This dissertation explores questions on the substance and methods of economic history. Chapter one studies a little-known policy change in the late 19th and early 20th centuries to explore the causal effects of political exclusion on the economic wellbeing of immigrants. Starting in the mid-19th century, twenty-four states and territories expanded their electorates to allow non-citizen immigrants the right to vote; from 1864-1926, however, these same jurisdictions reversed this policy, creating a mass disenfranchisement for which the timing varied across states. Using this variation as well as a discontinuity in nationalization proceedings of the era, I find that political exclusion led to a 25-60% reduction in the likelihood that affected immigrants obtained public sector employment. I also document significant negative intergenerational effects: individuals of immigrant parentage born around the time of disenfranchisement earned 5-9% less as adults than comparable individuals of native parentage. I am able to rule out as mechanisms for this intergenerational effect a variety of policy and spending channels, but find evidence for a reduction in English-language proficiency among disenfranchised immigrants, which may have adversely affected the human capital of their children. Chapter two explores the causes of the adoption and repeal of alien voting in the United States. This policy shift offers a valuable opportunity to understand the forces determining political inclusion and exclusion in a formative period of American democracy, and contributes to the broader literature on theories of democratization. I use qualitative evidence from the historical record to outline competing theories of both adoption and repeal of alien voting, and then rationalize these hypotheses within the context of a median voting model. Using a discrete time hazard specification, I find evidence consistent with the hypothesis that states used alien voting as a locational amenity, with the objective of inducing immigrant in-migration in order to foster agricultural development. The results indicate that the timing of repeal was driven by social costs, rather than economic or political factors, although there is evidence for heterogeneity in correlates of support for repeal across states. Finally, the costs of constitutional change were salient for both adoption and repeal: states for which it was less costly to re-write or amend the constitution were more likely to adopt and repeal alien voting. Chapter three is a co-authored methodological study intended to assess the efficacy of commonly used techniques to create name-linked historical datasets. The recent digitization of historical microdata has led to a proliferation of research using linked data, in which researchers use various methods to match individuals across datasets by observable characteristics; less is known, however, about the quality of the data produced using those different methods. Using two hand-linked ground-truth samples, we assess the performance of four automated linking methods and two commonly used name-cleaning algorithms. Results indicate that automated methods result in high rates of false matches – ranging from 17 to over 60 percent – and the use of phonetic name cleaning increases false match rate by 60-100 percent across methods. We conclude by exploring the implications of erroneous matches for inference, and estimate intergenerational income elasticities for father-son pairs in the 1940 Census using samples generated by each method. We find that estimates vary with linking method, suggesting that caution must be used when interpreting parameters estimated from linked data.
【 预 览 】
附件列表
Files
Size
Format
View
Three Essays on the Substance and Methods of Economic History
PDF
download
878987797
028-85220240
