Executive Summary

In 2009, two epidemiologists, Wilkinson and Pickett, published a book entitled

“The Spirit Level, Why More Equal
Societies Almost Always Do Better” in which they claim that inequality and its
acute perception by the average EU
citizen is a toxic element of today’s European societies and one that seems to be
associated with decreased levels of
trust, civic engagement and participation, as well as to a host of other social
challenges from poor health to crime, to
underage pregnancies. Despite Wilkinson and Pickett’s intuitively convincing
story of the link between higher
income inequality and worse social outcomes, the empirical tests are based
on simple bivariate correlations,
implying that the authors fail to control for all the other numerous factors, which
might have had an impact on both
the social outcomes and income inequality. In doing so, the empirical associations
reported in their book are likely
to lead to misleading causal inferences. Nonetheless, Wilkinson and Pickett’s book
attracted a lot of attention and
called for a more careful analysis of the consequences of rising income inequality

The aim of this report is, hence, to look into sound empirical studies - based on
multivariate analysis - which
examine the effect of income inequality on important social outcomes related to (i)
well-being, (ii) criminality, (iii)
health, (iv) social capital, (v) education, (vi) political participation and (vii) female
labor market participation. The
upshot of this literature review is that higher criminality, reduced political agency
and, to some extent, lower social
capital formation and well-being appear to be tangible illustrations of the
wastage produced by rising income
inequality. In addition, there are a number of self-reinforcing loops linked to
inequality. A clear illustration of this is
the role of inequality in reducing the voting participation of the low income
groups and the concomitant
consequences in terms of redistributive policies and therefore on income
disparities.

1. Introduction
In the past years, there has been a growing concern about the rising income inequalities with
particularly stark rises in inequality in the US and also substantial widening of the income gap in Europe.
The relevance of the widening of income has not only caught the political and scholarly attention but is
heavily discussed nowadays on the streets, with most prominent manifestation of such protests being
the ‘Occupy Wall Street’- movement. This movement and its widely-cited slogan “We are the 99%” (see
for instance the reporting of the movement by the New York Times, 2011, and also the web blog “We
are the 99 percent”, 2012) refers specifically to a growing unequal distribution of wealth.
The development of income inequality in the EU member states has been the subject of a recent
publication by the
OECD (2011). Surveying the development of income inequality over the past 3 decades reveals an
interesting
picture. In particular, there seems to be a general trend of widening the income gap starting in the
1980s. While in
the 1980s the Gini coefficient was 0.29 it markedly rose to 0.32 in the late 2000s (ibid., p. 22).
Particularly striking
is the increase in income inequality of former ‘equal societies’, such as the Nordic countries and
Germany. In
general, there seems to be a convergence trend towards a generally higher level of income inequality.

The causes of this rising income inequality in the past decades has also attracted much political and
scholarly
attention. The OECD’s (2011) report provides a wealth of explanatory mechanisms, ranging from
rising wage
inequality to different taxation policies and household structures
3. Literature review
3.1 Income inequality and happiness
3.1.1 Rationale
The discussion on whether income inequality affects an individual’s happiness dates back to
theoretical
considerations on relative deprivation and relative utility and refers to the idea that people’s utility
depends not only
on their own income but also on their relative position in the society (van de Stadt, Kapteyn and van de
Geer, 1985).
In addition, some scholars suggest that individuals can have a ‘taste for equality’. In particular,
Thurow (1971,
p.327) proposes that “the individual is simply exercising an aesthetic taste for equality or
inequality similar in
nature to a taste for paintings”.

An intuitive and comprehensive explanation of the impact of income inequality on individuals’

well-being is
provided by Hirschman and Rothschild (1973). These authors use the analogy of a traffic jam on a
two-lane
motorway to explain the effect of income inequality on happiness and call this the ‘tunnel effect’ (ibid,
p.545):
3.1.2 Measures of happiness and income inequality
The empirical studies measure happiness by relying on questions directly asking respondents on their
perceived
happiness or their life satisfaction. For example, the studies use the responses to the following
questions:
- “Taken all together, how would you say are things these days – would you say you are very happy,
pretty happy,
or not too happy?” (from the United States General Social Survey)
- “On the whole, are you very satisfied, fairly satisfied, not very satisfied or not at all satisfied with the
life you
lead?” (from the Euro-barometer Survey Series)
- “How satisfied are you with your life all things considered?” (from the German Socio-economics Panel
Study)

There are two different ways on how income inequality is measured in the studies focusing on
happiness. First,
most of the studies rely on the Gini coefficient or on alternative indices such as the Theil, Atkinson or
Stark index
(cf. Alesina et al., 2004, and Schwarze and Harpfer, 2007). Second, studies frequently employ the
‘reference group
income’, which is designed to capture the income of peers (defined by e.g. occupation, education level,
sex, age and
region), following the idea that people strive to ‘keep[..] up with the Joneses’ (cf. Hopkins, 2008, p.4).
For studies

employing the “reference group income” variable, see Luttmer (2005), Clark (2006), Grosfeld and Senik,
(2008).
3.1.3 Empirical evidence

There exists quite substantial empirical evidence on the impact of income inequality on
happiness or life
satisfaction, mainly covering the U.S., Europe and transition countries.

The empirical evidence confirms the relationship between income inequality, happiness and social
mobility. The
effect of income inequality on happiness critically depends on whether individuals perceive the society
to open to
upward mobility and on whether it is likely that they will eventually be able to reach higher income
levels. Evidence
can thus be divided into the low-mobile countries (typically European), where inequality has a negative
effect on
satisfaction, and the highly mobile society such as US and transition countries, where there seems to be
a greater
variability in the outcomes of income inequality
3.2 Income inequality and criminality

3.2.1 Rationale
The determinants of criminality, and in particular the role played by income inequality, has attracted the
attention of
scientists from various disciplines.

Economic theories for criminal activities date back to Becker (1968) and stress that a criminal act is the
result of a
rational decision based on a cost-benefit analysis. Individuals decide to participate or not in criminal
activities by
comparing the returns of criminal and legal activities. The net return of a criminal act is the difference
between the
loot and the associated costs such as the opportunity cost and the severity of punishment if the
individual is caught
while committing the crime. Income inequality should increase the potential gain derived from a
criminal act for
individuals situated at the bottom end of the income distribution because the gap between their
income and the
country mean income is larger, relatively to a situation in which the resources would be more evenly
distributed.

Sociological theories sustain that criminal activities result from a feeling of frustration of the less well-off
people
when they compare their situation with respect to the one of wealthier individuals. The higher is income
inequality,
the greater is the sentiment of unfairness of disadvantaged individuals. Economic deprivation and the
associated
feeling of resentment might spur criminal behaviors (Morgan, 2000, citing, in particular, Merton’s work,
1938).

3.2.2 Measures of criminality and income inequality

Typically, crime statistics used in empirical studies refers to homicide, robbery and property crime rates.
However,
most of these official data sources suffer from under-reporting, with some categories of crime more
particularly
afflicted by errors than others (MacDonald, 2002). Homicide and robbery rates tend to be more reliable
figures
since the violence associated with such criminal acts tend to increase the proclivity for the victim to
officially
declare the crime to the police. Cross-country comparisons are also often problematic because of legal
differences
across countries in the way crimes are defined. In addition, the quality of the data strongly depends on
the country-
specific police and justice systems (Fajnzylber et al, 2002a).

The Gini coefficient is the measure of economic inequality employed in all empirical papers reviewed
below.
However, some studies use additional inequality indices (ratio of income of the richest to the poorest
quintile,
proportion of the population with an income below a certain value) to check the robustness of the
findings (see
Nilsson, 2004, Brush, 2007, Fajnzylber et al, 2002a)
3.2.3 Empirical evidence

Empirical studies
Testing the causal effect of inequality on crime rates is not straightforward because several
socioeconomic factors
are likely to be simultaneously correlated with income inequality and criminal rates. If these
factors are not
controlled for in the multivariate setting, we cannot conclude that the estimated association
between economic
inequality and criminality is causal. The fact that crime rates are measured with errors also complicates
the work of
researchers in particularly if these measurement errors are not random but are, instead,
correlated with other
variables related themselves to income inequality. Dealing with the problems cited above, i.e.
with the
“endogeneity” of the inequality index, is critical to be able to say something about the causal
effect of income
inequality on crime rates.

Empirical papers examining the effect of income inequality on crime rates are based on (i) cross-country
data or (ii)
single country data. Country-specific studies can be of two types: the first one only relies on
cross-sectional
information (i.e. cross-region) whilst the second type of study combines cross-sectional data with
time series
information (i.e. cross-region observed over two or more periods of time).

Cross-country studies
Using data from the United Nations World Crime Surveys, Fajnzylber et al. (2002a) examine the
determinants of
national criminal rates across a sample of around 40 countries and on a 40 years period. The authors
show that
income inequality, as measured by the Gini coefficient, exerts a positive and significant effect on
homicide and
robbery rates, and these results are robust to the inclusion of a large set of control variables and to
alternative
econometric methods. In a companion paper, Fajnzylber et al. (2002b) find that the effect of income
inequality on
violent crime is robust to alternative measures of income inequality such as the ratio of income of the
richest to the
poorest quintile of the population, which is an index of income polarization.

Single-country studies
Country-specific studies have mainly been based on US data. Back in 1973, Ehrlich, in his analysis of
state crime
determinants, finds a positive association between property crime and inequality as measured by the
percentage of
the population with an income below one half of the median income. Recent studies on US data show
contrasting
results. Kelly (2000), using data from the 1991 FBI uniform crime reports on urban counties,
concludes that
inequality has a substantial positive effect on property crime but does not relate to violent crime. The
conclusions of
both studies must be treated with care because the empirical analyses are based on cross-sectional
data, which
prevents the authors to control for time-invariant local effect. If the time-invariant local effects are
correlated with
both criminality rates and income inequality, the estimated effect of income inequality might be
spurious. The two
studies cited below deal with this methodological issue (Brush, 2007 and Choe, 2008)

Conclusion
This report presents a critical reading of the literature on the impact of income inequality on
important social
outcomes related to (i) well-being, (ii) criminality, (iii) health, (iv) social capital, (v) education, (vi)
political
participation and (vii) female labor market participation. In particular, the aim of this report was to look
into sound
empirical studies - based on multivariate analysis - which examine the effect of income
inequality on these
important social outcomes. Thereby, this report provides a first step into understanding more clearly
how rising
income inequalities might affect societies and established widely accepted knowledge that inequality is
a toxic
element of today’s European societies.

The upshot of this literature review is that higher criminality, reduced political agency and, to some
extent, lower
social capital formation and well-being appear to be tangible illustrations of the wastage produced by
rising income
inequality. In addition, there are a number of self-reinforcing loops linked to inequality. A clear
illustration of this is
the role of inequality in reducing the voting participation of the low income groups and the
concomitant
consequences in terms of redistributive policies and therefore on income disparities.

In more detail, the literature review has highlighted the following elements:

The effect of income inequality on happiness critically depends on the perceived country level of
mobility. If
income mobility is high, such as in the USA, income inequality tends to be positively associated with
reported well-
being as individuals tend to consider that they will eventually reach a higher income. The opposite is
observed in
low mobile countries (i.e. typically in European countries) because in those countries individuals feel
that it is
impossible to reach a higher level of income.

The majority of the studies focusing on the relationship between the income distribution and criminality
conclude in
favor of a detrimental effect of income inequality on criminal behaviors. The rationale behind these
findings might
be based on economic considerations – income inequality increases the gain derived from a criminal act
–and/or on
a sentiment of frustration of the less well-off individuals when they compare their situation with
respect to the
wealthier ones.

Empirical analyses of the harmful effect of income inequality on health are usually not conclusive, at
least among
wealthier European countries. This goes in line with the fact that there is still not a widely accepted
rationale for
explaining why income inequality should impact on health and even, several scholars tend to
suggest that the
causality runs in the other way around, from health status to income inequality.

In virtue to the aversion to heterogeneity theory, heterogeneous societies should be characterized by

fewer contacts
and in consequence, by lower levels of social capital. This prediction, also confirmed by adjacent
theories, appears
to be empirically validated in cross-country studies as well as in papers focusing on the US context.
Findings
specific to EU countries are limited and less conclusive.

The relationship between income inequality and educational attainment might go in both directions. On
one hand,
rising inequality should encourage investments in education through increased returns to education. On
the other
hand, it might prevent these investments for those people belonging to the bottom of the income
distribution
because of resources’ constraints. Regarding the latter mechanism, the empirical studies reviewed
suggest a modest
effect or no effect of income on educational outcomes (attainment, enrollment, etc.). However, when
interpreting
these results caution is needed, because conclusions rely strongly on the econometric approach
used by the
researchers.

The relationship between turnout and inequality is likely to be mutually reinforcing because, according
to the class-
bias assumption, the benefits from voting are lower for the low-income group, reducing the incentive
for this fringe
of the population to vote. If voter turnout is skewed by income, the policies implemented with favor the
well-off
group (median voter hypothesis), thus participating to the intensification of income disparities. In
turn, rising
economic inequality will discourage participation among low-income groups, and so on. These
predictions are
confirmed by the majority of cross-country and single-country based studies.

There is neither a sound theoretical base nor empirical evidence of an effect of income
inequality on the
participation rate of women in the labor force. The causality is found to run instead from labor
participation of
women to income inequality.
This report is a first step of a more comprehensive project aiming at analyzing the socio-consequences
of rising
income inequalities in Europe, and will be complemented with quantitative bivariate and multivariate
analyses of
the relationship between income inequality and some of the social outcomes discussed. The empirical
work will
cover the 27 EU countries and will be carried out at the sub-national level (NUTS 1 level).

The results of the literature review offer important guidelines for the succeeding quantitative step to be
carried out.
In particular, while each of the seven social outcomes reviewed in this document are important
constituencies of a
‘healthy’ society, and hence worthy of being examined in a more thorough way, it seems reasonable
to restrict
further quantitative analyses to the most relevant ones. More precisely, while the bivariate analysis will
be done for
the all social outcomes, the multivariate analysis will focus on the harmful effect of income inequality on
political
agency and criminality.

The choice of these two social outcome variables for the multivariate analysis is based on two
arguments. First,
political participation and criminality constitute important proxies for the functioning of a society.
Secondly,
besides their political relevance, a clear causal relationship can be more easily postulated for these
two social
outcome variables. In particular, the multivariate analysis at the sub-national level could add
substantially to the
understanding of the social challenges caused by rising inequalities in Europe.
Bibliography
Acemoglu, D., & Pischke, J.-S. (2001, May). Changes in the Wage Structure, Family Income, and
Children's Education.
European Economic Review, 45(4-6), 890-904.
Akee, R. K., Copeland, W., Keeler, G., Angold, A., & Costello, E. J. (2010, January). Parents' Incomes and
Children's
Outcomes: A Quasi-Experiment. American Economic Journal: Applied Economics, 2(1), 85-115.
Alesina, A., & La Ferrara, E. (2002). Who trust the others? Journal of Public Economics, 85(2), 207-234.
Alesina, A., Baqir, R., & Easterly, W. (1999). Public Goods and Ethnic Divisions. Quarterly Journal of
Economics, 114(4),
1243-1284.
Alesina, A., Di Tella, R., & MacCulloch, R. (2004). Inequality and Happiness: Are Europeans and
Americans Different?
Journal of Public Economics, 88(9-10), 2009-2042.
Banerjee, A. V., & Newman, A. F. (1993, April). Occupational Choice and the Process of Development.
Journal of Political
Economy, 101(2), 274-298.
Becker, G. (1964). Human Capital. New York: Columbia University Press for the National Bureau of
Economic Research.
Becker, G. S. (1968). Crime and punishment: An economic approach. Journal of Political Economy, 76,
169-217.
Berggren, N., & Jordahl, H. (2006). Free to Trust: Economic Freedom and Social Capital. Kyklos, 59(2),
149-169.
Bjørnskov, C. (2006). Determinants of generalized trust: A cross-country comparison. Public Choice,
130(1-2), 1-21.
Bourdieu , P. (1977). Outline of a Theory of Practice. Cambridge: Cambridge University Press.
Brush, J. (2007). Does income inequality lead to more crime? A comparison of cross-sectional and time-
series analyses of
United States counties. Economic Letters, 96, 264-268