Immigrant Entrepreneurship in the United States, 1910-1940. [Draft]
(Job Market Paper)
This paper studies immigrant entrepreneurship in the early twentieth century. Using the full-count US Censuses, I document the widening gap between immigrants and native whites in self-employment from 1910 to 1940. Immigrants from Asia and Southern and Eastern Europe were the most entrepreneurial. I propose multiple factors to explain the immigrant-native difference in self-employment. Immigrants were more likely to become self-employed in cities where their wage expectations were lower and their ethnic enclaves were larger. A difference-in-differences design finds that self-employment among Southern and Eastern European immigrants rose by 0.8 percentage points relative to other immigrants after the Immigration Acts of the 1920s, but an event study shows this relative rise began before the acts. Linking fathers and sons, I find an intergenerational reversal: first-generation immigrants turned to self-employment when wage prospects were poor, whereas their sons, like native sons, did so when expected wages were high.
Wartime Discrimination and Ethnic Media Markets: Evidence from WWI. [Draft]
This paper estimates the effect of wartime discrimination on the ethnic newspaper market, using the collapse of the German-language press during World War I as a case study. Drawing on hand-digitized data from N.W. Ayer & Son’s American Newspaper Annual and Directory (1909-1920) and 1910 Census controls, I compare German newspapers to Scandinavian (Swedish and Norwegian-Danish) newspapers which faced a similar secular decline and the same wartime regulatory environment, but not the anti-German hostility that followed U.S. entry into the war. Using city-language-year level data to capture market-wide effects, I find that the number of German newspapers fell by roughly 43% and total German circulation by roughly 48% relative to the Scandinavian control, with effects beginning in 1917 and growing through 1919. Event-study estimates show no differential pre-trends from 1909 to 1913, supporting a causal interpretation. The results document how an external shock to public attitudes can reshape an entire ethnic business ecosystem.
Schooling and Community of Second-generation Immigrants in the Age of Mass Migration. [Draft]
To test whether large co-ethnic communities drew the children of immigrants into work and out of school, I use US full-count census data for the 300 largest cities during the Age of Mass Migration (1910-1930) and estimate the effect of co-ethnic adult community size with a shift-share instrument built from 1900 settlement patterns and national inflows. A model of the schooling-work decision implies that if a co-ethnic labor-network channel is present, a larger community would harm schooling once children reach the legal working age of 14. A doubling of the adult co-ethnic community lowered school attendance at ages 7-13 by 1.4 percentage points and teen employment by 3.2 percentage points. Pooling the two age groups, the community effect on schooling was more positive above the working age, rejecting the view that enclaves pulled the second generation into child labor through co-ethnic job networks.
Work in Progress
Immigrant and Native Disparity in Federal Program Access
I document immigrant-native gaps in New Deal work-relief participation, larger in the South and West than in the Midwest and Northeast, and examine whether they reflect eligibility rules, local program administration, or discrimination.
The Effects of Child Support Guidelines: Cognitive Outcomes of Young Children and Mother Behavior
Using state-level variation in child-support guideline adoption and NLSY79 data, I find that guidelines improved young children’s math scores and reduced divorce and maternal labor supply.