Northwestern University  
Yannay A. Spitzer


Contact Information

224-616-0143 (mobile)
847-491-7001 (fax)

Personal Webpage

Yannay A. Spitzer

Ph.D. Candidate
Department of Economics

Ph.D., Economics, Northwestern University, 2014 (expected)
MA, Economics, Northwestern University, 2008
BA, Economics and History, The Hebrew University of Jerusalem, 2004.

Fields of Specialization

Economic History, Applied Microeconomics

Curriculum Vitae

Pogroms, Networks, and Migration:
The Jewish Migration from the Russian Empire to the United States 1881–1914

This version: January 2, 2014 (frequent updates will follow)

(Job Market Paper)

Abstract: The migration of one and a half million Jews from the Russian Empire to the United States during the years 1881–1914 is commonly linked to the occurrence of pogroms, eruptions of anti-Jewish mob violence, that took place mainly in two waves in 1881–1882 and in 1903–1906. The common perception that pogroms were a major cause for Jewish migration is now doubted by historians, but little quantitative evidence exists to support or refute it. I use a new data set that matches hundreds of thousands of Jewish immigrants to their respective hometowns in the Russian Empire over the years 1900–1914, and traces the evolution of migration networks over the years 1861–1920 using incorporation records of 1,476 Jewish hometown-based associations in New-York. Additionally, the locations of hundreds of pogroms that occurred during the two waves were identified. Mapping the pogroms, as well as the yearly migration flows from more than 200 districts provides a first close look into the geographic evolution of the Jewish migration and the way it was affected by pogroms. I find no evidence that migration in its earliest stages was caused by the 1881–1882 pogroms; instead, post-1881 migration was a continuation of the spatial trend of pre-1881 nascent migration utilizing networks of previous emigrants. The second wave of pogroms, however, increased the rate of migration from pogrom-districts by at least 10–20 percent, although the evidence on whether it changed the demographic composition of the migrants is inconclusive. Above all, there was a dominant pattern of convergence in rates of migration across districts. I interpret the findings as an indication that the timing of the beginning and the intensification of Jewish migration were chiefly determined by spatial diffusion of chain-migration networks—victims of the first wave of pogroms could not react with migration because they were not personally linked to previous migrants. This supports the diffusionist view of European migration patterns, relating the late arrival of mass migration from southern- and eastern-Europe to slow spatial diffusion of migration networks. The general lesson for the economics of migration is that links to friends and relatives do not merely reduce the costs of migration; in certain circumstances they are a necessary condition for migration, their absence creating a bottle-neck delaying the evolution of mass migration by many decades.

The Dynamics of Mass Migration:
Estimating the Effect of Income Differences on Migration in a Dynamic Model of Discrete Choice with Diffusion

PDF of the work in progress paper available upon request

Abstract: The European Pattern of Transatlantic Mass Migration poses a puzzle: while time-series evidence shows that levels of European immigration to the United States were very volatile and highly sensitive to business-cycle fluctuations, there is little cross-section evidence for an effect of income on migration— poorer countries did not always send more emigrants than wealthier countries. I show that traditional attempts to estimate the effect of income on migration suffered from an acute problem of weak identification, due to data shortage and mis-specification of the dynamics of the individual migration decision. I address the identification problem, estimate the effect of income on migration, and explain the puzzle by using unique data and a new estimation model. I document yearly-district-cohort flows of Russian-Jewish migration to the United States during 1900–1914 from more than 200 districts, as well as district- and province-level panel data on demographics, income, and build-up of immigrants networks. I employ these data in a model of dynamic discrete choice with unobserved heterogeneity, that captures the way in which the option of prospective migrants to time their migration can generate large fluctuations in migration levels. An underlying diffusion process of migration "options" reflects the spatial evolution of chain-migration networks. I find that the strong sensitivity of migration to business cycles can be largely attributed to individuals optimally timing their migration—temporary shocks to migration were offset in the long run by delayed migration. The lack of systematic negative cross-section correlation between income levels and emigration is largely explained by a slow spatial diffusion of chain-migration networks.

Self-Selection of Immigrants on the Basis of Living Standards:
Evidence from Stature of Italian Immigrants at Ellis Island 1907–1925

Preliminary draft: currently making significant data extensions, results in future versions may vary.

with Ariell Zimran

Abstract: Are immigrants positively or negatively self-selected with respect to their living standards? We study this fundamental and persistent question of the economics of migration using data on one of the largest flows of free migration ever—that of Italians to the United States between 1907 and 1925. We exploit never-before-used stature data in the Ellis Island arrival records, Italian province-birth cohort height distributions, and our own geo-matching of millions of Italian passengers to their places of origin in order to construct a novel data set for our analysis. We transcribed the heights and other personal information of a random sample of 50,000 Italian passengers. Relying on the well-established relationship between population average stature and living standards, we test for self-selection with respect to standards of living by comparing the heights of migrants to the height distributions of their respective birth cohorts in their provinces of origin. This strategy constitutes the first large-scale comparison of a population at risk for migration with actual migrants based on a well-defined measure of well-being. This paper thus improves on previous studies that are limited to measures such as occupational status, or that compare migrants to samples that are not representative of the base population. We find evidence that the average Italian immigrant was shorter than the average Italian of the same birth cohort. Contrary to prevailing contemporary beliefs, however, our preliminary results indicate that immigrants were positively self-selected relative to their birth cohorts in their provinces of origin. This difference is mainly due to strong positive self-selection in shorter provinces and birth cohorts, which are primarily located in south Italy, and were the origins of a disproportionately large share of immigrants. Moreover, we find that the degree of self-selection was decreasing in the average stature of a province-birth cohort group and in the industrialization of a province, indicating that relatively more-deprived environments provided relatively higher-quality migrants to the United States. We also find some evidence that self-selection was weaker for immigrants with immediate family connections with persons already living abroad. These findings are consistent with positive self-selection being driven by liquidity constraints to migration.


Prof. Joel Mokyr (Committee Chair)
Prof. Igal Hendel
Prof. Joseph Ferrie

Back to Job Market Candidates