Research has shown that discrimination based on race, religion, or gender negatively affects the economic wealth of those discriminated against in areas such as the labour market (Bertrand and Mullainathan 2004) and mortgage lending (Ladd 1998). These forms of discrimination may also reduce the wealth of discriminators, who settle for worse employees or borrower matches.
An open issue is to understand the channels through which cultural discriminatory beliefs affect wealth. These channels have important policy implications – they may justify anti-discrimination policies and affirmative actions as a way to increase societal wealth, instead of only fostering the human rights of those discriminated against. Several countries and cultures do not support individual human rights, whereas maximising societal wealth is a widely accepted aim.
Jewish persecution and distrust in finance
In a recent paper (D’Acunto et al. 2014), we address this question in the domain of households’ investments, which determine the amount of wealth accumulated over time. We test whether discrimination has long-run negative effects on the wealth of households even after the discriminated population has mostly disappeared. We look at the geographic variation of historical Jewish persecution across German counties (Kreise) to proxy for households’ distrust in the financial sector. For centuries, Jews have been associated with financial services. This perception dates back at least to 1049 CE, when Pope Leo IX banned Christians from lending money at interest. Eight centuries later, in 1882, before credit unions and government-owned banks eliminated the Jewish monopoly on lending, 3% of the total German workforce was Jewish, but Jews accounted for 23% of those employed in the financial sector. Whereas the association of Jews with banking and lending services has gradually faded away due to the introduction of government-owned banks (Sparkassen) and local credit unions (Volksbanken), the Jewish population was accused of influencing the economic system through their predominance in the stock market by the German political discourse until the end of WWII. We run our own survey on a representative sample of Germans, and find that households in counties with higher historical Jewish persecution as far back as in the Middle Ages distrust the stock market more, irrespective of their education level.
Voigtländer and Voth (2012) show that antisemitism at the local level persisted for centuries from the Black Death through to the 20th century. Figure 1 depicts the concentration of historical violence against Jews across German counties (left), which we measure following Voigtländer and Voth (2012), and the ratio of households which invest in stocks in each county (right). In both cases, the darker a county is, the higher the value of the depicted variable. The maps show that the historical violence against the Jewish population was higher along the river Rhine and the river Mosel. In those wealthy areas there is also little stock market investment by households nowadays.
Figure 1. Historical Jewish Persecution (left) and stockholdings (right) across German counties
Jewish persecution and stockholdings
We find that a one-standard-deviation increase in the persecution of Jews during the Nazi period is associated with 1.2-percentage-point lower stock market participation today; that is, 7.5% of the average participation (Figure 2) once we take out variation coming from standard determinants of participation such as age, education, and income, as well as historical determinants of growth. Moreover, households in counties that experienced at least one pogrom in 1349 are 2 percentage points less likely to hold stocks today (12% of the average participation). To understand which variation drives these long-run effects, we run a set of tests comparing only geographically close counties. The baseline facts are robust if we only exploit variation in Jewish persecution across counties in the same states (Bundesländer), in the same occupation zone after WWII, or in the same virtual states. Virtual states are arbitrary partitions that address the endogeneity of state borders. Geographically close counties are very similar across several dimensions such as culture, land quality, and the quality of current and historical institutions. The effect is stable over time and similar across cohorts.
Figure 2. Jewish Persecution and stockholdings across German counties (residuals)
Voting for the Nazi party, economic crisis, and stockholdings
We also look at the electoral support for the Nazi party before the Third Reich, because Jewish persecution may have been triggered by incentives different from hatred and distrust against the Jews. Anti-Jewish sentiment was a crucial ingredient of Nazi ideology. But the support for the Nazis was also due to Germans’ discontent with their economic situation (see de Bromhead et al. 2013). Hence, the Nazi vote share should proxy for localised anti-Jewish sentiment in counties that were less affected by the economic crisis, whereas it should be a noisier proxy in counties that were affected the most. In Figure 3, we sort counties by their unemployment rate in 1933, and run our baseline analysis on different percentiles of this distribution. Consistent with our hypothesis, we only find a negative effect of the Nazi vote share in areas with low unemployment rates (left histograms in Figure 3), where voting for the Nazis was more likely driven by ideological rather than economic motives. The effect of the vote share for the Nazi party on present-day households’ stock market participation disappears in areas with high unemployment as of 1933.
Figure 3. Votes for the Nazis, economic crisis, and current stockholdings
Channels that transmit the effect over time
We consider three demand-side and two supply-side channels through which historical Jewish persecution may affect present-day households’ investments. First, on the demand side, Jewish persecution may correlate with the deep-rooted backwardness of households, whose ancestors were xenophobic, racist, and had other retrograde beliefs. Second, present-day households that are antisemitic may still associate financial markets with the Jewish population, and hence consciously stay away from stocks. Third, deep-rooted distrust in finance, proxied by the localised Jewish persecution, may have transmitted across generations independently from antisemitism, and may have lingered until the present day. The three channels all predict lower participation in counties with higher historical persecution, but they do not necessarily have the same predictions across the distribution of households by education. Schooling decreases backwardness and antisemitism, which are directly addressed by educational curricula, but the benefits of investing in stocks are very unlikely to be addressed in school. Hence, a deep-rooted norm of distrust in finance may have transmitted across generations whatever the education level of households.
Using novel data from the Friedrich Ebert Stiftung (Decker et al. 2012), we confirm that backwardness and the perceived connection of Jews with finance strongly decline with education. Consistent with the distrust in finance channel, we find that our effects are similar across households with very different levels of education. We also test for two supply-side channels – the diffusion of credit unions in the 19th century, and the depletion of human capital specialised in running the financial sector as a consequence of persecution – but we do not find evidence consistent with these channels.
What can we say about causality? Forced Jewish migrations in the Middle Ages
As an attempt to establish a causal link between historical Jewish persecution and stockholdings, we consider the forced migrations of the first German Jewish communities out of the Rhine Valley. Historical evidence shows that British knights on their way to the Holy Land during the First Crusade (1096 CE) perpetrated attacks against local Jewish communities. These attacks and those in the following centuries initiated a staggered migration pattern toward eastern and southern areas of Germany. Figure 4 depicts the dates when the first Jewish community was documented in each county – the darker a county is, the earlier the first documented Jewish community. Blank counties are those for which the data are not available. The farther away from the Rhine, the more recent the Jewish communities. Therefore, the distance of a county from the Rhine instruments for the existence of a Jewish community before 1349 CE, when we observe pogroms against local Jewish communities during the Black Death. We exploit this variation to instrument for Jewish persecution. The rationale is that in areas where a community did not exist, the probability of a pogrom against Jews must have been zero, since there were no targets for anti-Jewish violence. By contrast, wherever a community existed, there must have been a positive probability of a pogrom. Instrumental-variable results confirm that historical anti-Jewish violence reduces stockholdings.
Figure 4. First documented Jewish communities in current Germany
To our knowledge, these results are among the few to imply that policies designed to change deprecable stereotypes and norms, such as racial and religious hatred, should not only be inspired by the promotion of individual human rights, but also by the economic wealth of the broader population. Plausible policies to reduce the effects of a long-run subtle discrimination bias may simply include awareness (Pope et al. 2013). Anti-Jewish sentiment reduces stock market participation. Given the historically large equity risk premium, deep-rooted antisemitism arguably reduces the financial wealth of households, and it lowers the funds available for productive investments at the firm level. Hence, it lowers both individual and societal welfare.
Bertrand, M and S Mullainathan (2004), “Are Emily and Greg more employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination”, American Economic Review 94(4): 991–1013.
D’Acunto, F, M Prokopczuk, and M Weber (2014), “Distrust in Finance Lingers: Jewish Persecution and Households’ Investments”.
de Bromhead, A, B Eichengreen, and K O’Rourke (2013), “Political Extremism in the 1920s and 1930s: Do the German Lessons Generalize?”, Journal of Economic History 73(2): 371–406.
Decker, O, J Kiess, and E Braehler (2012), Die Mitte im Umbruch: Rechtextreme Einstellungen in Deutschland, Bonn: Verlag J H W Dietz Nachfolger GmbH.
Ladd, H (1998), “Evidence on Discrimination in Mortgage Lending”, Journal of Economic Perspectives 12(2): 41–62.
Pope, D, J Price, and J Wolfers (2013), “Awareness Reduces Racial Bias”, NBER Working Paper 19765.
Voigtländer, N and J Voth (2012), “Persecution Perpetuated: The Medieval Origins of Anti-Semitic Violence in Nazi Germany”, Quarterly Journal of Economics 127(3): 1339–1392.