The Selectivity of Migrants

In migration studies, being able to identify the characteristics of those who migrate to a given destination, in this case the United States, is very important. Commonly, when there are migration flows in significant amounts to a particular country, these are made up of people of all kinds in a socioeconomic sense. Men and women; youth and adults; people with many and few years of schooling; there are farm workers, workers and professionals, etc., but it is still important to be able to know if people with certain similar characteristics prevail in these migration flows, as this allows us to identify their profiles and thus we can relate some aspects of their contexts of origin, as well as the possibilities that they have or would have in the destination according to the characteristics of the labor markets they will arrive in. This has been termed “migration selectivity” and has been used as an important yet limited heuristic tool. It is assumed that under normal conditions when there is an emigration flow from a place, it is most likely that there will be no emigration of all or almost all of those who live in the areas of origin (this would only occur in cases of war or catastrophes), but mostly some people with certain characteristics would leave. Obviously, this is an assumption that must be confirmed with data, but it allows us to know to a certain extent some of the socioeconomic fragilities of the contexts of origin, as well as certain possibilities of insertion in the destination society (^{Browning & Feindt, 1969}).⁵

Main Characteristics of the Migration Flow to the United States (1993-2017)⁶

The composition of the outgoing flow⁷ has always been predominantly male, although the percentage of women has gradually increased from 7% in 1993-94, to figures averaging between 20% and 25% over the last 10-15 years.

Between 1993 and 2009, the average percentage of people aged 20 to 39 migrating to the United States was 65%. A few years later, between 2010 and 2017, that percentage would drop to 54%; the total outgoing and return flows were already declining in those years. In the entire period from 1993 to 2017, the average has been 61.5% with a standard deviation of

14.5 and a median of 62.5; clearly, these age groups have been the majority. It has been a young population at the best of their labor possibilities.

As for the average schooling level of this outgoing flow, in the 1990s those who had completed primary school or less were a majority of around 60-65%, and between 20-25% had some years of secondary school or had completed secondary school; those who had a few years of high school or perhaps a little more barely reached 10-15%. These trends gradually reversed, until the members of these flows appeared as follows in the years between 2010 and 2017: 37% having completed the lowest levels of primary school, or less; 32% at the high school level; and 31% those who had a few years of high school or more (these are averages). In fact, between 2010 and 2017, the average years of schooling increased from 7.4 to 10 years among this migrant population. Thus, it can be seen that the people in this flow reflected with certain fidelity the changes in this respect occurring in Mexico in general, although also maintaining the global characteristic of being a flow of people with low educational levels still within the context of the changes that were taking place in the country.

On the other hand, although we know that the origin of migrant workers from rural areas has changed over the years with significant increases in those who came from urban areas of the country,⁸ most declared having been employed as “field laborers” in their place of origin. The average in this regard for the last 24 years is 42.6% with a standard deviation of 7.4 and a median of 45.3%. However, with respect to this occupational characteristic, I believe that we must also consider that in the reality of the country’s communities, faced with an agriculture that for various reasons has had low productivity, the labor channel that most of the men in those places has had at their disposal has been that of construction activities, whether they carry it out in their own communities or in cities in the region.

As shown by various field studies, it is common for those who work their farming plots to at some point work in construction, either themselves or their children (for women, their option has been mostly housekeeping in the cities). Hence, I have decided to group together in one category those who said they were “field laborers” in Mexico and those who said they were “construction workers”, since these are two occupational strata that feed off each other (^{Arias, 1992}; ^{Lara Flores, 1996}; ^{Verduzco, 2007}). For the last 16 years of the Emif Norte, the average in this joint category has been 59.2%, with a standard deviation of 9.7 and a median of 62.6%. Occupationally in Mexico, this has been the predominant group in outgoing flows to the United States. It is interesting to note that laborers in the agricultural sector in Mexico have gone from 22% in 1990 to 16% in 2000, and only 11% in 2010, in relation to the total working population as a whole. That is the occupational stratum from which Mexican migration to the United States has been coming for many years, and so the drop in these numbers has to do not only with internal changes in rural activities in Mexico, but also with the departures to the United States of that population.

As for regional origins, we know that, although flows have come from all corners of the entire national territory, the majority of them originated in center-west areas and in some of Puebla-Oaxaca, as well as coming later from the center-north and other areas in the country over the years, until finally coming from all corners of all the regions of the territory. This has been another unusual fact in that during the first eighty years of the last century, migrations to the United States had been circumscribed to the great region of the center-west, plus some others from the center and the north, and from there they expanded to all the country, although in a restrained way (^{Zúñiga & Leite, 2006}).

With the above information, it is clear that those who have made up the majority of this flow have been predominantly young men, although with a growing presence of women. Most of the men worked in agriculture in Mexico, as well as in construction activities.⁹ As for schooling, these migrants have mostly been people with low schooling levels despite the fact that over the years these levels have been increasing in line with the general increases in this regard in Mexico. This highlights the importance of this type of profile as more specific to Mexican migration to the United States, which in turn is directly connected to the issue of “migration selectivity” that was already mentioned.

It is also important to highlight that these data from the Emif Norte, which have a wide coverage in space and time, match in general terms with what was shown in the dozens of community studies carried out in Mexico in those years, where it was clearly seen that the “typical migrant” (the one who has mostly migrated from the communities) has been above all a young man mostly occupied in rural activities, with little schooling (^{Calvo & López, 1988}; ^{Dinerman, 1982}; ^{Donato, 1994}, ²⁰⁰¹; ^{López Castro, 1986}; ^{Massey, Alarcón, Durán & González, 1990}; ^{Mines & Massey, 1985}; ^{Massey, Goldring & Durand, 1994}; ^{Verduzco, 1995}; ^{Verduzco & Unger, 1997}).

However, to complement the above, it is not superfluous to explain that when pointing out the main socioeconomic characteristics of the majority of those migrating to the United States, there have also been other migrants who, although in smaller numbers, have been part of that flow and displayed different socioeconomic profiles. This should lead us to be careful not to try to encompass or understand these flows as totally focused on what is now being pointed out as the most common characteristics. As stated above, migration selectivity is useful as a heuristic tool, but also limited.

It was already pointed out that such high migration flows began to decline as of 2007, until they became relatively small, hence the topic of the possible factors that have had to do with these changes becomes relevant.

A number of questions arise; among them: why has this migrant flow suddenly decreased according to what is presented in the figures from various sources?

It has been repeatedly stated that the decrease in migration trends from Mexico to the United States in recent years has been mainly due to the economic crisis of 2008-2009, although a demographic issue has also been brought up when pointing out that the general fertility rate in Mexico having dropped in previous years, this decline ultimately influenced the downward trend of Mexican emigration (^{Passel & Cohn, 2018}). Likewise, the greater strictness of border control by the United States has been mentioned as another cause (^{Anguiano & Trejo, 2007}).

As for the first point about the economic crisis, as has happened on other occasions, economic crises either in Mexico or the U.S. without a doubt have repercussions on migration flows: sometimes more Mexicans leave or more come back, yet those ups and downs have been circumstantial and have usually readjusted to general trends in a short time. However, in the 2008-2009 crisis, the previous flow intensity never reestablished itself, as can be seen in the first Table of this work, despite the fact that the crisis was overcome.

Suggested Hypotheses to Explain the Decrease in Flow

In addition to what was said in the above paragraph, I first propose a set of three hypotheses where one of them seems central to me, but which is accompanied by two others that would also be important, although less than the central one; then I will provide other complementary arguments. In addition, and this is relevant, I affirm that although there has been a significant decrease in Mexican migration to the United States, there have also been important changes in the characteristics of contemporary migrants that have implied an important continuity in labor flows.

The central hypothesis that I propose is that since for a little more than two decades it was mainly young people who were migrating en masse, in the end this source of young labor was exhausted, and it is this exhaustion of a specific and selective type of supply which has mainly been observed in the decrease of these migration flows. Note that I am not referring to the decline as an effect of the decrease in general fertility rates, as has been otherwise repeatedly proposed (^{Passel, Cohn, & González-Barrera, 2012}), but rather to the changes in the profile of specific age groups.¹⁰ Let us recall the data presented from the Emif Norte (^{El Colef et al., 2020}) where it is observed that the largest age group of the flow was made up of young people.

As we know, in the basic population equation the sources of population loss are deaths and emigration, so in order to know how much is due to emigration and how much to mortality, it would be necessary to monitor specific age cohorts.

The country’s population data shows, for example, that the following age groups lost significant volumes of population between 1990 and 2000, as follows:

Note especially the great loss of the youngest of all, who are those who were 15 to 19 years old and 20 to 24 years old in 1990. The total gross loss of those two groups was 2 199 300 people for those years alone. Mortality in these cases was 211 782. The mortality loss for the next three age groups was 386 053. Based on this information, the probable population loss due to emigration to the United States for those years would be 2 353 250,¹² counting only those age groups that we know have been the majority.¹³

The loss of the young population is undoubtedly a serious problem insofar as it implies, among other things, the effects of a future slowdown in aspects of economic growth by a reduction of the working population. As we will see later, this situation has clearer and more forceful effects at the local levels, where the shortage of labor is becoming noticeable.

The above data regard the country in general, but we know that emigration to the United States does not affect the different regions equally, since there are areas of a high migratory intensity and others where, although existent, the phenomenon has a very minor presence.

I summarize this information with data for the state of Michigan: in the case of an entity and unlike what was presented before for the country, here three effects are found together, that of mortality, that of internal migration that enters and leaves the state, and that of international migration; however: a) Michoacán has not been a state that attracts population from other states; b) Michoacán occupies the third place in the index of migratory intensity towards the United States with “very high” migration (Conapo, 2010); c) between 1990 and 2010, the original cohort of 15 to 19 years lost 32% and the initial cohort of 20 to 24 years lost 20%.

Certainly, such large losses of the young population have negative repercussions for future economic growth, as pointed out in a study on demography and migration by ^{Gómez de León & Tuirán (1998}).

Now let us see a local case in Michoacán known by the author and about which there is ethnographic information, that of the municipality of Chavinda.

It has been a municipality wherein emigration to the United States is typical. Between 1990 and 2010, not only did population did not grow, but it also had a net loss of 1 060 inhabitants, equivalent to 9.6% of its original population in 1990. Yet in specific age cohorts, between 1990 and 2010 it lost 34.7% of the original population between 15 and 19 years old, and 52.05% of the population that in 1990 was between 20 and 24 years old (see Table 2).

As in the case of Michoacán, here we can see the joint effects of mortality and those of internal and international migrations, yet from studies of that community we also know that although there are and have been migrations to the nearby city of Zamora and to other parts of the country, emigration to the United States has been of central importance (^{Massey, Alarcón, Durand & González, 1987}). One issue that I wish to highlight is that of the loss of local labor force, which, as can be seen, has been very acute, especially between 1990 and 2000. Let us also remember that these are selective migration flows that have mainly affected the local working population mainly dedicated to agricultural and construction activities, as can be read in the referred works. The local consequences of this phenomenon had clear positive and negative effects on productive conditions. On the one hand, when labor was scarce, it became more expensive, a situation that can be seen as positive as related to an increase in wages. On the other hand, this scarcity also led to necessary changes in the production structure, both with the introduction of machinery and with the introduction of different crops that require less labor. It should be noted that municipalities such as Chavinda cease to be a source of new migration flows simply because the local workforce that was “available” (selectively available) has been exhausted. This is seen from a local point of view, but at a global level, situations like this were also leading to new communities from other parts of the country to participate in the migration flow to the United States. This is why over the last 25 years migrations to the United States, which were at first very localized in the center-west and north of Mexico, have gradually reached more towns in states of the country that previously did not participate in that job market. This was and continues to be a logical way in which the labor supply of that specific type of labor continued and has continued to provide arms to the type of demand that partially continues to exist in the United States.

Now I propose a second hypothesis on the possible deterioration of wages in the United States in the labor niches where migrants usually position themselves. To begin with, I will cite the main conclusions of a study presented in the Binational Study (^{Mexican Ministry of Foreign Affairs & U.S. Commission on Immigration Reform, 1998}). ^{Davies, Greenwood, Hunt, Kohli, and Tienda (1998}) ask themselves whether low-skill Mexican workers are labor force substitutes or complements and, in each case, what would be their impact on wages? They conclude that this type of worker is not a substitute for the native labor force but that it does compete with the same type of worker, so one of the effects has been that when more low-skill Mexican workers arrive in labor markets where such type of workers already exists, this impacts on their wages negatively. This occurs, according to the authors, mainly in California and Texas, which are places where there is a great number of Mexican migrants.¹⁴ Given this, I interpret that surely because of this negative effect on wages in certain places, many of the migrants from Mexico in the United States began to explore new destinations in other states, as we later saw in both personal accounts¹⁵ and in U.S. statistics from subsequent years. Mexican migrants gradually switched destinations the more traditional migration locations (California, Texas, Chicago, New York and nearby) to others such as Arizona, Nevada, Kentucky, North Carolina, Georgia and some central and southern states (^{Giorguli, Gaspar, & Leite, 2009}). These changes took place over at least twenty years, from 1990 to 2010. It was a process in which, based on the knowledge acquired from their U.S. experiences, Mexican workers discovered other niches in places not known before, a situation that expanded the possibilities of that labor market a little more in time and space to the point where a certain limit was reached, at which labor demand began to weaken. On the other hand, the likely decrease in wages for this demand had the effect of wages no longer being able to compete with the costs of an increasingly expensive and risky trip.¹⁶ In addition, probably for many of the new potential labor suppliers in Mexico, this type of job offer was falling outside their expectations, having achieved higher schooling levels as of late.

These were processes that took place during those years in which, for demographic and economic reasons, and because of the peculiar characteristics of this type of selective migrant, an incidence occurred that later decreased the total flow from Mexico. Yet I must insist that it has oftentimes been easier to blame the most immediately apparent, as was in this situation the 2008-2009 economic crisis in the United States.

On the other hand, and to add more elements to the arguments proposed, other changes have taken place in Mexico that, although not leading the entire population to be in a situation of overarching well-being, have indeed represented certain improvements, such as an increase in the years of schooling of young people who, upon achieving more education, try to turn to other types of activities requiring higher qualifications. Likewise, there are other factors that may be less direct but that can still be forceful, and that in general terms have already been presented in other studies on migration in general. These are the most extensive changes that have taken place in the population distribution in Mexico. Although for a long time there has been a shared belief that it is a country with few large cities, much has changed; in a not so long period of time, the number of cities in Mexico with more than half a million inhabitants reached 26, nine of them with more than 1 million.¹⁷ These cities are spread over the various regions of the territory, so probably the migration flows that previously went to the three large cities and the United States have now been at least partially directed to the large cities of their own regions. This new situation has surely also had some impact on the flow previously bound to the United States.

Two issues are still to be addressed, having had a strong impact both on what has to do with Mexico’s labor contribution to the United States, and on various aspects of the relationship with that country. I will first address the issue of the border strip that has to do with both. Then we’ll take a look at recent changes in the rates of migration to the United States.

The Border Strip

The border strip of Mexico is that geographical area that, as the name expresses, goes around the border with the United States. It is not a geographically or economically homogeneous region, but perhaps it can be defined rather by that strong neighborhood with the country with which it shares that same space, the United States. Perhaps it is that intense relationship with the populations on the other side, altogether with that mixed culture, that would best define this sociodemographic space.¹⁸ It was populated and grew even more on the Mexican side during the 20th century based on very clear policies by several administrations, especially the federal government. ^{García (2010}), ^{Aboites (2010}) and ^{Cruz (2010)} clearly summarize the settlement processes of that large area from the time of the Mexican Independence to the present day and, on the other hand, ^{Valenzuela Arce (2002}, ²⁰⁰³) in several of his works exposes the cultural richness of those mixed sub-cultures that have been shaped over the years across the border. It is a region where the inhabitants at both sides of the border have daily binational experiences of diverse order. ^{Cruz (2010)} speaks of 35 to 38 municipalities on the Mexican side. ^{Alegría (1990}) has carried out important works on the northern border from the perspective of urban development. For this work I have made a count of the current population of those municipalities, to which I have added the population of the U.S. border counties, covering approximately 100 kilometers on each side. The total population that lives there reaches a little more than 15 million people. Along the border strip there are urban complexes only divided by the line marking the national limits, but still sharing important aspects of daily life and the natural resources in the surroundings. Many of the inhabitants of this large area have relatives on either side, whom they visit very frequently and with whom they share information on various topics pertaining jobs, the prices of various items, and job, educational and cultural options.

The U.S.-Mexico border is considered one of the most dynamic and complex in the world. According to the U.S. Bureau of Transportation, annual vehicle crossings between 1997 and 2000 averaged 277.5 million per year, equivalent to 754.7 thousand per day (^{Bureau of Transportation Statistics, n. d.}).

The border strip, with its very peculiar characteristics, has a great weight in the economy, politics and social relations, particularly when it comes to state entities in the two countries. The labor contribution of Mexicans, especially those living on the Mexican side of the border, to U.S. economic sectors is enormous, since thousands of them go to work there daily (Santibáñez & Cruz, 2000). This matter, although well known, has not been sufficiently accounted for as another aspect of the great Mexican labor contribution to the U.S. economy.

Changes in the Migration Modalities from Mexico to the United States

It has been a clear fact that the flow of Mexican laborers to the United States has dropped considerably; that is at least true for that traditional, mostly undocumented flow, of which we seek to account for here. However, there are other recent data that, while not contradicting the above, show that labor flows from Mexico continue in figures that are certainly lower than those of 2007 and previous years, but still very high and important; is this factual?

Let us remember that the bulk of the migrant flow from Mexico was made up primarily of young people with low educational levels long-experienced as agricultural laborers. In another section we attempted to show that probably for various reasons, such as the demographic depletion of these particular groups in thousands of locations in Mexico, as well as due to certain gradual decline in wages in the United States (for these groups of laborers), as well as because of the rise in transportation costs to destinations as an effect of the greater difficulties at crossing, it became more problematic for undocumented migrants to access that labor market in the U.S. In addition, it is likely that for today’s young people, in a situation of relatively increasing schooling, low-skilled jobs in the United States are now less attractive, to some of them at the very least.

New Migration Flows to the United States

For many years there has been a labor channel that allows the U.S. Department of Labor to employ foreign laborers in agriculture and some other services, in case of a shortage of these laborers in the various U.S. localities. These are temporary labor contracts through H2-A visas for agriculture and H2-B for services. In Mexico, ^{Trigueros (2015}) has been one of the pioneers in collecting data and analyzing aspects of these types of contracts. Unfortunately, until now, very little is known about the experiences of laborers who have arrived in the U.S. with these visas. In the field, we often find laborers who have crossed with these visas, but have only been able to obtain some rather general and anecdotal data from them. The Ministry of Labor and Social Welfare (Secretaría del Trabajo y Previsión Social) in Mexico City also lacks much more information other than that coming to them through some U.S. agricultural entrepreneurs looking for laborers. This according to interviews I had with the official in charge of these matters. Below I present the data on the granting of temporary work visas, obtained through various U.S. government portals; I accessed this information based on my long-time held suspicion of an increase in hiring with temporary work visas, that I was still to seek confirmation for.

Source: ^{U.S. Department of State (n.d.)}.

Graph 2 Granted visas for temporary workers 

Both H2-A and H2-B visa workers and professionals are referred to as temporary workers. They are people who applied for a temporary work visa, processed it and paid for it, and were granted it. I clarify this because there are other statistics that refer to admissions that are “events”, and this is not the case.

The increase in “granted visas” from 2010 to 2018 has been 173.7% for workers with H2- A and H2-B visas. As can be seen, 2014 has been the year in which the growth of these workers experienced a great rise, as some of us studying the topic had begun to suspect. For their part, professionals had a 167.5% growth in the same period. Now these data show that there are also many thousands of Mexican professionals who have been obtaining “temporary work visas” each year. Together, both types of temporary workers together reached 294 628 people in the last year, a figure that although lower than the annual averages of other times, remains quite large and, most importantly, these are legal annual flows; this situation has come to change the way migration had traditionally been undertaken before; also, the participation in this flow of so many professionals speaks to us of other types of circumstances under which labor migration from Mexico is taking place, that we will have to keep in mind in the future.