Content uploaded by Maximilian Stockhausen
Author content
All content in this area was uploaded by Maximilian Stockhausen on May 21, 2021
Content may be subject to copyright.
Wirtschaftliche Untersuchungen,
Berichte und Sachverhalte
An international comparison with a special focus on the case of Germany
Judith Niehues / Maximilian Stockhausen
Köln, 19.05.2021
Inequality revisited
1
Content
Executive Summary 2
1 Introduction 7
2 Global inequality 8
2.1 Income inequality 8
2.2 Extreme Poverty 14
2.3 Wealth inequality 17
3 The case of Germany 20
3.1 Income inequality 21
3.2 Wealth inequality 24
3.3 Middle class 26
3.4 Determinants of inequality 31
3.5 COVID-19 and income inequality 35
4 The inequality-growth-nexus 38
5 Conclusion 42
List of figures 45
References 46
Inequality revisited
2
Executive Summary
Both income and wealth are unequally distributed around the world and by far not all people
have equal access to education, health, and prosperity. Nevertheless, the world has made great
progress towards a better and more inclusive place in recent decades and was especially
successful in combating extreme poverty. But it is also true that positive changes have not oc-
curred everywhere, and many successes are now jeopardized by the Corona pandemic. The re-
port highlights some major global trends before analysing the situation in Germany in more
detail.
Global inequality
◼ Global income inequality has changed significantly over the past 200 years resulting in a
more equal distribution of global income and less extreme poverty. The 19th century is char-
acterized by the rise of Western countries through industrialization, while the 20th century
is marked by advances in South-East Asia. In the first period, global income inequality in-
creased, while it has decreased in the second period when populous countries such as China
and India began their catch-up process. Since the late 1980s, global income inequality meas-
ured by the Gini coefficient decreased from 0.68 in 1988 to 0.62 in 2013 due to narrowing
income gaps between countries around the world (the Gini coefficient can be expressed on
a 0 to 1 or 0 to 100 scale). At the same time, though, the contribution of within-country
inequality to global inequality has increased.
◼ Global earnings inequality follows similar patterns as global income inequality. The respec-
tive Gini coefficient has decreased from around 0.7 in the 1970s to around 0.6 by the year
2015. The largest reduction took place in the late 90s and 2000s. The main driver of equaliz-
ing global earnings was a reduction of earnings inequality between countries, while earnings
inequality within countries increased on average. The effects of the Corona pandemic are
uncertain yet. Although some vulnerable groups like low-skill workers are hit harder by the
crisis than other groups, government support can mitigate some of the negative effects on
incomes. Hence, it is not clear so far, how global earnings or income distribution will be af-
fected in the short- and long-run.
◼ Global net wealth inequality has constantly decreased since 2000 according to data from
the Credit Suisse Global Wealth Databook: the Gini coefficient decreased from around 91.9
in 2000 to 88.5 in 2019. At the end of 2019, North America and Europe accounted for 55 per
cent of total global net wealth, while they represented 17 per cent of the world adult popu-
lation. It is also most Europeans and North Americans who belong to the top 10 per cent of
the global rich. Nevertheless, although wealth is still far more concentrated than income,
the economic rise of Asia, and China in particular, has contributed to a discernible reduction
in global net wealth inequality: In 2000, the net wealth share of the top 10 per cent
amounted to 88.5 per cent but decreased to 81.7 per cent in 2019. In contrast, the net
wealth share of the top 1 per cent remained almost unchanged and has been varying around
45 per cent.
Inequality revisited
3
◼ Ending global extreme poverty by 2030 is the foremost goal of the United Nations and large
progress has been made during the last decades. Global extreme poverty decreased sub-
stantially from 42 per cent in 1981 to 10 per cent in 2015 if measured as the percentage of
the global population living on less than $1.90 a day at 2011 PPP dollar. This is slightly more
than 700 million people in 2015 living in extreme poverty compared to 1.9 billion people in
1981. Using a higher poverty threshold of $3.20 ($5.50) a day also shows a significant reduc-
tion of poverty from 57 (66) per cent in 1981 to 26 (46) per cent in 2015. China and India
contributed to a large extent to this success story, since economic integration and produc-
tivity growth were able to lift people out of extreme poverty. However, productivity growth
has slowed down since the financial crises and accompanied output losses imply large,
missed opportunities for more rapid poverty reduction. The COVID-19 pandemic will likely
further decelerate productivity growth and will be a threat to the achievement of the reduc-
tion of (extreme) poverty. Thus, the pandemic and the associated loss of income could even
increase global poverty for the first time in more than 30 years and reverse the progress of
previous years.
The case of Germany
◼ As in many developed countries, net income inequality measured by the Gini coefficient is
higher in Germany today than it was in the 1990s. It increased from 0.25 in 1991 to 0.29 in
2017 according to household panel data from the Socio-Economic Panel (SOEP). The rise in
net income inequality predominantly occurred between the late 1990s and 2005. Since
2005, inequality in net incomes remained almost unchanged, a year that represents a turn-
ing point in the development of income inequality in Germany. Compared to other countries,
Germany still exhibits a relatively equal distribution of net incomes. Among the member
countries of the Organisation for Economic Co-operation and Development (OECD) coun-
tries, values of the Gini coefficient vary between 0.25 in more egalitarian countries like the
Slovak Republic and 0.50 in more unequal countries like Costa Rica.
◼ Germany is certainly not a role model in every socio-political area and has some fairness
deficits, as well. When it comes to educational mobility, for example, pupils are segregated
relatively early in their school careers, which is likely to limit educational opportunities of
children from less well-off families who are comparatively less likely to acquire a university
degree. However, Germany manages to compensate for some of these disadvantages
through its dual vocational training system, which enables people without a college degree
to pursue well-paid professions. A career as a skilled worker is, thus, a good alternative to a
bachelor's degree. This is also a major reason why Germany performs rather poorly in terms
of educational mobility but shows significantly better results in terms of income mobility;
an often-used indicator of equal opportunities. According to comparative studies, Germany
ranks in the midfield among industrialized nations. In addition, it is more mobile in terms of
labour income than the United States (US) in both absolute and relative terms.
◼ Net wealth inequality has remained comparatively stable over the past decade in Germany,
too. The Gini coefficient of individual net wealth has been varying around 0,78 since 2002.
The stable trend in net wealth inequality is observed in a period of low interest rates and
Inequality revisited
4
rising asset prices, which mainly resulted from the loose monetary policy after the financial
and economic crisis in 2008/2009. In particular, the value of owner-occupied real estates has
strongly increased during the last ten years, especially in urban areas. Since owner-occupied
real estate is the major wealth component of the middle class, they were able to benefit
relatively strongly from rising real estate prices. In contrast, business assets are particularly
important at the top of the wealth distribution but are often not covered very well in house-
hold surveys. By now, there are various attempts to estimate the missing wealth at the top
from rich lists. The results show that the share of the top 10 per cent is underestimated,
while time trends are rather robust to top-wealth-adjustments. In international comparison,
wealth inequality in Germany is rather high. However, comparatively high wealth concen-
tration is rather typical for those European countries which are characterized by generous
welfare state, below-average net income inequality and high levels of living standard.
◼ Although there is no commonly accepted definition of the German middle class, most ap-
proaches that try to operationalize this group rely on income related measures and define
income boundaries relative to the median net income of the population. Using the IW-in-
come-classes it can be shown that the development of the middle class since reunification
can be divided into three phases: During the East German catch-up process, the share of the
middle class in the narrow sense initially increased from 50.4 to 54.7 per cent until its peak
in 1997. By 2005, its share had fallen again to 50.1 per cent, and barely changed since then:
The share of the population in the middle class in the narrow sense equals 49,4 in 2017 which
is very close to the middle class share in 2005.
Determinants of inequality in Germany
◼ Income and wealth inequalities are caused by many factors and originate from different
sources. Thus, the effect of each factor on the macro and micro level is difficult to identify
and many factors depend on each other. One important factor is the primary distribution
of income, namely the distribution between capital and labour. Contrary to the common
belief, the share of labour income is not constantly decreasing. In fact, after a sharp decline
from about 70 per cent in 2003 to 64 per cent in 2008, the labour income share recovered
after some ups and downs following the 2008 financial crisis and, with a share of around 72
per cent in 2019, is at a level similar to the 1990s. At the same time, there is no clear rela-
tionship between the evolution of the labour income share and the distribution of market
incomes, that is a higher labour income share does not automatically correspond with a
more equal distribution of market incomes. In contrast, the comparison of changes in the
distribution of market incomes and gross hourly wages over time reveals more similar
trends, but differences remain here, too. Inequality in both market incomes and gross hourly
wages increased between 1991 and 2005/2006. While gross wage inequality has slightly de-
clined in the following years, market income inequality remained almost unchanged.
◼ The introduction of the statutory minimum wage in Germany in 2015 increased gross hourly
wage rates for low-income earners, especially in Eastern Germany. This did not result in an
unambiguous lower level of market income inequality, though, merely due to working hours
adjustments. Compared to other EU countries, the low-wage sector is rather pronounced in
Inequality revisited
5
Germany, while youth unemployment and unemployment among the less-educated is com-
paratively low. Furthermore, even when accounting for purchasing power differences, wage
levels in Germany are comparatively high. With respect to time trends, there were no further
remarkable changes of the low-wage sector nor of atypical employment since 2007.
◼ Another factor that influences the trends in market and disposable income inequality is the
change in the composition of the population. For example, a rising number of single house-
holds in Germany has increased inequality in disposable incomes. Migration from Eastern
Europe since 2010 as well as the influx of refugees from 2015 onwards also had an impact
on the development of inequality and counteracted decreasing inequality trends in recent
years, since most migrants and refugees first belong to low-income groups when entering
the country. This is reflected, among other things, by the observation that at-risk-of-poverty
rates are higher for persons with a migration background and that poverty risks have in-
creased among this group in recent years. Meanwhile, the share of low-income earners with-
out migration background has been constant or has even declined in some age groups. Coun-
terfactual analyses reveal that when isolating the increase in employment since 2005, it
would have resulted in decreasing inequality.
COVID-19 and income inequality in Germany
It is yet not clear to what extent the Corona pandemic will change existing income inequalities.
With respect to the worldwide development, analyses project that the pandemic will likely in-
crease income inequality and poverty since job losses could disproportionally affect the income
and labour participation of low-skill workers. However, the impact of the pandemic also de-
pends on the measures taken by local governments to counteract the negative effects of the
crisis. Hence, market and disposable incomes can be affected very differently. Whether the
pandemic will result in long-lasting negative effects does also depend on its duration and
whether a quick economic recovery is feasible or not.
Despite these uncertainties, first simulation results for Germany show that market incomes per
capita could have decreased by around 6 per cent in 2020 compared to 2019 and that individuals
in lower income groups suffered the greatest losses in relative terms. However, the losses in
disposable incomes were much smaller – less than 1 per cent on average –, since the automatic
stabilizers of the social security system like unemployment benefits but also additional
measures like short-time work allowances (Kurzarbeitergeld) helped to cushion income losses.
Overall, while the Corona pandemic is expected to increase market inequality in Germany, sim-
ulation analyses suggest no change of disposable income inequality in the short run. However,
long-run effects of the Corona pandemic on income but also wealth inequality are still highly
uncertain. Similar results can be found for other European countries.
The inequality-growth-nexus
The relationship between inequality and growth regained renewed interest when the Interna-
tional Monetary Fonds (IMF) and the OECD closely in time published to studies on this topic in
the year 2014. Their results implied that economic growth is negatively affected by income ine-
quality and there is no trade-off between equity and efficiency. However, the results heavily
Inequality revisited
6
depend on model assumptions, for example by assuming a linear relationship between both
dimensions. If non-linear relationships are considered it becomes apparent that the negative
effect of increasing inequality on economic growth crucially depends on the initial level of ine-
quality, the level of economic development of countries, and the scope of redistribution by taxes
and transfers. According to a global comparison of 113 countries, up to a value of the Gini coef-
ficient of 0.35, rather a positive correlation between inequality and growth can be presumed. If
this threshold value of inequality is exceeded, rather negative consequences of increasing ine-
quality on economic growth can be expected.
Inequality revisited
7
1 Introduction
Many debates about the distribution of income and wealth are highly political and lack a com-
prehensive view of the facts. Two big challenges in international comparisons are the availability
of appropriate data and the great heterogeneity of countries, which differ in many ways: Be it
their general level of development in terms of GDP per capita, their population structures, or
the extent of social security systems. It is precisely these differences that highlight the need for
a differentiated debate guided by facts, since looking at different dimensions of inequality can
lead to very different assessments of the extent of inequality. Ultimately, however, the question
of how much inequality in income, wealth, health, education, or opportunity is optimal for a
society will always remain a normative one.
As Anthony Atkinson has made very clear in his book "Inequality. What can be done?", the ques-
tion to be answered first is which inequalities one is talking about and among whom (Atkinson,
2015). Different delineations of income can yield sometimes large differences in the extent and
evolution of inequalities. Accordingly, one can distinguish between market, gross, and net in-
comes of households or individuals, which reflect very different aspects of economic opportu-
nities and consumption possibilities. For example, if market incomes were considered alone,
redistribution through taxes and transfers would be neglected. But it is precisely for the very
young and the very old, or for the sick and those unable to work, that transfers represent es-
sential parts of their income from which they make their living, and which are an expression of
solidarity-based redistribution from the strong to the weak – at least in well-established welfare
states. Thus, from a welfare-theoretical perspective, an examination of net incomes provides a
better approximation of the distribution of economic resources and opportunities in a country
than market incomes. Since social security systems vary largely between countries, differences
in these systems, for example in pension systems, should always be considered carefully in ine-
quality analyses. Equally important is the question of whether to look at the individual or house-
hold level, since in many cases intrafamily redistribution of resources and mutual protection
against life risks already takes place in the household even before the state intervenes.
In this report, we attempt to provide an overview of the current trends in income and wealth
inequality in Germany and the world. In doing so, we not only describe the income and wealth
concepts used and their differences from one another, but also address the challenges of com-
paring different countries, which is difficult in many cases due to a lack of suitable and harmo-
nized data. In detail, we will first discuss the differences in the extent and evolution of inequality
in high-, middle-, and low-income countries and then turn to the specific developments in Ger-
many. Regarding low-income countries, it stands out that the fight against extreme poverty was
very successful in the decades before the Corona pandemic and large progress was made in
providing sufficient resources to the poor to cover basic needs. However, extreme poverty is
still not abolished and the fight against it should continue to have the highest priority to achieve
the number one goal of the United Nations (UN) Sustainable Development Goals (SDG) by 2030.
Action is especially needed in Sub-Sahara Africa and new challenges from the Corona pandemic
are threatening former achievements. Furthermore, there was large convergence between mid-
Inequality revisited
8
dle- and high-income countries starting around the 1990s, which led to a decline in global ine-
quality and lowered between country inequalities. The economic rise of China and other South
Asian countries are main drivers of this development. At the same time, income and wealth
inequalities within countries were on the rise in many but not all countries. In Germany, for
example, while income (wealth) inequalities have been rising until the early 2000s (until the
financial crisis), they have been almost unchanged since 2005 (2008). In general, it is difficult to
assess whether inequality is too high or too low when talking about developed countries with
high living standards and large economic resources. As a result, issues of equal opportunity are
becoming more of a focus in these countries since it is almost impossible for a society to agree
on optimal levels of outcome inequality.
2 Global inequality
2.1 Income inequality
With an increasing availability of appropriate data there is also an increasing literature (and in-
terest) on the development of global economic inequality. When mapping global economic in-
equality, incomes must be made comparable. Thus, approaches to measure global inequality
first convert incomes into so called international dollar, which is a hypothetical currency that
represents the amount of goods and services one could buy with the amount of one dollar in
the US in a certain year. In a second step, individuals are sorted in ascending order of their ad-
justed income. Figure 2-1 plots the resulting global income distribution for three different points
in time. The changes of the distribution have gained considerable attention and can be summa-
rized as follows:
1
Back in the 19th century, only a few countries have achieved economic growth. In contrast, the
majority of the worldwide population lived in conditions which could be referred to as extreme
poverty. Graphically this results in a frequency distribution of incomes with one hump, which
illustrates that the bulk of the global population had very low incomes at that time. Especially
in the aftermath of the Second World War, the shape of the global income distribution changed
substantially. With increasing economic growth in Northern America, Europe, Oceania and parts
of South America and East Asia (for example Japan) many people in those regions experienced
considerable income gains, thus, moving to the righthand-side along the income-axis. Therefore,
the graphical representation of the global income distribution changed into a bimodal distribu-
tion, with one hump below the international poverty line and a second hump representing peo-
ple with considerably higher incomes. Therefore, global inequality has increased strongly, the
world had divided into two clearly distinguishable regions – the developed world and the devel-
oping world. However, this also means that millions of people were lifted out of extreme pov-
erty by economic progress and even the poorer half was able to do slightly better in absolute
terms.
1
See https://ourworldindata.org/global-economic-inequality (by Max Roser) for detailed descriptions, further
references and illustrations about the development of global economic inequality.
Inequality revisited
9
Starting around the 1980s the global income distribution changed again. Due to the rapid eco-
nomic development in Asia, especially in China and India, incomes in some poor countries have
grown faster than incomes in rich countries – thus, the global income distribution has become
more equal. The distribution function has again changed to a “one-hump-world”, implying that
the clear-cut division into a developing and developed world no longer remained. However, the
far stretch between those world inhabitants with low incomes and those with high incomes
underlines that global incomes are still very unequally distributed.
Figure 2-1 focuses on a long-term – and even historical – perspective on the development of
global economic inequality. In a recent paper, Hellebrandt and Mauro (2015) show that after
the turn of the millennium, global incomes further equalized. According to their analysis, the
Gini coefficient of the global income distribution declined from 68.7 in 2003 to 64.9 in 2013.
2
What is more, global median income almost doubled from 1,090 international dollar per year
to 2,010 over this period (in 2011 international dollar), implying that a considerable share of
people with very low incomes have gained substantial income increases. On the basis of pro-
jected annual growth rates of different regions, the researchers also made a forecast of the
global income distribution up to the year 2035 and expect that by then, median income will
2
These numbers slightly differ to the results in Figure 2-2 due to different data sources.
Figure 2-1: Income per world citizen per year
In 1820, 1970, and 2000
Source: OECD (van Zanden et al., 2014, p. 281, Figure 11.1), data retrieved from
https://www.maxroser.com/roser/graphs/WorldIncomeDistribution1820to2000/WorldIncomeDistribu-
tion1820to2000.html [21.12.2020]
0
50.000
100.000
150.000
200.000
250.000
300.000
350.000
100 1000 10000 100000
Number of people in 1,000
Income per world citizen per year (in 1990 international Dollar)
1820 1970 2000
Inequality revisited
10
approximately double again and that the Gini coefficient will further decrease to 61.3. However,
although these estimations show substantial improvements of incomes, even in 2035 it can be
expected that the bulk of worldwide population will live with rather low incomes of less than
5,000 dollar per year (in 2011 international dollar) and that there is a large spread between low
and high incomes – illustrating that global inequality will still be rather high in 2035.
While increasing incomes in the course of continuing global integration raised the average living
standard in some of the poorest economies, some of those catching-up countries simultane-
ously experienced increasing inequality within their borders. In fact, global inequality depends
on both, differences in average incomes between countries and inequalities in the distribution
of incomes within countries. Global inequality can be decomposed to illustrate the contribution
of both explanatory factors. Therefore, in addition to the Gini coefficient, Figure 2-2 also in-
cludes the development of the mean-log deviation, which belongs to the group of the additively
decomposable inequality measures. Compared to the Gini coefficient, the mean-log deviation
is more sensitive to changes at the bottom of the income distribution. As Figure 2-2 shows,
between 65 and 80 per cent of total inequality can be attributed to differences in average in-
comes across countries. It also reveals that the reduction of inequality is driven by a conver-
gence of average incomes between countries. In fact, during the 1990s the reduction of global
inequality is in part counteracted by increasing inequality within countries, a development
which stabilized around the turn of the millennium. Overall, these trends implied that by 2013,
the contribution of within country inequality to total inequality has increased. Nevertheless, the
Figure 2-2: Global inequality
1988-2013
Source: World Bank, 2016, p. 81, Figure 4.5
0.2 0.22 0.25 0.27 0.27 0.27
0.8 0.73 0.67 0.66 0.63
0.49
0.68 0.68 0.67 0.67 0.65 0.62
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1988 1993 1998 2003 2008 2013
Gini index
Mean log deviation
Within-country inequality Between-country inequality Gini index (right-axis)
Inequality revisited
11
largest fraction of inequality is still explained by enormous differences in average incomes
across countries, emphasizing the importance of economic growth in poorer countries.
Hammar and Waldenström (2017) ran a similar analysis on the development of global earnings
inequality based on a unique earnings survey database run by UBS. The results basically confirm
the trends which have also been observed on the basis on income (or consumption) inequality.
In particular, they reveal that global earnings inequality was very high in the 1970s (with a Gini
coefficient of around 0.7) which has fallen to a level of around 0.6 by the year 2015, with the
main equalization taking place in the late 90s and 2000s. Beyond, they also found that decreas-
ing earnings inequality between countries was the main driver of equalizing global earnings,
while the contribution of within-countries earnings inequality has increased.
Their dataset is also unique in so far, that it allows to follow each occupational group in each
country over time. On this basis, Figure 2-3 plots the earnings-growth of each country-occupa-
tion since 1970 against its initial rank in the global earnings distribution. The illustration of this
so-called non-anonymous growth-incidence curve reveals that average earnings-growth has
Figure 2-3: Growth incidence of country-occupations (1970s-2010s)
Notes: Each point represents a country-occupation. Dashed line shows average growth for all observations, and solid
line a smoothed local polynomial with 95% confidence interval.
Source: Hammar/Waldenström, 2017, https://voxeu.org/article/new-data-global-earnings-inequality [21.12.2020],
Figure 2
Inequality revisited
12
been higher in the lower half of the global distribution, whereas in the upper half of the distri-
bution earnings growth often was below-average and for some country occupations it was even
zero or negative. The results are particularly interesting because while they confirm the increas-
ing earnings dispersion between managers and unskilled workers in the US, the results reveal
that both occupations experienced negative real PPP-adjusted earnings-growth over the ob-
served period. For France, the picture is even reversed. As the illustration suggests, unskilled
workers experienced higher earnings-growth than managerial occupations in France.
In this regard the results somehow differ from the results of the growth-incidence curve based
on the real income changes at various percentiles of the global income distribution, which has
become known as the so-called “elephant-curve” (its shape resembles the outline of an ele-
phant). The (anonymous) global incidence curves suggests that the largest gains between 1988
and 2008 were realized around the median of the global income distribution and among the top
1 per cent (Lakner/Milanovic, 2016). Income growth was rather negligible around the 80th and
90th percentile in these years. Further analysis reveals that seven out of ten people in these
percentiles are from the lower halves of ‘old rich’ OECD countries.
3
The picture becomes less
dramatic when a quasi-non-anonymous growth-incidence curve is applied. It reveals less growth
for the “former” top per cent of the income distribution than for the rest of the distribution.
Though, the general finding about rather stagnating middle class incomes in developed coun-
tries remain. Since Germany often holds as an exemplary case with merely stagnating middle
class incomes, the report will focus on the development of Germany in some more detail in
section 3.
The previous remarks have revealed that global income inequality decreased, and that this de-
velopment was mainly driven by a decline in between country inequality and, thus, by average
income convergence. Figure 2-4 represents the development of a key determinant of cross-
country per capita income convergence, which is labour productivity. According to the research
on the impact of productivity growth on cross-country differences in per capita income growth
rates, up to 60 to 90 per cent of the cross-country variations in per capita income can be at-
tributed to differences in productivity growth.
4
Thus, labour productivity is the main driver of
the catch-up process through which developing countries with lower-income per capita can
reach per capita income levels observed in advanced economies. Figure 2-4 shows the different
trends in productivity growth in advanced economies versus emerging markets and developing
economies (EMDE). The graphical representation reveals that in advanced economies produc-
tivity growth has experienced a long-run decline over the past 40 years. In contrast EMDE labour
3
New data covering the years from 2008 to 2013/14 shows that the elephant has lost its trunk (Milanovic, 2020). Real
income growth of the global top 1 per cent was remarkably lower in this period than for most other parts of the
global income distribution. And “it is these broad-based large differences in real growth that are the main engine
behind the reduction of global inequality” within this period (Milanovic, 2020, 36). But this also means that incomes
in developed countries, which comprise large parts of the upper tail of the global income distribution, recovered
relatively slowly after the financial crisis in 2008.
4
Easterly and Levine (2001) find that it is rather productivity growth than differences in schooling and capital accu-
mulation (as suggested by neoclassical growth theory) which drives differences in economic growth. According to
Klenow and Rodriguez (1997) the overwhelming majority of growth divergence can be explained by productivity
growth. See Mayer-Foulkes (2019, p. 39 f.) for further references.
Inequality revisited
13
productivity growth has rather trended up over the same time horizon. After severe declines in
productivity growth in the 1980s and early 1990s, growth rose sharply from the late 1990s on-
wards. Starting in 2000, average productivity growth in EMDE was larger than that in advanced
economies. In fact, in 60 per cent of EMDE productivity growth exceeded the average rate of
advanced economies over the past two decades. The observed differences in productivity trends
over the last decades built the ground for income convergence across countries and the remark-
able reduction of between country inequality. Nevertheless, the productivity gap between
EMDE and advanced economies is still extensive, with labour productivity in EMDE being less
than one-fifth of the average level of advanced economies and the pace of convergence is still
relatively small.
The highlighting of recessions reveals that global economic downturns are regularly followed by
decreases in labour productivity growth, with more pronounced declines in less advanced econ-
omies. The global financial crisis, though, marked a turning point in the global development of
labour productivity because in the aftermath of this recession global productivity growth slowed
down dramatically. In contrast to previous recessions, the deceleration of productivity growth
after the global financial crisis seems to be persistent. Thus, even before the emergence of the
COVID-19 pandemic there were concerns about productivity growth in EMDE. Given the expe-
rience with previous recessions it is likely, that the pandemic will lead to further losses in
Figure 2-4: Productivity growth in developed versus developing countries
In per cent
Notes: Productivity is defined as output per worker in US dollar (at 2010 prices and exchange rates). Sample of 29 ad-
vanced economies (AEs), and 74 emerging market and developing economies (EMDEs) including 11 low-income coun-
tries (LICs), as of 2019 World Bank classifications. Aggregate growth rates are GDP-weighted at constant 2010 prices
and exchange rates. Shaded regions indicate global recessions and slowdowns.
Source: World Bank 2020, Figure 1.1A (based on Conference Board; Penn World Table; World Bank, World Develop-
ment Indicators)
-2
-1
0
1
2
3
4
5
6
7
8
1981 1985 1990 1995 2000 2005 2010 2018
World Advanced economies EMDEs
Inequality revisited
14
productivity growth and will further slowdown global productivity and, thus, income conver-
gence.
Nevertheless, it is yet not clear, how the still ongoing Corona pandemic will affect global income
inequality in the short- and long-run. A recent study by Deaton (2021) shows that countries with
more deaths saw larger declines in income. For the period under observation, it was in fact a
number of higher-income countries who suffered more deaths per capita despite their better
health care systems (one reason for this might be the higher share of older people in high-in-
come countries who were more vulnerable to the virus than younger ones). As a result, the fall
in per capita incomes was more pronounced in higher-income countries and global income
inequality decreased. However, the results on global income inequality change if countries
are weighted by their population such that the influence of populous countries like India or
China increases. Thus, when considering population-weighted income changes, global in-
come inequality increased, “because Indian incomes fell, and because the disequalizing ef-
fect of declining Indian incomes was not offset by rising incomes in China, which is no longer
a globally poor country” (Deaton, 2021, p.1).
2.2 Extreme Poverty
The first and foremost goal of the United Nations is the eradication of extreme poverty by the
year 2030. Living in extreme poverty means that basic needs like health, education, and access
to water and sanitation are not fulfilled. For this reason, this goal is written in the first place of
the Sustainable Development Goals of the UN and is of the highest priority. Extreme poverty is
defined in terms of an absolute poverty threshold: Anyone who has less than $1.90 per day to
live on is considered as extremely poor (in 2011 PPP dollar). The threshold is expressed in pur-
chasing power parities (PPPs), which are the rates of currency conversion that try to equalise
the purchasing power of different currencies, by eliminating the differences in price levels be-
tween countries for a similar basket of goods and service. However, the level of the poverty
threshold is subject of constant debate, and it is often criticized as being too low to guarantee
a minimum subsistence level. Therefore, there are two other thresholds that are used in partic-
ular for more advanced countries with higher consumption levels: $3.20 per day (PPP) and $5.50
per day (PPP).
Available data from the World Bank suggests that the fight against extreme poverty has been
very successful since the 1980s (Figure 2-5). Global extreme poverty has decreased substantially
from 42 per cent in 1981 to 10 per cent in 2015 if measured as the percentage of the population
living on less than $1.90 a day at 2011 PPP dollar. This is slightly more than 700 million people
in 2015 living in extreme poverty compared to 1.9 billion people in 1981. If extreme poverty is
measured as having less than $3.20 ($5.50) a day, the share of people living in extreme poverty
declines from 57 (66) per cent in 1981 to 26 (46) per cent in 2015. Thus, large progress has been
made in fighting extreme poverty worldwide although there are still too many people who face
great challenges to make their daily living. Nevertheless, it seemed feasible to end extreme pov-
erty by 2030.
Inequality revisited
15
Especially China and India contributed to a large extent to this success story, since economic
achievements and inclusive growth were able to lift people out of extreme poverty. Building on
the importance of productivity for income convergence, Figure 2-6 illustrates its relevance for
poverty reduction. Those EMDE in the top quartile of productivity growth reduced their extreme
poverty rates by an average of more than one percentage point per year since 1981. In contrast,
in those countries in the lowest quarter of productivity growth poverty rates increased through-
out the same period. Further analyses show that the slowdown in productivity growth since the
financial crises and accompanied output losses implied large, missed opportunities for more
rapid poverty reduction. The COVID-19 pandemic will likely further decelerate productivity
growth and, thus, will be a threat to the achievement of development goals, in particular the
reduction of (extreme) poverty. Thus, the pandemic and the associated loss of income could
even increase global poverty for the first time in more than 30 years and reverse the progress
of previous years. Projections suggest that 71 to 100 million people may be pushed back into
extreme poverty in 2020 (Lakner et al., 2020). Those forecasts underline the importance to fa-
cilitate conditions to return to the previous productivity path as soon as possible.
Figure 2-5: Extreme poverty
Poverty headcount ratios for different poverty thresholds (in 2011 PPP dollar)
Source: World Bank DataBank, 2020; https://databank.worldbank.org/home.aspx
0
10
20
30
40
50
60
70
80
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014
$1.90 a day $3.20 a day $5.50 a day
Inequality revisited
16
Although the consequences of the COVID-19 pandemic trigger concerns about the further de-
velopment, the empirical numbers unambiguously show that extreme poverty has substantially
declined over the past decades. However, if people are asked about what they think about the
development of extreme poverty over the last two decades, half of those asked around the
world believe that extreme poverty has increased (Figure 2-7). Especially in advanced econo-
mies the overwhelming majority believes that poverty was on the rise. In Germany, for example,
only 11 per cent of respondents correctly guessed that the proportion of the world population
living in extreme poverty has declined over the last 20 years. The misperception of positive
trends is broad-based. According to a study by the Bertelsmann Stiftung, two-thirds of the EU
population believe that the world used to be a better place (de Vries/Hoffmann, 2018). The
pessimistic view on global development will likely further encourage negative views on globali-
zation, economic integration, and free-market systems. Processes and structures, which have,
in fact, helped to lift millions of people above the poverty line.
Figure 2-6: Poverty reduction and productivity
Annual change in poverty rates in EMDEs by productivity growth, in percentage points
Labour productivity is defined as output per worker in US dollar (at 2010 prices and exchange rates). Data is from a
sample of 74 EMDEs. Unweighted averages using annual data during 1981-2015. Fastest-growing EMDEs are those in
the top quartile by productivity growth; slowest-growing EMDEs are those in the bottom quartile of labour productiv-
ity growth. Poverty rate defined as the share of the population living on less than $1.90 a day (2011 PPP).
Sources: World Bank, 2020, Figure 1.2A (on basis of Conference Board; Penn World Table; PovcalNet; World Bank,
World Development Indicators)
-1.2
0.6
-1.25
-1
-0.75
-0.5
-0.25
0
0.25
0.5
0.75
Fastest-growing Slowest-growing
Inequality revisited
17
2.3 Wealth inequality
International comparisons of wealth are often associated with great difficulties regarding the
(time) consistent measurement of overall wealth and its components. Different national curren-
cies and purchasing powers have to be considered, too. For example, until 2010 there were no
harmonized data on the level and on the distribution of wealth in the Euro zone. It was only with
the Household Finance and Consumption Survey (HFCS) that this basis was created under the
supervision of the European Central Bank (ECB). Despite this progress, even these household
survey data are not perfect and, for example, do not fully represent all the wealth at the top of
the distribution. A global database on net wealth does not exist so far, although there are initi-
atives like the OECD Wealth Distribution Database to make progress on this issue, at least for
OECD countries. However, the actuality of the data is limited and likewise many gaps can be
found in them.
Figure 2-7: Believes on the development of extreme poverty
In per cent
Question: In the last 20 years, the proportion of the world population living in extreme poverty has …?
Source: Ipsos Perils of Perception – Global Impact of Development Aid, September 2017
49
37
36
35
35
31
30
25
24
21
21
20
19
17
16
15
15
14
14
12
12
11
11
11
9
9
9
9
9
21
55
36
51
41
44
39
64
39
51
43
52
43
48
68
51
38
56
64
58
65
67
63
56
68
32
67
68
60
30
8
28
14
24
25
31
11
37
28
36
28
38
35
16
34
47
30
22
30
23
22
26
33
23
59
24
23
31
010 20 30 40 50 60 70 80 90 100
China
Kenya
Peru
Senegal
India
Indonesia
Sweden
Nigeria
Poland
Brazil
Great Britain
Global
Australia
Canada
South Africa
United States
South Korea
Saudi Arabia
Turkey
Belgium
Russia
Mexico
Germany
Spain
Hungary
Japan
Argentina
France
Italy
Believe poverty decreased Believe poverty increase Don´t know
Inequality revisited
18
Although comprehensive harmonized data on net wealth is still missing on the global level, there
are also initiatives like the Credit Suisse Global Wealth Databook led by Anthony Shorrocks that
try to combine available data sources for different countries to monitor the trends of global
wealth inequality. For this purpose, various available national wealth data sources are combined
and supplemented with regression-based estimates whenever data is missing. The basis of this
approach is existing household survey data from industrialized countries that is extended by
information from household balance sheets, financial balance sheets, or a combination of them.
In the latest report, the determinants of per capita wealth were estimated with data from 53
countries, when at least one year was available. These estimates, then, were used to predict
missing wealth information for 119 additional countries. Hence, information on most countries
is not directly observed but estimated. Furthermore, since information on the extremely
wealthy is lacking in household survey data, additional information from the Forbes list and
other comparable sources is used to estimate missing information at the top of the wealth dis-
tribution. In addition, assets in different currencies are converted into current US dollar such
that the estimates also depend on the current national exchange rates to the US dollar, which
may be subject to large fluctuations, especially in developing countries. Therefore, short-term
changes should be treated with great caution. Differences in purchasing powers are not consid-
ered.
If one accepts the uncertainties of the data and looks at how the distribution of global net wealth
has developed since 2000 – net wealth is defined as the sum of all financial and non-financial
assets of a household minus its debts including private pension fund assets but excluding enti-
tlements to state pensions –, a decline in the inequality of net wealth can be observed at a high
level (Figure 2-8).
5
Overall, the Gini coefficient decreased from around 91.9 in 2000 to 88.5 in
2019. Accordingly. wealth is unequally distributed around the globe and far more concentrated
than net income. At the end of 2019 North America and Europe accounted for 55 per cent of
total global net wealth, while they represented 17 per cent of the world adult population. It is
also most Europeans and North Americans who belong to the top 10 per cent of the global rich.
In 2000, the net wealth share of the top 10 per cent amounted to 88.5 per cent but decreased
to 81.7 per cent in 2019. In contrast, the net wealth share of the top 1 per cent remained almost
unchanged and has been varying around 45 per cent. It was only temporarily affected by the
financial crisis back in 2008, where it decreased to 41.3 per cent but recovered relatively fast
afterwards. In 2019, it amounted to 45 per cent and, thus, is only slightly smaller than in 2000.
Nevertheless, the economic rise of Asia, and China in particular, has contributed to a discernible
reduction in global net wealth inequality. For example, net worth per adult in China increased
by 12.8 per cent in 2019, rising to an average of $70,962 per adult in current US dollar. Africa
also experienced an increase of 10.7 per cent in 2019. However, the average net worth per adult
was only $7,372. In comparison, net worth per adult in North America (Europe) grew by 11.4
per cent (6.1 per cent) in the same year, while remaining significantly higher at an average of
5
The Gini coefficient is not bounded to values between 0 and 100 in this case, since net wealth contains negative
values which change the codomain of the Gini.
Inequality revisited
19
$446,638 ($159,730) per adult.
6
Despite convergence between countries, existing differences in
levels between them remain pronounced (Shorrocks et al., 2020).
Figure 2-8: Global net wealth inequality
Gini coefficient and top net wealth shares (in per cent)
Source: Credit Suisse Global Wealth Databook, 2019, p. 143, Table 5-1
Similar to Credit Suisse, the Allianz also publishes a global wealth report focussing on the devel-
opment of gross and net financial assets at regular intervals (Allianz Research, 2020). Conse-
quently, only part of the total wealth of private households is considered and the methodolog-
ical challenges are the same as before. Nevertheless, Allianz also gives a positive assessment of
the development over the past two decades. For example, 600 million people have moved up
into the global wealth middle class and 2.5 billion people recently reached net financial assets
of around EUR 3,000. That is ten times more than at the turn of the millennium. As a key deter-
minant of this positive development, Allianz names open markets and free trade and highlights
the welfare increasing effect from globalization starting with the integration of former Soviet
Union member states and China 30 years ago.
The influence of the Corona pandemic on levels of net wealth and trends in wealth inequality
still remain uncertain for the time being. A first glimpse into possible consequences is given by
Shorrocks et al. (2020, p. 13): “The initial impact was felt through asset prices, causing global
household net worth to decline by USD 17.5 trillion during January-March 2020, a 4.4% reduc-
tion. Actions taken by governments and central banks then reversed this fall. By June, global
wealth was USD 1 trillion above the starting value. However, reduced GDP and rising debt will
6
The lower growth rate in Europe is partly due to the fact that the euro depreciated by 2,3 per cent against the US
dollar in 2019 (Shorrocks et al., 2020, p. 9).
0
10
20
30
40
50
60
70
80
90
100
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Gini top 10 per cent top 1 per cent
Inequality revisited
20
result in long-term damage, so wealth growth will be depressed for the next couple of years,
and likely longer.”
3 The case of Germany
The topics of inequality, wealth, and poverty have also frequently been the subject of heated
debates in Germany. This is exemplified by how often these terms have been used in speeches
in the plenary debates of the German Bundestag since 1991. The word poverty is an often and
increasingly used term and was mentioned around 8 times per 100,000 words in 2019. Certainly,
this was not always about income poverty and related combinations of the word are not in-
cluded (the ZEIT online tool only allows to use single words). In comparison, wealth was men-
tioned only 0.5 times per 100,000 words, while inequality was mentioned every 1.3 words per
100,000 words. Although poverty was frequently discussed in the Bundestag, other topics like
climate protection (Klimaschutz) were even more frequently used. After being highly debated
in 2007, climate protection has once again moved strongly into the centre of the debate and
was mentioned 16 times per 100,000 words in 2019. Thus, it occurred almost twice as often as
the topic of poverty in the same year.
Figure 3-1: This is what the German Bundestag is talking about
Number of mentions per 100,000 words
Source: ZEIT Online, 2020, 70 Jahre Bundestag: Darüber spricht der Bundestag | ZEIT ONLINE
0
5
10
15
20
25
0
2
4
6
8
10
12
14
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
Inequality Wealth Poverty Climate protection (right axis)
Inequality revisited
21
Unfortunately, the data series ends in the summer of 2019 and with the Corona pandemic it is
very likely that the questions of poverty and wealth will be discussed more frequently again.
Some parts of the left-wing parties are already considering a wealth tax or a one-time wealth
levy to finance the financial burdens of the pandemic. Although the financial consequences of
the Corona pandemic are still largely unclear and these claims are more political in nature, the
question remains how unequally income and wealth are distributed in Germany and how the
Corona pandemic will affect both dimensions.
3.1 Income inequality
As in many developed countries, net income inequality measured by the Gini coefficient is
higher in Germany today than it was in the 1990s.
7
As depicted in Figure 3-2, the level of the
Gini coefficient has increased from 0.25 in 1991 to 0.29 in 2017 according to household panel
data from the SOEP (Goebel et al., 2019). But the distribution of net incomes has not changed
uniformly over time: Between 1991 and 1999, the Gini coefficient initially varied between 0.25
and 0.26, despite the major upheavals after reunification. Between 2000 and 2005, the level of
income inequality increased, reaching a temporary peak of 0.29 Gini points in 2005.
Figure 3-2: Net income inequality over time
Gini coefficients of equivalized real disposable household income
Notes: The new OECD scale is used for equivalisation, dotted lines represent 95% confidence interval; Note on SOEP
data: *Time series break due to integration of subsample D (migration 1984-1994) and change in income retrieval/re-
cording; **Time series break due to integration of subsample M1 (migration 1995-2011).
Source: Stockhausen/Calderón, 2020
7
Net income (also known as disposable income) is defined as total market income (sum of gross earnings, self-em-
ployment income, capital income), plus the current private and public transfers received, less the taxes and social
security contributions paid. It also includes the imputed value of owner-occupied housing.
0.18
0.2
0.22
0.24
0.26
0.28
0.3
0.32
SOEP - Germany SOEP - Western Germany SOEP - Eastern Germany
Inequality revisited
22
However, inequality in net incomes remained almost unchanged since 2005, a year that repre-
sents a turning point in the development of income inequality in Germany. Thus, contrary to the
general perception, the so-called “Hartz” labour market reforms did neither lead to an increase
in income inequality nor in the low-wage sector. While the “Hartz” reforms were gradually im-
plemented between 2003 and 2005, the rise in net income inequality predominantly occurred
between the late 1990s and 2005. The “Hartz” reforms rather led to more flexibility in the labour
market and were not associated with a further increase in income inequality. While there were
initially signs of a slight decline in the Gini coefficient in the aftermath of the labour market
reforms, it moved slightly upward again in subsequent years. In the latest available income year
2017, the Gini coefficient was around 0.29 points according to the SOEP. Considering statistical
uncertainties that can be represented graphically by adding 95% confidence intervals, the data
does not suggest an increase in inequality in disposable incomes of private households in Ger-
many since 2005 (Stockhausen/Calderón, 2020).
Compared to other countries, Germany still exhibits a relatively equal distribution of net in-
comes and a high degree of redistribution via taxes and transfers is considerable large in Ger-
many. Values of the Gini coefficient vary between 0.25 in more egalitarian countries like the
Slovak Republic and 0.50 in more unequal countries like Costa Rica (see Figure 4-3). A regional
differentiation between Eastern and Western Germany also reveals that net incomes are more
evenly distributed in the Eastern part of Germany which includes the federal states of the for-
mer German Democratic Republic. However, absolute mean income levels are still significantly
lower in the East than in the West. Since most of the German population lives in the Western
federal states, the overall Gini coefficient is closer to the level of inequality in the West than to
the one in the East. Nevertheless, some convergence took place between both regions and over-
all trends have been quite similar in the past. With the exception that the rising trend in net
income inequality in Eastern Germany happened just until 2011 and did not reach its turning
point in 2005 as in the Western part of Germany.
Although the level of net income inequality is moderate, Germany is certainly not a role model
in every socio-political area and reveals some deficits in inequality-related areas. For example,
pupils are segregated relatively early in their school careers and despite some progress in the
last years there is still less early childhood support in kindergarten compared to other OECD
countries, which particularly limits educational opportunities of children from less well-off fam-
ilies. This is illustrated in a rather low share of university accesses from children from less-privi-
leged families. However, Germany manages to compensate for some of these disadvantages.
To stay with the example: Thanks to the dual vocational training system, which links theory and
practice, it is not always necessary to have a university degree to pursue a well-paid profession.
A career as a skilled worker is a good alternative to a bachelor's degree, especially for people
with greater practical talents. Advanced training such as master craftsmen and technicians of-
fers these dual-qualified individuals very good career opportunities, and many master craftsman
qualifications are accompanied with income opportunities similar to those offered by a univer-
sity degree. This is also a major reason why Germany performs rather poorly in terms of educa-
tional mobility and why the educational success of children depends more on that of their par-
ents than in the US or the United Kingdom (UK) (OECD, 2018; Neidhöfer/Stockhausen, 2018).
Inequality revisited
23
However, Germany shows significantly better results in terms of income mobility, which is often
used as an indicator of equal opportunities. According to comparative studies by Corak (2016),
Germany ranks slightly above average among industrialized nations. In addition, it is more mo-
bile in terms of labour income than the US in both absolute and relative terms (Schnitzlein, 2016;
Stockhausen, 2018a).
Surprisingly, the OECD's (2018) most recent findings have shown Germany to be less mobile.
However, these results are striking outliers compared to other existing literature. This is mainly
due to the fact that only dependent employees were considered in their main analysis and that
quite restrictive income definitions were applied in measuring permanent incomes of parents
and children. Accordingly, these results were sharply criticized in Germany (Hufe et al., 2018;
Stockhausen, 2018b). Overall, Germany is characterized by a comparatively low level of net in-
come inequality and an above average degree of income mobility measured by the intergener-
ational elasticity between father and son earnings. The relationship between both measures got
famous as the so-called Great Gatsby Curve. It generally shows that there is an inverse relation-
ship between inequality and mobility. In countries with low-income inequality, intergenera-
tional income mobility is higher and vice versa Corak (2016).
Figure 3-3: The Great Gatsby Curve
Intergenerational elasticity between father and son earnings; Gini coefficient of equivalized net incomes
Intergenerational Earnings Elasticity
Note: The higher the elasticity coefficient, the lower the income mobility. The higher the Gini coefficient, the higher
the income inequality.
Source: Corak, 2016
Inequality revisited
24
3.2 Wealth inequality
Although net wealth is generally more concentrated than disposable income, net wealth ine-
quality has also remained comparatively stable over the past decade in Germany (Figure 3-4).
The Gini coefficient of net wealth has been varying around 0,78 since 2002, if information on
individual net wealth is used from the SOEP. At the household level, net wealth inequality is
somewhat lower and is showing a downward trend rather than an upward trend. It slightly de-
creased from around 0,76 in 2002 to less than 0,74 in 2017. Similar results persist if micro data
from the German Bundesbank (PHF) or from the so-called “Einkommens- und Ver-
brauchsstichprobe” (EVS) are used. Both datasets provide wealth information on the household
level and have different limitations (see Stockhausen/Calderón (2020) for more details).
The stable trend in net wealth inequality is observed in a period of low interest rates and rising
asset prices, which mainly resulted from the loose monetary policy after the financial and eco-
nomic crisis in 2008/2009. In particular, the value of owner-occupied real estates has strongly
increased during the last ten years, especially in urban areas. Since owner-occupied real estate
is the major wealth component of the middle class, they benefited relatively strongly from rising
real estate prices. This has contributed to a stabilization of the net wealth distribution over time
in Germany (Deutsche Bundesbank, 2019).
8
However, all available micro data sets with infor-
mation on wealth suffer from the fact that they often fail to capture the top of the wealth dis-
tribution adequately. Therefore, there have recently been numerous efforts to add missing
wealth information at the top. Information from rich lists is used for this purpose mostly. In
general, this increases net wealth concentration among the top (see Westermeier/Grabka
(2015) or Bach et al. (2019) for earlier attempts).
A unique top-wealth sample was recently collected for Germany as part of the SOEP (Schröder
et al., 2020). This allows a detailed analysis of the wealthiest Germans for the first time. Unlike
in the middle of the net wealth distribution, business assets play a greater role at the top and
are a major source of wealth for the very rich. The wealth share of the top 10 per cent (top 1
per cent) increases from 58,9 per cent (21,6 per cent) to 67,3 per cent (35,3 per cent) in 2018
when both information from the top wealth sample and information from rich lists is used.
These results are as expected and are mostly in line with earlier work by Westermeier/Grabka
(2015) which shows similar changes in net wealth shares by adding information from rich lists.
In fact, even when adding top-wealth information – on a higher level – the development of net
wealth inequality remained rather stable for the years from 2002 to 2012.
Furthermore, the deficits in the wealth data do not only relate to the top. There are also regular
discrepancies between the wealth aggregates from micro data and from national accounts data.
This is due, on the one hand, to difficulties in defining individual wealth components and, on the
other hand, to the fact that some wealth components are simply under-reported in the micro
8
Research from the ECB (2021) also points into the direction that the easing of monetary policy through the ECB’s as-
set purchase programme (APP) is not associated with an increase in net wealth inequality. Increased house prices
and lower debt burdens, which are relatively more important in the wealth portfolios of the middle and lower clas-
ses, are likely to have counteracted the inequality-increasing effect of increased financial asset prices, which are
relatively more important in the net wealth portfolio of the upper wealth classes.
Inequality revisited
25
data. Examples are financial assets including, for example, private insurance assets, where only
40 per cent of the actual amounts are recorded in the micro data compared to national account
data. And it is these asset classes, which are particularly owned by the middle class.
Figure 3-4: Net wealth inequality in Germany
Gini coefficient
Notes: In the SOEP, persons aged 17 years and older in private households are considered (excluding persons from the
refugee samples M3 to M5), weighting factors include the first wave of the survey. In EVS and PHF, the net wealth dis-
tribution is determined at the household level. Inequality at the household level tends to be lower.
Source: Stockhausen/Calderón, 2020
Albers et al. (2020) are the first to ensure that the wealth aggregates of individual components
match between the micro and macro data, in addition to using information from rich lists to
correct the distribution of wealth at the top. With wealth shares of less than 25 per cent, result-
ing top 1 per cent shares are substantially lower than previous adjusted top wealth shares. Using
the share of total net wealth held by the top 1 per cent, they show that today's share of the top
1 per cent is much lower than in the early 20th century and at somewhat the same level as in the
mid-1960s. And here, too, there has been no clear change in the share since the financial crisis.
However, even this very comprehensive approach does not consider the impact of statutory
pension insurance. Although it is debatable whether this is a classic asset type, since the enti-
tlements are not freely transferable and cannot be liquidated at any time, state pension
schemes exert an influence on savings possibilities and diminishes incentives to save privately.
0.60
0.62
0.64
0.66
0.68
0.70
0.72
0.74
0.76
0.78
0.80
0.82
0.84
EVS
PHF
SOEP - individual level
SOEP - individual level (incl. cars/student loans)
SOEP household level
Inequality revisited
26
Thus, differences in the architecture of welfare states make international comparisons difficult,
as in many cases apples are compared with oranges. Among other things, this also leads to Ger-
many having comparatively high inequality in net wealth and low levels of median net wealth.
Irrespective of whether the top of the wealth distribution is correctly covered or not. In this
context, a paper by Bönke et al. (2019) shows that the Gini coefficient of net wealth decreases
by about one third, if claims against the statutory pension insurance system are included as an
asset. In general, pension wealth makes up to 61 per cent in Germany (compared to 48 per cent
in the US) and the equalizing effect is larger in Germany than in the US. Furthermore, higher
levels of net wealth inequality are rather typical for countries with generous welfare states, high
living standards, and comparatively low-income inequality: Accordingly, net wealth inequality is
also rather high in Scandinavian countries such as Sweden or Denmark with Gini coefficients of
respectively 0,867 and 0,838 (s. Figure 3-5).
Figure 3-5: Net wealth inequality by country
Gini coefficient, 2019
Source: Credit Suisse Global Wealth Databook, 2019
3.3 Middle class
As an anchor of stability between rich and poor, the middle class is often seen as a gauge of
societal and social cohesion. Especially in Germany the development of the middle class has
always gained special accordance. (Media) Reports on an eroding middle class are, therefore,
seen as worrying. However, even though the middle class is often cited in the political and public
spheres, it is by no means a self-explanatory term, nor is there a binding definition for the "mid-
dle class." Rather, it can be described in terms of different dimensions, such as socio-cultural,
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Slovakia
Belgium
Malta
Croatia
Romania
Greece
Bulgaria
Slovenia
Hungary
Lithuania
Italy
Poland
Portugal
Spain
France
Switzerland
Czech Republic
Austria
Finland
United Kingdom
Ireland
Norway
Cyprus
Germany
Denmark
United States
Sweden
Russia
Netherlands
Inequality revisited
27
financial, or subjective and value-oriented characteristics.
9
Most economic studies focus on a
purely income-related definition, whereas in social science, socio-demographic criteria such as
education, employment or concepts on subjective orientations and values dominate. So called
“milieus” combine vertical socio-economic status variables with common lifestyles and values
as a second dimension. This allows to define rather homogenous social groups, with similar val-
ues, perceptions, and a common identity. However, lifestyle and value orientations change over
time and, therefore, it is difficult to analyse the temporal development of such groups. Thus, to
consistently analyse the size of the middle class over time, a simplifying structuring of the soci-
ety is required. For this purpose, in the following we rely on a definition solely based on dispos-
able incomes, since this variable represents a central social status characteristic in which many
socio-cultural features such as education and employment status are reflected.
Income strata are usually defined in relation to the median income of the respective society.
That is the income that the population divides exactly in half: One half has a higher income, the
other half has a lower income. However, the boundaries between the lower income class, mid-
dle income class, and the rich cannot be unambiguously defined on the basis on income alone.
With the help of a multidimensional approach, it is possible to establish meaningful income
boundaries. To this end, first a socio-cultural middle class is defined and then it is examined
which income ranges households with typical middle class educational qualifications and occu-
pations predominantly occupy (Niehues et al., 2013). The distribution of the socio-cultural mid-
dle class yields an income-based definition that divides society not into the poor, the middle
class, and the rich – but into five groups: The at-risk-of-poverty group (below 60 per cent of
median income), the low-income or "lower" middle class (60 to 80 per cent of median income),
the middle class in the narrow sense (80 to 150 per cent of median income), a high-income or
"upper" middle class (150 to 250 per cent of median income), and the relative income rich (more
than 250 per cent of median income). In addition to providing additional social differentiation,
the use of five rather than three income groups also has the advantage that the broad definition
of 60 to 250 per cent provides a kind of upper bound for the middle class and the narrow defi-
nition of 80 to 150 per cent provides a lower bound compared with other income thresholds
often used in the literature.
Figure 3-6 illustrates the development of different income groups according to this income
strata definition. The analysis is based on household net incomes after the deduction of taxes
and social contributions plus state pensions and social transfer payments. In addition to labour
income, capital and property incomes are also taken into account, as are imputed rents of
owner-occupied housing. Furthermore, as common in distributional analyses, household in-
come is equivalised to account for different household sizes and economies of scale within
households. In Germany, 1991 is an obvious starting point for distributional analyses since the
reunification constitutes a striking structural break. Only from this year on harmonized micro
data are available for unified Germany. Since then, the development of the middle class can be
divided into three phases: During the East German catch-up process, the share of the middle
9
See Niehues et al. (2013) and Niehues (2017a) for a more detailed discussion about the definition of the German
middle class.
Inequality revisited
28
class in the narrow sense initially increased from 50.4 to 54.7 per cent until its peak in 1997. By
2005, its share had fallen again to 50.1 per cent. At the same time, the share of those at-risk-of-
poverty and of those with relatively high incomes has risen. The shrinking middle class is mir-
rored by a rise in inequality: The Gini coefficient of net income rose from 0.25 to 0.29 over the
same period (see Section 3.1). However, the decline of the middle class is by no means a con-
tinuous process. For more than one decade now, the stratification has changed only marginally:
The share of the population in the middle class in the narrow sense equals 49,4 in 2017 which
is very close to the middle class share in 2005. Additionally, it should be noted that the discern-
ible decline in the middle class between 2012 and 2013 is largely due to an additional migration
sample whose respondents are predominantly located in the lower income range.
10
All in all,
given the relative concept of the middle class definition, the development of the income groups
is unsurprisingly very similar to the development of income inequality.
Figure 3-6: The development of the middle class in Germany
Share of population, in per cent
Notes: Income classes in relation the median of nominal equivalised net incomes of the respective year (median in-
come in 2017: 1.946 Euro). *Time series break due to integration of subsample D (migration 1984-1994) and change in
income retrieval/recording; **Time series break due to integration of subsample M1 (migration 1995-2011).
Source: SOEP v35
10
See also Goebel et al. (2015, p. 582 f.). Additional migration samples are necessary to better reflect immigration,
which is naturally underrepresented in long-time panel studies. However, the majority of respondents in this migra-
tion sample immigrated to Germany before 2005 – so while the level of inequality may have been underestimated
before 2005, the timing of the structural effect on income stratification is questionable (see Niehues, 2017b, for a
further explanation).
13 12 11 12 11 11 10 10 11 12 12 13 13 14 14 14 14 15 15 14 15 14 15 16 16 16 16
18 19 18 17 17 18 17 18 17 17 17 17 17 16 17 17 17 16 16 18 17 18 17 16 16 16 16
50 52 53 53 54 55 55 54 55 55 54 52 52 51 50 51 50 50 50 49 48 49 47 48 48 48 49
16 15 16 15 14 14 15 15 15 14 14 15 15 15 15 15 15 16 16 15 16 15 17 17 17 17 15
2 3 3 3 3 3 3 3 3 3 3 3 3 3 4443334444333
0
10
20
30
40
50
60
70
80
90
100
Relative Poor Lower middle class Middle income class Upper middle class Relative rich
Inequality revisited
29
Besides the development of the size of the middle class, changes of middle incomes are often
in the focus of the discussion. In the context of the aforementioned “elephant curve” (see sec-
tion 2.1), it is often discussed that the minimum of the growth-incidence-curve represents
merely stagnating incomes in developed countries middle classes. In particular, Germany is of-
ten in the focus of the discussion. Therefore, Figure 3-7 represents the development of dispos-
able income in five exemplary income groups. For the bottom 20 per cent of households (1st
and 2nd decile), the middle-income group (5th decile) and the top 20 per cent (9th and 10th
deciles). Annual changes are indexed to the 1994 income year. The choice of the base year re-
duces the effect of the time-series break by the additional samples D1 and D2, which is particu-
larly pronounced for the lowest decile.
11
Consequently, across all income groups considered,
there is an increase in real disposable household incomes. However, the extent of the raise var-
ies considerably across income groups. While disposable household incomes in the 1st decile
increased by 6 per cent in real terms, they rose by around 9 per cent in the 2nd decile. In the
9th and 10th deciles, average disposable household income rose by 21 and 31 per cent respec-
tively.
Figure 3-7: Change of disposable household income by income deciles
Index: 1994 = 100; equivalized using the new OECD scale; decile means
Notes: *Time series break due to the integration of sub-sample D (migration 1984-1994) and change in income ques-
tion/recording; **Time series break due to the integration of sub-sample M1 (migration 1995-2011).
Sources: SOEP v35; own calculations
The comparatively low growth in the lowest income groups results in particular from a negative
development between 1999 and 2005, a period which was characterized by a sharp increase in
11
See Stockhausen/Caldéron (2020) for a further discussion.
80
90
100
110
120
130
140
1. Decile 2. Decile 5. Decile 9. Decile 10. Decile
Inequality revisited
30
unemployment and comparatively low economic growth. If we focus on the more recent devel-
opment of incomes since 2010, it becomes clear, that middle-income groups experienced the
largest increases in incomes (7,2 per cent increase in real incomes between 2010 and 2017).
With respect to low-income groups, the illustration reveals a marked decline since 2010. While
the remarkable decrease between 2012 and 2013 is, again, related to the inclusion of the addi-
tional migration sample (Niehues, 2017b), the overall development also roots in some structural
changes of the society which are discussed in the following chapter.
It is, however, important to note that the graphical representation of income development in
Figure 3-7 follows an anonymous approach, meaning, that people in the first decile in one year
are not the same people who are in the first decile in another year. In fact, the longitudinal
design of the SOEP allows to also investigate how the composition of income groups has
changed over time. To this end, only those respondents of the SOEP are considered, which take
part in the survey in all years between 2009 and 2017. Figure 3-8 suggests that around 60 per
cent of the people who have been initially in the lowest quintile have left the group of the lowest
20 per cent after eight years. Persistence is slightly higher in the upper quintile with a share of
about 60 per cent who do still belong to the highest quintile group after eight years. Changes
between income deciles can also be regarded as an indicator of income mobility.
Figure 3-8: Income mobility across quintiles
Equivalent household disposable incomes (modified OECD scale); Germany, 2009-2017
Notes: Q1-09: Fifth (quintile) with the lowest incomes in 2009, Q1-17: Fifth (quintile) with the lowest incomes in 2017.
Reading example: 40 per cent of the people who have been initially in the lowest income quintile in 2009 (Q1-09) have
remained in the lowest income quintile after eight years (Q1-17). Likewise, 26 per cent of the people who have been
initially in the lowest quintile group in 2009 (Q1-09) have moved to the second income quintile in 2017 (Q2-17). 7 per
cent of the people who belonged to the lowest quintile in 2009 (Q1-09) have been able to move to the highest income
quintile in 2017 (Q5-17).
Sources: Stockhausen/Calderon, 2020, Figure 5.1 (on basis of SOEP v35)
40%
39%
8%
9%
4%
26%
34%
25%
10%
5%
11%
14%
39%
24%
12%
17%
10%
19%
34%
20%
7%
2%
9%
22%
59%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Q1-09
Q2-09
Q3-09
Q4-09
Q5-09
Q1-17 Q2-17 Q3-17 Q4-17 Q5-17
Inequality revisited
31
3.4 Determinants of inequality
Income and wealth inequalities are caused by many factors and originate from different sources.
Thus, the effect of each factor on the macro and micro level is difficult to identify and many
factors depend on each other. On the one hand, macroeconomic developments like business
cycles, technological progress, but also political power relations determine the functional distri-
bution between labour and capital. On the other hand, individual decisions, and institutional
circumstances, among others, influence the distribution of personal income, whereby a distinc-
tion must be made between market, gross, and disposable income. For example, the amount of
labour earnings depends on individual decisions on the labour market, which in most cases can
be determined by the individual. These include decisions on occupational choice or working
hours. However, institutional factors also play a role, which the individual can only partially con-
trol or cannot significantly influence at all. For example, the availability of public childcare de-
termines how much working time (single) parents can offer at the labour market. Also questions
about the design of the welfare system and the extent of redistribution from high-income to
low-income earners are part of the equation. Other factors that cannot be affected individually
include the family into which one was born as a child, one's talents and how the family was able
to foster these talents. All these factors – and the list is not exhaustive – have an impact on
market outcomes and ultimately on disposable incomes.
First, we want to investigate the claim that a growing share of national income is going to capital,
less to labour, and that market income inequality has risen as a result. Figure 3-9 shows that the
labour income share, namely the share of compensation of employees in national income, var-
ied between 70 and 72 per cent between 1991 and 2003, before it decreased to just under 64
per cent at the beginning of the financial and economic crisis in 2008. Due to the collapse of the
capital markets and correspondingly lower capital incomes, the labour income share rose again
in the following years to around 68 per cent and remained at this level until 2016. Since 2016, it
has risen to around 72 per cent according to the latest, revised data from national accounts and
is back at its 1990 level.
In comparison, the distribution of market incomes has developed in a notably different way.
While the labour income share remained largely stable in the 1990s and varied around a level
of about 70 per cent, inequality in market incomes increased significantly during the same pe-
riod. Then in 2003 the labour income share started to decline sharply until the beginning of the
financial crisis in 2007. Market income inequality just started to decrease in 2005 and continued
to do so until 2009, when the labour income share was already on the rise again. The financial
crisis in 2007/2008 marked a turning point in the evolution of the labour income share. As cap-
ital income was particularly affected by the crisis, the share of labour income initially rose
sharply between 2007 and 2009, but also corrected downward again between 2009 and 2011,
without, however, falling back to its lowest level in 2007 (around 64.5 percent). An increase
from 2011 to 2012 was followed by a prolonged sideways movement until 2016, before the
labour income share most recently increased to around 72 percent in 2019, a level similar to
that in the 1990s. In contrast, market income inequality did also respond to the financial crisis
by showing an increasing trend between 2009 and 2013 but has decreased since then. Overall,
Inequality revisited
32
the data does not support the hypothesis that market income inequality decreases whenever
the labour income share is increasing.
Figure 3-9: Evolution of labour income share and market income inequality in
Germany
Gini coefficient of equivalent household market incomes (new OECD scale)
Notes on SOEP data regarding market incomes: *Time series break due to the integration of sub-sample D (migration
1984-1994) and change in income question/recording; **Time series break due to the integration of sub-sample M1
(migration 1995-2011).
Sources: Stockhausen/Calderón, 2020; VGR des Bundes, Fachserie 18, Reihe 1, 2, 3, Vierteljahr 2020
The level and development of market income inequality is also often associated with the size of
the low-wage sector in Germany and thus with the distribution of gross hourly wages. In EU-
wide comparison, the low-wage sector in Germany is relatively large. Although its size has de-
clined slightly in recent years, still around one in five employees work in the low-wage sector
(Fedorets et al., 2020). However, since low-wage incidence is measured in relation to national
earnings, it also noteworthy that the general wage level in Germany is relatively high compared
to other EU countries (Figure 3-10). Germany ranks 5th out of the 28 EU member states regard-
ing the level of mean hourly earnings adjusted for purchasing power. While mean hourly earn-
ings amounted to €17,52 in Germany in 2014, it was €19,85 in Denmark (highest value), €16,44
in Sweden, €15,44 in the United Kingdom or €4,90 in Bulgaria (lowest value). Similar patterns
hold for median hourly earnings at lower levels. Looking at the threshold to the lowest earnings
decile, however, Germany performs somewhat worse, ranking only 10th out of 28.
Besides, the German unemployment rate is comparatively low. In fact, in no EU member state
the youth unemployment rate is lower than in Germany. The observation that the unemploy-
ment rate of low-educated is also comparatively low, hints at a plausible trade-off between the
unemployment rate of low-skilled workers and hourly wages, implying that some of those who
60%
62%
64%
66%
68%
70%
72%
74%
0.300
0.350
0.400
0.450
0.500
0.550
1991
1992
1993
1994*
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013**
2014
2015
2016
2017
2018
2019
Share
Gini coefficient
Market income Labour income share
Inequality revisited
33
receive low hourly earnings in Germany are rather unemployed in other countries. Unlike the
contrasts with national accounts data, the comparison of changes in market income inequality
and gross hourly wages over time reveals slightly more similar trends but there are still differ-
ences. Inequality in gross hourly wages increased from the end of the 1990s until 2005/2006
(Figure 3-11). Since then, it first decreased merely in the lower half of the distribution and from
2013 onwards among the whole distribution (Fedorets et al., 2020). From 2015 onwards, the
downward trend is even stronger, which is likely to be due to the introduction of the statutory
minimum gross wage of €8,50 per hour. Although inequality in gross wages has, thus, declined
recently, inequality in market incomes as well as in disposable incomes has remained relatively
stable.
Figure 3-10: Hourly earnings adjusted for purchasing power in an EU comparison
Industry, construction, and services (excluding public administration), 2014
Source: Eurostat (Lohn- und Gehaltsstrukturerhebung)
The introduction of the minimum wage in Germany in 2015 did obviously not lead to an unam-
biguously lower level of market income inequality and the positive trend in gross hourly wage
inequality, which has already started before its introduction, did not automatically or fully trans-
late into lower market income inequality. This is likely due to an adjustment of working hours
as Grabka/Schröder (2018) can show. Monthly wages in the lowest decile hardly increased at all
after the introduction of the minimum wage because working hours among low-wage earners
fell. Thus, while the minimum wage has contributed to a significant increase in hourly wages in
some areas and occupations, especially in several parts of Eastern Germany and in low-wage
sectors, it has had little overall impact on the distribution of market and disposable incomes.
0
4
8
12
16
20
Denmark
Ireland
Luxembourg
Belgium
Germany
Sweden
Netherlands
France
Finland
United Kingdom
Austria
Italy
Spain
Cyprus
Malta
Greece
Slovenia
Poland
Portugal
Croatia
Czech Republic
Hungary
Slovakia
Estonia
Latvia
Lithuania
Romania
Bulgaria
In PPPs
Mean Median 1st decile
Inequality revisited
34
Moreover, Schröder/Kestermann (2020) point out that an increase in the minimum wage would
not necessarily lead to a reduction in the at-risk-of-poverty rate because the net income of a
single person working full-time at the present minimum wage level is already very close to the
at-risk-of-poverty threshold. Indeed, in most European countries full-time workers that receive
a wage as high as the national minimum wage already exceed their national at-risk-of-poverty
thresholds. Moreover, they highlight the importance of low working hours for falling below the
at-risk-of-poverty threshold.
Figure 3-11: Inequality of agreed gross hourly wage in main job
Percentile ratios (P90/P10)
Note: Gray areas show 95%-confidence intervals.
Source: Fedorets et al., 2020
Furthermore, another factor influencing the trends in market and disposable income inequality
is the increasing number of single households in Germany, which tends to raise inequality in
disposable incomes, as advantages through economies of scale within households cannot be
exploited (Peichl et al., 2011). This development is not only related to a society that relies more
on individualism, but also to an aging society. A third factor that drives inequality and counter-
acted decreasing inequality trends in recent years is the migration from Eastern Europe to Ger-
many since 2010 as well as the influx of refugees from 2015 onwards. This is reflected, among
other things, by the fact that at-risk-of-poverty rates are higher for persons with a migration
background and that poverty risks have increased significantly among this group in recent years.
Meanwhile, the share of low-income earners without migration background has been constant
or has even declined in some age groups (Grabka/Goebel, 2020). In this regard, counterfactual
analyses on behalf of the 6th poverty and wealth report of the Federal Government show, that
Inequality revisited
35
the observed employment gains since 2005 on its own would have resulted in lower inequality
levels (Kleimann et al., 2020, p. 275 ff.). It further reveals that the development of inequality in
Germany from 2005 onwards was mainly determined by changes in the composition of socio-
economic characteristics of the society.
3.5 COVID-19 and income inequality
In view of the Corona pandemic, which has already left deep marks on the economy and society,
the question of distributional effects is once again at the centre of attention. With respect to
the worldwide development, analyses project that the pandemic will likely increase inequality
and poverty since job losses could disproportionally affect the income and labour participation
of low-skill workers (see section 2.2). However, the impact of the pandemic also depends on the
measures taken by the government to absorb negative effects of the crisis. To avoid an eco-
nomic and social collapse, the German government decided on the first extensive aid packages
at the end of March 2020. In addition to simplified access to basic social assistance benefits or
the strengthening of short-time working allowances (Kurzarbeitergeld), extensive unconditional
financial aid (Überbrückungshilfen), loans and concessionary credits were made available to ail-
ing companies. With respect to the beginning of the COVID-19 pandemic, this prevented a rapid
rise in unemployment and a wave of corporate insolvencies.
Although the distributional effects of the COVID-19 pandemic are of huge interest for the public
debate, it is yet difficult to determine its effects on basis of available data. To gain insight into
what impact the pandemic had on the economic situation of people in Germany, the German
Economic Institute (IW) commissioned an online survey in August 2020 in which 1,202 people
were asked about their changes in income as a result of the pandemic and about how they had
been affected by short-time work (IW survey). In a second step, individuals from the IW survey
were matched to statistical twins in the SOEP using Mahalanobis distance matching. The infor-
mation from the matched IW survey observations is then used to simulate crisis-induced income
and status changes in the SOEP population with the help of a microsimulation analysis. Govern-
mental measures to combat the negative consequences of the lockdown are modelled insofar
they had been enacted by the end of November 2020. Resulting taxes and transfers are calcu-
lated using the IW's Tax and Transfer Microsimulation Model (STATS).
12
The vehemence of the COVID-19 pandemic becomes obvious when first considering its impact
on market incomes, meaning incomes before taxes, transfers, and statutory pensions.
13
On av-
erage, monthly market incomes per capita have fallen by six per cent compared with 2019 – the
comparative incomes without the effects of the crisis (Figure 3-12). Individuals in the bottom
half of the income distribution suffered the greatest losses in relative terms. In the lowest in-
come decile, per capita market incomes fell by an average of 12 per cent. The middle-income
decile was also hit hard, losing an average of 9 per cent of its market income due to unemploy-
ment, short-time work or a lack of profit income from self-employment or from capital income.
People from the upper part of the income distribution also suffered losses. In absolute terms,
12
See Beznoska (2016) for a detailed description of the IW Microsimulation Modell STATS.
13
The following results merely base on the detailed simulation analysis in Beznoska et al. (2020).
Inequality revisited
36
these losses were higher than in the lower half, but as a proportion of total income they were
smaller in the top decile (around 4 per cent) than in the other income groups. The overall pic-
ture, thus, shows that the crisis affected all segments of the population, but to different degrees.
Considering disposable incomes reveals the compensatory effect of social security systems,
which were strengthened during the crisis by additional temporary assistance payments such as
the child bonus. As a result, the relative losses in the middle and at the bottom of the income
distribution were reduced. Simplified access to basic social assistance benefits, the increase in
the tax allowance for single parents, raising statutory pensions and the expansion of housing
benefits and the child supplement – which were implemented independently of the Corona cri-
sis in 2020 – even result in a slight nominal increase in disposable household incomes for the
1st and 2nd income deciles. It should be noted, however, that this result is based on the as-
sumption that 100 per cent of social benefits are claimed.
Figure 3-12: Changes in household income through COVID-19 pandemic
Deciles of equivalent household net incomes in 2019, changes compared to 2019 in per cent
Notes: KUG = Kurzarbeitergeld (short-time allowance).
Source: Beznoska et al., 2020 (on basis SOEP v35 and IW-Survey)
Also, middle incomes benefited from welfare redistribution and assistance measures, in partic-
ular from short-time working benefits. While market incomes in the 5th decile are reduced by
an average of 9 per cent, the loss in disposable income is much smaller, averaging 0.7 per cent.
Without the short-time allowance, the loss would have been around 2 per cent. Looking at the
upper income groups, we see that while high earners experienced relatively the smallest losses
in market incomes during the crisis, they experienced the largest percentage losses in disposa-
ble incomes, as aid measures have a lower relative impact in upper income deciles.
1.4 1.3
0.0 -0.4 -0.7 -0.8 -1.0 -1.2 -0.8 -1.4 -0.7
1.4 1.0
-0.9 -1.6 -2.3 -2.2 -2.0 -3.1
-2.1 -2.0 -1.8
-12.4 -13.3 -12.2
-9.4 -9.0
-7.2
-6.0 -6.2
-4.1 -3.5
-6.0
-16
-14
-12
-10
-8
-6
-4
-2
0
2
4
1st decile 2 3 4 5 6 7 8 9
10 th
decile Total
Disposable Income Disposable Income without KUG Market Income
Inequality revisited
37
Figure 3-13 illustrates the simulated impact of the COVID-19 pandemic on income inequality as
measured by the Gini coefficient. When first considering income before governmental interven-
tion the results reveal that the Gini coefficient of market incomes increased from 0.510 in 2019
to 0.525 in the Corona year 2020. However, only considering market incomes would ignore the
equalizing effect of the welfare state, whose primary goal is to protect against different life risks.
In fact, the described income changes do not imply an increase in inequality in disposable in-
comes. Similarly, to the times of the financial crisis, there is even a weak decline in the Gini
coefficient from 0.293 in 2018 to 0.289 in 2020. Thus, disposable household incomes are not
only fundamentally more equally distributed than market incomes, but the simulation analysis
also suggests that the distribution of disposable incomes – at least in a short-term perspective
– is also not expected to become more unequal through the impact of the COVID-19 pandemic.
These general effects on inequality also hold if other inequality measures are used.
Figure 3-13: Changes in net and market income inequality through COVID-19
pandemic
Gini coefficient of equivalent household incomes (modified OECD scale)
Notes: KUG = Kurzarbeitergeld (short-time allowance); whiskers represent 95 per cent confidence intervals (bootstrap
procedure, n=100); the addition with/without virus indicates whether Corona-related income changes were modelled
or not.
Source: Beznoska et al., 2020 (on basis SOEP v35 and IW-Survey)
Obviously, the simulation results must be interpreted with the necessary methodological cau-
tion and only represent the impact of the pandemic as they were captured by the IW survey in
August 2020. However, the results represent an initial estimate of the general distributional
consequences of the Corona pandemic, and they show the effectiveness of the German social
system and importance of the prompt reaction by the government. The robustness of the results
2019
without
virus
2019 with
virus
2020 with
virus
0.495
0.5
0.505
0.51
0.515
0.52
0.525
0.53
0.535
0.54
2017 2018 2019 2020 2021
Market incomes
2019
without
virus
2019 with
virus 2020 with
virus
without
KUG
2020 with
virus
0.28
0.285
0.29
0.295
0.3
0.305
2017 2018 2019 2020 2021
Net incomes
Inequality revisited
38
is corroborated by a simulation analysis by Bruckmeier et al. (2020), who, using a different meth-
odological approach, also conclude that an increase in disposable income inequality is not to be
expected in the crisis year 2020.
14
The long-term effects of COVID-19 will especially depend on
the extent to which employment can be further maintained and the previous growth path may
be reached.
4 The inequality-growth-nexus
Discussions about the development of inequality often implicitly assume that lower inequality
levels are generally preferable. While this is plausible in a ceteris paribus perspective, theories
also suggest that smaller differences in income may imply working disincentives because asso-
ciated income gains are expected to be small. In this context, the optimal level of inequality
becomes a crucial point of the discussion which is conventionally analysed in the context of the
inequality-growth nexus. In fact, the relationship between inequality and growth regained re-
newed interest when the IMF and the OECD closely in time published two studies on this topic
in the year 2014. The analyses both came to the result that increasing inequality is accompanied
by lower economic growth and beyond, that raising redistribution will have no negative effects
on growth (Cingano, 2014; Ostry et al., 2014, later published as Berg et al., 2018).
In particular, the graphical representation of the main results of the OECD study, illustrated in
Figure 4-1, gained a lot of media attention. For Germany, the results imply that economic growth
between 1990 and 2010 could have been larger by 6 percentage points if there had been no
increase in inequality. The results are particularly attention-getting because they question the
widely assumed equity-efficiency trade-off.
15
Theoretically, there are likewise reasons to expect
positive effects from inequality on growth as well as a negative relationship. The positive rela-
tionship bases most notably on the idea that income differences build the ground for incentives
and innovations. A negative effect can be expected, when people with low incomes have no
access to the education system and, thus, preventing them from realizing their optimal devel-
opment of educational opportunities.
A growth-inhibiting effect can also arise if the level of inequality is so high that it is accompanied
by social unrest and political instability. The theoretical channels between inequality and growth
suggest that the impact depends centrally on country-specific characteristics. For example, a
negative effect is more likely in poor countries with comparatively low living standards, where
many people are denied access to the education system. Similarly, inequality-related, and
14
Early simulation studies by Brunori et al. (2020) for Italy or BMSGPK (2020) for Austria find similar patterns regarding
the distributional effects of the Corona pandemic.
15
This newly sparked facet of the distribution debate sometimes gave the impression that economists had only just
begun to devote attention to the relationship between inequality and growth. Yet this very link was already the
subject of a large number of empirical analyses in the 1990s. A meta-study by Neves et al. (2016) shows that the
estimated coefficient for the effect of increasing inequality on economic growth ranges in a fairly wide range be-
tween -0.135 and 0.156 percentage points. The effects found follow a temporal cycle with rather negative coeffi-
cients in the 1990s, positive effects at the beginning of the new millennium and then again more negative effects
starting in the 2010s.
Inequality revisited
39
growth-inhibiting social unrest and political instability should be more likely in countries where
the level of inequality is already comparatively high.
Figure 4-1: Impact of inequality changes on economic growth – OECD results
GDP per capita growth in per cent
Notes: The results for the Scandinavian countries are highlighted to hint at the combination of low inequality levels
and yet, comparatively large supposed negative impacts on economic growth.
Source: OECD Focus on Inequality and Growth, 2014
Against the background of these considerations, the results of the OECD study (as illustrated in
Figure 4-1) are rather surprising. The findings of the study indicate that the negative effects on
growth which go back to rising inequality were noticeably larger in Scandinavian countries such
as Sweden, Finland, and Norway than in the USA, for example. From a theoretical viewpoint this
is surprising in so far, as the United States belong to the group of OECD countries with a partic-
ularly high concentration of income. Whereas the Scandinavian countries, especially Norway,
are characterized by comparatively low inequality levels. Beyond, Scandinavian countries regu-
larly perform very well in analyses of educational mobility. One would, therefore, expect that
increasing inequality in these countries is less harmful on growth because the educational und
political unrest channels are less likely to be at work. However, since the OECD study only con-
siders linear effects of inequality on growth, the initial level of inequality does not play any role
within this analysis.
If non-linear relationships are also considered, it becomes apparent that the negative effect of
increasing inequality on economic growth crucially depends on the initial level of inequality
(Kolev/Niehues, 2016). According to a global comparison of 113 countries, up to a value of the
Gini coefficient of 0.35, a positive correlation between inequality and growth can be assumed.
4
-9 -5 -7 -9 -9 -2 -5 -3
-16 -6 -6 -1 -3 4-6
6
-7 -11
- 20
- 10
0
10
20
30
40
50
60
70 Without impact of inequality
Impact of inequality
Actual (↘)
Inequality revisited
40
If this threshold value of inequality is exceeded, rather negative consequences of increasing in-
equality on economic growth can be expected. The level of economic development of countries
also plays a decisive role. In less developed economies – or more precisely, in countries whose
GDP per capita does not exceed $9,000 – the estimates show a negative effect of increasing
inequality on economic growth.
16
As Figure 4-2 illustrates, in global comparison, Germany is
characterized by a rather low level of income inequality and above-average living standards.
Thus, Germany belongs to the group of countries where rather a positive correlation between
inequality and growth can be expected.
Figure 4-2: GDP per capita versus economic inequality
2017, worldwide comparison
Notes: Countries with more than 60,000$ GDP per capita are not shown (for graphical reasons).
Source: United Nations Population Division, 2019; PWT; SWIID
With respect to the impact of governmental redistribution on economic growth, which is also
discussed in context of the inequality-growth-nexus debate, it is also likely that, ceteris paribus,
potential negative incentive effects in response to governmental redistribution depend on the
existing level of taxes and transfers. Consistent with this hypothesis, the IMF analysis finds that
“when redistribution is already high (above the upper quartile), there is evidence that further
redistribution is indeed harmful to growth” (Berg et al., 2018, 276). According to their analysis
redistribution becomes growth-negative when the difference between pre- and post-govern-
ment Gini coefficients amounts to at least 13 Gini points. In this regard, the result is in line with
the widespread view that at least at some point there exists a trade-off between equity and
efficiency. If, in contrast, one generally assumes that there is a negative or no effect from redis-
16
According to Fuest et al. (2018) the threshold value is even below $5000 in most estimations.
Brazil
Chile
China
Germany
Nigeria
Poland
South Africa
United States
Vietnam
20
25
30
35
40
45
50
55
60
65
1000 11000 21000 31000 41000 51000
Gini coefficient
Output-side real GDP per capita (international $ (2011))
Inequality revisited
41
tribution on growth, besides a causally negative effect from inequality on growth, the govern-
ment could endlessly increase taxes, thereby reducing inequality and enhancing economic
growth.
Taking the theoretical channels and empirical estimations together, the findings suggest that in
countries with high levels of economic development and comparatively low levels of inequality,
a positive correlation between inequality and growth can be expected. With a Gini coefficient
of 0.29, the level of inequality in Germany is rather low by international standards, while the
level of prosperity is well above the average. Like many other EU countries, these indicators
place Germany in the group of countries in which a growth-inhibiting effect of rising inequality
is rather unlikely.
17
Figure 4-3: Income inequality and redistribution
Gini coefficient, 2017
Note. *Inequality levels refer to a year different from 2017.
Source: OECD Income Distribution Database
Figure 4-3 contrasts inequality before and after governmental redistribution across OECD coun-
tries. If, first, Gini coefficients of market incomes are considered, the values suggest only small
differences in the level of inequality between the US (0.505) and Germany (0.500). Market in-
comes describe incomes before redistribution by the state trough taxes and transfers. It includes
all earned income from self-employment and employment as well as capital and property in-
come. However, without considering institutional differences between countries, inequalities
17
At the same time, however, it must be emphasized that this is at best an observed empirical correlation and by no
means a causal relationship between the two variables. Even when non-linear relationships are taken into account,
many factors influencing inequality and growth remain unconsidered.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Slovak Republic
Slovenia
Czech Republic
Iceland
Belgium*
Norway
Denmark
Finland
Austria
Poland
Sweden
Netherlands*
Germany
Hungary
France
Ireland
Switzerland
Estonia
Canada
Greece
Portugal
Australia*
Luxembourg
Russia*
Spain
Italy
Japan*
Israel
New Zealand*
Romania
Korea
Latvia
United Kingdom
Lithuania
United States
Bulgaria
Turkey*
Mexico*
Chile
Brazil*
Costa Rica
India* (2011)
China* (2011)
South Africa*
Gini coefficient
After taxes and transfers Absolute Difference Before taxes and transfers
Inequality revisited
42
of market incomes can be hardly compared. To illustrate this: While many older Germans al-
ready receive a statutory pension, US workers generally work longer and earn a market income
– even at ages when German workers have already retired. Thus, stronger statutory pension
protection in old age tends to go hand in hand with greater inequality in market incomes. Re-
searchers from the Luxembourg Income Study have, therefore, additionally computed the ine-
quality of market incomes of the population under 60 (Gornick/Milanovic, 2015). With a Gini
coefficient of 0.47, concerning this indicator the US shows noticeably higher inequality than
Germany (0.41).
Even greater differences between Germany and the US can be seen when looking at the distri-
bution of wages. According to OECD data from 2018, the wage ratio between the bottom 10 per
cent and the top 10 per cent of all full-time employees is almost 5 in the US. In Germany, the
top 10 per cent earn 3.3 times as much on average – a value in the middle-range of the OECD
countries. For the welfare position of a household within the society, net incomes after taxes
and including transfers are relevant. After state redistribution – namely plus social transfer pay-
ments and pension payments and minus income-related taxes and social contributions – the
Gini coefficient in Germany is reduced to 0.29. In the US, inequality remains at a significantly
higher level of 0.39 even when net income is considered. The absolute difference between ine-
quality of market incomes and the inequality of net incomes is conventionally considered as an
indicator of effective governmental redistribution. As illustrated by Figure 4-3, only a few coun-
tries achieve higher state redistribution than Germany. This is in line with the IMF study which
also sorts Germany to those countries, where the level of redistribution is already high and fur-
ther increases are expected to be harmful to growth (Berg et al., 2018, Figure 5)
5 Conclusion
The first and foremost aim of the Sustainable Development Goals of the UN is the eradication
of extreme poverty and hunger by 2030. When considering the empirical development of pov-
erty over the last decades, in fact, remarkable progress in the reduction of worldwide poverty
has been made. Between 1981 and 2015 the share of the world population considered as ex-
treme poor decreased from 42 to 10 per cent. Given the substantial growth of the world popu-
lation this decrease implied that the number of people living in extreme poverty more than
halved. Despite the continuing major challenge of completely eradicating global poverty by
2030, extreme poverty has been significantly reduced in recent decades and millions of people
have been able to build modest prosperity. When people are asked about the perceived devel-
opment of global poverty rates, though, the view is far more pessimistic. Half of the respondents
from 28 countries believed in 2017 that extreme poverty has been rising, in Germany only 11
per cent of respondents correctly assumed a declining trend. The pessimistic view on global
development is worrying because if positive trends are not noticed, drivers of these trends are
likely to be misinterpreted.
Progress was also made regarding the development of global income inequality. It has de-
creased significantly over the past 200 years. Particularly large progress was made in the 20th
century by advances in South-East Asia, especially due to trade-induced income increases in
Inequality revisited
43
populous China and India. Since the late 1980s, global income inequality measured by the Gini
coefficient decreased from 0.68 in 1988 to 0.62 in 2013. While income gaps between countries
got smaller, though, the contribution of within-country inequality to global inequality has in-
creased.
Although global income inequality has decreased over the last decades, inequalities between
advanced economies, emerging markets, and developing economies are still extensive. Increas-
ing global trade and economic integration have been proven to be effective ways to reduce
disparities between countries. However, globalization and free-market systems have often neg-
ative connotations. The successes of such processes and systems are misjudged, and the desire
for protectionism and autarky increases. Trade constraints may though harm the catching-up
process of emerging markets and developing economies and hinder productivity to grow, which
is an important driver of poverty reduction.
Discussions about inequality often explicitly refer to Germany as an example of high (wealth)
inequality, which is likewise characterized by a shrinking middle class and large low-wage sector.
However, if we zoom in to the situation in Germany, the picture reveals to be far more positive
as many popular inequality narratives suggest. The comparatively high concentration of wealth
is not uncommon for well-established welfare states with generous redistribution schemes. In
a worldwide comparison, Germany is similar to the group of Scandinavian countries which are
also characterized by generous social security systems, low unemployment rates, comparatively
low net income inequality and high levels of net wealth inequality. Beyond, although the major-
ity of Germans in surveys regularly believe that income and wealth inequality is increasing, dis-
tributional indicators stabilized since more than a decade. In fact, given the overly positive de-
velopment of employment in Germany since 2005, this finding may not be surprising. However,
given that in an isolated view, changing household structures, ongoing demographic change and
increasing migration numbers would have rather resulted in increasing inequality, stabilized dis-
tributional indicators may well be seen as positive development.
Yet, this does not change the fact that, similar with other advanced economies, today’s income
and wealth inequality levels in Germany are higher than throughout the 1990s. Therefore, alt-
hough the comparatively high wealth concentration can be explained and income inequality is
rather low in a global comparison, the natural question arises whether inequalities may be too
large. In the context of the optimal level of inequality, the debates on the relationship between
inequality and growth play a central role. Recent studies suggested that lower inequality levels
would mechanically imply higher economic growth and that higher redistribution will not have
any harmful effects. Further analysis suggests, however, that the relationship between inequal-
ity and economic growth is far more complex than can be depicted by cross-country compara-
tive studies with few aggregate variables. Neither does higher inequality mechanically imply
higher economic growth, nor can this necessarily be achieved by reducing inequality. Further-
more, considering non-linear effects reveals that negative effects on growth can particularly be
expected in countries with high degrees of inequality and low economic development, whereas
in prosperous countries with low levels of inequality rather a positive relationship can be ex-
pected. With a comparatively low level of inequality and high living standards Germany clearly
Inequality revisited
44
belongs to the latter group. In addition, Germany is characterised by an above-average level of
governmental redistribution (even among advanced economies) so that cross-country inequal-
ity analyses suggest that further redistribution is expected to have harmful effects on growth.
Nevertheless, the observation, that a relatively high level of economic development tends to be
associated with a lower level of inequality, suggests that both variables can be achieved simul-
taneously. However, it is not only the design of the social security system that plays a decisive
role here, but also stable and credible institutions, an empowering education policy and a wise
and forward-looking investment policy.
Finally, whether a country's level of inequality leads to social tensions and political instability
also depends on how inequality is perceived within the country and which ideas of justice prevail
in society. In Germany, for example, more than 80 per cent consider a society to be just if hard-
working people earn more than others (Adriaans et al., 2019). Compared with other European
countries, approval of the principle of meritocracy is particularly pronounced in Germany. In-
come differences are, thus, desirable, provided they are justified by different efforts. Ad-
vantages that stem solely from exogenous privileges or from family circumstances, on the other
hand, are largely perceived as unfair. With respect to debates about inequality it would, there-
fore, be helpful if they would focus more explicitly on the reasons why inequalities emerged and
whether they are for example justified by different effort or different leisure-work-preferences.
In Hufe et al. (2020) this attempt at differentiation is made and Germany proves to be neither a
particularly unequal nor an unfair country.
Inequality revisited
45
List of figures
Figure 2-1: Income per world citizen per year .......................................................................... 9
Figure 2-2: Global inequality ...................................................................................................10
Figure 2-3: Growth incidence of country-occupations (1970s-2010s) ....................................11
Figure 2-4: Productivity growth in developed versus developing countries ..........................13
Figure 2-5: Extreme poverty ....................................................................................................15
Figure 2-6: Poverty reduction and productivity ......................................................................16
Figure 2-7: Believes on the development of extreme poverty ...............................................17
Figure 2-8: Global net wealth inequality .................................................................................19
Figure 3-1: This is what the German Bundestag is talking about ............................................20
Figure 3-2: Net income inequality over time ...........................................................................21
Figure 3-3: The Great Gatsby Curve ........................................................................................23
Figure 3-4: Net wealth inequality in Germany ........................................................................25
Figure 3-5: Net wealth inequality by country ..........................................................................26
Figure 3-6: The development of the middle class in Germany ...............................................28
Figure 3-7: Change of disposable household income by income deciles ................................29
Figure 3-8: Income mobility across quintiles ...........................................................................30
Figure 3-9: Evolution of labour income share and market income inequality in Germany ...32
Figure 3-10: Hourly earnings adjusted for purchasing power in an EU comparison ..............33
Figure 3-11: Inequality of agreed gross hourly wage in main job ...........................................34
Figure 3-12: Changes in household income through COVID-19 pandemic .............................36
Figure 3-13: Changes in net and market income inequality through COVID-19 pandemic ...37
Figure 4-1: Impact of inequality changes on economic growth – OECD results .....................39
Figure 4-2: GDP per capita versus economic inequality ..........................................................40
Figure 4-3: Income inequality and redistribution ....................................................................41
Inequality revisited
46
References
Adriaans, Jule / Eisnecker, Philipp / Liebig, Stefan, 2019, Gerechtigkeit im europäischen Ver-
gleich: Verteilung nach Bedarf und Leistung in Deutschland besonders befürwortet, in: DIW Wo-
chenbericht, No. 45, pp. 818–824, Berlin
Albers, Thilo N. H. / Bartels, Charlotte / Schularick, Moritz, 2020, Die Verteilung der Vermögen
in Deutschland von 1895 bis 2018, ECONtribute Working Paper, Bonn
Allianz Research, 2020, Allianz Global Wealth Report 2020, München
Atkinson, Anthony Barnes, 2015, Inequality. What can be done?, Cambridge, Massachusetts
Bach, Stefan / Thiemann, Andreas / Zucco, Aline, 2019, Looking for the missing rich: Tracing the
top tail of the wealth distribution, in: International Tax and Public Finance, No. 26, pp. 1234–
1258
Berg, Andrew / Ostry, Jonathan D. / Tsangarides, Charalambos G. / Yakhshilikov, Yorbol, 2018,
“Redistribution, Inequality, and Growth: New Evidence.”, in: Journal of Economic Growth, Vol.
23, No. 3, pp. 259–305
Beznoska, Martin, 2016, Dokumentation zum Steuer-, Abgaben- und Transfer-Mikrosimulati-
onsmodell des IW Köln (STATS), IW-Report, No. 27, Köln
Beznoska, Martin, 2020, Die Verteilung von Steuern, Sozialabgaben und Transfereinkommen
der privaten Haushalte, IW-Report, No. 6, Köln
Beznoska, Martin / Niehues, Judith / Stockhausen, Maximilian, 2020, Stabil durch die Krise?
Verteilungsfolgen der Corona-Pandemie – eine Mikrosimulationsanalyse, IW-Report, No. 65,
Köln
Bönke, Timm / Grabka, Markus M. / Schröder, Carsten / Wolff, Edward N., 2019, A Head-to-
Head Comparison of Augmented Wealth in Germany and the United States, in: The Scandinavian
Journal of Economics, Vol. 122, No. 3, pp. 1140–1180
Bruckmeier et al., 2020, Covid-19-Krise: Für das Jahr 2020 ist mit keinem Anstieg der Einkom-
mensungleichheit in Deutschland zu rechnen, ifo Schnelldienst Digital, No. 15, München
Brunori, Paolo / Maitino, Maria L. / Ravagli, Letizia / Sciclone, Nicola, 2020, Distant and Unequal.
Lockdown and Inequalities in Italy, in: Working Papers-Economics, wp2020_13.rdf, Universita'
degli Studi di Firenze, Dipartimento di Scienze per l'Economia e l'Impresa, No. 13, Florenz
Inequality revisited
47
BMSGPK – Bundesministerium für Soziales, Gesundheit, Pflege und Konsumentenschutz, 2020,
COVID-19: Analyse der sozialen Lage in Österreich Teil 1, preliminary version, Wien
Cingano, Federico, 2014, Trends in income inequality and its impact on Economic growth, in:
OECD SEM Working Paper, No. 163, Paris
Corak, Miles, 2016, Inequality from Generation to Generation: The United States in Comparison,
IZA Discussion Paper, No. 9929, Bonn
Credit Suisse, 2019, Global Wealth Databook 2019, Zürich
Deaton, Angus, 2021, COVID-19 and Global Income Inequality, NBER Working Paper Series, No.
28392, Cambridge
Deutsche Bundesbank, 2019, Vermögen und Finanzen privater Haushalte in Deutschland. Er-
gebnisse der Vermögensbefragung 2017, in: Monatsbericht April 2019, pp. 13–44, Frankfurt am
Main
de Vries, Catherine E. / Hoffmann, Isabell, 2018, Die Macht der Vergangenheit – Wie Nostalgie
die öffentliche Meinung in Europa beeinflusst, eupinions – what do you think?, Bertelsmann
Stiftung, No. 2, Gütersloh
Easterly, William / Levine, Ross, 2001, What Have We Learned from a Decade of Empirical Re-
search on Growth? It’s Not Factor Accumulation: Stylized Facts and Growth Models, in: World
Bank Economic Review, Vol. 15, No. 2, pp. 177–219
ECB – European Central Bank, 2020, Monetary policy and inequality, in: Economic Bulletin, Issue
2, Frankfurt am Main
Fedorets, Alexandra / Grabka, Markus M. / Schröder, Carsten / Seebauer, Johannes, 2020, Loh-
nungleichheit in Deutschland sinkt, in: DIW Wochenbericht, No. 7, pp. 91–97, Berlin
Fuest, Clemens / Neumeier, Florian / Stöhlker, Daniel, 2018, Ungleichheit und Wirtschafts-
wachstum: Warum OECD und IWF falsch liegen, in: ifo-Schnelldienst, Vol. 71, No. 10, pp. 22−25,
München
Goebel, Jan / Grabka, Markus M. / Schröder, Carsten, 2015, Einkommensungleichheit in
Deutschland bleibt weiterhin hoch: junge Alleinlebende und Berufseinsteiger sind zunehmend
von Armut bedroht, in: DIW-Wochenbericht, Vol. 84, No. 25, pp. 571–586, Berlin
Goebel, Jan et al., 2019, The German Socio-Economic Panel Study (SOEP). in: Jahrbücher für
Nationalökonomie und Statistik / Journal of Economics and Statistics, Vol. 239, No. 2, pp. 345–
360
Inequality revisited
48
Gornick, Janet C. / Milanovic, Branko, 2015, Income Inequality in the United States in Cross-
National Perspective: Redistribution Revisited, in: LIS Center Research Brief, No. 1, New York
Grabka, Markus .M / Schröder, Carsten, 2018, Ungleichheit in Deutschland geht bei Stunden-
löhnen seit 2014 zurück, stagniert aber bei Monats- und Jahreslöhnen, in: DIW Wochenbericht,
No. 9, pp. 157–166, Berlin
Grabka, Markus M. / Goebel, Jan, 2020, Realeinkommen steigen, Quote der Niedrigeinkommen
sinkt in einzelnen Altersgruppen, in: DIW Wochenbericht, No. 18, pp. 316−323, Berlin
Hammar, Olle / Waldenström, Daniel, 2017, Global Earnings Inequality, 1970-2015, CEPR Dis-
cussion Paper, No. 12019, London
Hellebrandt, Tomas / Mauro, Paolo, 2015, The Future of Worldwide Income Distribution (April
1, 2015), in: Peterson Institute for International Economics Working Paper, No. 15−7, Luxem-
bourg
Hufe, Paul / Peichl, Andreas / Weishaar, Daniel, 2018, Intergenerationelle Einkommensmobili-
tät: Schlusslicht Deutschland?, in: ifo Schnelldienst, Vol. 71, No. 20, pp. 20−28, München
Hufe, Paul / Kanbur, Ravi / Peichl, Andreas, 2020, Measuring Unfair Inequality: Reconciling
Equality of Opportunity and Freedom from Poverty, ifo Working Papers, No. 323, München
Ipsos MORI, 2017, Perils of Perception – Global Impact of Development Aid, September 2017
https://www.ipsos.com/sites/default/files/ct/news/documents/2017-09/Gates_Perils_of_Per-
ception_Report-September_2017.pdf [21.12.2020]
Kleimann, Rolf et al., 2020, Analyse der Einkommens- und Vermögensverteilung in Deutschland,
Begleitforschung zum sechsten Armuts- und Reichtumsbericht der Bundesregierung, Bundesmi-
nisterium für Arbeit und Soziales (Eds.), Bonn
Klenow, Peter J. / Rodriguez-Clare, Andrés, 1997, The Neoclassical Revival in Growth Economics:
Has it Gone too Far, in: Ben Bernanke and Julio Rotemberg (Eds.), NBER Macroeconomics Annual
1997, Cambridge, MA: MIT Press, pp. 73-103
Kolev, Galina / Niehues, Judith, 2016, The Inequality-Growth Relationship: An Empirical Reas-
sessment, IW-Report No. 7, Köln
Lakner, Christoph /Milanovic, Branko, 2016, Global Income Distribution: From the Fall of the
Berlin Wall to the Great Recession, in: The World Bank Economic Review, Vol. 30, No. 2, pp.
203–232
Inequality revisited
49
Lakner, Christoph / Mahler, Daniel G. / Negre, Mario / Prydz, Espen B., 2020, How Much Does
Reducing Inequality Matter for Global Poverty?, Global Poverty Monitoring Technical Note 13,
World Bank, Washington, DC
Mayer-Foulkes, David ,2019, Macroeconomics of poverty: Development and underdevelopment
under Globalization. Centro de Investigación y Docencia Económicas, México
Milanovic, Branko, 2020, After the crisis: the evolution of the global income distribution be-
tween 2008 and 2013, Stone Center on Socio-Economic Inequality Working Paper, No. 18, New
York
Neidhöfer, Guido / Stockhausen, Maximilian, 2018, Dynastic Inequality Compared – Multigen-
erational Mobility in the United States, the United Kingdom and Germany, in: Review of Income
and Wealth, Vol. 65, No. 2, pp. 383–414
Neves, Pedro Cunha / Afonso, Óscar / Silva, Sandra Tavares, 2016, A Meta-Analytic Reassess-
ment of the Effects of Inequality on Growth, in: World Development, Vol. 78, No. C, pp. 386−400
Niehues, Judith / Schaefer, Thilo / Schröder, Christoph, 2013, Arm und Reich in Deutschland:
Wo bleibt die Mitte?, in: IW-Analysen, No. 89, Köln
Niehues, Judith, 2017a, Die Mittelschicht in Deutschland – Vielschichtig und stabil, in: IW-
Trends, Vol. 44, No. 1, pp. 3–20
Niehues, Judith, 2017b, Einkommensentwicklung, Ungleichheit und Armut – Ergebnisse unter-
schiedlicher Datensätze, in: IW-Trends, Vol. 44, No. 3, pp. 117–135
Niehues, Judith, 2019, Subjektive Umverteilungspräferenzen in Deutschland, in: IW-Trends, Vol.
46, No. 1, pp. 79–98
OECD – Organisation for Economic Cooperation and Development, 2014, Focus on Inequality
and Growth - December 2014 https://www.oecd.org/social/Focus-Inequality-and-Growth-
2014.pdf [21.12.2020]
OECD, 2018, A Broken Social Elevator? How to Promote Social Mobility, OECD Publishing, Paris
Ostry, Jonathan D. / Berg, Andrew / Tsangarides, Charalambos, 2014, Redistribution, Inequality,
and Growth, IMF Staff discussion Note, February
Peichl, Andreas / Pestel, Nico / Schneider, Hilmar, 2011, Does size matter? The impact of
changes in household structure on income distribution in Germany, in: Review of Income and
Wealth, Vol. 58, No. 1, pp. 118–141
Inequality revisited
50
Schnitzlein, Daniel D., 2016, A New Look at Intergenerational Mobility in Germany Compared to
the U.S., in: Review of Income and Wealth, Vol. 62, No. 4, pp. 650–667
Schröder, Carsten et al., 2020, Verbesserung der Forschungsdateninfrastruktur im Bereich
Hochvermögender mit dem Sozio-oekonomischen Panel (SOEP), SOEPpapers on Multidiscipli-
nary Panel Data Research, No. 1084, Berlin
Schröder, Christoph / Kestermann, Christian, 2020, Mindestlohn und Einkommensarmut, in: IW-
Trends, Vol. 47, No. 2, pp. 107–127
Shorrocks, Anthony / Davies, James / Lluberas, Rodriogo, 2020, Global wealth report 2020,
Credit Suisse Research Institute
Stockhausen, Maximilian / Calderón, Mariano, 2020, IW-Verteilungsreport 2020. Stabile Ver-
hältnisse trotz gewachsener gesellschaftlicher Herausforderungen, IW-Report, No. 8, Köln
Stockhausen, Maximilian, 2018a, Like Father, Like Son? – A Comparison of Absolute and Relative
Intergenerational Labour Income Mobility in Germany and the US, SOEPpapers on Multidiscipli-
nary Panel Data Research, No. 989, Berlin
Stockhausen, Maximilian, 2018b, Ist der Traum vom sozialen Aufstieg in Deutschland ausge-
träumt?, IW-Kurzbericht, No. 48, Köln
van Zanden, Jan L. et al., 2014, How Was Life? Global Well-being since 1820, OECD Publishing,
Paris
Westermeier, Christian / Grabka, Markus M., 2015, Große statistische Unsicherheit beim Anteil
der Top-Vermögenden in Deutschland, in: DIW Wochenbericht, No. 7, pp. 123–133, Berlin
World Bank, 2016, Poverty and shared prosperity 2016: Taking on inequality, World Bank Group
World Bank, 2020, Global Productivity – Trends, Drivers, and Policies, Alistair Dieppe (eds.),
World Bank Group
