This project draws on a novel dataset that I constructed with Jenna Nobles (University of California-Berkeley), Peter Catron (University of Washington), and María Vignau-Loría (University of Washington). To build this dataset, we digitized a representative sample of administrative records that contain information on individuals who crossed into the U.S. from Mexico at official ports of entry between 1910-1952. The records include a wealth of information about individuals, including perceived skin tone, height, occupation, literacy, place of birth, place of last residence, intended destination, purpose of trip, etc. The dataset provides robust insight into Mexico-U.S. migration patterns in the early twentieth century.

Skin Tone & Economic Stratification
In the first of several papers using this data (published in Demography), we examine the role of skin tone as a source of intra-group difference that stratified outcomes among Mexican immigrants during this period. By linking individuals from the border records to their 1940 U.S. Census records, we create a longitudinal dataset that tracks individuals from their entry in the U.S. through settlement decades later. Using at-entry characteristics to predict income in 1940, we find that—in line with dominant assimilation theories—standard measures of capital were associated with intra-group differences in attainment. However, we also find that perceived skin tone was a source of within-group stratification: being perceived as having darker skin was associated with lower subsequent economic attainment compared to those with lighter skin. Furthermore, we find that these patterns vary across contexts such that skin tone had a stratifying effect in Texas, but not in California. We attribute this finding to the effects of context-specific racial regimes, since it is not explained by differences in the characteristics of immigrants arriving in those states. We conclude that skin tone stratification has been a long-run feature of the Mexican immigrant experience in the U.S.

Agrarian Reform and Mexican Emigration
A paper in progress asks how property redistribution shapes migration by combining these data with a complete set of data on collective agrarian land grants (ejidos) following the end of the Mexican Revolution in 1920. I asses both the direct and indirect community-level effects of redistribution on emigration flows. Preliminary results suggest time-variant direct effects of property redistribution on emigration: in the first two decades of agrarian reform, communities that received at least one land grant subsequently sent fewer emigrants, but by the 1940s receiving land grants was associated with a subsequent increase in community-level emigration to the United States. Indirect exposure to property redistribution—that is, communities near other communities that received land grants—reduced the magnitude of subsequent emigration during the 1920s and 1930s, but not during the 1940s. I also find evidence that agrarian reform altered the composition of emigration flows: those who arrived in the U.S. from communities that recently received land grants were older and more likely to be female, while those coming from communities that did not recently receive land grants were more likely to be fist-time migrants and less likely to be joining someone already in the United States. (Working paper available on request)

An example of the administrative records used to construct the dataset. This record, from 1920 at the El Paso border crossing, indicates that 20-year-old José Rojas was an unmarried laborer from Valparaiso, Zacatecas, who had never been in the U.S. before. José was headed to El Paso in search of work, where he intended to reside for one year. He traveled alone. José could read, but not write. Standing at 5' 8", the border crossing agent noted José's skin complexion as "medium." (Source: Ancestry.com)

Mexicans entering the United States in El Paso, Texas, 1938. (Source: Library of Congress)