Jewish communities in pre-industrial European societies were more likely to be vulnerable to persecutions during periods of economic hardship. That is the central finding of research by Robert Warren Anderson, Noel D. Johnson and Mark Koyama, published in the June 2017 issue of the Economic Journal.
Their study finds that colder springs and summers, which led to reduced food supply, were associated with a higher probability of Jewish persecutions. What”s more, the effect of colder weather on the probability of Jewish persecutions was larger in cities with poor quality soil and in states that were weaker.
Throughout most of history, religious minorities were the victims of persecution. Violence against religious and ethnic minorities remains a major problem in many developing countries today. This study investigates why some societies persecute minorities.
To answer these questions, the researchers focus on the persecution of Jews in medieval and early modern Europe. Violence against Jews was caused by a complex set of factors that have been studied intensively by historians. These include religiously motivated anti-semitism, the need to blame outsider groups and the economic role that Jews played in pre-industrial European societies.
The new study focuses on the hypothesis that Jews were more likely to be vulnerable during periods of economic hardship. The researchers test this hypothesis by combining two novel datasets.
The first dataset is drawn from the 26-volume Encyclopaedia Judaica and contains yearly information on 1,366 city-level persecutions of Jews from 936 European cities between 1100 and 1800. The location of these cities as well as the intensity with which they persecuted Jews is illustrated in Figure 1.
The second source contains data on yearly growing season temperature (April to September), which have been reconstructed from proxies including tree rings, ice cores and pollen counts (Guiot and Corona, 2010).
The first result is that colder springs and summers are indeed associated with a higher probability of persecution. A one standard deviation decrease in average growing season temperature in the previous five-year period (about one-third of a degree Celsius) raised the probability that a community would be persecuted from a baseline of about 2% to between 3% and 3.5% in the subsequent five-year period or a 50% to 75% increase in persecution probability.
To explain this effect, the researchers develop a conceptual framework that outlines the political equilibrium under which pre-modern rulers would tolerate the presence of a Jewish community. They argue that this equilibrium was vulnerable to shocks to agricultural output and why this vulnerability may have been greater in locations with poor quality soil and in polities where sovereignty was divided or which were more susceptible to unrest.
Consistent with their conceptual framework, the researchers find that the effect of colder weather on persecution probability was larger in cities with poor quality soil and in states that were weaker. Moreover, the relationship between colder weather and persecution probability was strongest in the late Middle Ages.
Furthermore, as Figure 2 illustrates, the relationship disappeared after 1600, which the researchers attribute to various factors: the rise of stronger states (which were better able to protect minorities); increased agricultural productivity; and the development of more integrated markets, which reduced the impact of local weather shocks on the food supply.
The researchers support their results with extensive narrative evidence consistent with these claims and with further evidence that the relationship between colder weather and higher wheat prices also diminished after 1600.
”Jewish Persecutions and Weather Shocks: 1100-1800” by Robert Warren Anderson, Noel D. Johnson and Mark Koyama is published in the June 2017 issue of the Economic Journal. Robert Warren Anderson is at the University of Michigan-Dearborn. Noel D. Johnson and Mark Koyama are at George Mason University.
Mark Koyama
mark.koyama@gmail.com
