What are the economic and political determinants of ethnic violence? Political scientists have extensively studied how political factors such as political instability and weakness of the state affect the likelihood of outbreaks of civil conflict (e.g. Fearon and Laitin 2003). Economists, in turn, have provided robust evidence that economic downturns are an important trigger of civil conflict and that ethnic violence is aggravated by economic competition between ethnic majority and ethnic minority (see Blattman and Miguel 2010 and Couttenier and Soubeyran 2015 for surveys of this literature).
In a new paper (Grosfeld et al. forthcoming), we examine how political and economic factors interact to drive pogroms in the environment of widespread antisemitism. Our analysis sheds new light on how ethnic minorities can be susceptible to ethnic persecution when they are segregated into economic niches complementary to the majority, and thus, are not in direct economic competition with the majority.
In particular, we examine the conditions under which anti-Jewish violence broke out in the 19th century and the early 20th century in the Russian Empire, which at that time was home for the majority of world’s Jews. Jews in the Russian Empire dominated moneylending and trading sectors. Despite high economic segregation between the Jews and other subjects of the empire, Jews were a target of persecution, which at its worst points resulted in mass anti-Jewish violence that brought the word pogrom into European languages.
We collect a novel data set that combines panel data on anti-Jewish pogroms as well as measures of shocks to agricultural output and crop prices, with detailed cross-sectional data on local ethnic composition by occupation and with historical time series of political crises throughout the period. Figure 1 presents the map of the locations of pogroms in the Pale of Settlement, the vast area of the Russian Empire to the east of which Jews were not allowed to live.
Figure 1 The Pale of Settlement and the geographic distribution of pogroms
Using these data, we first show that both economists and political scientists were only partly right in claiming that economic and political shocks are associated with ethnic violence. Pogroms indeed were more likely to occur during the times of economic crises and during political crises. Economic crises in the Russian empire’s agriculture-dominated economy were driven primarily by crop failures and grain shortages. Political crises, in contrast, varied in nature: they were driven by wars, violent political turnover, coups, revolutions, and civil wars.
Yet, neither economic crises (including the most severe grain shortages such as the Russian famine of 1891–92) nor political crises (such as the Napoleonic invasion or the Crimean war) caused pogroms alone. Pogrom waves took place when and only when economic shocks, caused by severe crop failures and grain shortages, coincided with political turmoil threatening the political and social order. This result is illustrated in Figure 2.
Figure 2 Pogrom waves, economic shocks, and political turmoil
(a) Pogrom occurrence
(b) Pogrom occurrence and period of grain shortages
(c) Pogrom occurrence and political turmoil
(d) Pogrom occurrence at the intersection of political turmoil with period of grain shortages
In panel (a), we present the number of pogroms over time. The vast majority of pogroms came in three waves, which has been well noted by historians (e.g. Klier and Lambroza 1992). In panel (b), we overlay this time series with economic shocks, i.e. the times of high grain prices and grain shortages, caused by severe crop failures in major grain-producing areas (such as Ukraine). In panel (c), we superimpose the timing of pogroms on episodes of political turmoil, i.e. the periods of extreme political uncertainty about the future, such as following the assassination of Alexander II, the Tsar-Liberator, when peasants thought serfdom would be reinstalled by the new tsar, or during wars that led to occupation of Russia’s territory, or the Russian revolutions. In panel (d), we show what happens when the economic shocks coincided with political turmoil.
Pogrom waves occurred every time economic shocks coincided with political turmoil. Importantly, the economic shocks that occurred outside political crises were no less severe than those that occurred during political crises.
Second, we show that the occupational segregation between the Jews and the majority played an important role in triggering pogroms when economic and political shocks coincided. In particular, pogroms occurred primarily in localities where Jews dominated intermediary occupations related to agriculture, i.e. moneylending and grain trading. Economic shocks together with political shocks did not result in pogroms in localities where the Jewish community specialised in other occupations, including middleman occupations unrelated to agriculture, such as trading in non-agricultural goods.
Figure 3 illustrates the Jewish domination over middleman occupations servicing agriculture as the second key driver of pogroms. Panel (a) shows that the frequency of pogroms in localities that suffered from local crop failures in times of political turmoil strongly depended on the share of Jews among local moneylenders: in times of political turmoil, local crop failures were more likely to trigger pogroms in places where most of the moneylenders were Jewish. Panel (b) shows that the share of Jews among local grain traders is a strong predictor of pogroms when political turmoil coincided with periods of high grain prices and grain shortages, due to crop failures in the major grain-producing areas.
Figure 3 Pogrom occurrence and the shares of Jews among creditors and grain traders
(a) Pogroms vs. share of Jews among creditors during political turmoil across localities hit by local crop failure
(b) Pogroms vs. share of Jews among grain traders during political turmoil and during period of grain shortages across all localities
These results highlight the economic and political factors that led to outbursts of violence under the conditions of inherent religious and ethnic animosity (e.g. Grosfeld et al. 2013). Prevalent antisemitism among non-Jews in the Russian Empire was, of course, a prerequisite to pogroms.
We consider a number of potential mechanisms that can explain these results. Several prominent hypotheses are rejected by the data. Our findings are inconsistent with the traditional ‘scapegoat’ theory, according to which Jews were blamed for all misfortunes of the majority (e.g. Girard 1986, Glick 2008). First, basic scapegoating theory would imply that Jews are targeted as a group, irrespective of their occupations.
Second, political turmoil could matter for scapegoating because it is associated with lower probability of punishment for potential perpetrators. However, in the first two waves of pogroms, political turmoil was not associated with a decrease in the probability of punishment; instead, it meant the increase in uncertainty about the future, affecting the relationship between creditors and debtors (which we describe below). Only in the third wave of pogroms that occurred after the 1917 revolution, ethnic violence was exacerbated by the absence of law enforcement in the midst of the civil war. We show that our results hold both before and after 1917