October 2014
Institut für Makroökonomie
und Konjunkturforschung
Macroeconomic Policy Institute
At a glance
„„ Voluntary household surveys
such as the German SocioEconomic Panel (SOEP) tend
to underestimate income and
wealth inequality. The research
methodology developed by
Thomas Piketty et al. therefore
analyses official tax statistics in
order to more accurately determine the inequality between the
people at the top of the income
and wealth distribution scale
and the rest of the population.
„„ However, Piketty’s methodology underestimates the rise in
inequality in Germany since the
turn of the millennium due to
the fact that companies are retaining a significant percentage
of their rising profits which are
therefore not recorded as private household income.
„„ More accurate indicators of inequality in Germany can be developed by means of combining
data from the available surveys
and national accounting systems.
„„ The (re-)introduction of a comprehensive income tax base
and a wealth tax would make
it significantly easier to measure high incomes and wealth in
Germany.
„„ The reduction of inequality in
Germany could help to reduce
the country’s high current account surpluses, thereby contributing to greater macroeconomic stability.
99e
Report
Income and Wealth Distribution
in Germany:
A Macro-Economic Perspective
Jan Behringer, Thomas Theobald, Till van Treeck1
Videostatement:
Till van Treeck about inequality
and macroeconomic imbalances
http://youtu.be/WOkbfMyGTYk
Corrected version (25/10/2014): In comparison to the version published on 23th of October, in Figure 4b, the erroneous „p90“ percentile
was changed to „p99“.
Contents
The new debate on economic inequality ������������������������������������������������������������������2
Inequality in Germany �����������������������������������������������������������������������������������������������������������������2
The SOEP approach to measuring inequality based on
income and wealth data ��������������������������������������������������������������������������������������������������������������������������2
The World Top Incomes Database as a new data source �������������������������������������4
The “corporate veil” as one of the reasons why inequality
is underestimated ������������������������������������������������������������������������������������������������������������������������������������������5
Alternative indicators of inequality ������������������������������������������������������������������������������������������������5
INFOBOX 1: On the relationship between functional and
personal income distribution and macroeconomic development �������������������6
Enhanced top income shares ������������������������������������������������������������������������������������������������������������8
Wealth-to-income ratios ����������������������������������������������������������������������������������������������������������������������� 10
INFOBOX 2: An illustration of Piketty’s “fundamental laws of
capitalism” and long-term trends in income and wealth inequality ����������11
Conclusions ����������������������������������������������������������������������������������������������������������������������������������������14
References ����������������������������������������������������������������������������������������������������������������������������������������14
1 University of Duisburg-Essen.
The new debate on economic
inequality
Thomas Piketty’s international bestseller “Capital in
the Twenty-First Century” has given rise to a new
debate regarding which data should be used as the
basis for measuring income and wealth inequality,
and what are the macroeconomic repercussions
of rising inequality. Even before the publication
of his book, the findings of research into income
distribution in Germany had made the headlines
on more than one occasion. Studies carried out by
the Organisation for Economic Co-operation and
Development (OECD 2008, 2011), for example, reveal that over the past decade and a half, income
inequality in Germany has risen faster than in virtually any other OECD nation. In recent years, the
debate has revolved around the question of whether
income inequality has fallen again following the financial and economic crisis of 2008/9 (Grabka and
Goebel 2013; Rehm et al. 2014). The data that these
discussions are based on is provided by the SocioEconomic Panel (SOEP) at the German Institute for
Economic Research (DIW).
During the course of the past year, there has also
been a furore over the findings of a household survey
(HFCN 2013) coordinated by the European Central
Bank (ECB), according to which wealth inequality
in Germany comes second only to Austria among
the EU member states. Moreover, the DIW concluded that, based on SOEP data, there has been “persistently high wealth inequality in Germany” between 2002-2012 (Grabka and Westermeier 2014).
The approach taken by Thomas Piketty and his
research team to analysing inequality differs from
the studies cited above in two respects. Firstly, he
questions the validity of the data that the studies are
based on. This is because both the SOEP and the
ECB’s Household Finance and Consumption Survey (HFCS) rely on voluntary household surveys
in which very rich people tend not to take part. In
some cases, this can cause income and wealth inequality between the upper end of the income distribution scale and the rest of the population to be
seriously underestimated. Consequently, Piketty
and his co-authors have been systematically analysing the official income and wealth tax statistics of
several different countries for more than ten years.
Their goal in so doing is to provide a more realistic estimate of high incomes as a percentage of all
household incomes. The results of their seminal research are summarised in the World Top Incomes
Database (WTID) which is now publicly available
(Alvaredo et al. 2012).
Secondly, in his most recent book (Piketty 2014),
Piketty sets his analysis of income and wealth in-
equality in a broader macroeconomic context. In
order to do this, he uses data from the National Account Systems (NAS) and Financial Accounts (FA).
He is thus able to link microeconomic distribution
research to macroeconomic issues, allowing him to
address questions such as the relationship between
rising economic inequality and increased macroeconomic instability (Piketty 2014; van Treeck 2014a).
This report aims to compare the key findings of
the two research methodologies outlined above (voluntary household surveys vs. official tax statistics
and macroeconomic accounting systems) and discuss the reasons for the significant differences that
sometimes occur between them. It will also identify
urgent research priorities and policy interventions
needed to improve the quality of the available data.
Furthermore, drawing e.g. on Piketty (2014), it will
set out a number of proposals for accurate indicators of inequality that could be developed using
data that is already available. Finally, it will outline
the macroeconomic repercussions of rising economic inequality in Germany.
IMK Report 99e
October 2014
Inequality in Germany
The SOEP approach to measuring
inequality based on income and wealth
data
The debate surrounding income inequality in Germany is profoundly shaped by its focus on the Gini
coefficient of net equivalised household income,
based on data provided by the SOEP (Figure 1a).
Having remained largely stable for many years since
the mid-1980s, the Gini coefficient for Germany
rose sharply in the first half of the 2000s. From 2006
on it then stabilised around its new, higher level,
albeit with a slight downward trend, before rising
once more in 2012 to a value of 0.288.
One possible explanation for the slight dip in the
Gini coefficient that occurred between 2006 and
2012 could relate to trends in unearned income
(Rehm et al. 2014, Horn et al. 2014). Compared to
earned income, unearned income is much more
heavily concentrated at the upper end of the distribution scale. As a result, the temporary fall in
returns on capital that occurred during the financial market crisis automatically led to a reduction in
income inequality. Furthermore, the SOEP data can
only record the distributed income of corporations,
e.g. in the form of private withdrawals in partnerships. It does not record retained earnings.
Since 2002, the SOEP has recorded wealth distribution data on a five-yearly basis. It registered a rise
in wealth inequality for the period 2002 to 2007,
page 2
Figure 1
Inequality measures: SOEP
1
IMK Report 99e
October 2014
Problems with illustrating inequality
Problems
illustrating
inequality
Problems
with with
illustrating
inequality
2
Problems with illustrating
inequality
b) SOEP
savings
rates
income quartiles
quartiles 2
b) SOEP
savings
rates
byby
income
a) income
Gini income
coefficients Germany
1
a)a)Gini
coefficients
Germany
1
2
1
2
Gini
income
coefficients
Germany
b) SOEP
by income
a) Gini
income
coefficients
Germany
b) SOEP
savingssavings
rates byrates
income
quartilesquartiles
1
2
a) Gini income
coefficients Germany
b) SOEP savings rates by income quartiles
0.55
%
0.55
0.55
%
%
0.55
% 15
0.5
Real market equivalised
15
0.50.5
RealReal
market
equivalised
income,
year
before
market
equivalised
15
4th quartile
15
0.5
Real
market
equivalised
income,
yearyear
before
income,
before
0.45
4th quartile
13
4th quartile
income, year before
4th
quartile
0.45
0.45
13
13
0.45
13
0.4
11
0.40.4
11
11
0.4
3rd quartile
11
0.35
9
3rd quartile
3rd quartile
0.35
0.35
9
3rd
quartile
9
0.35
Real net equivalised
9
0.3
RealReal
netincome,
equivalised
7
year before
2nd quartile
net equivalised
0.3
0.3
7
income,
year
before
Real net
equivalised
2nd quartile
7
income,
year before
0.3
2nd quartile
7
0.25
1st quartile
income, year before
2nd
quartile
51st quartile
0.25
0.25
5
1st quartile
0.25
5
1st
quartile
0.2
5
0.2
3
1984
1988
1992
1996
2000
2004
2008 32012
0.2
1996
2000
2004
2008
2012
0.21984 1988 1992 1996 2000 2004 2008 2012
19963 3
2000
2004
2008
2012
1984
1988
1992
1996
2000
2004
2008
2012
1984
1988
1992
1996
2000 3 2004
2008
2012
1996
2000
2004
2012
4 2008 4
3
2000
2004 Statistical
2012
c) Individual
net wealth
selected
percentiles
d) 1996
Wealthd)SOEP
vs.SOEP
Federal
Office 2008
c) Individual
netfor
wealth
for selected
percentiles
Wealth
vs.Statistical
Federal
Office
c)c)Individual
netwealth
wealth
selected
percentiles
d) Wealth
SOEP
vs.vs.Federal
Statistical
Office
Individual net
for for
selected
percentiles
d) Wealth
SOEP
Federal Statistical
Office
3
4
c) Individual net
wealth for selected percentiles
d) Wealth SOEP vs. Federal Statistical Office
%
EUR
Billion EUR
%
EUR
0.16
%
Billion
EUR
900,000
0.16
10,000
900,000 EUR
-3.04
0
Billion
EUR
90
% % 10,000
EUR
-3.04
Billion
EUR
0.16
0
Coverage
SOEP/Destatis
900,000
800,000
0.16
Coverage SOEP/Destatis
10,000
900,000
800,000
-3.04
9,000
-6.73
0
-6.90
10,000
-3.04
-5
-6.73
-6.90
CoverageSOEP/Destatis
SOEP/Destatis
80
-5 0 9,000
700,000
800,000
Coverage
800,000
700,000
9,000
-6.73
-6.90
-58,000
-6.73
-6.90
-109,0008,000
-5
600,000
700,000
-10
70
600,000
700,000
-8.12 -8.12
-9.99
8,000
-9.99
7,000
-10
8,000
500,000
-15
600,000
7,000
-10
500,000
-8.12
600,000
-15
-8.12
60
-9.99
-9.99
7,000
400,000
6,000
500,000
-15-207,000
400,000
Real Real
500,000
-20 -15 6,000
growth
50
growth
300,000
Real
400,000
6,000
300,000
5,000
Real
400,000
-20
5,000
-256,000
-25 -20
growth
growth
200,000
40
300,000
200,000
300,000
5,000
4,000
-25
4,000
-305,000
-30 -25
100,000-76.41 -76.41
100,000
200,000
200,000
30
4,000
3,000
3,000
-30
-354,000
-35 -30
-76.41
0 -76.41
0
100,000
100,000
20
p5 p25
p10 p50
p25p75
p50p90
p75p95
p90p99p95 p99 2,000
p1
p5p1 p10
3,000
2,000
-35-403,000
-100,000
-40 -35
-100,000
0 0
p5 p10 p25 p50 p75p90
p90 p95 p99
10
2002
p1 p1
p5
1,000
2,000
1,000
2,000
2002 p10 p25 p50 p75 2007
2007p95 p99
-40
-100,000
-40
-100,000
1
coefficients:
0.776
(2002) 0.799
0.799
(2007)
GiniGini
coefficients:
0.776
(2002)
(2007)
0.78
(2012)
0.78
(2012)
in billion EUR
Destatis
in billion EUR
2007
4
2007
2007
2007
2012
0
2012
2012
2012
%
90
%%
90
90 80
SOEP
in billion EUR
SOEP
in billion
EUR
in Destatis
billion SOEP
EUR
in billion EUR
Destatis
in billionDestatis
EUR
in billion EUR
in billion EUR
0 1,000 0
1,000
2002
2002
00
2002
2002
Destatis
SOEP
SOEP
in billion
in billion
EUREUR
inDestatis
billionSOEP
EUR
in billion EUR
Destatis
in billion
EUR
in billionDestatis
EUR
SOEPEUR
in billion
SOEP
in billion EUR
2012
Real Real
growth
(2005=100)
2012
growth
(2005=100)
2002
2007
2002
2007
Gini
coefficients:
0.776 (2002)
0.799
(2007)
0.78 (2012)
Gini coefficients:
0.776 (2002)
0.799 (2007)
0.78 (2012)
2012
Real
growth
(2005=100)
2012
Real
growth
(2005=100)
4
SOEP
Destatis
SOEP
in billion
EUREUR
in billion
in Destatis
billion SOEP
EUR
in
billion
Destatis
in billion EUR EUR
in billionDestatis
EUR
in billion
SOEPEUR
3
808070
707060
606050
505040
404030
303020
202010
10100
00
2002
2007
Monthly
savings as a % of monthly net income.
2007
3
2002
2002
2012
N.B.: Wealth
data are taken from DIW Wochenbericht 9/2014. For computation of real growth rates the GDP deflator is used.
2012
4
N.B.: The
SOEP wealth data are taken from the DIW Wochenbericht 45/2007, 4/2009 and 9/2014. The Federal Statistical Office
2007
2007
1wealth data are taken from the corresponding wealth report‘s position „Private Non-Profit Institutions and Private Households“.
2012
2012
N.B.:
1 The SOEP wealth data are taken from the
N.B.:Federal
The
SOEP
wealth4/2009
data are
taken from the
Sources:
SOEP;
Statistical
Office.
DIW
Weekly
Reports
45/2007,
and
1 1 DIW Weekly Reports 45/2007, 4/2009 and
9/2014.
Federal
Statistical
Office
wealth
data
N.B.:
The
SOEP
wealth
are
N.B.:
The
SOEP
wealthdata
data
aretaken
takenfrom
fromthe
the
9/2014.
Federal
Statistical
Office wealth data
are
taken
fromThe
its
wealth
report
"Non-Profit
DIW
Weekly
Reports
45/2007,
4/2009
DIW
Weekly
Reports
45/2007,
4/2009and
and
areGini
taken
fromHouseholds".
itsStatistical
wealth
report
"Non-Profit
Institutions
Serving
9/2014.
The
Federal
Office
wealth
data
when
the
coefficient
rose from
0.78
to 0.80
9/2014.
The
Federal
Statistical
Office
wealth
datainequality is already much higher and the accumuInstitutions
Serving
Households".
are
taken
report
are
taken
fromitsitswealth
wealth
report
"Non-Profit
(Grabka
and from
Westermeier
2014).
In"Non-Profit
other words,
lation of wealth through savings takes time. Over
Sources:
SOEP;
Federal
Statistical Office.
Institutions
Serving
Households".
Institutions
Serving
Households".
wealth
is
far
less
equally
distributed
than
income.
the longer term, however, wealth inequality can be
Sources: SOEP; Federal Statistical Office.
1
2
Imputed rent without
2002correction for public officials.
Between 2007 and 2012, however, the Gini wealth
Sources:
SOEP; Federal
Sources:
FederalStatistical
StatisticalOffice.
coefficient
fell SOEP;
back slightly
(Figure
1c). Office.
How are we to explain the fact that income inequality rose sharply from 2002-2012, while wealth
inequality remained more or less unchanged, albeit
at a high level?2
One economic explanation is that changes in income distribution initially only have a modest impact on wealth distribution, since the level of wealth
2 See also Brenke/Wagner (2013, p. 114), where it is
argued that if unearned income and the income of top
earners increase at an above average rate, and if these
same top earners have a relatively high savings rate
while low-income households save nothing or next
to nothing, it is inevitable that wealth will become
increasingly concentrated in the hands of the few.
expected to “catch up” and start having an increasing impact on income inequality (Infobox 2). This
is especially true since, according to SOEP data, the
savings rates of the upper and lower income groups
have diverged since the year 2000 (Figure 1b).3 The
trends witnessed in recent years (the rise in income
inequality between 2011 and 2012 and the growing
gap in savings rates since 2010) also point to the
3 However, the SOEP data on savings suffers from
a number of serious flaws and cannot easily be
compared against the NAS savings rate. The way that
the SOEP questions are phrased rules out negative
savings rates. Moreover, compared to the questions
on incomes, the number of respondents who fail to
provide data on savings or only provide inconsistent
data is relatively high.
page 3
likelihood of an increase in wealth inequality in the
future.
A number of criticisms can also be levelled at
the data that the SOEP is based on. It is, by its very
nature, difficult to record wealth (especially high levels of wealth) using voluntary household surveys
like the SOEP. Accordingly, the finding that real net
worth declined between 2002 and 2012 specifically
in the upper percentiles (Figure 1c) is somewhat
questionable. Over the same period, the SOEP’s coverage of the total net worth of private households
compared to the wealth report of the Federal Statistical Office (“Private Non-Profit Institutions And
Private Households”) fell sharply from around 85 %
to approximately 65 % (Figure 1d). It may therefore
be surmised that wealth inequality is not only being
underestimated in terms of its degree (Grabka and
Westermeier 2014), but also in terms of its rate of
change.
The World Top Incomes Database as a
new data source
As a result of the research carried out by Facundo
Alvaredo, Anthony B. Atkinson, Thomas Piketty and Emmanuel Saez, a new measure of income
inequality is now emerging alongside the established Gini coefficients. It assesses high income
groups’ share of the total (pre-tax) income of private households (top income shares). It is based on
data taken from official income tax statistics and the
National Account Systems.
Figure 2a shows the evolution of the top income share of the total pre-tax income for private
households in Germany according to the World Top
Incomes Database (WTID). Two things are particularly striking. Firstly, in contrast to most other
countries, the time series for Germany end as long
ago as 2007. Secondly, the data reveal only a slight
increase in the income share of the top 5 % and the
top 1 % of households as a percentage of the total
income for all households, whilst the share of the
top 10 % experienced a somewhat larger increase.
The most recent available time series are based on
the study carried out by Dell (2007) and no further
studies have been carried out since. The problem
confronting data for the years after 2009 is that, following the introduction of the flat-rate withholding
tax in that year, information on tax paid on capital
income is no longer recorded on an individual basis
as it used to be under the comprehensive income
tax system. Instead, financial institutions pay the
anonymous flat-rate withholding tax direct to the
taxman, meaning that no first-hand information
is available regarding the distribution of capital income. Because of the high concentration of wealth,
Other measures of inequality
2a) Top income shares incl. capital gains
IMK Report 99e
October 2014
F igu r e 2
Other measures of inequality
Inequality
2a) Topmeasures:
income shares incl. capital gains
Other measures of inequality
%
WTID
and
National Accounts
45
%
40
a)
Top
45
35
40 %
4530
35
4025
30
3520
25
3015
20
2510
15
20 5
10
15 0
5
1992
10
0
51992
2a) Top income shares incl. capital gains
Top 10%
income shares1
0
b)
1992
Top 10%
Top 5%
Top 10%
Top 5%
Top5%
1%
Top
Top 1%
1995
1995
Top 1%
1998
1998
2001
2001
2004
2007
2004
2007
Corporate and capital income, Germany
1995
1998
2001
2004
2007
b) Corporate and capital income, Germany
%
%
b)39Corporate and capital income, Germany
b) Corporate and capital income, Germany
%
3937
Disposable corporate
income as a percentage of
disposablecorporate
private income
Disposable
(right axis)
income
as a percentage of
disposable private income
Disposable
(right
axis) corporate
income as a percentage of
disposable private income
(right axis)
%
9
9
8
7
%8
3735%
6
9
39
7
3533
5
8
6
37
4
3331
57
Capital income as a
35
percentage of national 463
3129
income (left axis)
2
33
Capital income as a
35
2927
percentage of national
1
4
2
31
income (left axis)
0
Capital income as a
2725
13
29 1991 1994 1997 2000 2003
2006 of2009
percentage
national2012
2
25
0
income (left axis)
271991 1994 1997 2000 2003 2006 2009 2012
1
25 Net financial investment by sector
c)
2
1994 1997
2000 2003
2006
c)1991
Net financial
investment
by sector
c) Net financial investment by sector
12
%
2009
2012
0
Economy as a
whole
c)10%Net financial investment by sector
Economy as a
12
10
8
Corporate sector
whole
6%
Corporate sector
Economy as a
Private households
8
12
whole
4
6
10
Private households
2
Corporate
sector
48
0
26
Private households
-2
04
-4
-22
Public sector
-6
-40
-8
Public
sector
-6
-2 1991
1994
1997
2000
2003
2006
2009
2012
-8
-4
1991
1994
1997
2000
2003
2006
2009
2012
Public sector
-6
Sources: AMECO; Destatis; own calculations.
-81 Top income shares including capital gains.
1994
1997 Destatis;
2000
2003
2009
2012
21991
Sources:
AMECO;
own2006
calculations.
As % of GDP.
Sources: AMECO; Destatis; own calculations.
Sources: AMECO; Destatis; own calculations.
however, this unearned income is less equally distributed than earned income and therefore makes
a significant contribution to the top income shares.
It is necessary to provide an explanation for why
the available data points to a relatively modest increase in the top income shares in Germany. After all, based on the Gini coefficient for disposable
household income, Germany has experienced one
of the sharpest rises in income inequality of all the
OECD countries over the past few decades. Between
the mid-1980s and the mid-2000s, for instance, the
page 4
Gini coefficient in Germany rose by about the same
rate as in the US (OECD 2008, 2011). According to
the WTID, over the same period, the income share
of the top 1 % in the US grew from approx. 10 % to
20 % and the income share of the top 5 % rose from
approx. 20 % to 35 %. In other words, these increases were far more pronounced than in Germany.
In this context, it is important to remember that
the Gini coefficient’s mathematical design means
that it is less sensitive to changes at the extreme
ends of the distribution scale. This is compounded by the under-recording of high incomes by voluntary household surveys such as the SOEP.4 As a
result, it is actually not so surprising that although
the Gini coefficient rose sharply in Germany during
the first half of the 2000s, the top income shares
remained stable.5 A rising Gini coefficient is indicative of a tendency for the overall distribution of
income to become more unequal. A rise in the top
income shares, on the other hand, indicates a shift
of income distribution in favour of the top 10 %, 5
% or 1 % of households, to the detriment of the vast
majority of the population.
The “corporate veil” as one of the reasons
why inequality is underestimated
Notwithstanding the above, an exclusive focus on
top income shares à la Piketty results in the increase in inequality in Germany being underestimated.
When analysing a country’s top income shares, it is
of fundamental importance always to take the macroeconomic context into account and in particular
functional income distribution trends (Infobox 1).
Capital income has risen sharply as a percentage
of Germany’s national income since the year 2000
(Figure 2b). Conversely, the wage share – i.e. earned
income as a percentage of national income – has
fallen. Much of the huge rise in corporate earnings
has been retained by companies rather than being
passed on to private households. Consequently, the
top household incomes (and their share of total
household income) have not risen as much as they
would have done if instead of retaining these increased earnings companies had followed the lead of
e.g. the US by distributing a greater proportion of
them to senior executives and shareholders (who
mostly fall within the top income groups). In the
4 According to the SOEP, a net equivalised household
income of about 45,000 euros would be enough for
someone to be classified in the top 5 % of incomes for
2012.
5 Consequently, it is not possible to fully support
the proposal of e.g. Leigh (2009) to simply use top
income shares as a substitute for other measures of
inequality over periods when alternative income
distribution measures are unavailable.
US, the reason that the wage share fell far less sharply prior to the economic downturn is precisely because it was stabilised by the high salaries of senior
executives.
The significance of this “corporate veil” is often
overlooked by analyses of economic inequality. An
approach based on top income shares à la Piketty
underestimates economic inequality in countries
like Germany. This is because while companies’ owners predominantly belong to rich households, the
growth in their income is “veiled” by the fact that
their companies retain a significant proportion of
their earnings (Infobox 1).
This corporate veil is also making it more difficult to measure wealth inequality. According to
the SOEP, the average business assets reported
by households between 2002 and 2012 fell from
approx. 212,000 euros to approx. 191,000 euros
(Grabka and Westermeier 2014, p. 159). In view of
the fact that the National Account System indicates
a sharp rise in profits and wealth for the economy
as a whole, there would appear to be some doubt
regarding the validity of these figures. It is clear that
the value of stakes in partnerships and (unlisted)
joint-stock companies is not being accurately reported by the households in the survey (“corporate
veil”). Indeed, the leeway that exists in terms of how
a company’s valuation is reported means that this is
to some extent even true of listed companies.
IMK Report 99e
October 2014
Alternative indicators of inequality
Piketty (2014) identifies two empirical phenomena
that are indicative of a rise in economic inequality.
Firstly, in many economies, the top incomes’ share
of the total pre-tax income of private households
has risen significantly. Secondly, the ratio of private
wealth to national income has also risen in several
countries. This indicates a trend of growing inequality, since wealth is less equally distributed than income. The fact that the ratio of unearned income
from inheritances to earned income is also growing
only serves to consolidate this economic inequality.
There is no question that the data currently being
employed in Germany is unsatisfactory. However,
by making certain simplifying assumptions and
taking Germany’s specific circumstances into account, it is possible for Piketty’s indicators to be adapted in order to allow the key current and possible
future trends of inequality in Germany to be identified (Infobox 2).
page 5
I n f ob ox 1
On the relationship between
functional and personal income
distribution and macroeconomic development
There has been increasing discussion in recent
international research of the relationship between
high levels of inequality and macroeconomic instability (Rajan 2010, Stiglitz 2012, van Treeck
and Sturn 2012). Both functional and personal income distribution should in principle be included
in analyses of the macroeconomic implications of
income distribution. Functional distribution, which
is based on NAS data, breaks gross domestic
product down into wage income on the one hand
and entrepreneurial and capital income on the
other. Personal income distribution describes the
distribution of households’ pre-tax or disposable
income and is generally derived from voluntary
household surveys (e.g. SOEP) or tax statistics
(e.g. WTID). Different macroeconomic trends can
result depending on whether the predominant
changes relate to functional or personal income
distribution shocks. In particular, changes in distribution can result in either overindebtedness of
private households or current account imbalances, both of which can cause macroeconomic
instability over the longer term.
As a result, the Macroeconomic Policy Institute (IMK) carried out a research project1 to investigate the macroeconomic repercussions of
changes in personal and functional income distribution. A simplified description of the potential
impacts of these different shocks is presented
in Boxes 1.1a-1.1c. One of the key components
of the analysis is the relative income hypothesis
(Duesenberry 1949; Frank 2005), according to
which at least part of a household’s consumption
(“positional goods”) is dictated by the consumption of a reference group that generally comprises higher-income households (upward-looking
status comparisons). This type of consumption
is by no means irrational or solely restricted to
the consumption of luxury goods. One example
is provided by household spending on education
in countries where the “good” schools and universities are mostly private. If rising inequality drives
up the cost of a “good” education because the
upper income groups are spending more in this
area, then even many upper middle-class families can find themselves facing difficult decisions
about where there priorities lie (e.g. providing
their children with a relatively good education
F igu r e B ox 1.1
IMK Report 99e
October 2014
Macroeconomic effects of
a)distributional
Simplified baseline
scenario
changes
a) Simplified baselinescenario
scenario
scenario
Simplified baseline
baseline
120
a)a)Simplified
120
100
120
10
10
100
80
100
35
20
10
10
20
1010
80
60
80
35
35
20
20
30
2020
65
30
30
70
60
40
40
20
0
20
0
0
65
65
40
gDP
40
40
gni
gDP
gDP
gni
gni
7070
Domestic
demand
nx
Domestic
Domestic
demand
nxnx
demand
b) Simplified impact of a personal income
shock
Simplified
impact
a personal
income
b)b)Simplified
Simplified
impact
of
personal
income
shock
120
b)
impact
ofaaof
personal
income
shockshock
100
120
120
10
80
100
100
35
80
60
80
35
35
60
40
60
65
40
20
40
20
0
20
-200
0
-20
-20
20
10
10
20
20
40
40
40
30
65
65
10
20
10
10
20
20
80
80
80
30
30
gDP
gni
gDP
gDP
gni
gni
-10
Domestic
demand
Domestic
demand
Domestic
demand
-10
nx
-10
nx
nx
c) c)
Simplified
impact
of a of
functional
income
shock shock
Simplified
impact
a functional
income
c) Simplified impact of a functional income shock
c) Simplified impact of a functional income shock
120
120
100
120
100
80
100
80
60
80
60
40
60
40
20
40
20
0
20
0
10
10
30
10
30
47
47
47
30
30
53
53
10
10
20
2010
20
30
60
30
30
60
10
60
10
Domestic
nx
demand
Domestic
nx
10
demand
0
gDP
gni
Domestic
nx
1. pillar: GDP
3. pillar:
DD chosen for
Graphs
do not include primary income.
Values
demand
profits [P]
government spending [G]
Graphs
do[W]
not include primary income.
Values chosen for
illustrative
purposes.
+ wages
+ corporate investment [I]
illustrative
purposes.
= gross domestic
product [GDP]
+ private consumption [C]
Source:
calculations.
Graphs IMK
do not
include primary income.
Values
chosen
= domestic
demand
[DD] for
Source: IMK calculations.
53
gDP
gDP
30
gni
30
gni
illustrative purposes.
2. pillar: GNI
Source:
taxes IMK
[T] calculations.
+ corporate income [Y_F]
+ top household income
[Y_H_T]
+ middle and bottom household
income [Y_H_MU]
= gross national income [GNI]
4. pillar: NX
government financial balance
[T-G]
+ coporate financial balance
[Y_F-I]
+ household financial balance
[Y _H_MU + Y_H_T - C]
= net exports [NX]
Graphs do not include primary income and depreciation.
Values chosen for illustrative purposes.
Source: IMK calculations.
1 For more information, visit http://ineteconomics.
org/grants/income-inequality-household-debt-andcurrent-account-imbalances
page 6
I n f ob ox 1
vs. saving enough money for their retirement).
This example also demonstrates how very country-specific institutions such as highly developed
public services or restrictions on credit-based
consumption can temper the influence of reference groups on consumption and help to keep
household debt in check. “Positional arms races”
(Frank 2007), for instance, are thought to play a
greater role in the US than in Germany.
The analysis also drew on the “corporate veil”
concept described elsewhere in this report.
Whilst it is true that enterprises ultimately belong
to households, if a company’s value increases in
the form of retained earnings, for instance, this
increase tends not to be fully reflected in the consumer behaviour of its owners (top earners). Accordingly, the increase also has little influence on
the consumer behaviour of other households who
look to the top earners as a reference. The retention of corporate earnings also weakens domestic demand by inhibiting correspondingly higher
levels of private investment (Lindner 2014).
The IMK study also makes the assumption that
the observed changes in income distribution
mostly involve shocks to permanent (as opposed
to only transitory) income components – in other
words, that it is always the same households that
are affected by these changes in distribution.2
This precludes e.g. an explanation of increased
household debt purely in terms of consumption
smoothing.
The simplified impacts of personal and functional
income shocks are illustrated below using simple numerical examples (Boxes 1.1a-1.1c). In the
baseline scenario (Box 1.1a), the wage share is
65 % and the capital income share is 35 % (Column 1). The top income share is approximately
43 % (3/7) of disposable household income, while
disposable corporate income accounts for 20 %
of the national income and disposable household
income for 70 % (Column 2). Domestic demand is
equal to the national income (Column 3), in other
words the current account is balanced (Column 4).
Box 1.1b shows the simplified impacts of a personal income shock in an institutional setting where
a rise in the incomes of top earners (Column 2)
leads not only to an increase in their own consumption but also to higher levels of credit-based
consumption among lower-income groups for
2 Bartels and Bönke (2013, Figures 3 and 4) illust-
rate the development of transitory and permanent
changes in real income in Germany between 1985
and 2006. Whilst a more pronounced increase was
recorded in terms of permanent changes to gross
household income, the changes to both the permanent and transitory components of net household
income were relatively low. This would appear to
be at odds with the evolution of the Gini coefficient
over the same period.
whom the top earners act as a reference group
(Column 3). In this scenario, the smaller income
share of the lower and middle income groups
does not lead to a decline in the wage share as
long as the rise in top earners’ incomes is primarily accounted for by earned income (e.g. executive salaries, bonuses, etc.). The wage share in
this example remains the same, at 65 % (Column
1), as does the share of disposable household
income, at 70 %. However, the top income share
rises to around 57 % (4/7). The increase in
household debt is accompanied by a decline in
households’ financial balance, resulting in a current account deficit (Column 4).
Box 1.1c illustrates the simplified impacts of a
functional income shock. The decline in the wage
share to 53 % (Column 1) is due primarily to the
detriment of low- and middle-income households
(Column 2). This decline is accompanied by a
rise in disposable corporate income to 30 % (Column 2) and a corresponding fall in disposable
household income to 60 % of the national income.
The top income share of disposable household
income rises less sharply than in the previous
scenario, to a level of 50 % (3/6). If the institutional framework also curbs the incentives for
reference-based consumption, the savings rate
among lower income groups does not fall sufficiently to compensate for the drop in disposable
income. This in turn translates into weaker domestic demand (Column 3) and fuels an export-led
growth model with substantial current account
surpluses (Column 4). Demand for credit among
private households is correspondingly weak.
The interactions described above are based on
empirical evidence (Behringer and van Treeck
2013) as well as theoretical models (Belabed,
Theobald and van Treeck 2013). Box 1.2 provides a purely descriptive illustration of the correlation between measures of personal and functional income distribution for the G7 countries and
China. There is a trend towards a negative (positive) correlation between the financial balance
of corporations (the wage share) and the top income shares. The decline in the wage share has
been less pronounced in countries where a sharp
rise in top earners’ income since the beginning of
the 1980s has been accompanied by companies
saving less over the same period.
In summary, it can be seen that changes in both
personal and functional income distribution can
have destabilising macroeconomic impacts.
In the case of a personal income shock, rising
inequality in household incomes can lead to
an increase in household debt, a decline in private households’ financial balance and a current account deficit. In the case of a functional
income shock, a redistribution of income from
IMK Report 99e
October 2014
page 7
I n f ob ox 1
households to businesses can result in weaker
domestic demand, a rise in the financial balance
of corporations and a current account surplus. As
illustrated by the case studies in Behringer et al.
IMK Report 99e
October 2014
(2013), the US and Germany provided prime examples of these opposing trends in the run-up to
the financial crisis.
F igu r e B ox 1.2
Correlation between personal and functional income distribution
Correlation
Correlation between
between personal
personal and
and functional
functional income
income distribution
distribution
Top
income
share
Top
income
share
and corporate
corporate
saving
Top
income
share
and and
corporate
savingsaving
Top
income
share
and
Top
income
share
and wage
wage share
share
Top income
share
and wage
share
14
14
USA
USA
12
12
10
10
GBR
GBR
JPN
JPN
66
22
ITA
ITA
CHN
CHN
JPN
JPN
ITA
ITA
44
10
10
GBR
GBR
CAN
CAN
88
00
12
12
USA
USA
GER
GER
CHN
CHN
66
88
FRA
FRA
88
66
CAN
CAN
44
GER
GER
22
FRA
FRA
00
22
44
10
10
-20
-20
-15
-10
-5
-15
-10
-5
Change
Change in
in wage
wage share
share as
as aa %
% of
of GDP
GDP
Change
Change in
in corporate
corporate financial
financial balance
balance as
as aa %
% of
of GDP
GDP
00
00
Change in top 5% income share
Change in top 5% income share
14
14
The illustrations show the correlation between measures of personal and functional income distribution for the G7 and China. All figures relate to the
period 1980/3 – 2004/7 (four-year averages) except for the UK (1984/7 – 2004/7) and China (1992/5 – 2000/3).
= Canada; CHN = China; GER = Germany ; FRA = France; GBR = Great Britain; ITA = Italy; ; JPN = Japan;
The
TheCAN
illustrations
illustrations
show
show the
the correlation
correlation between
between measures
measures of
of personal
personal and
and functional
functional income
income distribution
distribution for
for the
the G7
G7 and
and China.
China.
USA = USA
All
All figures
figures relate
relate to
to the
the period
period 1980/3
1980/3 –– 2004/7
2004/7 (four-year
(four-year averages)
averages) except
except for
for the
the UK
UK (1984/7
(1984/7 –– 2004/7)
2004/7) and
and China
China (1992/5
(1992/5 ––
Source: IMK calculations.
2000/3).
2000/3).
Source:
Source: IMK
IMK calculations.
calculations.
Enhanced top income shares
Top income shares are intended to provide us with
an idea of income distribution across two social
groups: top earners on the one hand and all other
members of society on the other.6 However, the increasing divergence between these two groups that
can be seen in many places around the world is expressed differently in different countries. In the US
and the UK, for example, WTID figures show that
the top incomes’ share of total household income
has risen dramatically since the 1980s. In Germany,
on the other hand, the most striking phenomenon
is the increase in retained corporate earnings as a
percentage of total private income since the early
2000s. The WTID’s top income shares do not take
this type of income into account since it is not subject to personal income tax. As a result, they fail to
provide a complete picture of income polarisation.
In Figure 3, the WTID data on top income shares is therefore combined with the NAS data on
retained corporate earnings before being compared against total private income. This highly simplified approach, which is employed for illustrative
purposes, makes the assumption that all corporate
earnings can be allocated to the top earners. The il6 Kumhof et al. (2010) refer to “investors” (rich
households and corporations) and “workers”.
lustration reveals a much greater increase in top income shares when they are adjusted in this manner.
Figure 2c illustrates the macroeconomic context
in which the rise in retained corporate earnings in
Germany should be interpreted. The financial balanF igu r e 3
Top Topeinkommensanteile
income shares
adjusted
to
korrigiert
um einbehaltene
unternehmensgewinne
take account of retained corporate
earnings1
1
50
%
45
40
35
30
25
20
15
10
5
0
1
1995
1998
2007
Top 1% of households
Top 1% Haushalte
Top 5% - top 1% of households
Top10%
5%- top
- Top
Haushalte
Top
5% of1%
households
Retained
corporate
Top
10%
- Topearnings
5% Haushalte
The
graph shows the Unternehmensgewinne
evolution of top household incomes and
Einbehaltene
retained corporate earnings as a percentage of private pre-tax
income.
Die
Abbildung zeigt den Anteil der Top -
1
Haushaltseinkommen
und der
einbehaltenen
Sources: AMECO; World Top Incomes
Database;
IMK calculations.
Unternehmensgewinne
an den privaten
Vorsteuereinkommen im zeitlichen Verlauf.
page 8
Quellen: AMECO; World Top Incomes Database; Berechnungen des IMK.
ce for joint-stock companies has been positive since
2002 – in other words the corporate sector as a whole has been accumulating additional net financial
wealth on an annual basis over this period. This phenomenon is highly unusual both historically and internationally. Since neither the corporate sector nor
the State (owing to the zero-structural-deficit-rule)
is absorbing the traditionally positive net savings of
private households, Germany has a structural current account surplus. While this trend can clearly be
interpreted as an expression of problems in the realm
of income distribution, it also has a destabilising impact on the economy as a whole (Infobox 1).7
Wealth-to-income ratios
Figure 4a shows the ratio of net wealth to income
for different definitions of wealth and income. The
black line in Figure 4a shows the net wealth of private households as a percentage of the national income. Since 1991, this ratio (which Piketty refers to
as “β”, see Infobox 2) has risen from around 300 %
to over 400 %. Over the same period, the net wealth
of the economy as a whole experienced a far more
modest increase owing to the decline in net government assets. Since the mid-2000s, the net wealth of
the economy as a whole has risen sharply in relation
to national income as a result of the increase in the
net wealth of private households and private companies. It is also noticeable that, since the beginning
of the 2000s, the net wealth of private households
has risen far more sharply in relation to disposable household income than it has in relation to the
national income. Prior to this date, these two alternative ways of calculating β had largely tracked
each other. The divergence of these two ratios over
the past decade and a half once again highlights the
weak growth of household incomes in relation to
(retained) corporate earnings.
7 Wagner (2011) sums up this situation by pointing out
that the core of the problem is not so much government debt as the huge imbalances in international
trade: Germany’s economic model has contributed
significantly to the instability of the Eurozone. It
has pursued an excessive export strategy, supported
by stagnating real wages. If the incomes of the vast
majority of the population rise only slowly, then
domestic demand will also be weak. This also means
that people cannot buy more goods and services from
abroad. However, if a country keeps producing more
goods than it can consume itself, this inevitably leads
to other countries becoming indebted. The basis of
these countries’ economies – i.e. jobs and thus also
tax revenue – has been gradually eroded, while
Germany has continued to accumulate wealth.
From a distribution perspective, a rise in β is
relevant if wealth is less equally distributed than
income. In this context, one interesting indicator
is the ratio of the net wealth of a relatively well-off
household to the disposable income of an average
household. Figure 4b therefore estimates the net
wealth of the 99th percentile of wealth based on a
combination of data from the SOEP and the figures for the wealth of the economy as a whole (Gesamtwirtschaftliche Vermögensbilanz). The SOEP
wealth distribution data used were for the years
2002, 2007 and 2012. The change in the net wealth
of the 99th percentile post-2002 therefore corresponds to the average growth in net wealth as measured by the figures for the wealth of the economy
as a whole. The net wealth values calculated using
this method (“p99 Destatis”) were then compared
against the median equivalised household income
figures provided by the SOEP (“Median SOEP”).
The advantage of using median income values is
that they are less prone to underestimating high
incomes and they reflect the financial situation of
the average citizen. According to our calculations –
which are still likely to underestimate the net wealth
of the top percentiles, since this is under-recorded
by the SOEP – in 2012, the 99th percentile had a net
wealth 80 times higher than the annual income of
the median German household. Ten years earlier,
the same figure was only 50 times higher. If we were
to rely purely on SOEP data, then this trend would
apparently be far less pronounced (Figure 4b).
Another question that needs to be tackled is the
extent to which this increase in the ratio between
the top percentiles’ wealth and average incomes is
contributing or may contribute in future to the perpetuation of economic inequality down the generations. It is helpful to consider the age-wealth profile
when addressing this issue (Figure 4c). It is interesting to note that there is only an extremely modest
decline in people’s wealth in the years before they
die (as a result of them spending their savings). The
average net wealth for the over-81 age group is not
significantly lower than for people aged 65. In other
words, most wealth in Germany is inherited by the
next generation. Figure 4d shows that the proportion of Germany’s national income accounted for by
inheritances (Schinke 2012) has risen continuously
since 1960. As such, there is a danger that economic
inequality will be perpetuated from one future generation to the next (Infobox 2).
IMK Report 99e
October 2014
page 9
F igu r e 4
Wealth-to-income ratios
p99 Destatis
p90
p90Destatis
Destatis
p90
p99Destatis
SOEP
p90 SOEP
p90
SOEP
p90 SOEP
p90
SOEP/
p90
SOEP/
p90
SOEP/
p99
SOEP/
Median
SOEP
Median
SOEP
Median
SOEP
Median
SOEP
p90
Destatis/
p90
Destatis/
p90
Destatis/
p99
Destatis/
Median
SOEP
Median
SOEP
Median
SOEP
Median SOEP
90
90
90
90
80
80
80
80
70
70
70
70
60
60
60
60
50
50
50
50
40
40
40
40
30
30
30
30
20
20
20
20
10
10
10
10
00
00
2002
2002
2002
2002
2007
2007
2007
2007
2012
2012
2012
2012
2
EUR
EUR
1,600,000
EUR
EUR
1,600,000
1,600,000
1,600,000
1,400,000
1,400,000
1,400,000
1,400,000
1,200,000
1,200,000
1,200,000
1,200,000
1,000,000
1,000,000
1,000,000
1,000,000
800,000
800,000
800,000
800,000
600,000
600,000
600,000
600,000
400,000
400,000
400,000
400,000
200,000
200,000
200,000
200,000
0 0
00
d) Rising
Rising importance
importanceofofinheritances
inheritances
d)
d)d)Rising
Rising
importance
of
inheritances
d)
inheritances
Rising importance
importance
ofof
inheritances
c)c)Income
Incomeby
byage
agegroup
group
c)c)Income
Income
by
age
group
c)
Incomeby
byage
age group
group
EUR
EUR
EUR
EUR
200,000
200,000
WesternGermany:
Germany:
Western
200,000
200,000
Western
Western Germany:
Germany:
180,000
180,000
2002
2002
180,000
180,000
2002
2002
160,000
160,000
160,000
160,000
140,000
140,000
140,000
140,000
2007
120,000
2007
120,000
2007
2007
120,000
120,000
100,000
100,000
100,000
100,000
80,000
80,000
2012
80,000
80,000
2012
60,000
2012
2012
60,000
60,000
60,000
40,000
40,000
2007
40,000
40,000
2007
20,000
2007
2007
20,000
20,000
20,0000
00
26-30 36-40 46-50 55-60
0 <20
<20
26-30
36-40 46-50
55-60
<20
26-30
<20
26-30 36-40
36-40 46-50
46-50 55-60
55-60
2
p90
SOEP/
p90
SOEP/
p90
SOEP/
p99
SOEP/
MedianSOEP
SOEP
Median
Median
SOEP
Median
SOEP
p90
Destatis/
p90
Destatis/
p90
Destatis/
p99
Destatis/
MedianSOEP
SOEP
Median
Median
SOEP
Median
SOEP
800
800
Net wealth of economy as a
800
800
Net
wealth
of
economy
as
1
Netwealth
wealth
ofeconomy
economyas
asaaa
whole
700
Net
of
1
whole
700
whole11
700
whole
700
600
600
600
600
Net wealth of private
500
Net wealth1 of private
500 Net
wealth of
of
private
Net
wealth
500 households
1 private
500
households
households11
households
400
400
400
400
Net
Net wealth
wealth of
of private
private
300
Netwealth
wealthof
of
private
Net
private
300
households
as
aa %
households
as
% of
of
300
300
households
as
a % of
households
as
national
income
national incomea % of
Net
wealth
ofofjoint-stock
Net
wealth
joint-stock
200
national
income
200
national
income
1
Net
wealth
of
joint-stock
Net
wealth
of
joint-stock
companies
200
companies1
200
companies11
companies
100
100
1
100
100
Net
Netgovernment
government wealth
wealth11
Netgovernment
governmentwealth
wealth1
Net
00
001991 1994
1991
2009
2012
1994 1997
1997 2000
2000 2003
2003 2006
2006
2012
1991 1994
1994 1997
1997 2000
2000 2003
2003 2006
2006 2009
2009 2012
2012
1991
Ratio of
to median
income
b)b)Ratio
ofp99
p90wealth
wealth
to median
income
2
b)
Ratio
of
p90
wealth
b)Ratio
Ratioof
ofp90
p90wealth
wealthto
medianincome
income2 2
b)
totomedian
median
income
p90
SOEP/
p90
SOEP/
p90
SOEP/
p99
SOEP/
Median
SOEP
Median
SOEP
Median
SOEP
Median
SOEP
p90
Destatis/
p90
Destatis/
p90
Destatis/
p99
Destatis/
Median
SOEP
Median
SOEP
Median
SOEP
Median
SOEP
Wealth-income ratios
Wealth-income ratios
ratios
Wealth-income
ratios
a) Wealth by sector
Wealth-income
a) Wealth by sector
a)
Wealth
by
sector
a) Wealth
Wealth by
by sector
sector
a)
IMK Report 99e
October 2014
2012
2012
2012
2012
%
%
18
18 %%
18
18
15
15
15
15
12
12
12
12
Eastern
Eastern
Eastern
Eastern
Germany:
Germany:
Germany:
Germany:
2002
2002
2002
2002
66-70 76-80
66-70
76-80
66-70
66-70 76-80
76-80
99
99
Inheritances
a%
of of
Inheritancesasas
a%
Inheritances
asasaa%%ofof
Inheritances
national
nationalincome
income
national
nationalincome
income
66
66
3
3
33
0
0001910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
1910
1910 1920
1920 1930
1930 1940
1940 1950
1950 1960
1960 1970
1970 1980
1980 1990
1990 2000
2000 2010
2010
11
income. income.
asa %a of%national
of national
11 as
1 2as a % of national income.
as
income.
as
a%
% toof
ofthenational
national
income.
p99 a
refers
99th percentile
of the relevant distribution. „p99 Destatis“ was calculated by combining the SOEP wealth distribution figures
with the average wealth trends according to Federal Statistical Office data: p99 Destatis = (p99 SOEP/per capita SOEP)*per capita Destatis.
Per capita Destatis was calculated based on the population over 14 years of age according to AMECO data.
2
Schinke (2012).
p90 refers
to the 90th percentile of the relevant distribution. "p90 Destatis" was calculated by
22 Source:
2p90
relevant distribution.
distribution."p90
"p90Destatis"
Destatis"was
wascalculated
calculatedby
by
p90 refers
refers to
to the
the 90th
90th percentile of the relevant
p90 refersthe
to the
90th
percentile
of the figures
relevantwith
distribution.
"p90
Destatis"
was
calculated
by
combining
SOEP
wealth
distribution
the average
wealth
trends
according
to Federal
combining
distribution
figures
with
the
average
wealth
trendsDestatis.
according
toFederal
Federal
combining
the
SOEP
figures
with
average
wealth
according
toto
combining
theSOEP
SOEP
wealth
distribution
figures
withthe
the
average
wealthtrends
trends
according
Federal
Statistical the
Office
data:wealth
p90 Destatis
= (p90
SOEP/per
capita
SOEP)*per
capita
Per
capita
Statistical
Office
data:
p90
Destatis
(p90
SOEP/per
capita
SOEP)*per
capita
Destatis.
Per
capita
Statistical
Office
data:
=
(p90
SOEP/per
capita
SOEP)*per
capita
Destatis.
Per
capita
Statistical
Office
data: p90
Destatis
= (p90
SOEP/per
capita
SOEP)*per
capita Destatis.
Perdata.
capita
Destatis was
calculated
based
on the
population
over 14
years
of age according
to AMECO
Destatis
was
calculated
based
population
over
14
years
of
age
according
to
AMECO
data.
Destatis
was
calculated
on
the
population
over
14
years
of
age
according
to
AMECO
data.
Destatis was calculated based on the population over 14 years of age according to AMECO data.
Source: Schinke (2012)
Source:
Source:
Schinke
(2012)
Source:Schinke
Schinke(2012)
(2012)
page 10
I n f ob ox 2
An illustration of Piketty’s
“fundamental laws of capitalism” and long-term trends in
income and wealth inequality
Piketty (2014) proposes a simple model for describing the interaction between income inequality and wealth inequality. His model for what he
calls the “fundamental laws of capitalism” lives
up to its billing, comprising nothing more than
an identity equation (Equation 1) and a simple
arithmetical principle (Equation 2). However,
notwithstanding fundamental theoretical disputes (e.g. Keynesian vs. neoclassical macroeconomics), the model can be said to be valid as
long as a steady-state approach is supposed to
be acceptable.
The “first fundamental law” states that α (defined
as the ratio of capital income (P) to the national
income (Y)) is equal to the return on capital (r)
multiplied by β (defined as the net wealth of the
economy as a whole (W) divided by the national
income (Y)):
(1)
α = P/Y = r*β = rW/Y
According to the “second fundamental law”, in a
long-term steady state, β converges with the ratio
between the savings rate for the economy as a
whole (s) and the nominal growth rate of the national income (g):
(2)
β = s/g
Piketty makes two empirical observations that
highlight the importance of these relationships to
income and wealth distribution trends:
Firstly, high-income groups save a greater proportion of their income and bequeath a larger percentage of their income than low-income groups.
This plays a key role in causing wealth to be less
equally distributed than income and in ensuring
that the importance of inheritances relative to earned income increases over the course of time.1
Secondly, Piketty argues that historically, the return on capital (r) has often exceeded the rate
of economic growth (g). What this means is that
if the owners of capital save a sufficiently large
proportion of their income, capital will tend to
outpace earned income. Under certain circumstances, this results in a continuous rise in the
wealth-income ratio (β), meaning that capital income accounts for a greater and greater share of
the national income (α). Ultimately, this translates
1 The concentration of wealth is exacerbated by low
population growth.
into a constant growth in income inequality.2
In order to gain a better understanding of the interaction between income inequality and wealth
inequality, it may be helpful to illustrate how
Piketty’s model works using a few concrete numerical examples.3 Given the model’s simplicity
and the necessary simplifying assumptions, the
simulations outlined below are primarily for illustrative purposes. Nevertheless, the trends shown
by the processes that they describe are perfectly
realistic.
In Table 2.1 the model was “calibrated” so that
its key ratios and parameters in Period 0 essentially reflected the situation in Germany in the
early 2000s.
In Period 0, the model is in steady state. In other
words, as long as its parameters are not altered,
both the ratios α and β and the distribution of income (Y) and wealth (W) will remain unchanged.
Households are divided into three groups (T:
top, M: middle, U: lower). The simplifying assumption is made that the income and wealth
quantiles coincide and remain stable over time.4
In the interests of simplicity, it is also assumed
that the return on capital will be the same for all
households.5 Since the model does not include a
corporate sector, the top households represent
both wealthy households and businesses. Moreover, since the State is also not represented in
the model, no distinction is made between gross
and net income and pre- and after-tax rates of
return. The savings rates of the three income
groups are income-based and have been cho-
IMK Report 99e
October 2014
2 Formally speaking, β will continue to rise infinitely
if sProfitr > g, where sProfitr is the savings rate for
capital income. The reason that a high r-g ratio is
so explosive in income distribution policy terms is
because the different income groups have different
savings rates. If savings rates were unconnected
to income, the wealth-income ratio of individual households would not be dependent on their
income either. Furthermore, if savings rates were
uniformly distributed, then in the long term wealth
and income distribution would become identical to
wage distribution and the r-g ratio would be irrelevant to income distribution trends.
3 The excel file, on which this simulations are based,
is available online (van Treeck 2014b).
4 Empirically, a rise in income inequality can be
caused either by a rise in transitory or permanent
changes in income or by a combination of both
factors (Bartels and Bönke 2013).
5 In actual fact, households that have a lot of wealth
are typically able to obtain a better return on
wealth, since a large portfolio can be more easily
diversified and is better able to incorporate a
larger proportion of higher-risk investments that
also offer higher returns. Furthermore, wealthier
households tend to be better informed about attractive investment opportunities.
page 11
I n f ob ox 2
sen so that the β value for the economy as a
whole and the individual β values remain constant.6 In other words, wealth and income grow at
the same rate. This baseline period clearly demonstrates that – contrary to what is often claimed – if r is greater than g this in no way means
that both β and inequality will inevitably continue
to rise indefinitely.7
Table 2.1a illustrates Period 1, where a shock to
wage income distribution that benefits top earners is accompanied by a rise in returns on capital. This results in a direct rise in the top income
share from 25 % to 35 %, along the same lines as
in Figure 3. The capital income share rises from
27 % to 32 %, mirroring the trend shown in Figure 2b. It is interesting to observe how things
develop over the subsequent periods. Initially,
wealth inequality is largely unaffected by the increase in wage and income inequality. However,
since the top income groups save a relatively
high proportion of their increased income, wealth
inequality also gradually increases. This in turn
has the effect of exacerbating income inequality.
(The greater the differential between savings rates, the stronger the effect.) After 15 periods, the
top wealth share has risen from 60 % to 64 %,
after 30 periods it has reached 67 %, after 50 periods it stands at 70 % and in the new long-term
steady state the top wealth share climbs to 81
%. As a result, the top income share rises to 51
% over the long term, even though the top wage
share remains at 23 %. This demonstrates how
differences in the baseline wage and wealth distribution can be exacerbated over time as a result
of differences in the savings rates of the different
income groups.
In Table 2.1b, the overall rate of economic growth
(g) is reduced from 3 % in Period 0 to 1 % as
of Period 1. The assumption that the nominal in-
6 E.g.: β = s/g = 0.108/0.03 = 3.6;
βM = sM/g = 0.0897/0.03 = 2.99.
come growth rate will decline is in line with the
trend forecast by Piketty and several other economists and demographers who claim that we
can expect lower population growth (accompanied by lower income growth) and even “secular
stagnation” (Summers 2014) over the next few
decades. Whilst initially the simulation illustrated
in Table 2.1b develops almost identically to that
in Table 2.1a, over the longer term it displays a
much stronger tendency towards greater inequality. This is because the r-g ratio rises while savings rates remain unchanged. The result is that
wealth and capital income increase significantly
faster than the national income. Even after 50 periods, the top income share has reached 60 %,
while over the longer term, α, β and income and
wealth inequality all continue to rise indefinitely.
In Table 2.1c, the differential between the savings
rates of the top and middle income groups is also
increased. This is a trend that has been apparent
in Germany for some years as a result of a rise
in corporate saving and it can also be detected in
the SOEP household savings rates. This phenomenon further exacerbates the rise in inequality.
In Period 50, the top households already account
for 82 % of all wealth (as opposed to 73 % in Table 2.1b) and 71 % of all income (compared with
60 % in Table 2.1b).
Based on these simulations, the SOEP’s finding
that there has been a sharp rise in income inequality over the past decade but almost no change
in wealth inequality is unlikely to remain valid
over the longer term. By its very nature, wealth
inequality is initially slow to react to changes in
income distribution – not only is it starting at a
much higher level, but it also takes time to accumulate wealth through savings. Nevertheless, in
the long term both wealth inequality and income
inequality can be expected to keep rising unless
the appropriate economic policy measures are
taken to counter them.
IMK Report 99e
October 2014
7 The reason for this is that the savings rate for top
earners is “too low”.
page 12
I n f ob ox 2
IMK Report 99e
October 2014
Ta b l e B ox 2.1
Some simple simulations based on a version of Piketty‘s (2014) model1
Rise
in top
wage incomes
and capital
a) Rise in a)
top
wage
incomes
and capital
returnsreturns
Period
Period
Period
0
10
0
21
1
32
2
43
3
54
4
105
5
10
15
10
15
30
15
30
50
30
50
80
50
80
100
80
100
1000
100
1000
1000
alpha
alpha
(=P/Y)
(=P/Y)
alpha
0.27
(=P/Y)
0.27
0.32
0.27
0.32
0.33
0.32
0.33
0.33
0.33
0.33
0.33
0.33
0.33
0.33
0.33
0.33
0.34
0.33
0.34
0.35
0.34
0.35
0.37
0.35
0.37
0.4
0.37
0.4
0.42
0.4
0.42
0.43
0.42
0.43
0.48
0.43
0.48
0.48
a) Rise in top wage incomes and capital returns
a) Rise
and
capital
returns
Shareinoftop
l wage
r incomes
g
Savings
rates
T Share
M of lu
T Share
M of l u
0.13
0.45
0.43
T
M
u
0.13 0.45
0.45 0.32
0.43
0.23
0.13
0.45
0.43
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23
0.45
0.32
0.23 0.45 0.32
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.23 0.45
0.45 0.32
0.32
0.23 0.45 0.32
r
r
0.08
0.08
0.09
0.08
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
g
g
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
TotalSavings
T rates
M
TotalSavings
T rates
M
0.11
0.26
0.09
Total
T
M
0.11
0.26 0.09
0.09
0.13
0.26
0.11
0.26
0.09
0.13
0.26 0.09
0.09
0.13
0.26
0.13
0.26
0.09
0.13
0.26 0.09
0.09
0.13
0.26
0.13
0.26
0.09
0.13
0.26
0.09
0.13
0.26 0.09
0.13
0.26
0.09
0.13
0.26 0.09
0.09
0.13
0.26
0.13
0.26
0.09
0.13
0.26 0.09
0.09
0.13
0.26
0.13
0.26
0.09
0.13
0.26 0.09
0.09
0.14
0.26
0.13
0.26
0.09
0.14
0.26 0.09
0.09
0.14
0.26
0.14
0.26
0.09
0.14
0.26
0.09
0.14
0.26 0.09
0.14
0.26
0.09
0.14
0.26 0.09
0.09
0.15
0.26
0.14
0.26
0.09
0.15
0.26 0.09
0.09
0.15
0.26
0.15
0.26
0.09
0.15
0.26 0.09
0.09
0.16
0.26
0.15
0.26
0.16
0.26 0.09
0.09
0.16
u
u
0.02
u
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.26 0.09 0.02
b) Fall in economic growth rate
b) Fall in economic
growth rate
b) Fall in economic
growth rate
inofeconomic
rate Savings rates
Period alpha b) Fall
Share
l
r growth
g
Period
Period
0
10
0
21
1
32
2
43
3
54
4
105
5
10
15
10
15
30
15
30
50
30
50
80
50
80
100
80
100
1000
100
1000
alpha
(=P/Y)
(=P/Y)
alpha
0.27
(=P/Y)
0.27
0.32
0.27
0.32
0.33
0.32
0.33
0.34
0.33
0.34
0.35
0.34
0.35
0.36
0.35
0.36
0.4
0.36
0.4
0.44
0.4
0.44
0.57
0.44
0.57
0.74
0.57
0.74
1.02
0.74
1.02
1.23
1.02
1.23
3113
1.23
3113
T Share
M of lu
T Share
M of l u
0.13
0.45
0.43
T
M
u
0.13 0.45
0.45 0.32
0.43
0.23
0.13
0.45
0.43
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23
0.45
0.32
0.23 0.45 0.32
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.32
0.23 0.45
0.45 0.32
0.23
0.32
0.23
0.45
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.45
0.32
0.23 0.45
0.45 0.32
0.32
0.23
0.23
0.23 0.45
0.45 0.32
0.32
1000 3113 0.23 0.45 0.32
r
r
0.08
0.08
0.09
0.08
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
g
g
0.03
0.03
0.01
0.03
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
TotalSavings
T rates
M
TotalSavings
T rates
M
0.11
0.26
0.09
Total
T
M
0.11
0.26 0.09
0.09
0.13
0.26
0.11
0.26
0.09
0.13
0.26 0.09
0.09
0.13
0.26
0.13
0.26
0.09
0.13
0.26 0.09
0.09
0.13
0.26
0.13
0.26
0.09
0.13
0.26 0.09
0.09
0.13
0.26
0.13
0.26
0.09
0.13
0.26 0.09
0.09
0.13
0.26
0.13
0.26
0.09
0.13
0.26 0.09
0.09
0.14
0.26
0.13
0.26
0.09
0.14
0.26
0.09
0.14
0.26 0.09
0.14
0.26
0.09
0.14
0.26 0.09
0.09
0.16
0.26
0.14
0.26
0.09
0.16
0.26 0.09
0.09
0.18
0.26
0.16
0.26
0.09
0.18
0.26 0.09
0.09
0.22
0.26
0.18
0.26
0.09
0.22
0.26 0.09
0.09
0.26
0.26
0.22
0.26
0.09
0.26
0.26 0.09
0.09
661.18
0.26
0.26
0.26
661.18
0.26 0.09
0.09
661.18 0.26 0.09
s/g
s/g
s/g
3.6
3.6
4.36
3.6
4.36
4.37
4.36
4.37
4.38
4.37
4.38
4.39
4.38
4.39
4.4
4.39
4.4
4.45
4.4
4.45
4.5
4.45
4.5
4.64
4.5
4.64
4.78
4.64
4.78
4.95
4.78
4.95
5.04
4.95
5.04
5.39
5.04
5.39
5.39
beta (=W/Y)
Total beta
T (=W/Y)
M
Total beta
T (=W/Y)
M
3.6
8.53
2.98
Total
T
M
3.6
8.53
2.98
3.6
6.17
3.02
3.6
8.53
2.98
3.6
6.17
3.02
3.62
6.21
3.02
3.6
6.17
3.02
3.62
6.21
3.02
3.64
6.25
3.03
3.62
6.21
3.02
3.64
6.25
3.03
3.67
6.29
3.03
3.64
6.25
3.03
3.67
6.29
3.03
3.69
6.32
3.04
3.67
6.29
3.03
3.69
6.32
3.04
3.79
6.49
3.06
3.69
6.32
3.04
3.79
6.49
3.06
3.88
6.64
3.08
3.79
6.49
3.06
3.88
6.64
3.08
4.13
7.02
3.12
3.88
6.64
3.08
4.13
7.02
3.12
4.39
7.39
3.14
4.13
7.02
3.12
4.39
7.39
3.14
4.68
7.76
3.14
4.39
7.39
3.14
4.68
7.76
3.14
4.82
7.93
3.12
4.68
7.76
3.14
4.82
7.93
3.12
5.39
8.53
2.98
4.82
7.93
3.12
5.39
8.53
2.98
5.39
8.53
2.98
beta (=W/Y)
Total beta
T (=W/Y)
M
Total beta
T (=W/Y)
M
3.6
8.53
2.98
Total
T
M
3.6
8.53
2.98
3.6
6.17
3.02
3.6
8.53
2.98
3.6
6.17
3.02
3.69
6.29
3.09
3.6
6.17
3.02
3.69
6.29
3.09
3.79
6.41
3.16
3.69
6.29
3.09
3.79
6.41
3.16
3.88
6.52
3.23
3.79
6.41
3.16
3.88
6.52
3.23
3.97
6.63
3.3
3.88
6.52
3.23
3.97
6.63
3.3
4.44
7.16
3.65
3.97
6.63
3.3
4.44
7.16
3.65
4.91
7.64
4
4.44
7.16
3.65
4.91
7.64
6.32
8.84
5.084
4.91
7.64
4
6.32
8.84
5.08
8.24
10
6.78
6.32
8.84
5.08
8.24
10
6.78
11.34
11.2 11.69
8.24
10 11.69
6.78
11.34
11.2
13.64
11.7
28.07
11.34
11.2 28.07
11.69
13.64
11.7
34589 13.4
4.67
13.64
11.7 28.07
34589 13.4
4.67
0.02 #### 34589 13.4
4.67
u
u
0.02
u
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
s/g
s/g
s/g
3.6
3.6
13.1
3.6
13.1
13.2
13.1
13.2
13.3
13.2
13.3
13.3
13.3
13.3
13.4
13.3
13.4
13.9
13.4
13.9
14.3
13.9
14.3
15.9
14.3
15.9
18.2
15.9
18.2
22.4
18.2
22.4
25.9
22.4
25.9
####
25.9
####
Share of W
T Share
M of Wu
TShare
M of Wu
0.6 0.35
0.05
T
M
u
0.6 0.35
0.35 0.05
0.05
0.6
0.6 0.35
0.35 0.05
0.05
0.6
0.6
0.35 0.05
0.6 0.35
0.35 0.05
0.05
0.6
0.61
0.35 0.05
0.6 0.35
0.35 0.05
0.05
0.61
0.61
0.34 0.05
0.61 0.34
0.35 0.05
0.05
0.61
0.61
0.34 0.05
0.61
0.34 0.05
0.05
0.61 0.33
0.34
0.62
0.04
0.61 0.33
0.34 0.04
0.05
0.62
0.64
0.32 0.04
0.62 0.32
0.33 0.04
0.04
0.64
0.67
0.3 0.04
0.64 0.32
0.04
0.67
0.3 0.03
0.04
0.7 0.27
0.67
0.3 0.03
0.04
0.7 0.24
0.27
0.73
0.03
0.7 0.24
0.27 0.03
0.03
0.73
0.75
0.23 0.02
0.73
0.24
0.03
0.75
0.23
0.02
0.81 0.18 0.02
0.75 0.18
0.23 0.02
0.02
0.81
0.81 0.18 0.02
Share of Y
T Share
M of Yu
T Share
M of Y u
0.25
0.42
0.32
T
M
u
0.25 0.42
0.42 0.23
0.32
0.35
0.25 0.42
0.42 0.23
0.32
0.35
0.35
0.42 0.23
0.35 0.42
0.42 0.23
0.23
0.35
0.35
0.42 0.23
0.35 0.42
0.42 0.23
0.23
0.35
0.35
0.41 0.23
0.35 0.41
0.42 0.23
0.23
0.35
0.36
0.41 0.23
0.35 0.41
0.41 0.23
0.23
0.36
0.36
0.41 0.23
0.36 0.41
0.41 0.23
0.23
0.36
0.37
0.41 0.22
0.36 0.41
0.41 0.22
0.23
0.37
0.39
0.39 0.21
0.37 0.39
0.41 0.21
0.22
0.39
0.42
0.38 0.21
0.39 0.38
0.39 0.21
0.42
0.44
0.36 0.21
0.2
0.42 0.36
0.38 0.21
0.44
0.2
0.45
0.35 0.19
0.44
0.36 0.19
0.2
0.45 0.32
0.35
0.51
0.17
0.45 0.32
0.35 0.17
0.19
0.51
0.51 0.32 0.17
Share of W
T Share
M of Wu
TShare
M of Wu
0.6 0.35
0.05
T
M
u
0.6 0.35
0.35 0.05
0.05
0.6
0.6 0.35
0.35 0.05
0.05
0.6
0.6
0.35 0.05
0.6 0.35
0.35 0.05
0.05
0.6
0.61
0.35 0.05
0.6 0.35
0.35 0.05
0.05
0.61
0.61
0.34 0.05
0.61 0.34
0.35 0.05
0.05
0.61
0.61
0.34 0.05
0.61 0.34
0.34 0.05
0.05
0.61
0.63
0.33 0.04
0.61 0.33
0.34 0.04
0.05
0.63
0.64
0.32 0.04
0.63 0.32
0.33 0.04
0.04
0.64
0.68
0.29 0.03
0.64 0.29
0.32 0.03
0.04
0.68
0.73
0.25 0.03
0.68
0.29
0.03
0.73
0.25
0.03
0.8 0.18 0.02
0.73
0.25 0.02
0.03
0.8 0.14
0.18
0.85
0.01
0.8 0.14
0.18 0.01
0.02
0.85
1.35 -0.3
-0
0.85 0.14
1.35
-0.3 0.01
-0
0.87 1.35 -0.3
-0
Share of Y
T Share
M of Yu
T Share
M of Y u
0.25
0.42
0.32
T
M
u
0.25 0.42
0.42 0.23
0.32
0.35
0.25 0.42
0.42 0.23
0.32
0.35
0.35
0.42 0.23
0.35 0.42
0.42 0.23
0.23
0.35
0.36
0.41 0.23
0.35 0.41
0.42 0.23
0.23
0.36
0.36
0.41 0.22
0.36 0.41
0.41 0.22
0.23
0.36
0.37
0.41 0.22
0.36 0.41
0.41 0.22
0.22
0.37
0.39
0.4 0.21
0.37 0.41
0.22
0.39
0.4 0.21
0.41
0.39
0.2
0.39 0.39
0.4 0.21
0.41
0.2
0.48
0.36 0.16
0.41
0.39
0.2
0.48
0.6 0.36
0.3 0.16
0.1
0.48
0.36
0.16
0.6
0.3
0.1
0.81 0.18 0.01
0.6 0.18
0.3
0.1
0.81
0.99
0.07 0.01
-0.1
0.81 0.07
0.18 0.01
0.99
-0.1
3499 #### ####
0.99 ####
0.07 ####
-0.1
3499
Share of W
T Share
M of Wu
TShare
M of Wu
0.6 0.35
0.05
T
M
u
0.6 0.35 0.05
0.6 0.34
0.35 0.05
0.05
0.61
0.6 0.34
0.35 0.05
0.05
0.61
0.61 0.33
0.34 0.05
0.05
0.62
0.62
0.33
0.61 0.32
0.34 0.05
0.05
0.63
0.63
0.05
0.62 0.32
0.33
0.66
0.3 0.04
0.66
0.3 0.04
0.63 0.27
0.32
0.05
0.69
0.69 0.21
0.27
0.04
0.66
0.3 0.03
0.75
0.75
0.21
0.03
0.69 0.15
0.27 0.02
0.04
0.82
0.82
0.15
0.02
0.75
0.21 0.01
0.03
0.9 0.09
0.9
0.09
0.01
0.82 0.05
0.15 0.02
0.94
0.94
0.9 0.05
0.09
1.14
-0.1 0.01
-0
1.14
-0.1 0.01
-0
0.94 0.05
1.14 -0.1
-0
Share of Y
T Share
M of Yu
T Share
M of Y u
0.25
0.42
0.32
T
M
u
0.25 0.42 0.23
0.32
0.35
0.35
0.42 0.23
0.25 0.41
0.32
0.36
0.35 0.41
0.42 0.23
0.23
0.36
0.36 0.41
0.23
0.41 0.22
0.37
0.36 0.41
0.41 0.22
0.23
0.37
0.37
0.41 0.21
0.22
0.4 0.39
0.4 0.37
0.39
0.37
0.41 0.21
0.22
0.43
0.2
0.43
0.37
0.2
0.4 0.31
0.39 0.15
0.21
0.54
0.54
0.31
0.15
0.43 0.21
0.37 0.08
0.2
0.71
0.71
0.21
0.54 0.01
0.31 0.08
0.15
1.05
-0.1
1.05
0.01
-0.1
0.71 0.21
1.35
-0.2 0.08
-0.2
1.35
-0.2 ####
-0.2
1.05 ####
0.01
-0.1
####
####
1.35 ####
-0.2 ####
-0.2
#### #### ####
u
u
0.56
u
0.56
0.77
0.56
0.77
0.77
0.77
0.77
0.77
0.77
0.77
0.76
0.77
0.76
0.76
0.76
0.76
0.74
0.76
0.74
0.73
0.74
0.73
0.69
0.73
0.69
0.65
0.69
0.65
0.61
0.65
0.61
0.6
0.61
0.6
0.56
0.6
0.56
0.56
u
u
0.56
u
0.56
0.77
0.56
0.77
0.79
0.77
0.79
0.81
0.79
0.81
0.83
0.81
0.83
0.84
0.83
0.84
0.94
0.84
0.94
1.03
0.94
1.03
1.37
1.03
1.37
2.12
1.37
2.12
18.3
2.12
18.3
-2.4
18.3
-2.4
0.87
-2.4
0.87
3499 #### ####
c) Divergence of savings rates
c) Divergence of savings rates
c) Divergence of savings rates
c) Divergence
of lsavings
Period alpha
Share of
r rates
g
Period
Period
0
10
1
0
2
2
1
3
3
42
4
53
5
104
10
5
15
15
10
30
30
15
50
50
30
80
80
50
100
100
80
1000
1000
100
1000
alpha
(=P/Y)
(=P/Y)
alpha
0.27
(=P/Y)
0.27
0.32
0.32
0.27
0.33
0.33
0.32
0.34
0.34
0.33
0.35
0.35
0.34
0.36
0.36
0.35
0.4
0.4
0.36
0.45
0.45
0.4
0.59
0.59
0.45
0.81
0.81
0.59
1.22
1.22
0.81
1.57
1.57
1.22
45748
45748
1.57
45748
T Share
M of lu
T Share
M of l u
0.13
0.45
0.43
T
M
u
0.13 0.45 0.32
0.43
0.23
0.13 0.45 0.32
0.43
0.23
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
0.23 0.45 0.32
r
r
0.08
0.08
0.09
0.08
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
g
g
0.03
0.03
0.01
0.03
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
Savings rates
TotalSavings
T rates
M
TotalSavings
T rates
M
0.11
0.26
0.09
Total
T
M
0.11
0.26
0.09
0.13
0.3 0.05
0.11
0.26
0.09
0.13
0.3 0.05
0.13
0.3 0.05
0.13
0.3 0.05
0.13
0.14
0.3 0.05
0.13
0.14
0.3 0.05
0.14
0.15
0.3 0.05
0.15
0.3 0.05
0.14
0.18
0.18
0.3 0.05
0.15
0.22
0.22
0.3 0.05
0.18
0.31
0.31
0.3 0.05
0.22
0.39
0.39
0.3 0.05
0.31
10962
10962
0.39
0.3 0.05
10962
0.3 0.05
u
u
0.02
u
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
s/g
s/g
s/g
3.6
3.6
13
13
3.6
13.1
13.1
13
13.3
13.3
13.1
13.4
13.4
13.3
13.6
13.6
13.4
14.4
14.4
13.6
15.2
15.2
14.4
18
18
15.2
22.5
22.5
18
31.4
31.4
22.5
39.4
39.4
31.4
####
####
39.4
####
beta (=W/Y)
Total beta
T (=W/Y)
M
Total beta
T (=W/Y)
M
3.6
8.53
2.98
Total
T
M
8.53
2.98
3.6
6.17
3.02
3.6
6.17
3.02
8.53
2.98
3.69
6.31
3.06
3.69
6.31
3.06
3.6
6.17
3.02
3.79
6.44
3.1
3.79
6.44
3.1
3.69
6.31
3.06
3.88
6.57
3.14
3.88
6.57
3.14
3.79
6.44
3.1
3.97
6.7
3.19
3.97
6.7
3.19
3.88
6.57
3.14
4.46
7.3
3.41
4.46
7.3
3.41
3.97
6.7
3.19
4.95
7.84
3.65
4.95
7.84
3.65
4.46
7.3
3.41
6.55
9.16
4.51
6.55
9.16
4.51
4.95
7.84
3.65
8.99
10.4
6.58
8.99
10.4
6.58
6.55
9.16 195.40
4.51
13.58
11.7
13.58
11.7
195.40
8.99
10.4 -5.02
6.58
17.48
12.2
17.48
12.2
-5.02
13.58
11.7 195.40
508316 13.9
2.32
508316 13.9
2.32
17.48
12.2 -5.02
508316 13.9
2.32
u
u
0.56
u
0.56
0.77
0.77
0.56
0.79
0.79
0.77
0.81
0.81
0.79
0.83
0.83
0.81
0.84
0.84
0.83
0.94
0.94
0.84
1.04
1.04
0.94
1.43
1.43
1.04
2.64
2.64
1.43
-2.9
-2.9
2.64
-0.6
-0.6
-2.9
0.77
0.77
-0.6
0.77
1
N.B.: L=Wage income, P=Profits, Y=National income, W=Wealth, T=Top income households, M=Middle income households, U=Lower income
households, r=Return on capital, g=Growth rate of national income, s=Savings rate.
Source: IMK calculations.
page 13
Conclusions
The data on economic inequality that are currently
available for Germany are unsatisfactory, even compared to other countries. While the SOEP data are
without doubt extremely valuable for a wide variety
of analyses of economic inequality, they are not so
good at accurately recording very high incomes and
wealth. In principle, the WTID offers a valuable alternative for measuring top income shares based on
official tax statistics. However, the most recent figures for Germany are from 2007. There is an urgent
need for further research in this area. However, this
is currently complicated by the fact that tax paid on
investment income is not recorded on an individual
basis. This alone constitutes a strong argument for
abolishing the withholding tax and returning to the
old comprehensive income tax system.
It is also easier to measure wealth inequality in
other countries. In France, for example, the statistical basis for doing so has existed ever since the
introduction of the wealth tax during the French
Revolution of 1789. No reliable long-term data
series on wealth inequality currently exist for Germany, however. The fact that no wealth tax exists in
Germany and that the data sets based on gift and inheritance tax are incomplete is hindering research
in this area. Well-off households are reluctant to
provide details of their financial situation, making
it difficult to determine the true concentration of
wealth. The introduction of a low-rate wealth tax
would therefore allow significant progress to be
made in terms of the quality of the available data.
Indeed, even a 0 % wealth tax would constitute an
important step towards enabling the real distribution of wealth to be recorded.
However, any serious attempt at tackling the
negative consequences of rising inequality – both
from a distributive justice and a macroeconomic
stability perspective – will require more far-reaching fiscal policy interventions. The debate on the
reintroduction of a wealth tax (Bach and Beznoska
2012) and an increase in the income tax rate for top
earners needs to start placing far greater emphasis on the fact that reducing economic inequality
also diminishes the risk of future economic crises.
People who lived through the Great Depression
came to understand the relationship between inequality and macroeconomic instability, even all
those years ago. Indeed, the Wealth Tax Act that
formed part of US President Franklin D. Roosevelt’s
New Deal and was conceived as a response to the
global economic crisis of 1929 raised the top income
tax rate to 79 %.
References
IMK Report 99e
October 2014
All IMK publications are available online:
http://www.boeckler.de/imk_2733.htm
Alvaredo, F. / Atkinson, A.B. / Saez, E. / Piketty,
T. (2012): The world top incomes database. http://
topincomes.g-mond.parisschoolofeconomics.eu/;
accessed 24.08.2014.
Bach, S. / Beznoska, M. (2012): Vermögensteuer:
Erhebliches Aufkommenspotential trotz erwartbarer Ausweichreaktionen. In: DIW Wochenbericht,
Vol. 79, No. 42, pp.12-17.
Bartels, C. / Bönke, T. (2013): Can households
and welfare states mitigate rising earnings instability? In: Review of Income and Wealth, Vol. 59, No. 2,
pp. 250-282.
Behringer, J. / Belabed, C. A. /Theobald, T. / van
Treeck, T. (2013): Einkommensverteilung, Finanzialisierung und makroökonomische Ungleichgewichte. In: Vierteljahrshefte zur Wirtschaftsforschung/Quarterly Journal of Economic Research,
DIW Berlin, German Institute for Economic
Research, Vol. 82, No. 4, pp. 203-221.
Behringer, J. / van Treeck, T. (2013): Income
Distribution and the Current Account: A Sectoral
Perspective. INET Research Notes 35, Institute for
New Economic Thinking (INET).
Belabed, C./Theobald, T./van Treeck, T. (2013):
Income Distribution and Current Account Imbalances. INET Research Notes 36, Institute for New
Economic Thinking (INET).
Brenke K. / Wagner, G. G. (2013): Ungleiche
Verteilung der Einkommen bremst das Wirtschaftswachstum. In: Wirtschaftsdienst, Vol. 93, No. 2, pp.
110-116.
Dell, F. (2007): Top Incomes in Germany
Throughout the Twentieth Century: 1891–1998. In:
Atkinson, A./Piketty, T. (Eds.): Top Incomes over
the Twentieth Century: A Contrast Between Continental European and English Speaking Countries.
Oxford University Press, Oxford, pp. 365-425.
Duesenberry, J. S. (1949): Income, Saving and the
Theory of Consumer Behavior. Harvard University
Press.
page 14
Frank, R. H. (2005): Positional Externalities
Cause Large and Preventable Welfare Losses. In:
American Economic Review, Vol. 95, No. 2, pp.
137-141.
Frank, R. H. (2007): Falling Behind: How Rising
Inequality Harms the Middle Class. University of
California Press, Berkeley.
Grabka, M. / Goebel J. (2013): Rückgang der
Einkommensungleichheit stockt. In: DIW Wochenbericht, Vol. 80, No. 46, pp. 13-23.
Grabka, M. / Westermeier, C. (2014): Anhaltend
hohe Vermögensungleichheit in Deutschland. In:
DIW Wochenbericht, Vol. 81, No. 9, pp. 151-164.
HFCN (2013): The Eurosystem Household Finance and Consumption Survey: Results from the
first wave. ECB Statistical Paper Series No. 2,
Frankfurt am Main.
Kumhof, M. / Ranciere, R. (2010): Inequality,
leverage and crises. IMF Working Paper 10/268,
International Monetary Fund.
Leigh, A. (2009): How Closely Do Top Income
Shares Track Other Measures of Inequality? In:
Economic Journal, Vol. 117, No. 524, pp. F619–
F633.
OECD (2008): Growing Unequal? - Income Distribution and Poverty in OECD Countries, http://
www.oecd-ilibrary.org/social-issues-migrationhealth/growing-unequal_9789264044197-en,
accessed 24.08.2014.
OECD (2011): Divided We Stand: Why Inequality
Keeps Rising, http://www.oecd-ilibrary.org/socialissues-migration-health/the-causes-of-growinginequalities-in-oecd-countries_9789264119536-en,
accessed 24.08.2014.
Rajan, R. (2010): Fault lines: How hidden fractures still threaten the world economy. Princeton
University Press.
IMK Report 99e
October 2014
Rehm, M. / Schmid, K. D. / Wang, D. (2014):
Why has Inequality in Germany not Risen Further
After 2005? IMK Working Paper No. 137.
Schinke, C. (2012): Inheritance in Germany 1911
to 2009: A Mortality Multiplier Approach. SOEP papers on Multidisciplinary Panel Data Research, 462,
http://www.diw.de/documents/publikationen/73/
diw_01.c.407138.de/diw_sp0462.pdf ; accessed
21.08.2014.
Stiglitz, J. E. (2012): The price of inequality: How
today’s divided society endangers our future;
Norton and Company, New York.
Summers, L. H. (2014): U.S. Economic Prospects:
Secular Stagnation, Hysteresis, and the Zero Lower
Bound. In: Business Economics, Vol. 49, No. 2, pp.
65-73.
van Treeck, T. (2014a): Did inequality cause the
U.S. financial crisis? In: Journal of Economic Surveys, Vol. 28, pp. 421–448.
van Treeck, T. (2014b): On r > g and saving
rates. Accessing Piketty‘s ‚laws of capitalism‘ with
some simple simulations. http://verteilungsfrage.
org/2014/07/on-r-g-and-saving-rates/; accessed
24.08.2014.
van Treeck, T. / Sturn, S. (2012): Income inequality as a cause of the Great Recession?: A survey of
current debates. ILO Working Papers 470934, International Labour Organization.
Wagner, G. (2011): Nicht nur Griechenland, auch
die deutsche Wirtschaftspolitik steht vor einer
Wende. DIW Wochenbericht, Vol. 78, No. 45, p. 32.
Piketty, T. (2014): Capital in the twenty-first century. The Belknap Press of Harvard University Press;
Cambridge, Massachusetts, London.
page 15
IMK Report 99e
October 2014
Original German version
completed on 1 September 2014
Impressum
Published by:
Macroeconomic Policy Institute (IMK) at the Hans Boeckler Foundation,
Hans-Boeckler-Str. 39, 40476 Duesseldorf
Telephone +49 211 7778-331, Fax +49 211 7778-266,
IMK@boeckler.de, http://www.imk-boeckler.de
Managing Editor: Andrew Watt
Press officer: Rainer Jung, +49 211 7778-150
ISSN 1861-3683
Reproduction and other distribution, in full or in part, is only permitted if the source is quoted.
page 16