Policy, History, and Mexico-U.S. Migration

Not unlike his predecessor, Joe Biden began his presidency with a series of executive actions on immigration and border security, including halting construction of the border wall, preserving DACA protections, pausing deportations (or attempting to), ending the “Muslim ban” on travel, and ending new enrollment in the Migrant Protection Protocols policy (i.e., the “remain in Mexico” program). Political controversy over immigration policy and border security is obviously far from new in American politics. Before Donald Trump kicked off his presidential campaign with a promise to “build a great wall” along the southern border, former Presidents Obama, Bush, Clinton, and Bush all faced their own crises and controversies over migration and the southern border. Before that, major immigration reform laws passed under Reagan and Johnson fundamentally shifted Mexico-U.S. migration patterns in both predicted and unpredicted ways. Earlier controversies over immigration and border enforcement in the 1950s and the 1920s illustrate that many of today’s political debates about the southern border have been going on for a long time, over a century in fact.

What can be learned from earlier eras of Mexico-U.S. migration for today’s policy debates on border security or immigration policy? In this post, I review some research on the first period of large-scale Mexico-U.S. migration, from roughly the late 19th century until the Great Depression. This follows in a series of excellent migration-related posts on Broadstreet by Vicky Fouka (internal migration, too!), Volha Charnysh, and Mayya Komisarchik.

I focus on the following questions: What drove migration during this early period? What do we know about those who entered the U.S.? How do we know? What were the consequences?

The Early Period of Mexico-U.S. Migration

People have been crossing the southern border of the United States since long before the Mexico-U.S. migration corridor became the largest in the world. The current border (save some minor adjustments) dates from the Treaty of Guadalupe Hidalgo in 1848, which ended the Mexican-American War and resulted in the annexation of about half of Mexico’s territory to the United States, and the subsequent Gadsden Purchase of almost 30,000 square miles of land in what is now southern Arizona and New Mexico in 1854. Most of this land was sparsely settled, but the first wave of Mexico-U.S. migration in some sense occurred when the border moved south over existing settlements, leading to the “entry” of up to 50,000[1] Mexican citizens into the United States.

Informal border crossings never really stopped when the demarcation line between the countries moved—especially in communities like Nogales, El Paso, and Laredo that ended up split by the border—but what most people think of as the early era of Mexico-U.S. migration began later in the 19th century, driven by a series of push and pull factors operating on both sides of the border. Classic works on this period by Manuel Gamio (1930) and Lawrence Cardoso (1980) point to the roles of rapid economic growth and labor demand in the western United States, increasing population growth and land concentration in Mexico, and the development and expansion of railroads within and between these two countries (a process that was itself very political, as highlighted by Adam Slez’s recent post on U.S. railroad development and Coatsworth’s classic book on the impact of railroad construction in Mexico).

Policy and institutions also played a role. Policies adopted to curb immigration from Asia—the Chinese Exclusion Acts in the late 19th century and the 1907 Gentleman’s Agreement with Japan—encouraged the recruitment of Mexican labor instead.[2] Though the formal recruitment of immigrants with a promise of employment had been banned by the U.S. Immigration Act of 1885, an organized system of labor contracting grew up along the border and into the interior of Mexico (see Cardoso 1980, p. 27-30 for a summary). Often employing coercive practices, recruiters would lend money to prospective workers to subsidize their journey to the U.S. with the promise of high wages, forcing them to later pay back this loan with interest. As highlighted in a recent paper by David Escamilla-Guerrero and Moramay Lopez-Alonso, these  “debt peonage”-like set of institutions, known as the enganche, had long been used to allocate labor within Mexico and was partially repurposed to send labor northward in cooperation with U.S.-based labor recruiters.

This discussion already highlights a couple of features of Mexico-U.S. migration that continue to be relevant today. While U.S. policy choices have always played a role in structuring migration flows, there is an inherent limitation to how much authorities are willing and able to enforce existing laws. The persistence of the enganche despite its ostensible illegality is a good example. As Cardoso writes, efforts to crack down on this form of labor recruitment using laws that had been designed for immigrant arrival via ship were “hopeless” given that the 2,000-mile border between the countries was sparsely monitored and given the amount of rail traffic between the countries on daily basis (p. 28). The open practice of labor recruitment suggests that the perceived threat of repercussions was minimal. Powerful interests on both sides of the border benefited greatly from the lack of enforcement, a trend that would continue in later eras of Mexico-U.S. migration.

These implementation challenges affected the Mexican government as well. In his paper and book on the other side of migration policy, David FitzGerald describes the often ambivalent attitude of the Mexican government toward the emigration of its citizens and the various attempts that were made to restrict or regulate migration from that side of the border. As FitzGerald describes in his 2006 paper, federal, state, and local authorities in Mexico frequently clashed in attempts to regulate emigration given conflicting interests, and all of these actors were limited by the realities of geography and the draw of the larger economy to the north.

The Mexican Revolution and the 1920s

The start of the Mexican Revolution in 1910 began a decade of incredible political and economic upheaval in Mexico. As I noted in my previous post, the amount of emigration during the war remains the subject of some debate (a debate which is made more difficult by the unreliability of Mexico’s 1921 census), but hundreds of thousands and perhaps as many as a million Mexicans entered the U.S. during this decade. As Cardoso writes, the U.S. was often willing to take in these immigrants given wartime labor shortages related to World War I, making decisions to ease the inflow of immigrants at critical times (1980, Chapter 3).

Developments on both sides of the border led to a continued or even accelerated pace of emigration during the 1920s. On the U.S. side, the era of relatively open immigration policy for European immigrants ended, as Vicky Fouka describes in her earlier post. Because Mexican immigrants were largely exempted from this wave of restrictions—the 1917 literacy requirement and the restrictions based on national quotas passed in 1921 and 1924—these policy changes encouraged rather than discouraged immigration from Mexico. As Edward Kosack and Zachary Ward noted in their article on Mexican migration during the 1920s, more individuals arrived in the U.S. from Mexico than most European countries during this decade.

Ongoing political unrest and economic uncertainty in Mexico also contributed to this increase in migration. The Cristero conflict (1926-1929), which disproportionately affected previously high-emigration areas of the center-west region, led to an especially sizable increase in emigration. As Julia Young describes in her fascinating book, this essentially became a “transnational conflict” as thousands of refugees, exiles, and emigrants entered the U.S. while continuing to participate in the conflict in various ways. Emigration became a convenient “escape valve” in conflict-stressed areas, leading to acceleration of U.S.-Mexico migration (FitzGerald 2006, p. 268).

Interestingly, this massive population movement arguably happened in spite of official migration policy rather than because of it. FitzGerald notes that while emigration was a useful political tool in some ways, Mexican officials often sought to curtail migration flows, imposing regulations on exit. The federal government passed a law to prohibit workers from leaving without a signed labor contract approved by the municipal government. Coupled with the U.S. regulations in the 1885 Immigration Act, which prohibited entry into the U.S. with a contract, this meant that all Mexican labor migration was “illegal” from the perspective of one of the two countries for a time (2006, p. 267). Despite this, emigration continued largely unabated until the Great Depression.

Who crossed the border? How do we know?

It is challenging to assess the size and composition of Mexico-U.S. migration during this time for a few reasons. The first is related to the above discussion on enforcement of migration laws. Because of the geography between the two countries, a significant portion of this migration has always been irregular and thus difficult to observe in official records. Much of this migration was also seasonal or circular, so it is not necessarily reflected in U.S. census data, which is better suited to recording migrant stocks over the long term.

Both classic works in this field and more recent scholarship have addressed the problem of measurement in various ways. Some of the earliest quantitative evidence on this topic comes from evidence from official border crossing data, such as a Department of Labor Report by Robert Foerster in 1925 and the study by economist Paul Taylor in 1934. Mexican anthropologist Manuel Gamio (1930) takes a somewhat different approach, using data on money orders (i.e., remittances) sent from the United States to each Mexican state in the late 1920s via the postal service. As I discuss in some of my work, these measures pick up somewhat different things. In areas along the border, people tended to cross the border more frequently, often bringing money with them, while in other areas migrants tended to remain abroad longer and remit money back home

Some recent papers in economic history have built on these early works to obtain more microlevel data and better information on those who crossed. Kosack and Ward construct an original dataset using passenger manifests from trains crossing at important border towns in 1920. From these passenger lists, the authors obtain detailed data on the characteristics of those entering the country as immigrants. Their dataset indicates that most Mexican migrants were men of working age from the north or center-west parts of the county, they were usually recorded as literate (this may have been a function of prevailing immigration laws), and they were somewhat positively selected on height. In a series of papers, David Escamilla-Guerrero uses earlier data from Mexican Border Crossing Records completed at ports of entry to study the decade before the Revolution. He provides some rare quantitative evidence on the push and pull factors that inspired this earliest wave of Mexico-U.S. migration, challenging some of the existing work that largely relied on qualitative and anthropological methods.

One of the striking things about the geography of Mexico-U.S. migration is how persistent it’s been, despite many policy changes on either side of the border. The left map above represents the amount of money orders received from the U.S. during the 1920s, as recorded in Gamio’s 1930 book. The map on the right shows the percentage in households in each Mexican state that received remittances as of the 2000 Mexican census, the first to ask households about U.S. migration and remittances. As many have noted, the networks that were established in this early phase of migration ended up shaping migration flows over a very long period of time (see, for example, Durand, Massey, and Zenteno 2001). While net migration from Mexico has fallen sharply over the last 15 years, a sizable proportion of the Mexican-born population in the United States comes from these areas.

This geographic persistence is also evident looking at the destination of Mexican immigrants in the U.S. As several articles have highlighted, unlike European immigrants, Mexican immigrants during the early 20th century were highly concentrated in the U.S. Southwest, territory that used to belong to Mexico. While the geography of both origin and destination regions broadened somewhat in the 1990s and 2000s, census data continues to show a similar concentration of the Mexican-born population in this region, as illustrated in a 2014 article in the Economist titled, “Old Mexico Lives On.” Some of the same network effects that have led to persistence in the origin areas of Mexico-U.S. migration have also led to persistence in the destination.

Lessons and conclusions

One clear lesson from this early period of Mexico-U.S. migration is that recent policy debates over migration and border policy are nothing new. These policy shifts and political arguments have been going on in various forms for over a century, though this history is often left out of contemporary discussions.

There is a lot to learn from some of the early successes and failures of the earlier migration policies discussed above, on both sides of the border.  This history illustrates some of the limitations of policy in counteracting the forces of geography and history in determining migration patterns. Both countries have attempted to encourage, curtail, or otherwise shape Mexico-U.S. migration at different times in different ways. As scholarly work on U.S. immigration policy and Mexican emigration policy illustrates, these efforts have been limited by the realities of the border and the complexities of politics since the turn of the 20th century. The decline in migration to the U.S. from Mexico over the last two decades was driven not by policy alone, but by broader economic and demographic factors.

Finally, one thing that is clear from work on the early period of Mexico-U.S. migration and from the evident persistence of migration patterns over time is that simple models focusing only on wage differentials, for example, don’t always do a good job of predicting long- and short-run migration patterns. While the number of Mexican citizens crossing the border has fallen relative to people from Central America, these lessons are worth keeping in mind when thinking about policy responses to this migration as well.

[1] See Massey, Durand, and Malone. Beyond Smoke and Mirrors: Mexican Immigration in an Era of Economic Integration. New York: Russell Sage Foundation. 2002. p. 25

[2] See Massey et al. 2002, Ch. 3 and Cardoso 1980, Ch. 2 for an overview of these policies and their effect on migration patterns.



  • Sellars is an Assistant Professor in Political Science at Yale University, where she is affiliated with the MacMillan Center for International and Area Studies, the Leitner Program in Political Economy, and the Institution for Social and Policy Studies. Her research interests are at the intersection of comparative political economy, development economics, and economic history. She is particularly interested in the politics of emigration and in the historical political economy in Mexico and Central America. Her work has been published or is forthcoming in the Journal of Politics, the Journal of Development Economics, the Journal of Urban Economics, and the Journal of Political Institutions and Political Economy. Sellars received her Ph.D. jointly in Political Science and Agricultural and Applied Economics from University of Wisconsin-Madison.

Leave a Reply