511 is much greater than 59

Millionaires under the microscope: data gap for very high wealth closed - concentration higher than previously reported

  • Millionaires have so far been underrepresented in population surveys, so little was known about them and the exact concentration of wealth in Germany
  • New additional sample in the area of ​​high wealth closes the previous data gap - distribution of individual net wealth is now even more unequal than previously reported
  • The richest percent of the population brings together around 35 percent of the wealth, previously it was assumed to be just under 22 percent
  • More often than the population average, millionaires are men who are older, better educated, independent and more satisfied with their lives
  • Wealth creation by people from the lower half of the distribution could be promoted in the form of wealth accounts, to which the state also pays

“Thanks to the new data, we can for the first time carry out reliable and statistically meaningful studies in the area of ​​millionaires and the very wealthy. This also makes the exact extent of asset concentration in Germany clearer. Before that, people with very high wealth were underrepresented in the socio-economic panel. ”Johannes König


People with wealth in the millions have so far hardly been represented in population surveys - correspondingly little was known about them. The exact extent of the concentration of wealth also remained unclear. The Socio-Economic Panel (SOEP) has now integrated a special sample in which people with high wealth are strongly overrepresented in order to close the previous data gap. New calculations on this basis and with the addition of publicly accessible lists of the rich show that the concentration of individual net wealth in Germany is higher than previously reported: The top ten percent therefore own a good two thirds of the total individual net wealth, previously it was assumed at just under 59 percent. The richest percent of the population has around 35 (instead of almost 22 percent) of the wealth. Around 1.5 percent of adults have an individual net worth of at least one million euros. They differ from the rest of the population not only in terms of their wealth: they are more often men who are older, better educated, independent and more satisfied with their lives. The accumulation of wealth by people from the lower half of the wealth distribution could for example be promoted in the form of wealth accounts into which the state also pays.

The Federal Government's Poverty and Wealth Report Federal Ministry of Labor and Social Affairs (2017): Living conditions in Germany. The Federal Government's Fifth Report on Poverty and Wealth (available online; accessed on July 7, 2020. This also applies to all other online sources of this report, unless otherwise noted). repeatedly states that the data situation in Germany in the area of ​​high wealth is inadequate. The available representative surveys of the population cover assets in the low single-digit millions. From so-called lists of the rich, however, it is known that in 2017 there were around 700 millionaires (families or individuals) with assets of at least 250 million euros. Info A threshold of 250 million euros was chosen because this was the upper limit of the wealth information in the regular SOEP was.

From now on, the entire wealth distribution of the population can be described

The data gap for net wealth of around three million to a quarter of a billion euros (Figure 1) was closed in the Socio-Economic Panel (SOEP) with the help of a new sub-sample (SOEP-P) (Box 1). SOEP-P includes people residing in Germany who are noteworthy in at least one company worldwide, meaning more than 0.1 percent of the shares in a company. Hold shares. The focus on company shares is based on the empirical regularity that as the amount of net assets increases, at least some of the assets are held in the form of company shares. Holding at least part of the assets in the form of company shares is also attractive from a tax point of view, as there are various tax aspects There are exceptional stocks for business assets and related expenses. There are also tax incentives for investments in real estate. SOEP-P, together with the regular SOEP and publicly accessible lists of wealthy information, should be interpreted with caution, as these are usually based on estimates based on freely accessible information and total wealth is often overestimated. For the first time to describe the complete distribution of wealth of the population in Germany. So far, this has only been possible by making assumptions about the distribution of wealth in the data gap, which were associated with a corresponding uncertainty. Christian Westermeier and Markus M. Grabka (2015): Great statistical uncertainty regarding the proportion of the top wealthy in Germany. DIW weekly report no. 7, 123-133 (available online). Cf. also the mentioned international literature in Carsten Schröder et al. (2020): Improvement of the research data infrastructure in the area of ​​the high-net-worth with the Socio-Economic Panel (SOEP). SOEPpapers on Multidisciplinary Panel Data Research No. 1084 (available online). Because with SOEP-P the number of millionaires questioned in the SOEP has also increased significantly, their characteristics can now also be statistically examined more closely.

Box 1
Data and methodkeyboard_arrow_up

The basic idea behind the drawing of the new SOEP sample SOEP-P is based on the empirical regularity, according to which top wealthy people hold at least part of their assets in the form of investments in companies. Companies, in turn, are obliged to publish information about ownership structures. According to the database Orb of the service provider Bureau van Dijk (BvD) There are around 1.7 million people residing in Germany who hold significant shares in at least one company worldwide. From this population, 1,956 households selected at random were surveyed using the standard SOEP survey instruments - including the “Your personal balance sheet” module.info For additional information on both sampling and data collection, see Schröder et al. (2020), loc. Cit.

A comparison of the regular SOEP with the SOEP-P sample shows that people in the latter have on average around 21 times higher net wealth than the previous SOEP population. This is not due to individual observations: rather, the wealth in SOEP-P is far above those in the SOEP across the entire net wealth distribution. In SOEP-P there are a total of 881 unweighted cases with an individual net worth of at least one million euros, 353 people have at least three million euros and 185 people have at least five million euros.

Recording of assets through surveys

Analyzes of the distribution of wealth on the basis of population-representative microdata are confronted with a number of methodological and statistical problems. In population surveys, assets are usually recorded at household level. For example, compare the results based on the income and consumption sample (EVS) of the Federal Statistical Office or the study "Private households and their finances (PHF)" by the Deutsche Bundesbank. The SOEP has a methodological peculiarity here, since the individual assets of each respondent from the age of 17 are surveyed. The assets held by children (persons under 17 years of age) are neglected, although it can be assumed that this is only a very small one Share of total assets. This also shows differences within households or partnerships compared to household wealth.

A comparison of aggregated wealth stocks on the basis of the SOEP with the sectoral and macroeconomic balance sheets of the Federal Statistical Office is made more difficult by a number of delimitation and definition differences. First, the Federal Statistical Office shows private households together with private non-profit organizations. Secondly, in addition to usable assets, other types of assets are shown that are not recorded in the SOEP. These include cash, the value of livestock and crops, equipment, intangible assets, claims against private health insurers, commercial loans and commercial shares in residential buildings. Thirdly, the SOEP generally inquires about the current market value, while the Federal Statistical Office assesses real estate according to its replacement value. However, the market value of existing properties deviates significantly from the replacement value.

Estimating the market value of a property in a survey is difficult, especially if the property was inherited or bought a long time ago and the respondents do not have sufficient current market knowledge. The valuation of business assets is also particularly difficult. In contrast to regular income, assets can be very volatile and thus make valuation even more difficult. This in turn leads, in addition to the general sensitivity of this topic, to increased refusals to answer or missing information on questions related to wealth.

In addition to a comprehensive consistency check of the individual information, all missing assets are replaced in the SOEP by means of multiple imputation. Markus M. Grabka and Christian Westermeier (2015): Editing and Multiple Imputation of Item-Non-Response in the Wealth Module of the German Socio-Economic Panel. SOEP Survey Papers Series C., No. 272 ​​(available online). The quality of the imputation is better due to the use of longitudinal data as part of the repeated measurement of the wealth record in 2002, 2007, 2012 and 2017 than is the case with only one-time survey. Since the persons in SOEP-P were interviewed for the first time, no past values ​​can be used. For this reason, a new type of imputation procedure was used that can better take into account the logical structure of the questionnaire design. Schröder et al. (2020), loc. Cit.

Integrated database "from top to bottom"

The present report is based on detailed documentation of a project financed by the Federal Ministry of Labor and Social Affairs (BMAS) to improve the research data infrastructure in the area of ​​high wealth. Schröder et al. (2020), loc. Cit. The central empirical basis is provided by DIW Berlin in cooperation with Cantar The socio-economic panel (SOEP) is a representative annual repeat survey of private households that has been carried out in western Germany since 1984 and in eastern Germany since 1990. See Jan Goebel et al. (2019): The German Socio-Economic Panel (SOEP). Journal of Economics and Statistics, 239 (2), 345-360 (available online). An integrated data set was created for the regular SOEP sample from 2017 - the last year to date for which processed asset information is available for the SOEP respondents - and the new additional sample SOEP-P from 2019 owed to the extensive data preparation. Although the wealth situation was also surveyed among regular SOEP respondents at the same time in 2019, at this point in time only processed information is available for the new SOEP-P sample. The associated restrictions in the comparability of the asset information are to be regarded as rather minor and should only marginally affect the general statements in this report. Finally, these integrated data were supplemented by the loud Manager magazine 700 richest Germans residing in Germany in 2017 Manager Magazin (2017): Reichenliste Germany 2017. Edition 10/2017 (available online). The highest 700 entries of the rich list in Manager Magazin were used, which were then broken down to person level by researching or statistically imputing a number of people according to the entry in order to be able to describe the entire distribution of wealth.

In contrast to other population surveys in which wealth is recorded at household level, cf. the sample income and expenditure (EVS) of the Federal Statistical Office or the study “Private households and their finances (PHF)” ​​by the Deutsche Bundesbank. are the individual Assets of all persons aged 17 and over in a private household asked separately - in several categories (Box 2). These individual assets form the basis of the following investigations. Persons from the IAB-BAMF-SOEP sample of refugees (M3 to M5) for whom no asset information has yet been collected are excluded.

Box 2
Categories of assets and debts in the Socio-Economic Panel (SOEP)keyboard_arrow_up

A total of eight types of gross wealth are recorded in the SOEP:

  • Owner-occupied housing,
  • other real estate holdings (including undeveloped land, holiday and weekend apartments),
  • Financial assets (savings, savings and mortgage bonds, shares and investment units),
  • Assets from private insurance (life and private pension insurance including so-called Riester contracts),
  • Home savings balance,
  • Business assets (ownership of sole proprietorships and participation in partnerships or corporations, after deducting business liabilities),
  • Real assets in the form of valuable collections such as gold, jewelry, coins or art objects as well
  • the value of vehicles.

On the liabilities side, a distinction is made between:

  • Mortgage loans on owner-occupied real estate,
  • Mortgage loans on other real estate,
  • Consumer credit and
  • Student Loans.

If the liabilities are deducted from the gross assets, the result is the total net assets, which are usually used for analyzes of the personal distribution of assets. The following asset components are not included in the net assets analyzed here: cash, the value of household effects, the value of farm animals and crops, equipment, intangible assets , Claims against private health insurances, liabilities due to commercial loans and commercial shares in residential buildings as well as quantitatively significant entitlements to old-age pension systems. On the relevance of the last-mentioned aspect, see Timm Bönke et al. (2018): The joint distribution of net worth and pension wealth in Germany. Review of Income and Wealth (available online).

Asset concentration higher than previously reported

The distribution of wealth can be described with different measures. In this report, percentile values ​​including the median and the mean (average) are used. The median is the value that separates the richer from the poorer half of the population, i.e. it describes wealth exactly in the middle of the wealth distribution.

The analysis of the complete wealth distribution is built up step by step: The first step is based on the regular SOEP and therefore describes the distribution up to the previous data gap. Then the new SOEP-P sample is integrated so that the previous data gap is closed. In order to describe the complete distribution, in a third step the rich list of the Manager Magazin is then integrated (Table 1). The data of the Manager Magazin contain some estimated wealth positions for individuals as well as for larger families. In the last case, a per capita conversion was made. For details see Schröder et al. (2020), loc. Cit.

Table 1: Distribution of net wealth in the SOEP and after adding additional data

In Euro

SOEPSOEP + additional sample in the area of ​​high wealth (SOEP-P)SOEP + SOEP-P + Reichenliste Manager Magazin
Percentile values ​​and meanlower limitestimateupper limitlower limitestimateupper limitlower limitestimateupper limit
Gini coefficient0,7690,7830,7960,7980,8090,8200,8160,8260,836

Note: The “lower bound” and “upper bound” are the range of a bootstrap confidence interval with 500 replications. The information from the rich list of Manager Magazin relates to the 700 richest individuals in Germany.

Reading example: "p95" stands for the 95th percentile and thus for the lowest asset in the group of the five percent of people with the highest wealth. With the addition of additional data (additional sample SOEP-P and rich list of the Manager Magazin), this percentile value is significantly higher than in the SOEP.

Sources: Socio-Economic Panel (soep.v35), SOEP-P (provisional weights and provisional wealth data for 2019), Manager Magazin (Reichenliste 2017); own calculations with weighted values.

The integration of the new additional sample hardly changes the individual net wealth in the lower wealth groups (quantiles). The median in the regular SOEP sample is 22,000 euros and, with the addition of SOEP-P, 22,800 euros. This is primarily due to the fact that the median, because it describes exactly the middle of the wealth distribution, is hardly influenced by changes - in this case better data - in the case of the very high wealth in contrast to the mean. With a look at this mean value, there are also clearer differences: It rises from around 108,000 euros in the SOEP to 127,000 euros in the integrated data set with SOEP-P. This shows that SOEP-P - in accordance with the objective of the sampling - includes many people with high wealth. As a result, the value for the 95th percentile - i.e. the lowest asset of the five percent of people with the highest wealth - rises to around 438,000 euros in the integrated database, which is around 30,000 euros (or around seven percent) above the SOEP reported so far -Value. At the 99th percentile this difference grows to almost 27 percent, at 99.9. Percentile to almost 37 percent.

With SOEP-P, the measured wealth in the upper range and thus the measured inequality increase significantly. This can also be seen in a standard measure for measuring wealth inequality, the Gini coefficient: the higher its value, the more unevenly the wealth is distributed. Info Due to negative wealth, its range of values ​​is not limited to values ​​between 0 and 1 here. Based on the regular SOEP population, the Gini coefficient is 0.78. With SOEP-P the index rises to 0.81. After additional consideration of the top 700 according to Manager Magazin, the index rises again and is then 0.83.info According to the Household Finance and Consumption Survey (HFCS) of the European Central Bank, the Netherlands had a Gini coefficient of 0.78 in 2017 the highest wealth inequality in the euro area. In contrast to the concept of individual net worth used here, the HFCS is based on household net worth, which leads to somewhat lower values ​​than for individual assets. When comparing wealth inequality internationally, it should also be noted that the upper end of the distribution is not well covered in all countries. This means that wealth inequality in Germany is also high in an international comparison. It should be noted without a doubt that the motive for building up wealth depends largely on the design of the country-specific social security system: In countries with a well-developed system of social security, the need to build up private assets as a provision is completely different than, for example, in the Anglo-Saxon countries that rely much more on their own provision.

The high degree of wealth concentration is confirmed by another indicator, the proportion of total wealth that a certain part of the population holds in total individual wealth (Table 2). In the regular SOEP, the top ten percent have almost 59 percent of total wealth, the top five percent hold around 44 percent, the wealthiest one percent of the population around 22 percent and the top 0.1 percent still around seven percent. After integrating SOEP-P, the measured asset concentration increases significantly to around 64 percent (top tenth of the distribution), 51 percent (top five percent), 29 percent (top one percent) and just under 13 percent (top 0.1 percent). A further increase in the measured concentration of assets can be seen after taking into account the cases from the list of riches in Manager Magazin. Then the values ​​are around 67, 55, 35 and 20 percent, respectively. In other words, the bottom 90 percent only own around a third of the total individual net wealth (in the regular SOEP alone it was 40 percent so far).

Table 2: Share of total net assets in the SOEP and after adding additional data

In percent

SOEPSOEP + additional sample in the area of ​​high wealth (SOEP-P)SOEP + SOEP-P + Reichenliste Manager Magazin
lower limitestimateupper limitlower limitestimateupper limitlower limitestimateupper limit
Top 10 percent (top ten percent of the wealth distribution)57,158,961,262,364,166,165,667,368,9
Top 5 percent41,443,946,748,050,653,252,654,956,9
Top 1 percent18,521,625,125,829,032,432,135,337,9
Top 0.1 percent4,67,310,79,812,815,817,520,423,0

Note: The “lower bound” and “upper bound” are the range of a bootstrap confidence interval with 500 replications. The information from the rich list of Manager Magazin relates to the 700 richest individuals in Germany.

Sources: Socio-Economic Panel (soep.v35), SOEP-P (provisional weights and provisional wealth data for 2019), Manager Magazin (Reichenliste 2017); own calculations with weighted values.

It can therefore be stated that the integration of SOEP-P (closing the data gap) and the list of the rich (top end of the distribution) leads to a significant increase in the measured concentration of wealth in Germany compared to its measurement based solely on the regular SOEP sample This also applies to the other two data sources available in Germany (EVS and PHF), as there is no comparable additional sample of high-net-worth individuals in either data source. However, there are a particularly large number of cases in PHF in regions with above-average tax revenues.

Millionaires invest their wealth differently

In order to carry out a comparative analysis of socio-demographic, economic and qualitative characteristics along the distribution of individual net wealth, the adults in the SOEP (including SOEP-PinfoDa) are not given any harmonized socio-economic characteristics for the people included in the list of riches in Manager Magazin, these are not taken into account in the following. ) divided into four groups: The lower half of wealth distribution are persons with an individual net worth below the median; the upper middle class are people from the median to the 75th percentile (from 22,800 euros to 126,000 euros), Wealthy are people from the 75th to just below the 99th percentile (from 126,000 euros to less than one million euros). The fourth group consists of the Net wealth millionaires (They make up the richest 1.5 percent of the distribution). The separate observation of millionaires is much more meaningful after the integration of SOEP-P into the SOEP, since the number of observations of wealth millionaires increases from just over 300 to almost 1200 and the associated statistical uncertainty decreases.

Millionaires not only have higher wealth - they also invest them differently (Table 3). The lower half of the wealth distribution has an average gross wealth of around 11,000 euros. Of this, more than 29 percent (around 3200 euros) is attributable to vehicles, around a quarter (around 2700 euros) to owner-occupied residential property and around 15 percent each (around 1600 and 1500 euros) to financial investments and private insurance.

Table 3: Wealth and debts of persons in the SOEP after integration of the additional sample SOEP-P

In Euro

lower half of wealth distributionupper middle class WealthyMillionairesA total of
Owner-occupied housing27335673818868057463768180
Other real estate88758805005579178425011
Private insurance14748320176611032558452
Building society savings78939945946103372942
Business assets283145514269125588721814
Property, plant and equipment1147802779242651253
Gross wealth11139963953276903126997151502
Mortgages on owner-occupied properties−2219−18598−19996−29641−10895
Mortgages on other real estate−897−2064−8642−140907−5012
Consumer credit−3964−1731−2000−14567−3094
Student Loans−378−104−47−0−226
Net wealth (gross wealth less Liabilities)3682738992970042941882132274

Note: The lower half corresponds to the range up to the 50th percentile of the net wealth distribution, the upper middle class ranges from the 51st to the 75th percentile, and the wealthy cover the range from the 76th to 98.5. Percentile down and millionaires make up the top 1.5 percent of the wealth distribution.

Sources: Socio-Economic Panel (soep.v35), SOEP-P (provisional weights and provisional wealth data for 2019); own calculations with weighted values.