The panorama of Mexican migration changed substantially during the 21st century with decreasing emigration flows, a near-zero net migration rate, and increasing return migration. According to the Mexican Census, between 1995 and 2000, about 1.6 million people left Mexico. However, between 2015 and 2020, about 803,000 people emigrated from Mexico. The current dynamics starkly contrast with the decades of high emigration rates from Mexico during the 20th century.In this dissertation, I study one of the changing components: emigration. Understanding the size and composition of emigration is essential as it has unique implications on the migrants, their communities at their origin, their destinations, and repercussions over generations. Migration research is often complex as there is limited detailed data, and many times, it only pertains to a specific stage of migration. These issues usually prevent researchers from analyzing more detailed determinants of migration or from making claims about a general population. I focus on three important elements in migration research: availability of detailed data, understanding migration as a process, and differences by sex.
In the first chapter, I show that an underused high-quality dataset from Mexico, the Encuesta Nacional de Ocupación y Empleo (ENOE), can be used to understand migration in Mexico. The ENOE is the Mexican Labor Force Survey. Although it does not track migration directly, its survey structure allows it to identify migrants (both emigrants and immigrants) along with a rich set of covariates. I validate the ENOE by carefully comparing it to gold-standard data such as the Census and the Encuesta Nacional de la Dinámica Demográfica (ENADID), a demographic survey. My results find that immigrants and emigrants from the ENOE match standard data across key demographic characteristics.
Moreover, in the aggregate, the ENOE produces migration rates comparable to official demographic estimates. The ENOE may be preferred over other data because it has economic and demographic variables of migrants before they leave or after they enter Mexico. Also, it is a frequent and ongoing panel survey, which allows for timely estimates. This chapter also provides guidelines for the practical use of the ENOE for research on migration. Using the ENOE, in addition to official estimates, can provide researchers with an updated view of migration trends and identify areas of research. Due to its richness, the ENOE is the main data for the next two chapters.
In the second chapter, I collaborate with Rui F. Carvalho to analyze an understudied stage before emigration: preparing to emigrate. International emigration has been identified as a stepwise process entailing the formation of aspirations to emigrate, making preparations for the move, and eventually realizing those intentions and plans. Extant research has either focused on the aspirational phase or on the actual determinants of emigrant abilities, with less attention paid to the phase of preparing or planning for the move. Using the ENOE, we investigate (i) the relationships between emigration preparations and actual international emigration; (ii) whether preparations to emigrate depend on specific demographic and structural characteristics; and (iii) how preparations may affect employment outcomes. We uncover that preparing to emigrate is a rare event. We also find variations in sociodemographic features related to emigration preparations, which are different from predictors of actual emigration. Further, utilizing event study analysis we examine the relationships among employment behaviors, preparations, and emigration. We find that differences between whether preparations are materialized into emigration or not are associated with changes in income, hours worked, and being employed. Overall, the results provide a better understanding of emigration preparations and how they relate to actual emigration and to employment behaviors.
In the third chapter, I consider how international emigration from Mexico differs in a key demographic trait: sex. The share of female emigrants varies across countries: in many countries, emigration is composed of mostly men, while in others, female emigration is slightly above 50%. Smaller flows of female emigrants indicate more extensive structural conditions that prevent the mobility of women. This immobility is a form of inequality between male and female emigrants. This chapter analyzes differences between male and female Mexican emigrants during the 21st century, and explores how they translate into the share of female emigrants. In doing so, I revisit the question of the feminization of international emigration. Between 2000 and 2020, the share of female emigrants increased from 25% to 33%, as a result of decreasing male emigration with a steady female emigration. Using decomposition analysis, I find that this slight feminization process is due to changes in sex-specific patterns of determinants of emigration. This work adds to the extensive literature on Mexican migration by contrasting recent patterns of female selection into migration rather than focusing on male emigration.
Overall, my dissertation fills these gaps in data and substantive comparisons by providing a rich understanding of the composition of migrants before they leave, and if there is a process of feminization of international emigration in Mexico.
