Is the German Middle Class Crumbling? Risks and Opportunities

Report

21 December 2021

2. The German middle class – a statistical profiling

Sebastian Königs and Anna Vindics

Abstract

This chapter examines trends in the size and composition of Germany’s middle class, proxied by the middle‑income group, i.e. people living in households with disposable incomes between 75 and 200% of the national median. The chapter starts by looking at the share of people belonging to the middle‑income group, at how this share changed over time, and at income trends for middle‑income people. It then analyses changes in the composition of the middle‑income group along various socio‑demographic dimensions, including by age, household type, level of educational attainment, migrant background, and region of residence.

2.1. The economic, political and social importance of a thriving middle class

Strong and thriving middle classes support healthy economies and prosperous societies. Middle‑class citizens strongly contribute to the foundations of inclusive societies, to social and political stability, and to economic growth. They consume goods and services, accumulate savings, invest in human capital, safeguard democratic institutions, and support good-quality public services. Societies with a strong middle class experience greater levels of social trust, but also better educational outcomes, less crime, better health outcomes, and higher life satisfaction (Kelly, 2000[1]; Lynch and Kaplan, 1997[2]; Thorson, 2014[3]). In particular, strong middle classes promote:

Education: The middle class forms the bulk of the population, and typically invests more in their own education and that of their children than the low-income group. By doing so, it increases current and future human capital (OECD, 2019[4]) and supports well-being and economic growth. Human capital accumulation is an important mechanism through which changes in the income share held by the middle class affect economic growth (Brueckner et al., 2017[5]). As populations achieve higher levels of educational attainment, citizens become more productive, healthy, resilient and engaged in society.
Consumption and entrepreneurship: The middle class have played an essential role in the expansion of market economies, trade, and hence economic growth. They have more disposable income than the low-income group, and consume a higher share of it than the high-income group, hence fuelling demand for goods and services and creating jobs (Pressman, 2007[6]). People living on middle incomes contribute to physical capital accumulation essential for investment, and the small and medium enterprises that they establish and run constitute the backbone of strong economies. Economic growth and incomes are higher in countries with greater middle‑class income shares (Easterly, 2001[7]). Strong middle classes increase social trust, which reduces transaction costs and promotes innovation (Gould and Hijzen, 2016[8]).
Social and political stability: The middle class champions political stability and good governance through their involvement in public affairs and the power of the mass of their voices. Their intolerance of corruption and trust in their fellow citizens and in democratic institutions are essential for functioning liberal democracies. A strong middle class prevents political polarisation that can be harmful for government performance (Madland, 2015[9]). Middle classes provide a solid basis for democratic governance by having capacity to demand regulation, enforcement of contracts, and the rule of law (Birdsall, 2010[10]). Strong middle classes are the main financers of social protection systems through their taxes and contributions, which support stable and just societies.

However, there are signs that, in many OECD countries, the middle classes are not as stable as they used to be. The recent OECD flagship publication Under Pressure: The Squeezed Middle Class (OECD, 2019[4]) documented three main challenges facing middle classes in many countries (see Box 2.1): (1) socio‑economic outcomes are increasingly perceived as unfair. Middle incomes have been growing much more slowly than higher incomes in many OECD countries for more than three decades. (2) The middle‑class lifestyle is becoming more expensive. The cost of essential parts of a middle‑class lifestyle has increased faster than income in many countries, notably of housing and higher education. (3) Life outcomes have become more uncertain. Job polarisation and insecurity has risen in the context of rapidly transforming labour markets.

This report, which provides a comprehensive assessment of the situation of the German middle class, focuses on the first and the third of these challenges, i.e. on trends in disposable incomes and on labour market developments. It does not provide detailed evidence on the second challenge, i.e. changes in middle‑class expenditure patterns, hence also not covering the importance of changes in the cost of housing or energy for middle‑class households.

Box 2.1. Main insights from Under Pressure: The Squeezed Middle Class

The OECD flagship publication Under Pressure: The Squeezed Middle Class (OECD, 2019[4]) provided a comprehensive assessment of the economic situation of the middle classes in OECD countries, looking at trends in incomes, consumption behaviour, and labour market outcomes. It documented that middle classes in many OECD countries find themselves increasingly under pressure facing stagnating incomes, rising expenditures, and greater labour market uncertainty.

Socio‑economic outcomes are increasingly perceived as unfair

Middle incomes have been growing much more slowly than high incomes for more than three decades. Over the past 30 years, across the OECD, median incomes increased by a third less than the incomes of the richest 10%. Moreover, in many countries, incomes at the very top have surged. In the United States, for example, the share of the top 1% in total income almost doubled over the past three decades, from about 11 to 20%; almost half of all income growth over this period accrued to this group (Förster, Llena-Nozal and Nafilyan, 2014[11]; Saez, 2018[12]). In Germany, the top‑1% income share rose from around 12% in 2000 to a post-war high of more than 14% in 2008; it then dropped to 13% during the financial crisis, and remained stable thereafter until 2014, the year with the latest available data (Bartels, 2019[13]). The middle classes have lost in economic influence as a result. Three decades ago, the aggregate income of all middle‑income households was four times that of upper‑income households across the OECD on average; today, this ratio is less than three. Meanwhile, many people, especially in lower middle‑income households, feel that they contribute much more to the welfare state, through taxes and contributions, than they receive in return in benefits and services. Opportunities to climb up the social ladder have become rarer for middle‑class people and their children, while social risks have increased.

The middle‑class lifestyle is becoming more expensive

This sluggish income growth coincided with an increase in the costs of a middle‑class lifestyle. Prices for housing, health, and education increased faster than inflation across the OECD. Ageing and new medical technologies have driven up the cost of health services. The widespread trend towards tertiary education is pressing parents and young people to invest more in education, while education has become more costly in many countries. Also in Germany, the cost of education increased faster than median incomes, by 23 vs. 9% between 1995 and 2015. This likely reflects primarily a rise in the cost of attending higher education. Meanwhile, the geographic polarisation of jobs is pushing up housing prices in urban areas. Housing constitutes the largest expenditure item for middle‑income households in Germany, at around one‑third of disposable income relative to one‑quarter in the mid‑1990s. This is the second largest increase among the countries with available data. Such rising expenditures squeeze middle‑class households’ finances and reduce their ability to save. More than one‑in-five middle‑income households spend more than they earn across the OECD in the mid‑2010s.

Labour market trajectories have become more uncertain

People in middle‑class households are concerned that the digitalisation and automation of the economy will destroy their jobs. One‑in-six middle‑income workers are employed in jobs that face high risk of automation across the OECD on average, close to the corresponding share among low-income workers (one‑in-five). Rapid integration of global supply chains, fast and transformative technological change, and population ageing have resulted in labour market polarisation, i.e. a shift in employment towards high‑skilled and low‑skilled non-routine jobs and a hollowing-out of middle‑skilled jobs. In many cases, having medium skill level no longer guarantees the way up into the middle‑income group. Almost half of middle‑income workers across the OECD are in high-skilled occupations, compared to one‑third two decades ago. Meanwhile, job security and the level of income support in case of job loss have declined in many countries, up until the COVID‑19 crisis.

Source: OECD (2019[4]). Under Pressure: The Squeezed Middle Class, https://dx.doi.org/10.1787/689afed1-en.

This chapter describes income trends for the German middle class, and provides a socio‑economic profile of middle‑income individuals in Germany. It provides evidence on the size of the German middle‑income group, its composition, and the most important trends over the past two‑and-a-half decades since the mid‑1990s. It sets the scene for Chapter 3, which discusses trends in labour market outcomes for middle‑income workers in Germany.

2.2. Who counts as “middle class” in Germany?

2.2.1. The middle class is a complex social construct

The middle class is a complex social construct and has no unique definition. Indicators used to define, and study, the middle class vary significantly, within and across disciplines. Some measures and indicators are based on occupation and employment status, relating, for example, to certain aspects of job quality (Goldthorpe, 2016[14]). Others are based on social, cultural, or economic criteria, such as income level, educational attainment, as well as values and attitudes (Savage et al., 2013[15]; Pressman, 2007[6]). Yet other definitions are consumption-based, with households being defined as middle class if they are able to buy a home, afford certain type of cars, or holidays (Reeves, Guyot and Krause, 2018[16]). In a comprehensive analysis of social class in Germany, Reckwitz (2019[17]) argues that besides the old middle class – a group of middle‑skilled workers that is shrinking in size and losing prestige – a new middle class has emerged made up of academics working in knowledge‑based jobs in urban areas. Recent empirical studies of the economic situation, social mobility and social risks of the middle class in Germany have for example defined middle‑class households based on occupation class (Lengfeld and Ordeman, 2016[18]) and income (Burkhardt et al., 2013[19]; Niehues, 2017[20]; Zucco and Özerdogan, 2021[21]).

An alternative approach is to define the middle class based on subjective self-identification. On average around two‑thirds of the population in OECD countries consider themselves as belonging to the middle class (Figure 2.1). In Germany, 73% of people self-identify as middle class, around 10% more than belong to the group according to the income‑based definition used in Under Pressure. In some other countries, the difference between self-perceived and income‑based middle‑class status is even larger. The fact that many people self-identify as middle class even though their socio‑economic circumstances may suggest otherwise is referred to as “middle‑class bias” (Evans and Kelley, 2004[22]; Bellani et al., 2021[23]). One reason can be that people compare themselves to their immediate peer group, i.e. family members, neighbours, and co-workers, who tend to have a similar living standard, rather than to look at living standards across society more broadly. It can therefore be problematic to rely (alone) on self-identified middle‑class status when studying the economic well-being of middle‑class households.

Figure 2.1. More people self-identify as middle class than belong to the middle‑income group

Percentage of the population in the middle‑income group and percentage share of people considering themselves as “middle class”, 2018 or latest available year

2.2.2. An income‑based definition of the middle class

There is a strong rationale for looking at incomes when defining people’s middle‑class status for the purpose of comparative statistical analysis. Income is a key determinant of people’s living standards and of many aspects of their well-being. It permits households to consume goods and services, including housing, education and health care, and to build up savings as an insurance against shocks. Income also strongly correlates with other determinants of social class, such as occupation, education, and self‑perceived class.1 From an analytical point of view, the benefit of employing an income‑based definition of the middle class is that income data are readily available across countries for long periods. The same does not necessarily hold for information on self-perceived class or other socio‑economic outcomes.

This review therefore uses an income‑based definition of the middle class, i.e. it focuses on the middle‑income group as a proxy for the middle class. Following the approach used in Under Pressure, it defines the middle‑income group as people living in households with disposable incomes between 75 and 200% of the national median. In 2018, this corresponded to a monthly disposable income of around EUR 1 500 to EUR 4 000 for a single person, and EUR 3 000 to EUR 8 000 for a couple with two children. Within the middle‑income group, the review further distinguishes the lower middle (75 to 100% of the median), the mid middle (100‑150%) and the upper middle (150‑200%). Disposable household incomes are adjusted for household size using the square‑root method to account for within-household sharing of resources. This approach follows the methodology used for inequality and poverty measurement, including in the OECD Income Distribution Database (OECD, 2021[24]).2

The review defines low- and high-income groups by analogy. People in households with incomes below 75% of the median are considered “low-income”. This group is, in some parts of the analysis, further subdivided into two groups: people living in households on equivalised household incomes less than 50% of the median are classified as “poor”, following standard OECD practice;3 those with incomes between 50% and 75% of the median as “vulnerable”. At the other end of the income distribution, people with incomes above 200% of the median are considered as “high-income”.

Figure 2.2. The middle‑income group is slightly larger in Germany than across OECD countries on average

When applying this income‑based definition,

The bulk of the population is part of the middle‑income group in almost all OECD countries. Across the OECD, almost two‑in-three (62%) people belonged to the middle‑income group in 2018 (Figure 2.2 Panel A). In Germany, the share was slightly higher, at 64%. In many Central and Eastern European and the Nordic countries, the middle‑income group is larger than in Germany, while it is smaller in many English-speaking, Southern European, Latin American and Asian OECD countries.
Around one‑in-three people are part of the low-income group across OECD countries. Again, the figure for Germany lies close to the OECD average, with 30% of people living on low incomes, and 10% living below the poverty line.
The high-income group is usually smallest, but varies most strongly in size: it makes up 7% of the population in Germany, and 8% in the OECD on average. In some OECD countries, the high‑income group makes up only a few percent of the population, while it reaches 14% or more in Mexico, Chile, and the United States.

The country ranking in Figure 2.2, Panel A is relatively robust to using other thresholds to define the income groups (OECD, 2019[4]).4 For an overview of the nominal values of the income thresholds for a selection of countries expressed in USD, see Figure 2.2, Panel B.

The size of a country’s middle‑income group closely relates to the level of income inequality. This is because income groups are defined using thresholds expressed relative to the median income. In OECD countries where inequalities are large, such as Mexico, Chile and the United States, incomes are widely dispersed around the median, and the middle‑income group is consequently smallest (Figure 2.3). Meanwhile, in the more egalitarian countries in Central and Eastern Europe (e.g. the Czech Republic, the Slovak Republic, and Slovenia) and Northern Europe (Denmark, Finland and Norway), a greater share of households live on incomes close to the median, and the middle‑income group is larger.

Figure 2.3. The size of countries’ middle‑income group is closely tied to the level of income inequality

Percentage share of the middle‑income group against the Gini coefficient of disposable household income, 2018 or latest available year

2.3. Trends in middle‑class incomes in Germany

Under Pressure documents that middle classes in many OECD countries have been suffering from stagnating real incomes over the last decades – a trend that often brought about a decline in the size of middle‑income groups. This section updates and revisits this analysis with a particular focus on the developments in Germany. It demonstrates that also the German middle‑income group has shrunk since the mid‑1990s, and more rapidly so than middle‑income groups in other OECD countries. This has entailed a decline in the income share of the middle‑income group out of total income in Germany, implying shrinking economic influence. However, most of these changes have occurred prior to 2005.

2.3.1. German middle‑income households have experienced very modest income growth since the mid‑1990s, and income disparities in Germany have widened

Middle‑income households in Germany experienced only modest income growth since the mid‑1990s. The median disposable household income, i.e. the income for the household precisely in the middle of the income distribution, has stagnated in real terms in Germany for nearly one‑and-a-half decades (Figure 2.4, Panel A). This implies that many middle‑income households experienced essentially no rise in living standards between 2000 and 2014. The trend has been much more positive since around 2015, when the real median income started growing.

Income disparities have widened over the observed period. In the late 1990s, top and bottom incomes still grew in lockstep with the median. From 2001, a gradual decoupling took place, with top incomes steadily growing while median incomes stagnated. Households in the bottom of the income distribution even experienced a decline in real incomes after 2001, up to a rebound from around 2017. In 2018, the household incomes of the top 10% of households were 28% higher in real terms than in 1995, compared to a real income growth of 17% at the median, and only 7% for the bottom 10% of households. Slow growth and widening disparities in disposable incomes reflect a very similar trend for earnings, see Chapter 3, Figure 3.9. Also, levels of subjective well-being appear to have diverged between people of different income groups in Germany, see Box 2.2.

While the developments in Germany mirror a trend towards rising income inequality across the OECD more broadly, the pattern has not been uniform across all countries (Annex Figure 2.A.1). In some countries, such as the Netherlands, the Czech Republic and Turkey, middle‑income households did better than households in the top and bottom of the income distribution. Meanwhile, in Italy, Greece, and Japan, households all across the income distribution experienced stagnating or declining incomes. Some other countries, such as the Nordics and the United Sates, show patterns similar to those in Germany, in some cases even more pronounced.

Germany holds a mid-table position in median income growth among OECD countries since the global financial crisis (Figure 2.4, Panel B). Thanks to the uptick in income growth since 2015, median incomes grew on average by a bit more than 1% per year over the period 2008‑18. This is slightly above the OECD average (0.7%). Annualised median income growth during this shorter period has been higher in some Central and Eastern European countries (the Czech Republic, the Slovak Republic, Poland and some of the Baltics), but also in Canada, Israel, Ireland, and Sweden. In a number of countries heavily affected by the global financial crisis, median household incomes remain substantially below their pre‑financial-crisis values, notably in Greece, Spain, Iceland and Italy.

By contrast, median income growth in Germany has been among the lowest across OECD countries when considering the entire period since the mid‑1990s. Among the 21 countries with income data for that period, only Japan, Greece, Italy and Belgium recorded lower real median growth rates.

Figure 2.4. Median income growth in Germany picked up in 2015 after a long period of income stagnation

Box 2.2. Subjective well-being by income level in Germany

Income is a key determinant of people’s well-being, as noted in Section 2.2, and higher‑income people generally fare better along many dimensions of subjective well-being (OECD, 2020[25]). The increase in income inequality in Germany since the mid‑1990s, shown in Figure 2.4, is also mirrored in growing inequalities in subjective well-being between people from different income groups. This holds true for different dimensions of subjective well-being, as illustrated in Figure 2.5:

Life satisfaction: the share of people in middle‑income households in Germany reporting high life satisfaction, i.e. a level of 7 or higher on a scale from 1 (dissatisfied) to 10 (satisfied), was 84% in 2017, compared to 90 and 67% among high- and low-income people. Disparities between people in the high- and low-income group have widened from 5 to 23 percentage points between 1997 and 2017.
Trust in other people: less than half (45%) of people in the German middle‑income group say that “most people can be trusted”. High-income people are much more likely to report high levels of trust (61%), while the opposite is true for low-income people (33%). The share of people who say they trust in others has greatly increased over time in Germany, but only for high- and (to a lesser degree) middle‑income people.
Freedom of choice and control: measured again on a scale from 1 (none at all) to 10 (a great deal) the share of high-, middle‑, and low-income people in Germany who reported high perceived freedom of choice and control over their lives stood at 78, 71 and 65% in 2017. Disparities by income group substantially widened after the late‑1990s, but seem to have narrowed again in the most recent data for 2017.

Figure 2.5. Disparities in subjective well-being by income group have grown in Germany

Subjective well-being by income group (percentage of respondents), Germany, 1997‑2017

2.3.2. The German middle‑income group has shrunk in the late‑1990s and early‑2000s, mostly at the lower end, and has not recovered since

The German middle‑income group is smaller than it was in the mid‑1990s. Between 1995 and 2018, it shrunk by 6 percentage points, from 70 to 64% of the population. Most of this decline occurred in the early 2000s, at the time of widening income disparities (Figure 2.6). In spite of the positive employment growth after 2005, the German middle‑income group did not recover afterwards, as a result of stagnating disposable incomes for lower and middle‑income households. This decline largely reflects a shrinking of the lower middle‑income group, i.e. the share of households with incomes of 75 to 100% of median. Their share declined by 4 percentage points, to 21% in 2018. Meanwhile, the mid middle‑income group (100‑150% of the median) and the upper middle‑income group (150‑200% of the median) remained broadly stable, at 32% and 11% of the population.

The shrinking of the middle‑income group is mirrored by a growing share of low- and high-income households. In particular, the share of households living below the poverty line (50% of median income) has increased since the mid‑1990s, by 3 percentage points. Similarly, the share of vulnerable households, i.e. those who live on incomes above the poverty line but not enough to reach the lower middle‑income group (50‑75% of the median), increased by 2 percentage points. At the upper end of the distribution, the share of high‑income households (above 200% of the median) expanded by 1 percentage point. As before, these changes largely occurred up to 2005, although the share of poor households continued to rise thereafter.

Figure 2.6. The German middle‑income group shrank in the late‑1990s and early 2000s, and has not recovered since

Percentage share of the population by income group, Germany, 1995‑2018

2.3.3. Germany experienced a faster decline in the middle‑income group than most other OECD countries

Relative to the mid‑1990s, the middle‑income group has declined faster in Germany than in most OECD countries.5 Across all 26 OECD countries with available data, the size of the middle‑income group even slightly expanded on average, by 0.3 percentage points (Figure 2.7). This reflects a strong growth in the share of middle‑income households in such countries as Ireland, Hungary, Belgium, and Chile. Compared to Germany, the shrinking of the middle‑income group was more pronounced only in Sweden, Finland and Luxembourg. Countries, where the middle‑income group declined – such as the Nordics, some Central and Eastern European and Northern American countries – nearly all experienced income polarisation, i.e. an expansion of both the low- and high-income groups. In countries where the middle‑income group grew – such as Latin American, Southern European and European English-speaking countries – income inequality usually declined.

Figure 2.7. Relative to the mid‑1990s, the middle‑income group shrank faster in Germany than in most other OECD countries

Changes in the population shares of the middle‑income groups in OECD countries, mid‑1990s to 2018 or latest year, in percentage points and percentages

2.3.4. The German middle‑income group accounts for a declining share of total household income

In line with its shrinking size, also the income share of the middle‑income group has fallen in Germany relative to total population income, indicating declining overall economic influence. While total disposable income across all middle‑income households was – in real terms – 2.1 higher in 2018 than in 1995, total population income grew by a factor of 2.3 (Figure 2.8). Consequently, the income share of the middle‑income group declined from 74 to 67% of total income between 1995 and 2018. Again, this decline occurred over the first decade of the observation period, between 1995 and 2005. Almost all of it is due to the shrinking population share of the middle‑income group, as opposed to a decline in the relative average income of middle‑income households in comparison to the other income groups.6

Figure 2.8. The middle‑income group accounts for a declining share of total income in Germany

Income shares of the different income groups in Germany, 1995‑2018

2.3.5. The middle‑income group pays more in income taxes than it receives in social benefits, but most redistribution takes place within the middle‑income group

There is a widespread sentiment in many OECD countries, including in Germany, that taxes are too high for lower- and middle‑class households, see also the discussion in Chapter 5, Section 5.4. OECD calculations based on data from the International Social Survey Programme for 2016 show that around half of people in Germany, and across OECD countries, find that taxes for middle‑income households are “too high” or “much too high” (48% of respondents in Germany, 51% across 25 OECD countries on average). Even around three‑in-four respondents (75% in Germany, 72% in the OECD on average) find taxes too high for low-income households.7 Meanwhile, recent results from the OECD Risks that Matter Survey indicate that, on average, 58% of middle‑income households in OECD countries consider that they do not receive a fair share of public benefits for the taxes and social-security contributions that they pay (OECD, 2019[26]).

Across OECD countries, the middle classes indeed play a crucial role for the financing public expenditures (and are main beneficiaries of these expenditures), and often pay on average more in direct taxes than they received in cash social benefits. However, substantial redistribution takes place also within the middle‑income group:

Across income groups: while people the mid and upper middle‑income group in Germany indeed pay on average more direct taxes than they receive in cash benefits, the lower middle is a net beneficiary of the tax-benefit system (Figure 2.9, Panel A). Overall, across the entire middle‑income group, the sum of taxes and social‑security contributions paid on income exceeded the total value of social cash benefits received by 12% of disposable income in 2018.8 The middle‑income group’s net contribution slightly declined between 1995 and 2018 (Figure 2.9, Panel B).
Across age groups: redistribution through income taxes and social transfers has a strong intertemporal component. The social contributions paid by workers on today’s earnings given them entitlements to future social transfers in case of job loss or sickness, after childbirth and in old age. Substantial redistribution therefore takes places over the life course, and notably between working‑age people and seniors. Indeed, 18‑to‑64‑year‑olds in the middle‑income group are heavy net contributors, with taxes and social-security contributions exceeding benefits received by on average 33% of disposable income. Meanwhile, middle‑income people over 65+ years are clear net beneficiaries: they derive nearly all of their disposable income from social transfers, notably public pensions, and pay only little taxes and social contributions (Figure 2.9, Panel C).

All of these calculations do not account for any indirect taxes paid (incl. VAT), nor for in‑kind transfers received, for example, in form of public services, such as health care or education.

Figure 2.9. The lower middle is a net beneficiary of the tax-benefit system, but most redistribution takes place *within* the middle‑income group, between working-age people and seniors

Income components relative to disposable income (=100), as percentages

2.4. Changes in the composition of the middle‑income group in Germany

Against the context of a decline in the German middle‑income group compared to the mid‑1990s, this section presents evidence on the changing socio-demographic composition of middle‑income households, and on the implied chances for people from different population groups of making it into the middle‑income group. Specifically, it provides evidence on two different but related questions:

1. What share of the middle‑income group does population group X – say young people, or working‑age couples – account for, and how has this evolved over time? This first question is concerned with the composition of the middle‑income group.

2. What is the likelihood for a person from population group X to be in the middle‑income group, as opposed to the low- or high-income group, and how has this likelihood changed over time? This second question is about people’s chances of making it into the middle‑income group, or the group‑specific size of the middle‑income group.

The two questions are clearly related, but can yield at times, what seems like conflicting conclusions. This is because the socio-demographic composition of the overall population has changed. For example, as shown below, the share of working singles in the German population has strongly increased over the observation period, and consequently working singles make up a growing share of middle‑income households. (This is the response to Question 1 above, about the composition of the middle‑income group). However, the share of working singles grew more slowly in the middle‑income group than in the overall population. As a result, the chances for working singles of making it into the middle‑income group have declined. (This is the response to Question 2 above, about the odds of making it into the middle‑income group). Indeed, working singles have become more likely to be in the low-income group.

This section provides evidence on the above two questions about the composition of the middle‑income group and the group-specific chances of making it into the middle along different dimensions: by age group, household type, level of educational attainment, migrant background, and region of residence.9

2.4.1. The German middle‑income group has aged more quickly than the general population, and young people find it difficult to secure their place

The German middle‑income group has aged significantly since the mid‑1990s. In 2018, older working age adults (45‑64 years) and seniors (65 years and older) made up more than half (55%) of middle‑income people, up from around 40% in 1995 (Figure 2.10, Panel A). Meanwhile, prime working-age adults (30‑44 years), young people (18‑29 years) and children (below 18 years) account for declining shares.

This trend does not simply reflect general population ageing: While the share of people living on middle incomes has shrunk in each of the age groups, young people have been disproportionally affected. They are on average 10 percentage points less likely to be in the middle‑income group than in the mid‑1990s, a decline nearly twice the population average (‑5.7 percentage points; Figure 2.10, Panel B). However, this drop occurred entirely in the years up to 2005, while the middle‑income share among young people remained stable thereafter. Meanwhile, among older working-age adults and seniors, the size of the middle‑income groups has declined by less than average (‑3.8 and ‑4.3 percentage points), and just over the period since 2005 it has even expanded for older working-age adults. Particularly for young people and prime working-age adults, the shrinking of the middle‑income group coincided with a strong expansion of the low‑income group. Meanwhile, for older working-age adults, the high-income group expanded most strongly.

Figure 2.10. Older generations account for a growing share of Germany’s middle‑income group

Trends in labour force participation behaviour over people’s lives can partly explain the changes in the age structure of the middle‑income group. Young people increasingly obtain tertiary education and thus remain in education for longer than they did two decades ago. They are therefore older when entering the labour market and earning their first own income, which can explain part of the decline in the size of the middle‑income group among young people. Meanwhile, older workers retire later and often continue earning good incomes up to a higher age, which can explain the relative rise in the size of the middle‑income group – and the high-income group – among older working-age adults.

Yet, the changing age composition of the middle‑income group certainly also reflects trends in the relative financial well-being of different demographic groups. Today’s senior generation is, on average, financially much better off than previous generations of elderly (OECD, 2017[27]). This applies particularly to the baby boomers, many of whom have been able to contribute to the pension system throughout their entire working life and to accumulate, sometimes significant, private wealth, often in the form of housing (Balestra et al., forthcoming[28]). This is reflected in the growing share of older working-age people and seniors in the middle‑income group.

Indeed, since the baby boomers, the size of the middle‑income group has continuously declined from one generation to the next (Figure 2.11). When aged in their 20s and 30s, 71% of the baby boomers belonged to the middle‑income group. For the Generation X (i.e. people born in the mid‑1960s to early 1980s) and the Millennials (born in the early 1980s to mid‑1990s), the shares had declined to 68% and 61% at the same age. The consistent decline in the size of the middle‑income group from one generation to the next also holds when looking at people’s income status in their childhood and youth (0‑10s, 10‑20s and 20‑30s). This confirms that senior’s growing relative representation in the middle‑income group, and the decline of the middle‑income group for young people (as observed in Figure 2.10), is not alone a consequence of changing labour force participation behaviour.

Figure 2.11. Since the baby boomers, the size of the middle‑income group has declined generation by generation

Percentage share of the population belonging to the middle‑income group, by cohort and age, Germany

2.4.2. Working couples, and particularly one‑and-a-half earner households, have been much less affected by the decline in the middle‑income group

Household composition and living arrangements have considerably changed over the past decades, in Germany as in other OECD countries. The share of single‑person and senior households increased because of later family formation and gains in longevity. The share of households with children declined, as more couples decide to have children later, or not at all. These changing family structures are closely tied to, and interact with, the two larger trends of population ageing and rising female labour market participation.

These trends have also changed the composition of the German middle‑income group since the mid‑1990s (Figure 2.12, Panel A), and changed the likelihood for different types of households of belonging to the middle‑income group (Panel B):

Working couples – with one, one and a half, or two earners – make up nearly half (44%) of middle‑income households, and they were more likely in 2018 than in the mid‑1990s to live on middle incomes. Indeed, while their share among middle‑income households declined, it declined by less than in the overall population. In particular, the one‑and-a-half earner couples have greater chances of being in the middle‑income group. “Traditional” one‑earner couples still make up the majority of working couples in the middle‑income group, but are increasingly found in the high‑income group; also two‑earner couples are increasingly found among high-income earners.
Working singles are less likely than in the mid‑1990s to make it into the middle‑income group, even as their share among middle‑income households has been growing. The reason is that their share in the overall population has grown by even more, and particularly so among low‑income households. In 2018, working singles made up about one‑in-five (22%) middle‑income households.
Workless households, many of whom are households of seniors, account for a stable share of about one‑in-three middle‑income households. They have been somewhat less affected by the decline of the middle‑income group than the overall population, i.e. their relative chances of making it into the middle‑income group have risen. This is in line with seniors’ relative increase in the chance of being in the middle‑income group (see Figure 2.10).

The increasing share of one‑and-a-half-earner couples in the middle‑income group points to the growing importance of having a second earner in the household for generating the income necessary to make it into the middle‑income group. It also reflects the rapidly risen labour force participation among women in Germany, with nearly every second working woman in the German middle‑income group working part-time (see Chapter 3, Figure 3.8).10 Households with two full earners increasingly make it into the high‑income group. Meanwhile, there also appears to be a growing share of one‑earner couples with earnings high enough for high-income status.

Trends for the presence, or absence, of children in middle‑income households are less striking (results not shown). About half of middle‑income households have children: 45% coupled households with children, plus 6% single parents in 2018. While this is less than in the mid‑1990s (51% vs 60%), this decline fully reflects the growing share of childless households in the overall population. For both couples and singles with children, the likelihood of being in the middle‑income group declined to similar extents, by less than the population average (by ‑3.4 and ‑2.0 percentage points). Couples with children have become somewhat more likely to be in the high-income group, while single parents face a greater risk of finding themselves in the low-income group.

Figure 2.12. Working couples make up nearly half of all middle‑income households, and their chances of being in the middle‑income group have increased relative to working singles

2.4.3. People with upper- and post-secondary education still make up the bulk of the middle‑income group

Adults with upper-secondary or post-secondary non-tertiary education remain the backbone of the middle‑income group. In 2018, they accounted for 58% of all middle‑income adults; 33% of adults in the middle‑income group had completed tertiary education; a small minority, the remaining 9%, had less than upper-secondary education (Figure 2.13, left Panel A).11

The level of educational attainment in the middle‑income group has clearly risen since the mid‑1990s, over and above what would be implied by rising educational attainment in the population overall. People with upper‑secondary and post-secondary, non-tertiary education were affected more strongly than those with tertiary education by the decline in the middle‑income group since the mid‑1990s (‑7 and ‑5 percentage points; Figure 2.13, Panel B). The likelihood of being in the middle‑income groups dropped dramatically for people with less than upper‑secondary education (‑15 percentage points).

Figure 2.13. Adults with upper-secondary education still make up most of the middle‑income group, but the share of tertiary-educated has been rising particularly among young people

These trends are even more marked for young people. Among them educational attainment has risen much faster, and having tertiary education has become key for making it into the middle‑income group. Among 25‑35 year‑olds in the middle‑income group, those with at-most upper‑ or post-secondary education only make up 47%, much less than among adults more generally and down from 68% in 1995. Meanwhile, those with tertiary education account for 44%, substantially more than in the general population and up from 23% in 1995. The likelihood of belonging to the middle‑income group decreased heavily for young adults with at-most upper‑ or post-secondary education (‑12 percentage points) or less then upper‑secondary education (‑27 percentage points). By contrast, young adults with a tertiary degree were somewhat shielded by the decline in the middle‑income group: for them, the middle‑income share declined by 5.2 percentage points, less than the population average.

2.4.4. Immigrants are much less likely than in the mid‑1990s to make it into the middle‑income group in Germany

Immigrants are much less likely to be part of the middle‑income group compared to the mid‑1990s. While the share of people born abroad in the middle‑income group is slightly higher than in the mid‑1990s, the immigrant share in the overall population rose by twice as much (+2 vs. +4 percentage points, not shown). As a result, immigrants’ likelihood of being in the middle‑income group declined by 9.4 percentage points since the mid‑1990s (Figure 2.14). Meanwhile, their share among low-income people rose significantly (7.7 percentage points). In 2018, about one‑in‑eight (12%) people in the middle‑income group had been born abroad (not shown). The LIS data used for this analysis do not permit separately identifying people born in Germany to immigrant parents.

Figure 2.14. The likelihood of belonging to the middle‑income group declined considerably for immigrants over the past decades

Percentage‑point change in the population share belonging to income groups 1995‑2018 (left axis) and percentage belonging to the middle‑income group in 2018 (right axis) by immigrant background, Germany

Immigrants’ lower chances to be part of the middle‑income group likely reflects compositional changes in the immigrant population in Germany over the last decades. At the start of the observation period, in 1995, one‑in-four immigrants originated from Central and Eastern European countries (incl. Poland, the Czech Republic, and the Russian Federation), and another one‑in-five from either Turkey or Italy. The composition of the immigrant population changed over the observation period: notably, the inflow of humanitarian migrants into Germany from 2015 led to an increase in the share of immigrants from Syria, Iraq, and Afghanistan. Migrants from such countries may take longer to become part of the middle‑income group even if they possess similar skills as migrants from European countries. Main challenges are a lack of German language skills, different work habits, uncertainty regarding the length of stay in Germany, and difficulties in having their qualifications recognised (Degler and Lieblig, 2017[29]).

2.4.5. The size of the middle‑income group varies substantially between regions in Germany, and it has declined much more strongly in cities than in rural areas

Regional differences in the size of the middle‑income group in Germany are substantial, including, but not only, differences between east and west. The middle‑income group is smallest in the northern city‑state of Bremen, at only 54% (Figure 2.15). Also four of the six eastern German states (Berlin, Mecklenburg‑Vorpommern, Saxony-Anhalt, and Thuringia) have middle‑income groups of below 60% of the regional population. Meanwhile, over 65% of people in southern Germany (Baden‑Württemberg and Bavaria) and the two remaining eastern German states (Brandenburg, Saxony) belong to the middle‑income group. Trends in change in size of the regional middle‑income group since the mid‑1990s vary substantially across the states, ranging from stable or even increasing middle‑income shares in eastern German Brandenburg and Saxony, to hefty two‑digit declines in the western German Saarland, the two city-states of Berlin and Hamburg, and eastern German Saxony-Anhalt (results not shown).

Cities tend to have somewhat smaller middle‑income groups than rural areas (at 63% vs 65%; Figure 2.16). This reflects a much more pronounced decline of the middle‑income group in urban than in rural areas since the mid‑1990s. Among people living in urban areas – a little more than two‑thirds of the population in Germany – the share living in middle‑income households declined by 6.9 percentage points relative to 1995. This compares to a minus of 3 percentage points among the remaining one‑third of people living in rural areas. Trends differed mainly in the period after 2005, when the middle‑income share in rural areas increased. These trends partly reflect demographic changes, with younger people and prime working-age people increasingly living in cities, and older working-age people and seniors increasingly in rural areas. However, those trends appear to hold also within demographic groups: among older working-age people and seniors living in rural areas, the share belonging to the middle‑income group increased over the past decades, while it decreased for those living in urban areas.

However, one needs to be somewhat careful in interpreting these regional and rural-urban middle‑income shares simply as measures of regional living standards, because they do not account for geographic disparities in the cost of living. Middle‑income groups tend to be smaller in lower-income regions, because they are measured against the national median household income, which can lie substantially above the regional median (Königs and Vindics, forthcoming[30]). However, also the cost of living – and notably housing – is often lower in those regions, and in some cases substantially so. Similarly, the costs of living are usually much higher in urban than in rural areas, and have risen sharply in some of the most dynamic areas in recent years (Fink, Hennicke and Tiemann, 2019[31]). A lower middle‑income household in a high‑income region or city may hence have a lower living standard than a household classified as “vulnerable” (based on the lower nominal income) living in a low‑income region or rural area.

Figure 2.15. Regional differences in the size of the middle‑income group are substantial

Percentage share of the population belonging to the middle‑income group, German states, 2018

Figure 2.16. The middle‑income group has declined more strongly in urban than in rural areas

Percentage change in the share of households belonging to income groups 1995‑2018 (left axis) and percentage belonging to the middle‑income group in 2018 (right axis) by degree of urbanity, Germany

2.5. Conclusions

Middle‑income people make up nearly two‑thirds (64%) of the German population, a slightly larger share than across OECD countries on average. After having substantially shrunk in times of rising income inequality in Germany in the late 1990s and early 2000s, the German middle‑income group did not recover again after 2005 in spite of the positive employment growth. Since 2015, middle‑ and lower-income households in Germany experienced the first growth in real disposable household incomes in nearly two decades.

The composition of the German middle‑income group has changed over the last decades as a result of demographic change, rising educational attainment, and changes in labour force participation. It has aged more quickly than the general population, as particularly young people find it difficult to secure their place. Working couples, and particularly one‑and-a-half earner couples, have increased their chances of being in the middle‑income group relative to working singles. Educational attainment is rising faster in the middle‑income group than in the overall population, and particularly for young people holding a tertiary qualification is becoming important for securing a place in the middle‑income group. Regional disparities in the size of the middle‑income group are large in Germany, and cities have been more strongly affected than rural areas by the decline in the middle‑income group.

References

[28] Balestra, C. et al. (forthcoming), “Inequalities in household wealth: Trends, drivers and policy implications”, OECD Publishing, Paris.

[13] Bartels, C. (2019), “Top Incomes in Germany, 1871–2014”, The Journal of Economic History, Vol. 79/3, pp. 669-707, http://dx.doi.org/10.1017/s0022050719000378.

[23] Bellani, L. et al. (2021), Wenn alle Teil der Mittelschicht sein wollen, https://www.progressives-zentrum.org/wp-content/uploads/2021/05/Policy_Paper_06_Bledow-Bellani.pdf.

[10] Birdsall, N. (2010), The (Indispensable) Middle Class in Developing Countries; or, The Rich and the Rest, Not the Poor and the Rest., http://www.cgdev.org/content/publications/detail/1423994.

[32] Bönke, T., A. Harnack and M. Wetter (2019), Wer gewinnt? Wer verliert? Die Entwicklung auf dem deutschen Arbeitsmarkt seit den frühen Jahren der Bundesrepublik bis heute, Bertelsmann Stiftung, Gütersloh, http://dx.doi.org/10.11586/2019014.

[5] Brueckner, M. et al. (2017), The Rise of the Middle Class and Economic Growth in ASEAN, World Bank, Washington, DC, http://dx.doi.org/10.1596/1813-9450-8068.

[19] Burkhardt, C. et al. (2013), Mittelschicht unter Druck?, Bertelsmann Stiftung.

[29] Degler, E. and T. Lieblig (2017), “Finding their Way: Labour Market Integration of Refugees in Germany”, OECD, Paris, https://www.oecd.org/germany/Finding-their-Way-Germany.pdf.

[7] Easterly, W. (2001), “The Middle Class Consensus and Economic Development”, Journal of Economic Growth, Vol. 6/4, pp. 317-335, http://dx.doi.org/10.1023/a:1012786330095.

[22] Evans, M. and J. Kelley (2004), “Subjective Social Location: Data From 21 Nations”, International Journal of Public Opinion Research, Vol. 16/1, pp. 3-38, https://doi.org/10.1093/ijpor/16.1.3.

[31] Fink, P., M. Hennicke and H. Tiemann (2019), Unequal Germany – Socioeconomic Disparities Report 2019, Friedrich-Ebert-Stiftung, Bonn, http://library.fes.de/pdf-files/fes/15766.pdf.

[11] Förster, M., A. Llena-Nozal and V. Nafilyan (2014), “Trends in Top Incomes and their Taxation in OECD Countries”, OECD Social, Employment and Migration Working Papers, No. 159, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jz43jhlz87f-en.

[14] Goldthorpe, J. (2016), “Social class mobility in modern Britain: changing structure, constant process”, Journal of the British Academy, Vol. 4, pp. 89-111, http://dx.doi.org/10.5871/jba/004.089.

[8] Gould, E. and A. Hijzen (2016), Growing Apart, Losing Trust? The Impact of Inequality on Social Capital, https://www.imf.org/external/pubs/ft/wp/2016/wp16176.pdf.

[12] Grusky, D. and J. Hill (eds.) (2018), Striking It Richer: The Evolution of Top Incomes in the United States, Routledge, http://dx.doi.org/10.4324/9780429499821.

[1] Kelly, M. (2000), “Inequality and Crime”, Review of Economics and Statistics, Vol. 82/4, pp. 530-539, http://dx.doi.org/10.1162/003465300559028.

[30] Königs, S. and A. Vindics (forthcoming), “The geography of income inequalities in OECD countries: Evidence from national register data”, OECD Social, Employment and Migration Working Papers, OECD Publishing, Paris.

[18] Lengfeld, H. and J. Ordeman (2016), “Die Angst der Mittelschicht vor dem sozialen Abstieg revisited: eine Längsschnittanalyse 1984-2014”, SOEPpapers on Multidisciplinary Panel Data Research, Deutsches Institut für Wirtschaftsforschung, Berlin, http://hdl.handle.net/10419/146117.

[2] Lynch, J. and G. Kaplan (1997), “Understanding How Inequality in the Distribution of Income Affects Health”, Journal of Health Psychology, Vol. 2/3, pp. 297-314, http://dx.doi.org/10.1177/135910539700200303.

[9] Madland, D. (2015), Hollowed out : why the economy doesn’t work without a strong middle class, University of California Press, https://www.jstor.org/stable/10.1525/j.ctt14qrz80.

[20] Niehues, J. (2017), “Vielschichtig und stabil: Die Mittelschicht in Deutschland”, IW Trends: Vierteljahresschrift zur empirischen Wirtschaftsforschung, No. Jg. 44, No. 1, Institut der Deutschen Wirtschaft, Köln, https://www.iwkoeln.de/fileadmin/publikationen/2017/322410/IW-Trends_1_2017_Mittelschicht.pdf.

[24] OECD (2021), OECD Income Distribution Database (IDD), http://oe.cd/idd.

[25] OECD (2020), How’s Life? 2020: Measuring Well-being, OECD Publishing, Paris, http://dx.doi.org/10.1787/9870c393-en.

[26] OECD (2019), Main Findings from the 2018 Risks that Matter Survey, OECD Publishing, Paris, https://dx.doi.org/10.1787/9266e48a-en.

[4] OECD (2019), Under Pressure: The Squeezed Middle Class, OECD Publishing, Paris, https://dx.doi.org/10.1787/689afed1-en.

[27] OECD (2017), Preventing Ageing Unequally, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264279087-en.

[6] Pressman, S. (2007), “The Decline of the Middle Class: An International Perspective”, Journal of Economic Issues, Vol. 41/1, pp. 181-200, http://www.jstor.org/stable/25511161.

[17] Reckwitz, A. (2019), Das Ende der Illusionen, Suhrkamp Verlag, Berlin.

[16] Reeves, R., K. Guyot and E. Krause (2018), Defining the middle class: Cash, credentials, or culture?, https://www.brookings.edu/research/defining-the-middle-class-cash-credentials-or-culture/.

[15] Savage, M. et al. (2013), “A New Model of Social Class? Findings from the BBC’s Great British Class Survey Experiment”, Sociology, Vol. 47/2, pp. 219-250, http://dx.doi.org/10.1177/0038038513481128.

[3] Thorson, G. (2014), The Rise of Inequality, the Decline of the Middle Class, and Educational Outcomes, http://www.umdcipe.org/conferences/DecliningMiddleClassesSpain/Papers/Thorson.pdf.

[21] Zucco, A. and A. Özerdogan (2021), Verteilungsbericht 2021 - Die Einkommenssituation und Abstiegsängste der Mittelschicht, Wirtschafts- und Sozialwissenschaftliches Institut (WSI), Düsseldorf, https://www.wsi.de/fpdf/HBS-008182/p_wsi_report_69_2021.pdf.

Annex 2.A. Income trends in selected OECD countries

Annex Figure 2.A.1. Trends in top, median and bottom incomes across selected OECD countries

Real average equivalised disposable household incomes by income level (1995=100), 1995 to 2018

Notes

← 1. Some other important aspects of people’s well-being, such as their work-life balance or the quality of their social connections, relate less directly to income (OECD, 2020[25]).

← 2. Unless specified otherwise, the results presented in this chapter pertain to the whole population irrespective of their age or labour market status. Children and young people have the income status of the household they live in. Chapter 3 of this review, which presents evidence on the labour market trends for middle‑income workers, focuses on working-age people (18 to 64 years) who are in employment.

← 3. Some other organisations, such as Eurostat, use a poverty threshold of 60% of the median income.

← 4. Indeed, earlier studies from Germany have used different income thresholds to define the middle class. A recent study on the income situation and perceived social risks of middle‑class households (Zucco and Özerdogan, 2021[21]) and an earlier comprehensive review of the economic well-being of the German middle class (Burkhardt et al., 2013[19]) both use a narrower definition focusing on households with incomes between 70% and 150% of the median. In another earlier review of the size and composition of the German middle class, Niehues (2017[20]) distinguishes between a more narrowly defined middle class (80‑150% of the median), lower-income / lower middle‑income households (60‑80%), and higher income / higher middle‑income (150‑250%).

← 5. Germany holds a mid-table position when looking at the change in the size of the middle-income group since the mid-2000s, just below the OECD average.

← 6. This decline of the middle‑income share in Germany reflects two factors: First, for a given income level, the income share of the middle‑income group declines as its population share declines (“population-share effect”). Second, for a given population share, the income share declines if incomes of the middle‑income group decline relative to those of other groups (“relative‑income effect”). A decomposition shows that the former effect is more important: out of the 7 percentage point decline in the middle‑income groups’ income share, 6 percentage points reflect the decline in its population share, and the remaining 1 percentage point the decline in its relative income, notably with respect to the high-income group.

← 7. Only a minority (10% in Germany, 20% in the OECD) find taxes too high for high-income households.

← 8. Disposable income is the sum of labour income, capital income, and social transfers, minus income taxes and social-security contributions. Labour income is larger than disposable income, because it is measured before income taxes.

← 9. The section does not provide a gender breakdown. The reason is that the analysis focuses on household incomes, such that any gender differences would largely reflect income differences between male and female single‑person households. Chapter 3 provides detailed analysis of gender differences in labour market developments for middle‑income workers, and Chapter 5 discusses the importance of raising female employment in Germany for boosting middle‑class disposable incomes.

← 10. Labour force participation of women in West Germany nearly doubled between 1973 and 2012, from 6 to 12 million people. Meanwhile, the number of weekly hours worked only increased by 50% (Bönke, Harnack and Wetter, 2019[32]).

← 11. This part of the analysis is restricted to adults aged 25 years and older. Levels of educational attainment are difficult to interpret for younger people, many of whom will not yet have completed their studies.

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Countries

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Publications

Publications

Featured publications

Data

Data

Featured data

News & Events

News & Events

Featured Events

About OECD

About

Engage with us

Work with us

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment