can be used as potential collateral for credit.
By increasing nominal spending, ULMP can also
have an indirect effect on equity prices, as companies face more demand and increase their profits, which in turn drives the more favourable
macroeconomic environment.
3.1 Academic research on the impact of ULMP
on the macroeconomy
Research on the macroeconomic impact of the
monetary policy measures implemented since
the beginning of the crisis has generally produced
consistent results: most papers find a significant

positive impact on inflation and GDP.

In terms of empirical evidence from past assetpurchase programmes in other major advanced
economies, Chung et al (2011) found that the
large-scale asset purchase programme by the US
Fed had significant benefits for the macroeconomic situation in the US. Using an internal Federal Reserve Bank model, the authors found that
asset purchases reduced long-term interest rates
on treasuries by up to 50 basis points, while the
unemployment rate was about 1.5 percentage
points lower, GDP about 3 percentage points
higher and core inflation about 1 percentage point
higher than the counterfactual scenario without
Fed purchases.
Wu and Xia (2014) develop a so-called shadow
rate an interest rate that captures all the effects
of the Feds unconventional monetary policy, even
if the Federal Funds Rate (FFR) is constrained by
the zero lower bound. They find that the shadow
rate is a good representation of monetary policy
in the pre-crisis period, because the shadow rate
tracks the actual FFR very closely. The shadow
rate turns strongly negative as a result of policies
to ease credit and expand the Feds balance sheet.
Using a Factor Augmented VAR model, they construct counterfactuals in which the shadow rate is
set to the zero lower bound, thus negating the
effects of unconventional monetary policy. They
find that industrial production is more than 5 percent higher and unemployment 1 percent lower
than in a scenario with no unconventional policies.
Their model also predicts that forward guidance
the policy of communicating the path of future
interest rates was also successful. In their

Figure 8: Annual house price growth in selected countries (%)

20

25

A: Euro area

B: Global comparison

20

15

15
10
10
5

5
0

-5
-5
-10

Italy

Germany

Source: Thomson Reuters Datastream.

Spain

Japan

United States

United Kingdom

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

France

2000

-10

-15

BRU EGE L

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

POLICY

CONTRIBUTION

09
model, a 1-year extension of the expected zero
lower bound period in the future reduces the
unemployment rate by 0.25 percentage points.
Kapetanios et al (2012) at the Bank of England
found that GDP was boosted by about 2 percent,
and at its peak, CPI inflation was about 4 percentage points higher than would otherwise have been
the case, averting a situation of outright deflation.
The authors use three different vector autoregressive models, which allow for time-varying parameters. They construct their estimates of the effects
of QE by creating carefully designed counterfactual scenarios in which there is no effect of QE on
government yields. Hence, in their model, the primary effect of QE is through lower interest rates,
and the second-order effects on output and inflation happen entirely through the effect on interest
rates.
Similarly, Baumeister and Benati (2010) found
that the compression in the long-term yield spread
has had a strong positive effect on output and
inflation in both the UK and US. They use Bayesian
time-varying parameter structural VAR, and investigate the effects in reducing yield spreads
(assuming a fixed short term rate to simulate the
zero lower bound). In the US they find that the
yield-compression seen as a result of asset purchases increased growth by about 2 percent and
increased inflation by about 1 percent. Results for
Japan and the UK are quantitatively similar. It
should be noted that the Fed engaged in substantial rounds of further asset purchases after this
point.
Focusing on the euro area, Lenza et al (2010) provide evidence, again using counterfactuals via a
VAR model, that the ECBs early measures to ease
credit in the euro area helped reduce spreads in
money markets, which in turn had positive effects
on output and inflation. Darracq-Paries and De
Santis (2013) specifically focused on the ECBs
LTROs of December 2011 and February 2012.
They found, using Bank Lending Survey (BLS)
data, that the LTROs substantially boosted euroarea lending, and through their VAR model, that
GDP was 0.6 percentage points above its counterfactual level by 2013, inflation about 0.2 percentage points higher and outstanding loans 2
percentage points higher.

3.2 Implications for inequality

Recessions could potentially increase inequality
through two channels: (i) the composition of
income, and (ii) the differing impact on employment according to skill levels.
Since the poor rely much more heavily on wages
for their income, any change in employment
levels will affect them much more than the rich,
who accrue income through more diverse channels, such as capital gains. If ULMP is successful
in stimulating the economy, this will have net benefits for the poor and low-skilled relative to the
rich, and will result in a reduction in inequality.
Furthermore, evidence from the literature shows
that the poor and low-skilled are the most likely to
lose their jobs in recessions. While Figure 9 on the
next page and Figure A9 in the Annex indicate a
structural change in the composition of employment, whereby the low-skilled employment
tended to decline and high-skilled employment
increased already before the crisis in almost every
country, during the crisis low-skilled workers
(which are at the bottom of the income distribution) suffered much more relative to higher-skilled
workers. It is interesting to highlight that employment of high-skilled workers (those with tertiary
education) continued to increase throughout the
crisis, even in countries suffering from large
increases in unemployment like Cyprus, Italy, Ireland, Lithuania, Portugal and Spain, while their
employment remained broadly stable in Estonia,
Latvia and Greece.
Bitler and Hoynes (2015), using data from the
United States, show that those on lower incomes
experience much greater income cyclicality than
higher earners. Furthermore, this differential effect
of recessions on low earners was steeper in the
great recession compared to the previous 1980s
recession. Therefore, any policy that helps to prevent or alleviate recessions will help to keep those
at the bottom end of the income and wealth distribution in jobs and will therefore avoid a further
widening of inequality.
The academic literature confirms that monetary
policy might in fact reduce inequality. For example, Coibion et al (2012), taking a historical per-

BRU EGE L

POLICY

CONTRIBUTION

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

10
spective and not considering unconventional policies specifically, document that contractionary
monetary policy typically increases inequality,
while accommodative monetary policy reduces
inequality. Bivens (2015) argues that the view
that ULMP benefits only the rich through higher
asset prices is not correct. Although stock and
house prices rose as a result of the Feds policy
measures, helping people who own their home or
hold stocks, to the extent that the policies helped
maintain employment and output, the Fed's measures reduced inequality. Bivens concludes that in
the absence of the Fed's ULMP, wage growth would
have been lower and more unequal. For the UK, the
Bank of England (2012) makes a very similar case
to Bivens (2015) in a review of the effects that its
policy had on the distribution of wealth and
income, arguing that ULMP in the UK benefited various segments of society through its impact on
general economic conditions.
Yet the literature is not unanimous. For example,
Saiki and Frost (2014) conclude, using impulse
response functions from a VAR model with the Gini
coefficient included, that ULMP increased inequality in Japan. Meanwhile Philippon and Reshaf

(2009) have shown that remuneration in the

financial sector is extreme, even when one takes
into account technological progress and the skill
and education levels of employees. Therefore, to
the extent that ULMP benefited the financial
sector, it also benefited the wealthy owners and
employees of the financial sector (Acemoglu and
Johnson, 2012).
4 CONCLUSIONS
The widening of income and wealth inequality
observed in many advanced countries in recent
decades is a long-term trend and primarily the
result of deep structural changes. Nevertheless,
there are some concerns that current ultra-loose
monetary policies (ULMP) could amplify that
trend, at least in the short- and medium-term.
Since 2008, most major central banks have implemented various monetary easing measures. Given
the macroeconomic situation in advanced
economies and in the euro area in particular, these
measures were justified and in fact the European
Central Bank should have acted earlier. However,
some of these measures and the unusual length

Figure 9: Employment (in millions) by educational attainment in the four largest euro-area
countries, 1992-2014
30

14

Germany

France

12

25

10

20

8
15
6
10

4
2

14

1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014

1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014

10

Italy

12

Spain

10
8

4
2

1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014

0
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014

Less than primary, primary and lower secondary education

Upper secondary and post-secondary non-tertiary education
Tertiary education

Source: Eurostat Employment by sex, occupation and educational attainment level (1 000) [lfsa_egised] dataset.

BRU EGE L

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

POLICY

CONTRIBUTION

11
of the monetary accommodation could have side
effects on income and wealth distribution:

because wages are the primary source of revenues for poorer and lower-income people.

The impacts of ULMP through increases in

financial asset prices tend to increase inequality between the wealthy and poor, between the
young and old, and also between regions when
they have different financial structures.
Increases in the value of assets such as equities and government and corporate bonds will
tend to favour the rich who hold them in higher
proportions. Since older people tend to have
higher savings and may sell them in the future
in order to maintain their consumption, while
younger households are usually the ones that
will buy these assets in the future in order to
save for retirement, ULMP may have distributional consequences across generations. ULMP
can benefit households differently depending
on the structure of their financial assets, since
certain households could make better use of
the opportunity offered by low-interest rate borrowing than others.

The primary mandate of the European Central

Bank is to maintain price stability, and considerations of inequality are not within its purview,
unless inequality should prevent the transmission
of monetary policy in some way. The ECB should
focus on its price stability mandate and thereby
support the fragile recovery now taking place in
the euro area. This is the best way for monetary
policy to contribute to the avoidance of an
increase in inequality in times of recession. Yet we
recommend the ECB to monitor the side effects of
its monetary policy measures, including the
potential distributional effects. The ECB has
detailed internal datasets which should allow a
comprehensive assessment.

The impacts of ULMP through an increase in

housing prices and a fall in interest rates tend
to decrease inequality. Housing is the main
asset of the middle class and therefore housing price increases will tend to compress the
wealth distribution. A fall in mortgage interest
rates tends to benefit low-income people, who
spend a larger share of their income on servicing their debts.
The impacts of ULMP through stimulating the
economy tend to reduce inequality. A large literature concluded that ULMP boosts inflation,
output and employment. In the absence of
ULMP, unemployment would be higher, which
would lead to higher income inequality,
because the poor and low-skilled are the most
likely to lose their jobs in recessions and

The main policy question is how to tackle inequality in general, and whether governments should
design special measures in a deep recession or in
a situation in which central bank actions widen
inequality. For example, in the United States, policies such as the Housing Affordable Refinance
Programme (HARP), which helped homeowners
with negative home equity to refinance their mortgages, might have helped dampen the rising
inequality that resulted from the housing slump.
Fiscal and social policies are the right tools to fight
inequality. As documented by Darvas and Wolff
(2014), there are huge differences in the efficiency of social redistribution systems in EU countries. For their levels of social expenditure and
personal income taxes, several southern European countries and Belgium achieve a much
smaller reduction in inequality than other EU
countries. Revising national tax/benefit systems
for improved efficiency, intergenerational equity
and fair burden sharing between the wealthy and
poor is the right way to fight inequality.

BRU EGE L

POLICY

CONTRIBUTION

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

12
REFERENCES
Acemoglu, Daron and Simon Johnson (2012) Who Captured the Fed?, New York Times, 29 March
Adam, Klaus and Panagiota Tzamourani (2015) Distributional Consequences of Asset Price
Inflation in the Euro Area, unpublished manuscript
Bank of England (2012) The Distributional Effects of Asset Purchases, Quarterly Bulletin 2012 Q3
Bank of Spain (2009) Fixed and variable rate mortgages, business cycles and monetary policy,
Working Paper N. 0903
Baumeister, C. and L. Benati (2010) Unconventional monetary policy and the great recessionEstimating the impact of a compression in the yield spread at the zero lower bound, Working Paper
Series No 1258, European Central Bank
Beraja, M., A. Fuster, E. Hurst and J. Vavra (2015) Regional heterogeneity and monetary policy,
Brookings Institution Working Paper
Bitler, M. and H. Hoynes (2015) Heterogeneity in the Impact of Economic Cycles and the Great
Recession: Effects Within and Across the Income Distribution, American Economic Review, Papers
and Proceedings 2015, 105(5): 154-160
Bivens, Josh (2015) Gauging the impact of the Fed on inequality during the Great Recession,
Brookings Institution Working Paper
Claeys, Grgory (2014) The (not so) Unconventional Monetary Policy of the European Central
Bank since 2008, paper for the Monetary Dialogue discussions in the Economic and Monetary
Affairs Committee (ECON) of the European Parliament, available at
http://www.bruegel.org/publications/publication-detail/publication/837-the-not-so-unconventionalmonetary-policy-of-the-european-central-bank-since-2008/
Claeys, Grgory and Zsolt Darvas (2015) The financial stability risks of ultra-loose monetary
policy, Policy Contribution 2015/03, Bruegel
Claeys, Grgory, Zsolt Darvas, Silvia Merler and Guntram B. Wolff (2014) Addressing weak inflation:
The European Central Bank's shopping list, Policy Contribution 2014/05, Bruegel
Cohen, William (2014) How Quantitative Easing Contributed to the Nations Inequality Problem,
New York Times, 22 October
Coibion, O., Y. Gorodnichenko, L. Kueng and J. Silvia (2012) Innocent bystanders? Monetary policy
and inequality in the US, Working Paper 18170, National Bureau of Economic Research
Chung, Hess, Jean-Philippe Laforte, David Reifschneider and John C. Williams (2011) Estimating
the Macroeconomic Effects of the Feds Asset Purchases, FRBSF Economic Letter
Darracq Paries, M. and R. A. De Santis (2013) A non-standard monetary policy shock: the ECBs 3year LTROs and the shift in credit supply, Working Paper Series No 1508, European Central Bank
Darvas, Zsolt and Guntram B. Wolff (2014) Europe's social problem and its implications for
economic growth, Policy Brief 2014/03, Bruegel
Dobbs, Richard, Tim Koller and Susan Lund (2014) What effect has quantitative easing had on
your share price?, McKinsey on Finance Number 49, McKinsey Global Institute.
ECB (2013) Household Finance and Consumption Survey,
https://www.ecb.europa.eu/home/html/researcher_hfcn.en.html
European Mortgage Federation (2012) Study on Mortgage Interest Rates in the EU
Gagnon, Joseph, Matthew Raskin, Julie Remache and Brian Sack (2011) The Financial Market
Effects of the Federal Reserves Large-Scale Asset Purchases, International Journal of Central
Banking 7, no. 1: 344
Joyce, M., A. Lasaosa, I. Stevens and M. Tong (2011) The financial market impact of quantitative
easing in the United Kingdom, International Journal of Central Banking, 7(3), 113-161
Kapetanios, G., H. Mumtaz, I. Stevens and K. Theodoridis (2012) Assessing the Economywide
Effects of Quantitative Easing, The Economic Journal, 122(564), F316-F347
Lenza, M., H. Pill, H. and L. Reichlin (2010) Monetary policy in exceptional times, Economic Policy,
25(62), 295-339

BRU EGE L

THE EFFECTS OF ULTRA-LOOSE MONETARY POLICIES ON INEQUALITY

POLICY

CONTRIBUTION

13
Meier, Andr (2009) Panacea, Curse, or Nonevent? Unconventional Monetary Policy in the United
Kingdom, Working Paper 09/163, International Monetary Fund
Okun, A. M. (1975) Equality and efficiency, the big tradeoff, Brookings Institution Press
Ostry, M. J. D., M. A. Berg and M. C. G. Tsangarides (2014) Redistribution, inequality, and growth, Staff
Discussion Note 114/02, International Monetary Fund
Philippon, T., and Reshaf, A. (2009) Wages and Capital in the US Financial Industry: 1909-2006,
Quarterly Journal of Economics, November
Piketty, Thomas (2014) Capital in the 21st Century, Harvard University Press
Rajan, R. (2010) Fault Lines: How Hidden Fractures Still Threaten the World Economy, Princeton
University Press
Research Institute For Housing America (2010) International comparison of mortgage product
offerings, Special Report, September
Saiki, A. and J. Frost (2014) Does unconventional monetary policy affect inequality? Evidence from
Japan, Applied Economics, 46(36), 4445-4454
Skott, P. (2013) Increasing inequality and financial instability, Review of Radical Political Economics,
vol. 45(4), 478-488
Stiglitz, J. (2015) New theoretical perspectives on the distribution of income and wealth among
individuals: Part IV: Land and Credit, Working Paper 21192, National Bureau of Economic Research
Van Treeck, T. (2014) Did inequality cause the US financial crisis? Journal of Economic Surveys,
28(3), 421-448
Vandemoortele, M. (2009) Within-Country Inequality, Global Imbalances and Financial Instability,
ODI Research Report, commissioned by the Dutch Ministry of Foreign Affairs
Wu, J. C. and F. D. Xia (2014) Measuring the macroeconomic impact of monetary policy at the zero
lower bound, Working Paper 20117, National Bureau of Economic Research

ANNEX: COUNTRY-SPECIFIC DATA

This annex shows country-specific data for all euro-area countries (whenever available) for the figures
reported in the main text. Figure numbering in this annex corresponds to numbering in the main text,
eg Figure A1 in the annex reports country-specific data relating to Figure 1 in the main text.
Figure A1: Measures of inequality
A) Gini coefficient of income inequality (after taxes and transfers), 1960-2013
45

45
Austria

35

Belgium
Cyprus

40
35

Estonia

25

25

20

20

15

15

45

45
40
35

1960
1969
1974
1979
1984
1989
1994
1999
2004
2009

30

1960
1969
1974
1979
1984
1989
1994
1999
2004
2009

30

Latvia
Lithuania
Luxembourg
Malta

40
35

30

30

25

25

20

20

15

15
1960
1969
1974
1979
1984
1989
1994
1999
2004
2009

Netherlands
Portugal
Slovenia
Spain

1960
1969
1974
1979
1984
1989
1994
1999
2004
2009

40

France
Germany
Greece

Source: Standardised World Income Inequality Database

Note: The Gini coefficient ranges from 0 to 100, with 100 indicating complete inequality. It is a function of the surface between the Lorenz curve (which is
the cumulative distribution function of the probability distribution of income) and the line of equality.

14

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex

B) Share of income going to the 1% (before taxes and transfers), 1946-2012

Source: Top World Incomes Database (http://topincomes.parisschoolofeconomics.eu/)

Note: series for Finland contains break that merges two different data sources

Figure A2: Net wealth by wealth percentiles

Source: ECB HFCS (2013)

Note: Net Wealth is the difference between total household assets and total household liabilities. Total assets include real
and financial assets
15

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex

Figure A3: Home ownership across income percentiles (percent)

Source: ECB HFCS

Note: the bars indicate the percent of households in the income group owning their main residence.

Figure A4: Home ownership across wealth percentiles (percent)

Source: ECB HFCS

Note: the bars indicate the percent of households in the wealth group owning their main residence.

16

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex

Figure A5: Home ownership by age of the head of the household

Source: ECB HFCS

Note: the bars indicate the percent of households in the age group owning their main residence.

17

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex

Figure A6: Debt and debt service

A) Share of households with mortgage debt, by income (% of households)
80
70
60
50
40
30
20
10
0

Bottom 20%

20-40%

40-60%

60-80%

80-90%

90-100%

Source: ECB HFCS

Note: Data on Finland not available
B) Share of households with other debt, by income (% of households)
70
60
50
40
30
20
10
0

Bottom 20%

20-40%

40-60%

60-80%

80-90%

90-100%

Source: ECB HFCS

Note: other debt denotes all debt other than mortgage debt. Data on Finland not available

18

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex

C) Median value of mortgage debt among those who have mortgage debt, by income ( thousands)
Euro
Area
Bottom
20%
20-40%
40-60%
60-80%
80-90%
90-100%

Bottom
20%
20-40%
40-60%
60-80%
80-90%
90-100%

Austria

Belgium

Cyprus

France

Germany

Greece

43
47
55
67
86
100

21
33
39
36
67

45
67
63
76
86
69

84
69
81
110
89
131

28
39
47
57
56
91

44
29
78
69
92
116

32
36
33
50
49
46

Italy

Lux'bourg

Malta

N'lands

Portugal

Slovakia

Slovenia

99
79
131
131
156
160

38
55
41
43
65
65

38
50
50
60
70
75

133
114
114
149
240

55

Spain
47
50
57
58
89
83

23
27
25
18

Source: ECB HFCS

Note: Data on Finland not available. Empty cells indicate missing data.

D) Median value of other debt among those who have other debt, by income ( thousands)

Bottom
20%
20-40%
40-60%
60-80%
80-90%
90-100%

Bottom
20%
20-40%
40-60%
60-80%
80-90%
90-100%

19

Euro
Area

Austria

Belgium

Cyprus

France

Germany

Greece

3
3
5
6
6
8

3
1
2
3
3
6

2
3
5
7
8
8

6
6
8
13
11
19

2
4
5
7
8
12

2
2
4
4
4
5

4
4
5
5
4
4

Italy

Lux'bourg

Malta

N'lands

Portugal

Slovakia

Slovenia

Spain

5
4
6
7
5
8

6
9
12
11
19
16

3
4
5
8
6

10
10
12
16
32
18

3
2
2
4
4
6

0
1
2
1
1
3

3
5

4
5
6
8
10
13

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex

E) Debt service as a share of household income (%)

30%
25%
20%
15%
10%
5%
0%

Bottom 20%

20%-40%

40%-60%

60%-80%

80%-90%

Top 10%

Source: ECB HFCS

Note: other debt denotes all debt other than mortgage debt. Data on Finland is not available.

Figure A8: House Price Growth (annual percent change)

30
25

25

20
15

15

10
5

-15

Belgium

Finland

France

Germany

Greece

1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014

-25

Austria

-10
-15
-20
-25

Ireland
Italy
Netherlands
Portugal
Slovenia
Slovakia
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014

-5

-5

Source: Thomson Reuters EIKON and Datastream

Note: All countries with available data are shown

20

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex

Figure A9: Employment by educational attainment (thousand people), 1992-2014

21

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex

22

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex

Source: Eurostat Employment by sex, occupation and educational attainment level (1 000) [lfsa_egised] dataset.

23

The effects of ultra-loose monetary policy on inequality | Bruegel Policy Contribution 2015/09 Annex