96 10 3176 https://dh.scu.edu/exhibits/files/original/91/2916/2013_207_1_001-copy.jpg 233e1825fbfc93d405ebbe92d3591924 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Women's History 2022 Subject The topic of the resource Women's History Contributor An entity responsible for making contributions to the resource Ally Green, Kelsey Parker, Paisley Hou Creator An entity primarily responsible for making the resource Green, Ally Hou, Paisley Parker, Kelsey Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Sojourner Truth Description An account of the resource Cabinet card of Sojourner Truth Date A point or period of time associated with an event in the lifecycle of the resource 1864 Publisher An entity responsible for making the resource available Library of Congress Source A related resource from which the described resource is derived Collection of the National Museum of African American History and Culture https://dh.scu.edu/exhibits/files/original/2923/376B8FB4-BEE1-E366-27DB9B52EE980164.jpg 834266933310d761268904e0cf4da0fd Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Lucretia Mott Description An account of the resource Half-length & seated portrait of Lucretia Mott Subject The topic of the resource Lucretia Mott-Woman Abolitionist Publisher An entity responsible for making the resource available Library of Congress Date A point or period of time associated with an event in the lifecycle of the resource 1850 Format The file format, physical medium, or dimensions of the resource Photograph Source A related resource from which the described resource is derived Chicago Historical Society https://dh.scu.edu/exhibits/files/original/91/2924/10.2307_community.12248656-1.jpg 5516649ab657d23861620af230e1349f Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Women's History 2022 Subject The topic of the resource Women's History Contributor An entity responsible for making contributions to the resource Ally Green, Kelsey Parker, Paisley Hou Creator An entity primarily responsible for making the resource Green, Ally Hou, Paisley Parker, Kelsey Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Helen O. Storrow standing in front of a tent at the Girl Scout training school. Creator An entity primarily responsible for making the resource unknown Subject The topic of the resource Storrow, Helen Osborne, 1864-1944 Girl Scouts of the United States of America Camp sites, facilities, etc. Clothing and dress Flags--United States Portraits Tents Uniforms Women women women's studies American photo archives women's rights feminism history Description An account of the resource Inscription: Verso: Mrs. Storrow Commandant at Winsor School 1917. [in a different hand:] W-127, wife of Storrow, James Jackson, 1864-1926 Date A point or period of time associated with an event in the lifecycle of the resource 1917 Type The nature or genre of the resource Image Format The file format, physical medium, or dimensions of the resource Medium: non-projected black and white photograph Dimensions: 3.875x5.125 in. Source A related resource from which the described resource is derived Photograph Number: MC263-143-1, Arthur and Elizabeth Schlesinger Library on the History of Women in America (Radcliffe Institute for Advanced Study, Harvard University), Data From: The Schlesinger History of Women in America Collection, Folder Number: 143, Folder Title: Photographs: Storrow continued'1 photograph of 3 men by camp truck; 1 unidentified troop in Lincoln;1 photograph with unidentified others; 1 photograph with husband, James J. Storrow, and 1 cropped dulpicate; 1 photograph with Lincoln Golden Eaglets, unidentified woman, and flag; 1 photograph of HOS with movie camera at Cedar Hill Junior Camp; 1 photograph of statuette of HOS made by niece; 3 photographsHOS with canoe, 1 photograph HOS on dock, 1 photograph unidentified woman with HOS canoe'all originally filed together and labeled "Helen Storrow and Canoe at Pine Tree Camp", Collection Number: MC263, Collection Name: Massachusetts Girl Scouts, Collection Title: Records, 1915-1967 (inclusive) Rights Information about rights held in and over the resource This image has been made available by the Schlesinger Library, Radcliffe Institute, Harvard University solely for noncommercial educational and scholarly purposes. Your use of this image is restricted to those permitted uses specified in the ARTstor Digital Library Terms and Conditions of Use (http://www.artstor.org/info/about/terms_conditions.jsp). To request permission for any other use, please contact the Schlesinger Library at slref@radcliffe.edu. https://dh.scu.edu/exhibits/files/original/91/2925/10.2307_community.12232574-1.jpg 71f3aab4578e23b5665c732ca27f1f56 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Women's History 2022 Subject The topic of the resource Women's History Contributor An entity responsible for making contributions to the resource Ally Green, Kelsey Parker, Paisley Hou Creator An entity primarily responsible for making the resource Green, Ally Hou, Paisley Parker, Kelsey Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Helen O. Storrow looking at her "Silver Fish" award, given to her by the Girl Guides. Creator An entity primarily responsible for making the resource unknown Subject The topic of the resource Massachusetts Girl Scouts Girl Guides Girl Scouts of the United States of America Awards Portraits Women women women's studies American photo archives women's rights feminism history Date A point or period of time associated with an event in the lifecycle of the resource 1921 Type The nature or genre of the resource Image Format The file format, physical medium, or dimensions of the resource Medium: non-projected black and white photograph Dimensions: 8.125x10 in. Source A related resource from which the described resource is derived Photograph Number: MC263-147-1, Arthur and Elizabeth Schlesinger Library on the History of Women in America (Radcliffe Institute for Advanced Study, Harvard University), Data From: The Schlesinger History of Women in America Collection, Folder Number: 147, Fo Rights Information about rights held in and over the resource This image has been made available by the Schlesinger Library, Radcliffe Institute, Harvard University solely for noncommercial educational and scholarly purposes. Your use of this image is restricted to those permitted uses specified in the ARTstor Digital Library Terms and Conditions of Use (http://www.artstor.org/info/about/terms_conditions.jsp). To request permission for any other use, please contact the Schlesinger Library at slref@radcliffe.edu. https://dh.scu.edu/exhibits/files/original/86/2926/Nick_pellerano.jpg 863c987a9e04fdc7b5e28878aa5f2d3b Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Santa Clara College Course Catalog 1890-1891 Subject The topic of the resource Nicholas Pellerano in the 1890-1891 student list https://dh.scu.edu/exhibits/files/original/86/2927/college_58_full_[1].jpg 22aa78047ebf2b55aa9959270b23c728 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Business Classroom in Commercial Building, 1890 Creator An entity primarily responsible for making the resource Hill, Andrew Putnam, 1853-1922 Subject The topic of the resource Commercial Building Publisher An entity responsible for making the resource available Santa Clara University Library, Archives & Special Collections Date A point or period of time associated with an event in the lifecycle of the resource 1905-03-04 https://dh.scu.edu/exhibits/files/original/86/2929/classroom_pellerano.jpg 2f3a97a12271fe557ed45b430c0dfb36 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Business Classroom in Commercial Building https://dh.scu.edu/exhibits/files/original/86/2933/Vari_Newspaper.jpg ef048e9fcc369d14681d44ccb085c458 Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Text Any textual data included in the document "[...] Vari definitely is a man interested in promoting Italian culture and the University of Santa Clara. Recently, the professor of Modern Languages was honored for his work in promoting Italian Culture." Original Format The type of object, such as painting, sculpture, paper, photo, and additional data The Santa Clara Newspaper Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Victor Vari Spotlight in The Santa Clara https://dh.scu.edu/exhibits/files/original/90/2935/Mexican_Migration_to_the_United_States_A_Critical_Review.pdf e958007cbc35149ef39b31edfc5966d9 PDF Text Text Mexican Migration to the United States: A Critical Review Author(s): Jorge Durand and Douglas S. Massey Source: Latin American Research Review , 1992, Vol. 27, No. 2 (1992), pp. 3-42 Published by: The Latin American Studies Association Stable URL: https://www.jstor.org/stable/2503748 JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms The Latin American Studies Association is collaborating with JSTOR to digitize, preserve and extend access to Latin American Research Review This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE UNITED STATES: A Critical Review Jorge Durand, Universidad de Guadalajara Douglas S. Massey, University of Chicago Although social scientists usually do not speak in terms of laws, they believe they are at least able to make valid empirical generalizations. In studies of Mexican migration to the United States, for example, generalizations drawn from the research literature abound. Thus the problem is not that generalizations are lacking but that they are frequently inconsis- tent and contradictory. Often such inferences are simply invalid because they are based less on evidence than on the investigator's own preconcep- tions. As a result, the field of Mexican migration studies has been plagued by a fragmented debate that seems to go on and on without resolution. The intractability of this debate has been magnified by the inher- ently political nature of migration between two countries with very different cultures, histories, and levels of wealth and also by the fact that much of the movement is clandestine and therefore unobservable. Adding to these problems are the inevitably divergent interests, perceptions, and languages of researchers from north and south of the border, a combination that makes the lack of a clear consensus hardly surprising. We believe nevertheless that the inconsistency of research findings on Mexican migration to the United States is more apparent than real. In an effort to bring some order to this disparate body of research findings, we have undertaken a critical review and synthesis of research carried out in Mexico and the United States. We will begin by examining two particu- larly salient issues-the number of Mexican migrants to the United States and the quantity of their monetary remittances to Mexico-and will then suggest that once rhetoric is separated from fact and analysis from opinion, the various estimates are actually relatively consistent. We will also examine the growing number of case studies of Mexican communities that send migrants to the United States. These investi- gations have yielded apparently contradictory generalizations on a vari- ety of topics: the class composition of U.S. migration, the legal and demographic background of U.S. migrants, the economic effects of emi3 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review gration on the community, the effects of Mexican agrarian reform and agricultural modernization on U.S. migration, and the prevalence of dif- ferent strategies of migration. Most of the generalizations suffer the basic limitation of being based on isolated case studies. We will argue that such apparently inconsistent generalizations about Mexico-U. S. migration are not necessarily contradictory when they are examined in comparative perspective. Rather, diverse outcomes occur in various communities when common processes of migration are shaped and differentiated by structural variables operating at the community level. Because these structural variables exert their influence between communities rather than within them, such variables cannot be studied effectively through individual case studies. They must instead be observed and analyzed comparatively across communities. We have therefore undertaken a systematic and critical review of all the community studies that we uncovered in a survey of the research literature. Our goal is to identify the community-level factors that are most important in determining the expression of basic migratory outcomes. Our review suggests that only a few community factors account for the diversity of conclusions in different case studies: the age of the migration stream; the degree to which productive resources are equitably distributed; the quality of local resources, especially land; the niche in the U.S. industrial structure where the community's migrants first became established; and the geographic, political, and economic position of the com- munity within Mexico. In addition, our review suggests that life-cycle factors-particularly the relative numbers of workers and dependentsexplain much of the heterogeneity in outcomes at the household level. In undertaking this critical review, we are not recommending that community studies should be avoided-quite the contrary. We neverthe- less suggest that researchers should be sensitive to the peculiarities of the communities they are studying, particularly to the influence of the community-level factors identified above, before generalizing from a single case study. Ultimately, however, we argue for a research design that would incorporate the study of many different communities into a common analytic framework. Such a design would be capable of analyzing simultaneously the operation of variables at three levels-the individual, the household, and the community. It would also permit analyzing how community social and economic structures affect migration and how migration affects the social and economic organization of communities. THE NUMBERS GAME: HOW MANY MIGRANTS The debate over how many persons migrate between Mexico and the United States is obviously a key issue, but figures advanced by investigators on opposite sides of the border have always differed sub4 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. stantially. As early as 1930, Manuel Gamio noted the profound disparity between estimates prepared in Mexico and the United States. He sought to rebut the view widely held in the United States that more than a million Mexicans were living in the United States by 1929. He argued that it was impossible to determine the number of Mexican immigrants from the U. S. Census because it was then taken in July, when the number of Mexican migrants was at its peak. Because Mexican migration is highly seasonal, the number of Mexicans enumerated would have been substantially lower if the census had been taken in December (Gamio 1930a; 1930b). Gamio showed that monthly data on remittances declined sharply during the winter months, as the number of U.S. exit forms filed by Mexicans rose (Gamio 1930b). Gamio's contemporary in the United States reached a different conclusion, however. Paul Taylor argued that Mexican migration was largely "permanent" and that the winter decline in remittances was explained by the fact that there was less work during this season and migrants did not have the extra income to send home (1929, 237). Taylor examined national census figures, school statistics, and mar- riage registrations and also undertook special surveys in source and destination areas (P. Taylor 1930, 302-8; Taylor 1931, 48-56; Taylor 1933). This early "numbers debate" was never resolved. It was still being argued when the Great Depression caused the mass deportation of Mexicans from the United States. This event relegated the counting of immigrants to the background, as scholars began to worry instead about the counting of deportees, setting off yet another debate. According to Mercedes Carreras, 311,717 Mexicans were deported from the United States between 1930 and 1933 (Carreras 1974, 145). Ralph Guzman, however, believes that the figure was at least half a million, a number he considers conservative (Guzmain 1979). Abraham Hoffman put the figure at 319,673 between 1930 and 1933 and at 458,039 for the period from 1929 to 1937 (Hoffman 1974). The deportations also called attention to the large number of women migrants. Carreras (1974) reported that among Mexicans deported between 1931 and 1933, some two-thirds were women, a finding that questioned the prevailing stereotype of Mexican migrants as single males traveling alone. After a long hiatus during the depression years of the 1930s, Mexico-U.S. migration was revived in 1942 by the Bracero Program, a temporary worker program initiated by the U.S. government. Although it was begun as a temporary wartime measure, the program was successively extended by the U.S. Congress and did not end until 1964 (see Craig 1971). In contrast to the disputed estimates of the number of immigrants and deportees, widespread agreement exists as to the number of braceros. Because each bracero contract generated its own administrative record, official statistics are generally considered to be reliable by Mexicans as well as U.S. researchers. During the twenty-two years that the 5 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review program lasted, some 4.6 million braceros entered the United States (Morales 1981, 148). During the Bracero Program, however, another international flow arose that was much harder to measure and far more intractable as a subject of study: undocumented migration. From 1942 to 1964, official statistics indicate, nearly 5 million Mexicans were apprehended and deported from the United States, a fact widely reported on both sides of the border (see Samora 1971; Cornelius 1978; Morales 1981). Since 1964 an additional 17 million Mexicans have been arrested and expelled from the United State (U.S. INS 1989). But although observers may agree on the number of apprehensions, they disagree profoundly about the relationship between this figure and the number of illegal migrants. Because undocumented migration is by definition clandestine, it cannot be measured directly, a situation that has forced researchers to rely on indirect measures. The resulting estimates can be characterized as either analytic or speculative (Passel 1986). Analytic estimates are based on quantitative procedures whose assumptions can be scrutinized and whose biases can be systematically evaluated, whereas speculative estimates are based on little more than opinion or conjecture. In general, speculative procedures yield estimates that are much larger than those developed by using analytic methods, and they have therefore received far more attention from the U. S. press and from U. S. politicians. Speculative estimates generally put the number of undocumented Mexicans at 5 million or more. For example, Lesko Associates (1975) as- sembled a panel of 'experts" to guess the number of undocumented Mexicans in the United States in 1975 and averaged panel members' guesses to derive an estimate of 5.2 million undocumented Mexicans. In 1976 the Commissioner of the U.S. Immigration and Naturalization Service (INS), Leonard Chapman, estimated that 10 percent of Mexico's entire population (about 6 million persons) were living in the United States illegally (Chapman 1976). He arrived at this number by adding up "lestimates" provided by INS district offices. Arthur Corwin put the figure at 7 million in 1981 (Corwin 1982), and Edwin Reubens estimated the total net flow of undocumented migrants to be 600,000 in 1978 (Reubens 1980). None of these figures were derived from analytic procedures, however, and they have been widely criticized on both sides of the border (Bustamante 1979; Cornelius 1979; Massey 1981; Passel 1986). These estimates are significantly larger than the ones derived by U. S. demogra- phers using quantitative methods. For example, Frank Bean, Allan King, and Jeffrey Passel estimated that the maximum number of Mexicans that could possibly be in the United States in 1980 was 3.8 million, and most of their estimates ranged from 1.5 to 2.5 million (Bean, King, and Passel 1983). Passel and his colleagues have demonstrated that 1.1 million undocumented Mexicans were counted in the 1980 U.S. Census taken on 6 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. 1 April (Passel and Woodrow 1984; Warren and Passel 1987). Subsequent research has suggested that only one-half to two-thirds of all undocumented Mexicans were enumerated in the 1980 census (Heer and Passel 1985), yielding a range of 1.7 to 2.2 million for the total undocumented Mexican population in the United States in April 1980. Passel's reanalysis of final data from the 1980 Mexican Census yielded a best estimate for 1980 of 1.9 million, far fewer than the 5 to 7 million put forth by more speculative observers (Passel 1985). Analytic methods have also yielded estimates of net flow that are well below the speculative figures of 500,000 to 600,000 per month. For example, Jeffrey Passel and Karen Woodrow estimate that during the early 1980s, the net flow from Mexico was about 200,000 persons per year, a rate that would yield a total U.S. population of 3.1 million undocumented Mexicans by 1986 (Passel and Woodrow 1987). An opportunity to confirm this number independently occurred with the passage of the Immigration Reform and Control Act (IRCA) in 1986. It offered legalization to all undocumented migrants who had entered the United States before 1 January 1982 and a special amnesty for those who had worked for ninety days in agriculture during the year preceding 1 May 1986. In all, some 1.2 million Mexicans applied for the general amnesty, and another 1.1 million sought the special amnesty (Bean, Vernez, and Keely 1989). The total of 2.3 million applicants represents 75 percent of the estimated total popula- tion of undocumented Mexicans in 1986, suggesting that Passel's analytic estimates were close to the mark and that more speculative estimates were far too large. Undocumented migration has also become a salient issue in Mexico, and Mexican investigators have made their own attempts to measure the phenomenon. As in the United States, movement across the international border is not captured well by official statistics (Garcia y Griego 1987), and Mexican investigators, like their U.S. counterparts, have had to rely on indirect methods of estimation. Juan Diez Canedo, for example, tabulated the total number of dollars sent to persons with Spanish surnames in Mexico via checks or money orders drawn on U. S. banks (Diez Canedo 1984). By assuming an average size of remittances (taken from North and Houstoun 1976), Diez Canedo estimated the number of undocumented Mexicans in the United States to be 481,000 in 1975. Other investigators have employed the Mexican Census in attempting to assess the magnitude of undocumented migration. By using data from a census question about foreign residence during the past year and making assumptions about the sex ratio among undocumented migrants, Rodolfo Corona derived a statistical model that permitted estimating the number of migrants who had returned from the United States (Corona 1987). He calculated that 64,000 to 83,000 undocumented Mexicans had lived in the United States for at least six months by June 1979 but had 7 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review returned to Mexico by May 1980. The rather narrow target population limits the utility of these estimates, however. Gustavo Lopez and Sergio Zendejas have concluded that the 1980 Mexican Census offers little basis for accurate estimation (Lopez and Zendejas 1988). Jorge Bustamante approached the problem from a different angle in his 1990 study. Three times a day, he took a photograph of undocumented migrants gathered at two staging points on the international border near Tijuana. He then counted the number of migrants captured on film and computed a daily total of gross undocumented migration that was subsequently tabulated by month. This approach, however, cannot measure the extent of return migration nor can it capture the flow along the entire border, limitations that undermine its utility in specifying the size or rate of change in the undocumented Mexican population. The best source of Mexican data on undocumented migration is a survey carried out during 1978-79 by CENIET (Centro Nacional de Informacion y Estadisticas del Trabajo), a branch of Mexico's Secretaria de Trabajo y Prevision Social. Called the Encuesta Nacional de Emigracion a la Frontera Norte y a los Estados Unidos (ENEFNEU), the survey was commissioned to obtain "basic information that would be representative of the phenomenon in its entirety" (Bustamante 1979, 30). This survey administered a short questionnaire to a stratified prob- ability sample of 62,500 households in 115 localities during December 1978 and January 1979. The questionnaire asked about absent household members age fifteen and over who were working in the United States and about persons age fifteen and over who had worked there during the past five years but had returned to Mexico. The results indicated that 519,300 Mexicans were working in the United States as of January 1979 and another 471,400 had worked there between January 1974 and January 1979 but had returned before the survey was taken. Some 91 percent of the returned migrants reported that they had entered the United States without documents, and the combined population of absentee and returned migrants was 84 percent male (CENIET 1982; Garcia y Griego 1987). These percentages suggest that 472,000 undocumented Mexican workers were residing in the United States in 1978, of whom 443,700 were male. Although these figures appear to conflict with the 1.1 million undocumented Mexicans counted in the 1980 U.S. Census, they are not necessarily contradictory because the ENEFNEU sample excluded two important segments of the undocumented population covered by the U.S. Census estimates: persons under the age of fifteen and adult nonworkers, the latter category being composed primarily of married women. Robert Warren and Jeffrey Passel (1987) estimate that only 44 per- cent of the undocumented Mexicans counted in the U.S. Census were males over the age of fifteen, but the ENEFNEU survey yielded an un8 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. documented population that is 94 percent male, all over the age of fifteen. Moreover, the Mexican survey was carried out in December and January, when the undocumented population is at its minimum, whereas the U. S. Census took place in April, when seasonal migrants begin to return to the United States. Finally, the Mexican survey provided estimates for 1978, whereas the U. S. Census was carried out in 1980. Thus the ENEFNEU survey estimates the number of undocumented Mexican males residing in the United States as of January 1978, while the U.S. Census counts the entire undocumented Mexican population in the United States during April 1980. When these basic differences are taken into account, the two estimates turn out to be fairly close. The ENEFNEU survey indicates that there were 444,000 undocumented Mexican males age fifteen or over living in the United States in 1978, whereas Warren and Passel (1987) estimate that 484,000 such men were counted in the U.S. Census in 1980 (.44 times 1.1 million). Adjusting for underenumeration yields a total estimate of 836,000 undocumented Mexican men age fifteen and over in 1980 (.44 times 1.9 million). Depending on one's assumptions about net undocumented migration between 1978 and 1980, the two estimates either agree closely or differ modestly, but they are not wildly contradictory. If net undocumented Mexican male migration is assumed to be about 200,000 men per year, then the CENIET figure is slightly larger than the U. S. Census figure (844,000 versus 836,000). If net male migration was only 100,000 between 1978 and 1980, then the CENIET estimate is somewhat smaller (644,000 versus 836,000). Manuel Garcia y Griego and Francisco Giner de los Rios updated the CENIET estimates and suggest that by April 1984 there were 758,000 undocumented workers in the United States (1985, 230), a figure again within the range of estimates developed by Passel and his colleagues (assuming that "undocumented workers" refers mainly to men over the age of fifteen). Thus analytic estimates prepared in Mexico and the United States do not contradict one another and are consistently smaller than the speculative estimates put forth in the U.S. media. The U.S. estimates have one weakness not shared by the ENEFNEU data, however. Although Passel and his colleagues could estimate the composition of the undocumented population by age and sex, their indirect methodology did not permit analyzing other migrant characteristics. For this purpose, the CENIET data are considerably more useful because that survey asked returned U.S. migrants a detailed set of questions that enable describing them along a variety of dimensions. Several studies have taken advantage of these data to describe various aspects: the characteristics of undocumented workers (Ranney and Kossoudji 1983), the traits of undocumented women (Kossoudji and Ranney 1984), the process of border crossing (Kossoudji 1990), and the 9 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review process of labor-force adjustment (Kossoudji and Ranney 1986). Lamentably, however, data from the ENEFNEU survey have not been analyzed in depth within Mexico (see the final report, CENIET 1982). Apparently, a change in presidential administrations impeded the project's continuation within the Secretaria de Trabajo y Prevision Social. Yet ten years later, the survey remains the most reliable Mexican source on undocumented migration to the United States (Garcia y Griego 1988, 134). Despite the progress achieved in estimating the number of undocumented migrants, the "numbers game" has once again become a public issue as a result of the 1986 Immigration Reform and Control Act (IRCA). The new debate is focusing on whether IRCA's enforcement provisions have reduced the net flow of undocumented Mexican migrants to the United States. On 15 January 1988, the New York Times published a report by INS Director Alan Nelson, who stated that the number of undocumented entrants had been lowered as a direct consequence of IRCA.1 Three days later, Jorge Bustamante questioned this conclusion in Excelsior because it was based on INS apprehensions during December, when the flow of migrants traditionally diminishes.2 A subsequent analysis by Bean et al. (1990) attempted to control for the seasonality of undocumented migration, economic conditions in Mexico and the United States, and the ongoing legalization process. This study concluded that IRCA had reduced the flow of undocumented migrants, but it has been contradicted by other researchers' data indicating that IRCA has had little or no effect (Cornelius 1989; Massey, Donato, and Liang 1990; Bustamante 1990). Thus the debate on the number of Mexican migrants has continued in one form or another for fifty years, often using the arguments and counterarguments first broached by Manuel Gamio and Paul Taylor in the 1930s. The debate reveals the limitations of social science in studying a subject that is by definition unobservable, and it also underscores the political nature of a process linking two countries with different interests, perceptions, and histories. Indeed, the debate frequently has had little to do with facts. As Frank Bean, Edward Telles, and Lindsay Lowell have noted, "the heated discussions during the [U.S.] Congressional debates of the last few years on immigration legislation have shown that the exagger- ated figures are often still taken seriously" (Bean, Telles, and Lowell 1987, 674). Likewise, voting by U.S. Congressional representatives on immigration issues is determined more by party politics than by a rational assess- ment of interests and effects (Lowell, Bean, and de la Garza 1986). Thus the 'numbers game" often resembles a dialogue among the deaf rather than a reasoned debate about facts. 1. "Administration Calls Law on Aliens Effective," New York Times, 15 Jan. 1988, p. A-32. 2. Jorge Bustamante, "Manipulacion de la Migra," Excelsior, 18 Jan. 1988, pp. A6-7 10 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. "MIGRADOLARS: REMITTANCES TO MEXICO Since Mexican migration to the United States began, investigators have attempted to measure the size and effect of migrants' remittances. At the turn of the twentieth century, money orders coming from the United States attracted the attention of Mexican journalists, who reported in various Mexican newspapers the arrival of large quantities of dollars in rural communities (see Durand 1986, 1988). The first systematic study of remittances was carried out by Manuel Gamio, who compiled information on the origins, destinations, and size of postal money orders sent to Mexico from the United States (Gamio 1930a). These data provided Gamio with a reliable indicator of the main destination areas in the United States and the major Mexican sending communities. Gamio also attempted to infer the number of migrants from the size and distribution of remittances, an undertaking requiring a number of tenuous assumptions that were strongly criticized by his contemporary, Enrique Santibaniez. The latter stated that "we do not doubt the existence of these postal money orders; but it is another thing to attribute them to the savings of Mexican workers, when they could have had other origins" (Santibaniez 1930, 98). Paul Taylor also studied remittances but focused his attention locally, on the town of Arandas in Los Altos, a region of Jalisco, where he counted 7,678 remittances by telegraph or postal money order between 1922 and 1931 (P. Taylor 1933). From the number and denomination of the checks, he computed the total inflow of dollars into the community and used this figure to quantify the economic effect of international migration. This estimate had one weakness, however: Taylor could not determine the number or size of remittances that arrived in private letters. He simply assumed that the remittances he counted represented the total arriving in the forms of telegraphs and money orders. As the dispute between Gamio and Taylor demonstrates, differences concerning the size of remittances frequently follow national lines. In general, Mexican researchers have tended toward sins of omission and U.S. investigators toward sins of commission. Reports from Mexico have systematically understated the quantity of U.S. remittances, and the Mexican government frequently ignores the subject entirely (apparently the current policy). This official reticence has not always been the case, however. In their final reports, Presidents Ruiz Cortines and Miguel Aleman both offered specific estimates of the foreign exchange contrib- uted by migrants (Morales 1981). In the United States, meanwhile, researchers have tended to exaggerate remittances by multiplying a large number of undocumented migrants (often derived by extrapolating ap- prehension figures) by an average remittance derived from some field study (which may or may not be representative). 11 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 0001 Jan 1976 12:34:56 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review A recent example of the ongoing debate is that between Wayne Cornelius and his former student Juan Diez Canedo. Cornelius (1978) estimated that remittances to Mexico totaled more than 3 billion dollars in 1975, but Diez Canedo (1984) put the figure at only 317 million in the same year. Cornelius computed an average size of remittances based on his survey of migrants in Jalisco and then multiplied it by the number of Mexicans he assumed were working seasonally in the United States. Diez Canedo, in contrast, totaled checks or bank orders made out to persons with Spanish surnames in Mexico drawn on U.S. banks. Both estimates are open to criticism: that of Cornelius was made essentially by assumption whereas Diez Canedo's ignored remittances in the form of cash and goods. A more careful but heavily qualified estimate was prepared by Garcia y Griego and Giner de los Rios using data from both countries and information from the CENIET survey (1985, 236). They proposed a figure of 1.8 billion dollars in remittances for 1984. The analysis was replicated in a later article (Garcia y Griego 1988), which stated that "even though the estimates are not exact," they provide 'adequate" information to counter alarmist claims of millions of undocumented migrants entering the United States and billions of dollars leaving the country. Charles Keely and Bao Nga Tran used data tabulated by the International Monetary Fund on "unrequited private transfers" (IMF 1987) to examine trends in remittances for various countries from 1970 to 1985 (Keely and Tran 1989). They argue that these figures understate the funds remitted by migrant workers because they include only money trans- ferred through monitored channels and do not incorporate noncash remittances. The IMF data show that in 1975, 59 million dollars were privately transferred from the United States to Mexico, far less than the amounts reported earlier by Cornelius or Diez Canedo. But the amounts increased substantially during the late 1970s and again after the onset of the economic crisis in 1982. By 1984, private transfers totaled 2.3 billion dollars, exceeding the estimate of 1.8 billion put forth by Garcia y Griego and Giner de los Rios (1985). It is thus clear that migrant remittances represent a significant source of foreign exchange for Mexico, some 2 billion dollars in 1984. COMMUNITY STUDIES: FROM THE SPECIFIC TO THE GENERAL Given the problems encountered in estimating the number of migrants and the size of their remittances, researchers have turned to studying particular Mexican communities as a means of analyzing the social and economic processes that underlie the aggregate statistics. These basic processes are difficult to analyze using large-scale surveys for several reasons. First, processes of migration are developmental and longitu12 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. dinal, and surveys taken at a single point in time cannot capture this inherent dynamism. Second, migration is more than an individual or family decision: it is also a response to larger structural conditions in sending and receiving societies, factors that cannot be studied easily by using surveys of individuals or households. Finally, migrants and their families inevitably make decisions within a local setting. National, state, and regional conditions are important, but their effects are mediated through local social and economic institutions that are difficult to study by using survey data alone. Community studies provide a tool for analyzing the migration process in a way that is historical, developmental, and sensitive to the effects of local conditions as well as to those of the national political economy. Community studies suffer from one serious problem, however: it is difficult to generalize about broader patterns and processes of migration from a single isolated case. Yet this inherent weakness has not stopped most researchers, who find the temptation to generalize too great to resist. Thus the field is replete with general statements about the nature of Mexico-U. S. migration based on the experience of a single community. For example, Joshua Reichert considered his case of Guadalupe, Michoacan, sufficient to infer a "migrant syndrome" characteristic of the entire Mexican central plateau (1981). Similarly, although Cornelius points out that he refers only to his own study (1979), he frequently contrasts his findings with those of other investigators in order to make broader statements. Yet Cornelius does not provide the information needed to know whether the comparisons are appropriate: his sampling methodology, which communities he studied, or the overall sample size. He states only that the survey was carried out in Los Altos, Jalisco. At this juncture, a variety of community studies have been com- pleted and their findings introduced into the research literature. Table 1 lists seventeen case studies carried out within specific Mexican communities, and table 2 lists eight other comparative studies that have con- trasted several communities at once, yielding information on twenty additional settings. Together these studies provide data of varying detail and quality on thirty-seven separate Mexican migrant communities. The authors of these studies have put forth a host of generalizations about the nature of Mexico-U.S. migration. But most investigators have made few attempts to consider the peculiarities of their settings or to situate their studies within the broader research literature. As a result, inconsistent conclusions have proliferated on a variety of topics: which social classes migrate, the demographic and legal composition of the flow, the economic effects of emigration, the effects of Mexico's agrarian reform and agricultural modernization, and the relative importance of different strategies of migration. In the ensuing sections of this article, we will review findings on these issues and will argue that discrepancies between 13 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review TA B L E 1 Community Studies of Mexican Migration to the United States, 1929-1990 Community State Investigator Acuitzio del Canje Michoacan Wiest (1973, 1979, 1984) Aguililla Michoacan Rouse (1989) Arrandas Jalisco P. Taylor (1929, 1931, 1932, 1933) Alvaro Obregon Michoacan Trigueros and Rodriguez (1988) Copandaro Michoacan Rionda (1986, 1988) Gomez Farias Michoacan Lopez (1986a, 1986b, 1988) Guadalajara Jalisco Escobar, Gonzalez, and Roberts (1987), Escobar and Martinez (1990) Guadalupe Michoacan Reichert (1979, 1981, 1982), Reichert and Massey (1979, 1980) Jalostotitlan Jalisco Gonzalez and Escobar (1990) Jaripo Michoacan Fonseca (1988), Fonseca and Moreno (1988) Las Animas Zacatecas Mines (1981, 1984), Mines and de Janvry (1982) Los Altos villages Jalisco Cornelius (1976a, 1976b, 1978) Patzcuaro villages Michoacan Taylor (1986, 1987), Adelman, Taylor, and Vogel (1988), Stark and Taylor (1989) San Jeronimo Oaxaca Stuart and Kearney (1981) San Francisco del Rincon Guanajuato Durand (1989b, 1990) Santa Ines (Municipio Michoacan Fernandez (1988) de Villamar) Villa Guerrero Jalisco Shadow (1979) studies occur frequently because investigators fail to consider the impact of basic structural factors operating at the community level. Which Classes Migrate One of the most widely studied aspects of Mexican migration to the United States is the socioeconomic selectivity of migration-that is, the question of which social classes migrate. Various studies have reported on this issue, and the prevailing wisdom is that migrants come from the lower-middle segments of the income or wealth distribution (see Portes and Rumbaut 1990, 10-11). The rationale for this observation is that the rich have little incentive to migrate whereas the poor lack the resources to cover the costs and risks of a trip to the United States. For example, Gustavo Lopez concluded, based on his study of Gomez Farias, Michoacan, that "Imigration is most common among those owning irrigated land, in comparison with those who possess rain-fed fields; and among those in the first category, the more land they own, the 14 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. TA B L E 2 Comparative Community Studies of Mexican Migration to the United States Community State Investigator Jaral del Progreso Guanajuato Grindle (1988) Tepoztlan Morelos Union de San Antonio Jalisco Villamar Michoacan Gomez Farias Michoacan Cornelius (1990a, 1990b) Las Animas Zacatecas Tlacuitapa Jalisco El Bajio Guanajuato K. Roberts (1982, 1984) Las Huastecas San Luis Potosi Mixteca Baja Oaxaca Valsequillo Puebla Altamira Jalisco Massey, Alarc6n, Durand, Chamitlan Michoacan and Gonzailez (1987) Santiago Jalisco San Marcos (Guadalajara) Jalisco Guadalupe Michoacan Mines and Massey (1985) Las Animas Zacatecas Hecorio Michoacan Dinerman (1982) Ihuatzio Michoacan G6mez Farias Michoacan Goldring (1990) Las Animas Zacatecas Mexicali Baja California Selby and Murphy (1984) Mazatlan Sinaloa Tampico Tamaulipas Queretaro Queretaro San Luis Potosi San Luis Potosi Corralillos Jalisco Orozco(1990) El Refugio Jalisco Los Dolores Jalisco more they migrate" (Lopez 1986b, 574). His facts appear to coincide with those of Kenneth Roberts (1982), who found a connection between com- mercialized agriculture and migration, and with those of Richard Mines, who affirmed that migrants were more likely to own land than nonmigrants (Mines 1981). Likewise, Ina Dinerman reports that landowner- ship was more common among migrants in Huecorio than among nonmigrants (1982). Despite these studies confirming the prevailing wisdom, other investigators suggest considerable diversity in the class background of 15 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review U.S. migrants. Cornelius reports that in Los Altos, Jalisco, "those who migrate illegally are among the poorest residents of the community- especially the sons of landless workers and ejidatarios" (Cornelius 1976b, 7). He nevertheless allows that "those at the very bottom of the local income distribution are not likely to migrate because they lack . . . the resources to cover the costs of transportation and the fees charged by the coyotes... ." Likewise, although Raymond Wiest found that the very poorest households were unlikely to support a migrant, he could never- theless "state unequivocally from domestic group histories that nearly all the households involved in migration to the United States had incomes well below the median before sending members to the United States" (Wiest 1973, 197). Other researchers have located U.S. migration even more decisively among the poor and landless. In Guadalupe, Michoacan, Reichert (1979) argues, migration arose among the landless precisely because they had no other option but to migrate. James Stuart and Michael Kearney called migration "the only alternative to starvation" in San Jeronimo, Oaxaca, where 90 percent of the families did not possess enough land to support a family (Stuart and Kearney 1981). In two other rural communities (Altamira, Jalisco, and Chamitlan, Michoacan), U.S. migration was most prevalent among landless day laborers (Massey et al. 1987). Just to complicate matters further, one study reported finding no relationship between either income or wealth and the likelihood of U.S. migration among villagers living near Patzcuaro, Michoacan (E. Taylor 1986). Relatively few studies have examined international migration emanating from urban areas, and their results are also inconsistent. Agustin Escobar and Maria Martinez indicate that U. S. migration is quite common among manual workers in Guadalajara: 17 percent of those in their sample had worked in the United States, but the percentage was higher (up to 25 percent) among workers in small firms, suggesting that emigration was most common among the poorest and least stable segments of the working class (Escobar and Martinez 1990). In contrast, Henry Selby and Arthur Murphy surveyed five Mexican cities and found that migrants consistently came from households that were better off than others (Selby and Murphy 1984). Two other surveys carried out in Guadalajara and Santiago (a nearby factory town) found that U.S. migrants were concentrated among skilled manual and service workers rather than among unskilled laborers (Massey et al. 1987). Thus depending on which study one consults, U.S. migrants are drawn either from the landless or the landed, from the skilled working class or unskilled laborers, from the stable middle class or the poorest segments of society. One study has even suggested that class plays no role at all. It is therefore difficult to make sense of these diverse findings without knowing more about the communities involved because factors 16 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. operating at the community level play a large role in determining which classes migrate and when. Several of the studies listed in table 2 have undertaken a comparative analysis of communities in an effort to determine which factors affect the socioeconomic selectivity of international migration. These studies sug- gest that class background is not a fixed characteristic of migrants that invites simple generalization. Rather, the class composition of migration at any point in time is a variable determined by two key community characteristics: the age of the migration stream and the degree of inequality in the distribution of productive resources. The first factor recognizes that migration is a dynamic, developmental process in which decisions made by migrants at one point affect the course and selectivity of migration in later periods. For example, in their comparative study of four migrant communities, Massey et al. (1987) documented that U. S. migration generally begins in the middle segments of the local occupational or wealth distribution (consistent with the common wisdom) but that over time it becomes progressively less selective and eventually incorporates all segments of the socioeconomic hierarchy Ultimately, the flow becomes dominated by the poorest class of landless campesinos and unskilled workers rather than by the middle class of small property owners or skilled laborers. A similar conclusion was reached by Mines and Massey (1985) in comparing Guadalupe, Michoacan, with Las Animas, Zacatecas. The changing selectivity of migration results from the growth and elaboration of migrant networks, which are composed of ties of kinship, friendship, and paisanaje (shared community origin) between migrants and nonmigrants located in the United States and Mexico. The first migrants who leave for the United States have no social ties to draw on, and for them migration is very costly and risky, especially if they have no legal documents. As Cornelius (1976b) and others have argued, these costs and risks tend to exclude the poorest segments of society. After the first migrants have arrived in the United States, however, the costs of migration are substantially lowered for friends and relatives living in the same community of origin. Because of the nature of kinship and friendship structures, each new migrant creates a set of individuals with social ties to the United States and its labor market. Migrants are invariably linked to nonmigrants through bonds of kinship and friendship, and the latter draw on obligations implicit in these relationships to gain access to employment and assistance at the point of destination, thus reducing their costs substantially. Once the number of network connections in an origin area reaches a critical threshold, migration becomes self-perpetuating in creating the social structure needed to sustain it. Every new migrant reduces the costs of subsequent migration for a set of friends and relatives. Some of those 17 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review left behind are induced to migrate, which further expands the set of persons with ties abroad, which in turn reduces the costs and risks for a new set, causing some of them to migrate, and so on. Over time, migration becomes progressively less class-selective and ultimately a mass movement. Thus depending on when migration began and how developed the networks have become, the class composition of migration from a community may be dominated by the middle class, the poor, the landless, or the landed. That is to say, the effect of migrant networks is quite powerful and ultimately comes to dominate class factors in predicting migration, as has been demonstrated by quantitative studies carried out at community and national levels (Massey 1987; Massey and Garcia EspaAa 1987). The reason that Edward Taylor found no relationship between economic class and migration in his 1986 study was because he controlled for the effect of networks, which dominated all other variables in his statistical models. The age of the migration stream is important because it indicates the maturity of migrant networks but also because widespread migration has the power to transform the class distribution itself. In rural com- munities, money earned through U.S. wage labor can finance acquiring farmland, thus enabling migrant families to move from the class of landless workers to small landowners. If enough families acquire land through migration, it may appear at some point that migration results from landownership. It is therefore crucial to sort out the temporal sequence of migration and class membership, given that U.S. migration is a primary avenue of social mobility in Mexican sending communities. Unfortunately, however, when reporting a positive association between landownership and U.S. migration, many studies do not undertake this task. For example, in the two Michoacan communities of Gomez Farias and Huecorio, Lopez (1986a) and Dinerman (1982) report an apparent correlation between landownership and U. S. migration and conclude that the former causes the latter. It is nonetheless likely that the order of causality is reversed. Studies done in two neighboring Michoacan communities, Guadalupe and Chamitlan, showed that migration often preceded landownership, the likelihood of landownership rose sharply as migrant experience increased, and it was tied directly to U.S. earnings (Reichert 1981, 1982; Massey et al. 1987). The association between migration and landownership is also affected by the degree to which land is available for purchase. This linkage illustrates the effect of a second community-level factor: the structure and distribution of local productive resources. In communities where land is unequally distributed and held in large tracts by a few landowners, the possibilities for acquiring farmland are limited, reducing the likelihood of class mobility through migration (Reichert 1981; Grindle 1988; Gonzalez and Escobar 1990). Moreover, even at the outset of migration, class selec18 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. tivity is strongly shaped by land-tenure arrangements. For example, Reichert's town of Guadalupe was made up entirely of landless families when recruitment for the Bracero Program began in 1942, thereby guaranteeing that all migrants were landless day laborers (Reichert 1979). Luin Goldring has compared Las Animas, Zacatecas, and Gomez Farias, Michoacan, to illustrate the important role played by land tenure in determining the class composition of migrants (Goldring 1990). In Las Animas, land from a nearby hacienda was gradually sold to townspeople beginning at an early date, creating a property distribution dominated by small landowners. In Gomez Farias, however, land remained under the control of the hacienda of Guaracha until the late 1930s. Although the agrarian reform movement eventually provided small plots to a minority of households, they were generally insufficient to support families and were quickly leased to absentee landlords in nearby cities (Cornelius 1990a). Thus when the United States began recruiting migrants during the 1940s, these two Mexican communities sent very different classes of workers, not because processes of migration differed in the two communities but because common processes of recruitment drew on divergent class structures. The Demographic and Legal Composition of Migration Another common topic of generalization concerns the age, sex, and legal composition of Mexican migrants to the United States. Studies carried out during the 1920s revealed that migrants were demographically diverse, incorporating not only single men but women, children, and frequently entire families (Gamio 1930a, 1931; P. Taylor 1932; Carreras 1974). This heterogeneous pattern was effaced by the Great Depression and the mass deportations that followed. When the migration stream began again under the aegis of the Bracero Program, it was renewed principally as a male phenomenon. Very few braceros were female (Hancock 1959; Galarza 1964; Samora 1971). The prevailing view about contemporary patterns was established in the mid-1970s by Cornelius (1976a, 1976b), based on his study of rural communities in Los Altos, Jalisco. He found that U.S. migrants were overwhelmingly males of working age. Although most were single when they migrated for the first time, the migrant population generally was made up mostly of married men who traveled without their wives and children. The vast majority were undocumented. This view of Mexican migration has apparently been confirmed by studies of migrants apprehended by the INS, which reveal a great preponderance of men among Mexicans (Samora 1971; North and Houstoun 1976). This profile persisted remarkably among politicians, researchers, and the public. Thus when CENIET designed its 1978 national survey of 19 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review emigration, it did not ask households about persons in the United States but about workers age fifteen and over (CENIET 1982). The survey was therefore guaranteed to confirm the stereotype of Mexican migrants as males of working age, despite the fact that the undocumented population counted in the 1980 U.S. Census was 45 percent female and 21 percent under the age of fifteen (Warren and Passel 1987). Community studies reveal great diversity in the demographic and legal background of migrants, although enough individuals adhere to the classic stereotype of undocumented male workers to confirm its basis in fact. Field studies carried out in several rural communities (Acuitzio, Michoacan; San Jeronimo, Oaxaca; Huecorio, Michoacan; and Alvaro Obregon, Michoacan) report that virtually all U.S. migrants are males of working age who go without documents (see Wiest 1973; Stuart and Kearney 1981; Dinerman 1982; Trigueros and Rodriguez 1988). Yet other investigations of different communities (Guadalupe, Michoacan; Las Animas, Zacatecas; Altamira, Jalisco; Chamitlan, Michoacan; Santiago, Jalisco; Santa Ines, Michoacan; Tlacuitapa, Jalisco; Jalostotitlan, Jalisco; and Gomez Farias, Michoacan) report substantial participation by women and children and large numbers of legal migrants (see Reichert and Massey 1979; Mines 1981; Massey et al. 1987; Fernandez 1988; Cornelius 1990a; Gonzalez and Escobar 1990; Goldring 1990). This contradiction stems from the effect of two key factors: the age of the community's migration stream (which determines the maturity of the networks and the particular U.S. immigration policies to which migrants were exposed) and the niche in the U.S. occupational structure in which the community's migrants first became established. The operation of these two community-level factors create from a common process of labor migration a multiplicity of demographic and legal compositions. Communities sending large numbers of women and children in- variably have long histories of migration. When migrants are grouped according to the date they left on their first U.S. trip, the participation of women and children typically grows over time (see Reichert and Massey 1980; Massey, Donato, and Liang 1990; Fonseca and Moreno 1988; Gon- zalez and Escobar 1990). The first migrants are invariably men, and for a long time the flow continues to be dominated by males visiting the United States temporarily for limited periods of wage labor. At this stage, migration is dominated by economic motivations and imperatives. Over time, however, social processes come into play. As migrants acquire increasing experience in the United States, gain familiarity with U.S. employers and settings, and expand their network connections to other migrants, the costs and risks of migration progressively drop. Al- though men are at first reluctant to expose their wives and children to the hazards and hardships of migration, the life of a solitary migrant worker eventually becomes difficult to sustain. As the costs and risks drop with 20 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. the expansion of the networks, men increasingly bring their wives and children into the migratory process and the demographic base of migration broadens. The incorporation of women and children often happens rapidly. In Guadalupe, Michoacan, for example, virtually no women or children participated in U.S. migration during the twenty-five years from 1940 to 1965. But in the next fifteen years, the flow came to be dominated by women and children, and in the most recent period (1975-1978), 53 percent of new migrants were women and 67 percent were younger than fifteen (Reichert and Massey 1980). Thus in communities with a long migratory tradition, the average age of migrants tends to fall over time and the proportion of women rises (Massey, Donato, and Liang 1990; Gonzalez and Escobar 1990). The emergence of family migration is facilitated by another process in the development of migrant networks: the formation of branch communities in the United States. As migrants increasingly specialize in U.S. labor to the exclusion of other economic pursuits, the constant shuttling back and forth becomes more difficult to sustain, and some migrants ultimately settle in the United States, typically in urban areas. The settlement of other migrants from the same Mexican town greatly strengthens the networks and focuses migration increasingly on specific destinations, making subsequent settlement by additional townspeople more likely and considerably easier. This process provides a secure, Spanish-speaking environment in which dependents can adjust and adapt to life in the United States. The age of the migration stream also determines one more factor that influences the demographic and legal composition of the flow: the rules and regulations governing the acquisition of legal papers in the United States. In general, U.S. immigration policy has evolved in the direction of making it increasingly difficult for Mexicans to acquire legal status in the United States (Jasso and Rosenzweig 1990). Prior to 1965, Mexican immigration was not restricted numerically, and until 1976, Mexico had no specific quota, being subject only to the overall hemispheric ceiling of 120,000 immigrants. After 1976 all nations in the Western Hemisphere, including Mexico, were placed under a quota of 20,000 immigrants each (exempting immediate relatives of U.S. citizens). Then in 1978, the age at which a child could sponsor the immigration of his or her parents was raised to twenty-one, and hemispheric quotas were abolished in favor of a single worldwide ceiling of 290,000. In 1980 the ceiling was reduced to 270,000. All these changes reduced the number of U.S. immigrant visas available to Mexicans. The increasing restrictiveness of U. S. immigration policy means that communities with older migration streams are more likely to be dominated by legal migrants because applications for visas were made 21 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review under more liberal policies. In general, communities that began sending migrants during the bracero years (1942-1964) are characterized by flows that now contain a high proportion of legal migrants (Reichert 1979; Mines 1981; L6pez 1986a; Massey et al. 1987). When it became clear that the Bracero Program was going to end, U.S. employers began to actively sponsor legalization of their workers under the liberal provisions prevailing in U.S. immigration before 1965 (Reichert 1979; Mines 1982). During the late 1960s and early 1970s, much of the legal migration that occurred consisted of wives and children of former braceros and undocumented migrants who had received their documents in the 1950s and early 1960s. Until the IRCA amnesty was declared in 1986, flows from communities that began to send migrants during the 1970s were domi- nated by undocumented migrants, and the later a community's entry into the migration process, the more its flow was dominated by working-age men. Early investigations indicate, however, that the IRCA legalization program may have far-reaching effects in changing this situation. During a brief period, IRCA legalized the status of more than 2 million formerly undocumented migrants (Bean, Vernez, and Keely 1989). One community study indicates that this amnesty caused a pronounced shift in the legal composition of the flow to the United States (Cornelius 1989), transforming it from a stream dominated by illegals into one composed mainly of legals and rodinos, as those receiving amnesty are popularly called (being beneficiaries of the Simpson-Rodino Law cosponsored by Senator Alan Simpson of Montana and Representative Peter Rodino of New Jersey) (Durand 1989a). Although it is not known how many women and children were legalized under the terms of the amnesty, another field study suggests that rodinos have been overwhelmingly male (Goldring 1990). But even if they are mainly men, the amnesty will inevitably be followed by a wave of subsequent legalizations as wives, children, and other relatives use the family reunification provi- sions of U.S. immigration law to obtain their own documents. Thus the composition of migration to the United States is critically determined by the age of the migration stream. Communities that began sending migrants early now contain large numbers of women and children and a high proportion of legal migrants. Yet even among communities that began participating in U.S. migration at the same time, flows can display substantial differences in demographic and legal composition. For example, although migration began at about the same time in Las Animas, Gomez Farias, Guadalupe, and Tlacuitapa, only 31 percent of Animenos between the ages of fifteen and sixty-four are legal migrants whereas 42 percent of Guadalupenos possess documents (Mines and Massey 1985). Similarly, although 35 percent of all migrants are female in Gomez Farias, only 23 percent of U. S. migrants in Las Animas are women (Goldring 1990). And whereas 57 percent of all women between the ages 22 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. of twenty-five and twenty-nine have migrated to the United States in Gomez Farias, only 38 percent have done so in Tlacuitapa (Cornelius 1990a). These differences among communities with similar histories of U.S. migration suggest that other community-level factors come into play. Goldring suggests, based on her comparative analysis of Las Animas and Gomez Farias (1990), that one of the most important is the location of migrants within the U.S. occupational-industrial structure. The centrality of this factor is frequently overlooked by field researchers, who tend to focus on conditions in Mexico and ignore those in the United States (see, for example, the studies by Wiest 1973, 1979; Stuart and Kearney 1981; Dinerman 1982; E. Taylor 1987, 1988; Trigueros and Rodriguez 1988; Fernandez 1988; Escobar and Martinez 1990). In contrast, the earliest studies of Mexico-U.S. migration clearly recognized the importance of U.S. social and economic conditions and gathered data on both sides of the border (see P. Taylor 1929, 1930; Gamio 1930a, 1930b; Fabila 1929). Santibaniez stated, "To treat with any seriousness the issue of Mexican migration to the United States, it is absolutely necessary to consider it in its bilateral condition, since we Mexicans cannot understand it unless we study it in context and take into account U.S. interests. Likewise, the United States cannot determine anything more than what is convenient to its interests if it does not take account of the circumstances that have fatally determined the human current that leaves Mexico . . ." (Santibaniez 1930, 16). In considering the effect of U.S. employment on the demographic and legal status of migrants, the crucial distinction lies between agri- cultural and urban occupations, with the former leading to legalization and participation by women and children and the latter being associated with undocumented migration by working-age men. To a large degree, this choice is a matter of chance-of being in the right place at the right time. But once a few migrants from a community become established in a particular occupational-industrial niche, they acquire the ability to obtain jobs for friends, relatives, and other townspeople. Over time, therefore, migration from particular communities tends to focus not only on specific geographic destinations but on particular niches in the U.S. occupationalindustrial structure: busboys and dishwashers in a certain restaurant; janitors in a certain building; operatives in a particular factory; workers in a certain car wash; fruit pickers for particular growers. For example, migrants from Santiago, Jalisco, went to a specific neighborhood in Los Angeles because one townsman became the union representative in a lamp factory there. Migrants from Chamitlan, Michoacan, focused on two other California cities where a townsperson had obtained the position of foreman on a ranch and a paisano had became headwaiter in a restaurant (Massey et al. 1987). 23 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review Goldring (1990) compares the cases of Gomez Farias and Las Animas to demonstrate the role played by the U.S. occupational niche in shaping the composition of the migrant streams. Around 90 percent of migrants from Gomez Farias work in the strawberry fields in Watsonville, California, living mainly in one labor camp, because several Gomenios developed early connections with particular growers based in that city (Lopez 1986a; Cornelius 1990a, 1990b; Goldring 1990). In contrast, migrants from Las Animas go to urban destinations because one Animeino owns a construction business in Orange County and several others have acquired stable jobs and settled in South San Francisco and Los Angeles (Mines 1981; Cornelius 1990a, 1990b). As a result of these occupational differences, the migrant stream is primarily legal and family-oriented in Gomez Farias but is diverse and weighted more toward undocumented males in Las Animas (Goldring 1990). For various reasons, agricultural labor is more conducive to a pattern of migration by families than is urban employment. First, farm labor has historically provided a more secure path to legal status, which is then passed on to family members to enable their migration without risk. Over the years, growers have worked hard to ensure the continued flow of Mexican labor, often by arranging for legalization of their farm workers. Growers were responsible for initiating and perpetuating the Bracero Program, which gave migrants legal work permits. When this program ended, growers actively sought permanent resident status for their work- ers. As legal status became increasingly difficult to acquire, growers fought to maintain the "Texas Proviso," which exempted them from prosecution for hiring illegal workers. When growers lost this battle in 1986, they successfully lobbied the U.S. Congress for a "Special Agricultural Worker" amnesty that was made available to farm workers on extremely generous terms. Through the growers' efforts, IRCA also arranged for a "Replenishment Agricultural Worker" program to admit additional workers "to prevent farm labor shortages" (Martin 1990). As a result of these efforts, farm workers have consistently been more able than urban workers to obtain legal status for themselves and their families, making seasonal family migration an attractive and viable strategy The diversity of farm work also provides a range of light and heavy tasks that all family members can undertake to earn income. Moreover, growers often provide temporary housing at very low cost (sometimes even free), lowering the costs of family migration. Thus family migration for U.S. farm labor represents a rational strategy for maximiz- ing family income (Reichert 1979). In virtually all cases where a pattern of "legal shuttle" or "recurrent" migration has emerged, workers are employed primarily in agriculture (Reichert 1979; Mines 1981; Lopez 1986a). In contrast, urban employment discourages, although it does not prevent, legalization and family migration. Urban employers generally do 24 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. not act collectively to ensure their labor supply and have not made intensive efforts to secure legalization for their workers. Rather, they rely on the self-reinforcing nature of network migration to provide them with undocumented workers. In general, therefore, it has been significantly more difficult for urban-based migrants to acquire legal papers. Urban workers never had a Bracero Program, and the general amnesty program was far more restrictive in its criteria of eligibility than the special agricultural worker program. Consequently, many families who have resided for years in urban areas of the United States are ineligible for amnesty (Chavez, Flores, and Lopez-Garza 1990), while farm workers who only recently crossed the border have qualified, often with credentials of dubious validity (Martin 1990). Undocumented status, in turn, discourages family migration because of the hazards and costs associated with crossing the border and living clandestinely in the United States. Aside from such legal problems, urban employment discourages the migration of women and children in other ways. In most entry-level urban jobs, wages are low and city rents are expensive. Housing is rarely, if ever, subsidized or arranged by the employer. Moreover, in urban areas children are less able to work and contribute to family income. Thus the migration of wives and children, rather than maximizing family income, tends to lower net earnings and make it harder to accumulate savings quickly (Massey et al. 1987). In the long run, however, solitary migration is difficult to sustain, and as the length of stay and the number of trips increase, family migration eventually occurs, leading to greater diversity in the composition of migrant streams directed to urban areas. The Economic Effects of Emigration Probably the most widely studied aspect of migration to the United States is its effect on local economic development within Mexico. Community studies are remarkably unanimous in condemning international migration as a palliative that improves the material well-being of particular families without leading to sustained economic growth within migrant communities. Thus Reichert refers to migration as an illness or "syndrome" whereby migration undermines local development. Wiest (1979) calls migration an "addiction" that townspeople must satisfy to fulfill their need for consumer goods and an elevated standard of living. Stuart and Kearney (1981) consider U.S. migration to reflect a "dangerous dependence." The description of emigration and its effects in these ominous terms stems largely from the finding that U.S. earnings are spent overwhelmingly on current consumption, leaving little money for productive investment. Community studies consistently report that U. S. remittances 25 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review and savings are directed to a few nonproductive ends: family maintenance and health; the purchase, construction, or remodeling of homes; and the purchase of consumer goods. For example, in Huecorio, Michoacan, Dinerman (1982) found that 67 percent of U.S. income went to one of these ends. In Jalostotitlan, Jalisco, the figure was 93 percent (Gonzalez and Escobar 1990). Similarly, in the three communities studied by Cornelius, 92 percent of remittances and 66 percent of savings were spent in these ways (1990a, 1990b); and Lopez reports that 83 percent of U.S. earnings went to family maintenance, housing, health care, or parties (1986a). Although additional studies do not report specific figures, similar spending patterns were reported by Shadow (1979), Reichert (1981), Stuart and Kearney (1981), Mines (1984), Fernaindez (1988), and Wiest (1979, 1984). As a result, most observers have concluded with Reichert that "although out-migration has generated higher per capita income and increased rates of consumption, it has not led to the development of the town economy in ways that have stimulated production or generated new employment opportunities" (Reichert 1981, 63). Richard Mines and Alain De Janvry also point out that in rural areas, successful migrants "buy up village lands and businesses when they become available, [but] they have little incentive to invest time or money in improving their investments since a semi-skilled job in the U.S. provides much more income . (Mines and De Janvry 1982, 452). Despite the apparent unanimity of opinion about the ill effects of migration to the United States, several studies provide counter-examples that question the mainstream view. In the town of Alvaro Obregon, for example, Paz Trigueros and Javier Rodriguez report that 30 percent of U. S. earnings were spent on land, tools, or livestock (Trigueros and Rodriguez 1988). Escobar and Martinez (1990) surveyed manual workers in Guadala- jara and found that 31 percent of migrants used their savings to set up a business. In their survey of one neighborhood in Guadalajara, Massey and his colleagues found that 21 percent of migrants used their savings productively and another 10 percent spent their savings on something that could be considered productive under some circumstances (Massey et al. 1987). In addition, several community studies report that migrant earnings have been crucial in community efforts to improve local infrastruc- ture, providing a key infusion of capital to build roads, install electricity, extend sewer lines, provide potable water, and restore public buildings. For example, even as Reichert (1981) lamented the "migrant syndrome" in Guadalupe, Michoacan, he reported that contributions by legal migrants accounted for 84 percent of all funds contributed by townspeople toward six capital improvement projects. Likewise, in Chamitlan, Michoacan, migrant earnings represented a "significant portion" of private contribu26 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. tions to recent public works (Massey et al. 1987). Goldring indicates that townspeople in Las Animas and Gomez Farias have generally been in a better position to make capital improvements than inhabitants of surrounding communities because of the high concentration of U.S. migrants, who are expected to contribute dollars for community improve- ment projects (Goldring 1990). Thus although most studies find a lack of development and high consumer spending resulting from U.S. migration, a few report productive investment and local economic growth. This discrepancy suggests caution in accepting generalizations about a "migrant syndrome" of dependency without considering the characteristics of the communities involved. In fact, most places that have been studied exhibit a geographic location and structural position in the Mexican political economy that are unsuitable for productive investment. Typically, the sending communities are isolated rural villages that are removed from natural markets; they frequently lack basic infrastructure like paved roads, electricity, telephones, running water, and sewage; and they often lack ready access to an appropriate labor force. Even in the realm of agriculture, many communities are poorly suited for investment because of poor quality land, lack of water, a fragmented land tenure system, and highly unequal prop- erty distribution. Under these circumstances, productive investment is not only unlikely but unwise because the chances of business failure are extremely high. Even so, most communities contain some business enterprises, and when the effect of U.S. migration is examined from the viewpoint of businesses rather than from the perspective of migrants, U.S. earnings are found to play a much more important role. For example, among businesses that Cornelius identified in his survey of three rural communities, 63 percent were owned by migrants and 61 percent were founded with U.S. earnings (1990a, 1990b). Of 19 enterprises in Guadalajara studied in detail by Escobar and Martinez, 32 percent were owned by migrants and 10 percent were capitalized with U.S. earnings (1990). The Massey team uncovered 162 business ventures in their survey of four communities, of which 39 percent were owned by migrants and 12 percent were established with U.S. earnings. Migrant-owned businesses accounted for 39 percent of business employment in these communities (Massey et al. 1987). Rather than concluding that migration inevitably leads to depen- dency and a lack of development, it is more appropriate to ask why productive investment occurs in some communities and not in others. In general, a perusal of the above communities suggests that the highest levels of business formation and investment occur in urban communities, rural communities with access to urban markets, or rural communities with favorable agricultural conditions. The effect of such structural factors 27 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review cannot be studied in the isolated case but can only be examined across a range of different communities. Douglas Massey and Laurence Basem developed a multivariate statistical model to analyze investment in four Mexican communities (Massey and Basem n.d.). They found that one of the strongest determinants of productive investment was the community to which one belonged, although the limited number of communities prevented them from identifying which community factors were specifi- cally responsible for differences in the tendency to invest productively. In addition to neglecting the effect of structural factors operating at the community level, studies of spending patterns generally do not take into account the life-cycle stage of the households involved. Active U.S. migrants are heavily concentrated in the age group of twenty to thirtyfour years of age, a time when most people are marrying, forming families, and raising children. During this phase of the life cycle, demands for family maintenance, housing, and medical care are greatest, and it is not surprising that migrants channel most of their earnings into current consumption. As migrants progress through the life cycle, however, spending on consumer goods falls off sharply and productive investment rises (Massey et al. 1987). Moreover, as families age and their migrant experience grows, they become increasingly likely to invest in agricultural inputs that raise productivity, such as machinery, fertilizers, insecticides, and improved seeds (Massey et al. 1987). Finally, patterns of migrant spending also depend in part on historical conditions that present different opportunities for investment in different eras (Durand 1988). The purchase of land was feasible up through the 1930s but became increasingly difficult thereafter, as private land was increasingly redistributed as state-sponsored ejidos. Competition for the remaining private parcels brought about a rapid inflation of prices that precluded further investment in land for most migrants. During the 1940s and 1950s, those who had received land under the government's redistribution program had the opportunity to invest in inputs such as machinery and fertilizers to make their parcels produce, but this avenue of investment too was eventually exhausted. At last, during the 1970s and 1980s, the extension of urban infrastructure and services to rural areas and the increasing congestion of urban metropolises made industrial investment attractive in the countryside and initiated a new wave of rural industrialization (Arias 1988; Durand 1989b, 1990). Thus although most observers have concluded from Mexican com- munity studies that U.S. migration leads to dependency and a lack of development, we argue that this conclusion does not account adequately for the structural, life-cycle, and historical factors that operate to constrain individual patterns of spending and investment. A careful look at the evidence suggests not that a "migrant syndrome" or an "addiction" exists but that under structural circumstances where productive investments are 28 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. likely to be met with success and during phases of the life cycle when current needs are at a minimum, spending for production is quite likely An important task for future research is to identify more specifically those community and household characteristics that are conducive to the productive use of income earned abroad. The Effect of the Reparto Agrario and Agricultural Modernization During the regime of President Lazaro Cardenas in the 1930s, Mexico embarked on an ambitious program of land reform and redistri- bution known as the Reparto Agrario (Yates 1981, 140-43). During this period, millions of hectares formerly held by wealthy landowners in large haciendas were broken into smaller plots and given over to the campe- sinos who traditionally had worked the lands. This fundamental shift in the political and economic organization of Mexico touched virtually all rural communities, and most historical case studies have considered the effects of the agrarian reform movement on U.S. migration. In general, the prevailing view is that the Reparto Agrario dis- couraged migration for several reasons: faced with the prospect of acquiring land, potential migrants stayed to fight for their allocations; the rules of the Reparto Agrario required that land remain in production, forcing campesinos to stay and care for it; and the acquisition of land provided a means of support to rural families that obviated the need for international migration. Consistent with these reasons, several studies indicate that the Reparto Agrario indeed acted as a brake on migration abroad (Lopez 1986a; Fonseca 1988; Fonseca and Moreno 1988). A broader survey of studies, however, reveals that agrarian reform had diverse effects in different communities, depending upon local social, economic, and political conditions. In the towns of Altamira, Jalisco, and Chamitlan, Michoacan, for example, the Reparto Agrario provided an incentive to migrate rather than the reverse. In these communities, campesinos were given land but not the capital or credit they needed to acquire tools, seeds, fertilizers, and other productive inputs. The only pragmatic solution for many families was U.S. migration, a traditional source of ready cash (Massey et al. 1987). This solution was facilitated by the Bracero Program, which arranged for recruiting, transporting, and housing migrants beginning in the 1940s. In other communities, it was the absence of an agrarian reform program that led to widespread emigration. Such was the case in Guadalupe, Michoacan, where townspeople under the influence of a local priest opted for entrenching against agrarian activists. Because of their actions, the town did not receive an allocation of ejido lands from the neighboring hacienda, even though other nearby communities got generous allot- ments. Because they had no land, men from Guadalupe were forced to 29 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review migrate at an early date and soon built relationships with U.S. employers that allowed them to acquire legal status. U.S. earnings provided them with markedly higher incomes than the ejidatarios in surrounding communities, and gradually they began to acquire lands from neighboring ejidos. In the end, families from Guadalupe gained control of most of the land in the area. In Copandaro, Michoacan, an even more peculiar outcome followed the Reparto Agrario (Rionda 1986, 1990). Here a group of peasants bought parts of the hacienda before the reform began and became small landowners with an interest in preserving private property. With the advent of the agrarian reform movement, these new landowners were destroyed politically and economically by agrarian activists who came to political power under President Lazaro Cardenas. The ruined property owners found relief from their travails only through U.S. migration. Years later, however, the agrarian activists who had expropriated the lands found the small plots insufficient to support their families and were ironically forced to borrow money from the U. S. migrants whose property they had taken earlier. Thus U. S. migration became widespread in Copandaro despite an active agrarian reform program. In the hacienda of Jalpa, Guanajuato, yet another sequence of events transpired when U.S. migration preceded the Reparto Agrario and obviated the need for it. In this community, an enlightened hacienda owner launched his own private agrarian reform program and sold a large part of his property to local campesinos and to poor families in neighboring municipios. According to Paul Taylor, many migrants from neighboring Arandas bought land from the hacienda with U.S. earnings (1933, 34). An ejido was never formed, and the process of buying and selling land stopped only when U.S. remittances declined with the onset of the depression in the 1930s. In sum, it is impossible to conclude from the study of any single community how the Reparto Agrario affected the process of migration to the United States because its effects were contingent on community factors that are difficult to analyze in an isolated instance. Judging from the studies accumulated so far, these structural factors include the extent of political divisions between agrarian activists and landowners, the dis- tribution of land prior to the reform movement, and the degree to which a community participated in U.S. migration before initiation of the Reparto Agrario. During the 1960s and 1970s, another upheaval affected rural com- munities when a wave of agricultural modernization introduced commercial crops, capital-intensive production methods, fertilizers, and mechanization to many areas of Mexico (Yates 1981; Barkin and DeWalt 1988). Community studies have generally shown that this wave of modernization set in motion a series of changes that displaced farm laborers and 30 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. increased the pressures for out-migration (Lopez 1986a; Massey et al. 1987; Fonseca 1988; Goldring 1990; Grindle 1988). Specifically, agricul- tural modernization consolidated property holding, substituted machines for hand labor, introduced capital inputs such as herbicides and fertilizers, and expanded work opportunities for women in factories and canneries. A few studies, however, have examined the effects of agricultural modernization across a range of different communities and have reached a more guarded tentative conclusion. Kenneth Roberts compared four communities in the states of Guanajuato, San Luis Potosi, Oaxaca, and Puebla and found that the effects of agricultural modernization depended greatly on the distribution of property and the quality of land within the community (K. Roberts 1982, 1984). When commercial crops and capitalintensive production methods were introduced into areas with good soil, irrigated land, and an even distribution of property, farm incomes rose and the risks to household income fell, thereby reducing the need for migration. But when commercialized agriculture was introduced under unfavorable agricultural conditions and an unequal distribution of land, farm incomes fell and risks rose, leading families to diversify their sources of support through internal and international migration. These results accord with those of Jesus Arroyo (1989). His study showed that the introduction of commercial agriculture into poorly developed rural areas was strongly linked to out-migration, but its application within welldeveloped rural areas and semi-urban areas did not yield large outflows of labor (see also Arroyo, de Leon, and Valenzuela 1990). Again, a uniform process can produce very different outcomes depending on structural characteristics of the local community. Migrant Strategies One last area of debate in the literature on migration concerns the relative importance of different strategies that migrants employ while working in the United States. Investigators have generalized from the experience of diverse case studies to develop various typologies that attempt to summarize the motivations and behaviors employed by migrants in traveling and working in the United States. In the early period of U.S. migration, for example, Gamio identified two kinds of migrantspermanent and temporary (1930a). More recently, Mines used migrants' legal status, duration of trip, and commitment to U.S. work to classify migrants into four basic types: undocumented shuttles, beginner permanents, long-term permanents, and legal shuttles (Mines 1981). Similarly, based on the frequency and duration of U.S. trips, Massey et al. (1987) defined three basic strategies of migration: temporary, recurrent, and 31 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review settled. Cornelius (1978) provided the simplest typology of all, dividing migrants into sojourners and settlers. Often these strategies are treated as if they were fixed characteristics of migrants and their communities (see Cornelius 1978, 1990c; Mines 1981). Some investigators, however, have argued that migrant strategies are not reified traits but fluid patterns of behavior that shift ov time in response to changing household circumstances and community conditions. For example, Goldring (1990) links the prevalence of different strategies of migration in Las Animas and Gomez Farias to the contrasting niches that their migrants occupy within the U.S. occupational-industrial structure. Those from Gomez Farias specialize in shuttle or recurrent migration because they are overwhelmingly employed in agriculture, which is more conducive to legalization and has a built-in seasonality that encourages moving back and forth frequently. Las Animas, in contrast, is dominated by migrants employing a settled strategy because their employment in urban areas and their undocumented status are more com- patible with longer periods of residence. Consistent with these results, Massey et al. (1987) found that migrants from rural communities tend to work in U.S. agriculture and employ temporary or recurrent strategies, whereas those from urban areas worked in U.S. cities and tended to employ settled strategies. In addition, Massey and his colleagues also found that strategies varied systematically over the life cycle. Early phases of the life cycle, before marriage and family formation, were associated with a settled strategy; middle phases of the life cycle, during childbearing and childrearing, were associated with a temporary strategy; and late phases of the life cycle, after most children are grown, were correlated with either a recurrent or settled strategy Finally, Massey et al. (1987) found that the choice of strategies shifted as households and communities accumulated increasing amounts of U.S. migrant experience. During early phases of the process, just after the initiation of migration, temporary patterns were most common; but as the amount of U.S. experience grew, recurrent and settled strategies increasingly predominated. SUMMARY AND CONCLUSION In this article, we have summarized research-classic as well as recent-on a variety of salient issues related to Mexican migration to the United States. We have considered the sixty-year-old running debate about the number of Mexican migrants who enter and reside in the United States and reviewed the controversy about the size of their remittances. Disagreements on these issues stem primarily from the fact that much of the migratory flow is undocumented and hence unobservable. We have also considered results from two decades of community studies, which 32 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. provide detailed and reliable information on undocumented and legal migration to the United States. The main weakness of community stud- ies-and the source of much confusion-is that individual case studies do not readily sustain generalization to the whole of Mexico. Our review suggests that public debate over the number of undocumented migrants has been characterized by a great deal of rhetoric centered on highly speculative estimates and by considerable confusion over the terms of reference. When attention is restricted to estimates based on analytic methods and when one understands that some studies focus on undocumented workers while others consider undocumented migrants, the range of estimates is actually quite small. It leads to figures of under 2 million undocumented Mexicans in the United States in 1980 and just over 3 million in 1986. Similarly, migrant remittances appear to have totaled some 2 billion dollars in 1984. Community studies have attempted to analyze the social and economic processes underlying these aggregate statistics, but generalizations have been plagued by inconsistencies. These discrepancies indicate the hazards of generalizing from single cases. Yet the diversity of outcomes from different settings does not imply that it is impossible to generalize about U.S. migration. It simply reflects the fact that the pattern and course of U.S. migration are strongly shaped by factors acting at the community level. Generalizations can be made, but they must incorporate the con- tingent effects of community-level variables. These effects cannot be observed in single cases but only across a range of community types. Our review of findings from some thirty-two different communities suggests that four leading factors are crucial in determining patterns and processes of migration from Mexican communities to the United States. First, the age of the migration stream is crucial in determining the composition of the migrant flow. The length of time a community has been sending migrants to the United States determines the maturity of its networks, the immigration policies that applied to its migrants, and the total amount of migrant experience that has accumulated among its people. In general, the earlier that migration began, the more developed a community's networks will be, the more legalizations will have occurred, and the more experience community members will have accumulated in the United States. As a result, the longer a community has been involved in the migration process, the more likely that migrants will come from the poorest segments of the socioeconomic hierarchy, the greater the participation of women and children, the higher the proportion of documented migrants, and the greater the reliance on settled and recurrent migrant strategies. A second important community factor that shapes and conditions the process of migration is the niche in the U.S. occupational-industrial structure where migrants first became established. In general, agricultural 33 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review employment is associated with higher rates of legalization, greater participation by women and children, and a prevalence of temporary or recurrent migrant strategies. Although the occupational niche of the first migrants from a community is often a matter of chance, once migrants become established in a particular occupational-industrial position, they tend to channel other townspeople into the same structural location, thereby determining the character and composition of the subsequent stream. The position of a community within Mexico's political economy is also a crucial determinant of the composition and behavior of migrants to the United States. Obviously, rural agrarian communities will tend to send agricultural workers whereas urban communities send industrial or service workers, thereby conditioning the class composition of the migrant flow. But a community's position in Mexico's political economy also plays a strong role in determining migrant spending patterns. Migrants living in communities with access to urban markets, well-developed roads, accessible power supplies, good communications, and an even distribution of productive resources tend to spend larger shares of their U.S. earnings on productive ends than do migrants from isolated, poorly developed communities that lack this infrastructure. Finally, the distribution and quality of agricultural land plays a crucial role in shaping the processes of migration from rural communities. The degree of inequality in landholding obviously affects the class com- position of migration at the point of its initiation: recruitment from a community characterized by extreme inequality yields a stream of landless day laborers whereas recruitment from a town where land is more evenly distributed yields a flow of small landholders and ejidatarios. Inequality in landholding likewise affects the extent to which migrants are able to use their earnings to purchase land, and the quality of the land determines migrants' incentives for investing their earnings productively in farming. The local pattern of landholding also mediates the effect of structural shifts in the Mexican political economy, such as the Reparto Agrario, and strongly conditions the role they have played in generating migrants to the United States. The redistribution of land under the Reparto Agrario, rather than discouraging out-migration, actually encouraged international movement in most communities through several distinct mechanisms. In communities where land was owned by a large hacienda, the Reparto Agrario led to out-migration because it provided small parcels of land that were insufficient to support a family or gave land but no capital to enable farming. In communities where some households owned parcels but others owned nothing, the resulting political strife brought about the economic and political ruin of small landowners, who turned to U.S. migration as a means of supporting themselves. In other communities, 34 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. landowners and priests frightened townspeople into not applying for land under the Reparto Agrario, leaving them landless and open to re- cruitment by U.S. employers during the 1940s. The quality and distribution of land have also played important roles in conditioning the effects of the wave of agricultural modernization that swept Mexico during the 1960s and 1970s. Although the common wisdom is that adopting cash crops and capital-intensive production meth- ods led to greater out-migration via displacement of rural workers, comparative studies suggest that the process was more complex. Displacement did occur in communities with poor quality land and in places where land was unequally distributed, but in communities where land was of high quality and more evenly distributed among families, the advent of commercialized farming increased rural incomes, lowered risks to farm households, and thereby reduced the pressures for migration to the United States. The difficulty of generalizing about Mexico-U.S. migration has led many investigators to speak of "myths," "fallacies," "false assumptions," "exaggerations," or "games." The Colegio de Mexico published a book entitled Undocumented Migrants: Myths and Realities (1979). An article by David North discusses "five myths about illegal migration to the United States" (North n.d.). Cornelius (1979) speaks of the "new mythology of undocumented Mexican migration to the United States." Corwin calls estimates of undocumented migration a "numbers game" (1982). Garcia y Griego captions his book with the subtitle "Three False Assumptions about Emigration to the United States" (1988). Gustavo Verduzco com- ments on "the false assumptions about the Simpson-Rodino Law" (1987), and Margarita Nolasco refers to "the wishful calculations of returned migrants."3 Still others call for a "new focus" (Diez Canedo 1984) and a "new vision" (Garcia y Griego 1988). We have argued that this state of affairs follows from two factors: the political nature of the debate, which obscures the real consistency of facts about the number of migrants and the size of their remittances; and the fact that most case studies have failed to recognize the effect of structural variables operating at the community level, which strongly shape and condition the underlying process of migration. Our comparative analysis of communities suggests that a fruitful approach to developing more general statements about Mexico-U.S. migration is to focus on the ways in which community variables interact with individual and household processes to produce the manifold outcomes that we observe. 3. Margarita Nolasco, "Los alegres calculos de regresados," El Sol de Mexico, 4 July 1984, p. A-2. 35 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review REFERENCES ADELMAN, IRMA, J. EDWARD TAYLOR, AND STEPHEN VOGEL 1988 "Life in a Mexican Village: A SAM Perspective." Journal of Development Studies 25:5-24. ARIAS, PATRICIA 1988 "La pequefia empresa en el occidente rural." Estudios Sociologicos 6, no. 17:405-36. ARROYO, JESUS 1989 El abandono rural. Guadalajara: Editorial Universidad de Guadalajara. ARROYO, JESUS, ADRIAN DE LEON, AND BASILIA VALENZUELA 1990 Patterns of Migration and Regional Development in the State of Jalisco, Mexico. Commission for the Study of International Migration and Cooperative Economic Development, Working Paper no. 42. Washington, D.C.: Commission for the Study of International Migration and Cooperative Economic Development. BARKIN, DAVID, AND BILLIE R. DEWALT 1988 "Sorghum and the Mexican Food Crisis." LARR 23, no. 3:30-59. BEAN, FRANK D., BARRY EDMONSTON, AND JEFFREY S. PASSEL, EDS. 1990 Undocumented Migration to the United States: IRCA and the Experience of the 1980s. Washington, D.C.: Urban Institute Press. BEAN, FRANK D., THOMAS J. ESPENSHADE, MICHAEL J. WHITE, AND ROBERT F. DYMOWSKI 1990 "Post-IRCA Changes in the Volume and Composition of Undocumented Migration to the United States: An Assessment Based on Apprehensions Data." In BEAN, EDMONSTON, AND PASSEL 1990, 111-58. BEAN, FRANK D., ALLAN G. KING, AND JEFFREY S. PASSEL 1983 "The Number of Illegal Migrants of Mexican Origin in the United States: SexRatio-Based Estimates for 1980." Demography 20:99-109. BEAN, FRANK D., EDWARD E. TELLES, AND B. LINDSAY LOWELL 1987 "Undocumented Migration to the United States: Perception and Evidence." Population and Development Review 13:671-90. BEAN, FRANK D., GEORGES VERNEZ, AND CHARLES B. KEELY 1989 Opening and Closing the Doors: Evaluating Immigration Reform and Control. Santa Monica, Calif., and Washington, D.C.: Rand Corporation and the Urban Institute. BUSTAMANTE, JORGE 1979 "Emigracion indocumentada a los Estados Unidos." In CENTRO DE ESTUDIOS INTERNACIONALES 1979, 23-67. 1990 "Undocumented Migration from Mexico to the United States: Preliminary Find- ings from the Zapata Canyon Project." In BEAN, EDMONSTON, AND PASSEL 1990, 211-26. CALVO, THOMAS, AND GUSTAVO LOPEZ, EDS. 1988 Movimientos de poblacio'n en el occidente de Me'xico. Zamora: Colegio de Michoacan. CAMARA, FERNANDO, AND ROBERT VAN KEMPER, EDS. 1979 Migration across Frontiers: Mexico and the United States. Contributions of the Latin American Anthropology Group, vol. 3. Albany, N.Y.: Institute of Mesoamerican Studies, State University of New York. CARRERAS, MERCEDES 1974 Los mexicanos que devolvio la crisis 1929-1932. Mexico City: Secretaria de Relaciones Exteriores. CENIET (CENTRO NACIONAL DE INFORMACION Y ESTADISTICAS DEL TRABAJO) 1982 Los trabajadores mexicanos en Estados Unidos: resultados de la Encuesta Nacional de Emigracion de la Frontera Norte del Pais y a los Estados Unidos. Mexico City: Secretaria del Trabajo y Prevision Social. CENTRO DE ESTUDIOS INTERNACIONALES 1979 Indocumentados: mitos y realidades. Mexico City: Colegio de Mexico. CHAPMAN, LEONARD F., JR. 1976 "Illegal Aliens: Time to Call a Halt!" Reader's Digest, no. 109:188-92. CHAVEZ, LEO R., ESTEVAN T. FLORES, AND MARTA LOPEZ-GARZA 1990 "Here Today, Gone Tomorrow? Undocumented Settlers and Immigration Reform." Human Organization 49:193-205. 36 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. CORNELIUS, WAYNE A. 1976a "Outmigration from Rural Mexican Communities." Interdisciplinary Communications Program Occasional Monograph Series 5, no. 2:1-39. Washington, D.C.: Smithsonian Institution. 1976b Mexican Migration to the United States: The View from Rural Sending Communities. Migration and Development Monograph C/76-12. Cambridge, Mass.: MIT Center for International Studies. 1978 Mexican Migration to the United States: Causes, Consequences, and U.S. Responses. Migration and Development Monograph C/78-9. Cambridge, Mass.: Center for International Studies, Massachusetts Institute of Technology. 1979 "La migracion ilegal mexicano a los Estados Unidos: conclusiones de investigaciones recientes, implicaciones politicas y prioridades de investigacion." In CENTRO DE ESTUDIOS INTERNACIONALES 1979, 69-110. 1989 "Impacts of the 1986 U.S. Immigration Law on Emigration from Rural Mexican Sending Communities." Population and Development Review 15:689-706. 1990a Labor Migration to the United States: Development Outcomes and Alternatives in Mexican Sending Communities. Final Report to the U.S. Commission for the Study of International Migration and Cooperative Economic Development. La Jolla: Center for U. S.-Mexican Studies, University of California, San Diego. 1990b Labor Migration to the United States: Development Outcomes and Alternatives in Mexican Sending Communities. Commission for the Study of International Migration and Cooperative Economic Development Working Paper no. 38. Washington, D.C.: Commission for the Study of International Migration and Cooperative Economic Development. 1990c "From Sojourners to Settlers: The Changing Profile of Mexican Migration to the United States." Manuscript, Center for U.S. -Mexican Studies, University of California at San Diego, La Jolla. CORONA, RODOLFO 1987 "Estimacion del nuimero de indocumentados a nivel estatal y municipal." Mimeo, Centro de Estudios sobre Identidad Nacional en Zonas Fronterizas, Universidad Nacional de Mexico, Cuernavaca, Morelos. CORWIN, ARTHUR F. 1982 "The Numbers Game: Estimates of Illegal Aliens in the United States, 1970-1981." Law and Contemporary Problems 45:223-97. CRAIG, RICHARD B. 1971 The Bracero Program: Interest Groups and Foreign Policy. Austin: University of Texas Press. DIEZ CANEDO, JUAN 1984 La migracion indocumentada de Mexico a los Estados Unidos: un nuevo enfoque. Mexico City: Fondo de Cultura Econ6mica. DINERMAN, INA R. 1982 Migrants and Stay-at-Homes: A Comparative Study of Rural Migration from Michoacan, Me'xico. La Jolla: Center for U. S.-Mexican Studies, University of California, San Diego. DURAND, JORGE 1986 "Circuitos migratorios en el occidente de Mexico." Revue Europeanns des Migrations Internationals 2:45-67 (published in Poipiers, France). 1988 "Los mivrad6lares: cien afios de inversi6n en el medio rural." Argumentos: Estu- dios Criticos de la Sociedad 5:7-21. 1989a "Los hijos de Rodino." L'Ordinaire Mexique Amerique Centrale 122:57-60 (published by the Universite de Toulouse-Le Mirail). 1989b "Guanajuato, tierra de migrantes." Paper presented at the seminar "International Migration: Current State and Perspectives," Cocoyoc, Morelos, 4-6 October. 1990 "Guanajuato, cantera de migrants." Encuentro 4, no. 4:49-62. ESCOBAR, AGUSTIN, MERCEDES GONZALEZ, AND BRYAN ROBERTS 1987 "Migration, Labor Markets, and the International Economy: Jalisco, Mexico, and the United States." In Migrants, Workers, and the Social Order, edited by Jeremy Eades, 42-64. London: Tavistock. 37 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review ESCOBAR, AGUSTIN, AND MARIA DE LA 0 MARTINEZ 1990 Small-Scale Industry and International Migration in Guadalajara, Mexico. Commission for the Study of International Migration and Cooperative Economic Development, Working Paper no. 53. Washington, D.C.: Commission for the Study of International Migration and Cooperative Economic Development. FABILA, ALFONSO 1929 El problema de la emigracion de obreros y campesinos mexicanos. Mexico City: Talleres Graficos de la Nacion. FERNANDEZ, CELESTINO 1988 "Migracion hacia los Estados Unidos: caso Santa Ines, Michoacan." In LOPEZ, ED., 1988, 113-24. FONSECA, OMAR 1988 "De Jaripo a Stockton, California: un caso de migracion en Michoacan." In CALVO AND LOPEZ 1988, 359-72. FONSECA, OMAR, AND LILIA MORENO 1988 "Consideraciones historico-sociales de la migracion de trabajadores michoacanos a los Estados Unidos de America: el caso de Jaripo." In LOPEZ, ED., 1988, 65-84. GALARZA, ERNEST 1964 Merchants of Labor: The Mexican Bracero Story. Santa Barbara, Calif.: McNally and Loftin. GAMIO, MANUEL 1930a Mexican Immigration to the United States. Chicago, Ill.: University of Chicago Press. 1930b Numero, procedencia y distribucion geogrdfica de los inmigrantes mexicanos en los Estados Unidos. Mexico City: Talleres Graficos y Diario Oficial. 1931 The Mexican Immigrant: His Life Story. Chicago, Ill.: University of Chicago Press. GARCIA Y GRIEGO, MANUEL 1987 "International Migration Statistics in Mexico." International Migration Review 21: 1245-57. 1988 "Hacia una nueva vision del problema de los indocumentados en los Estados Unidos." In Mexico y Estados Unidos frente a la migracio'n de indocumentados, edited by Manuel Garcia y Griego and Monica Verea, 123-53. Mexico City: Miguel Angel Porru'a. GARCIA Y GRIEGO, MANUEL, AND FRANCISCO GINER DE LOS RIOS 1985 "ZEs vulnerable la economia a la aplicacion de politicas m denses?" In Me'xico-Estados Unidos, edited by Manuel Garcia y Griego and Gus- tavo Vega. Mexico City: Colegio de Mexico. GOLDRING, LUIN 1990 Development and Migration: A Comparative Analysis of Two Mexican Migrant Cir- cuits. Commission for the Study of International Migration and Cooperative Economic Development, Working Paper no. 37. Washington, D.C.: Commission for the Study of International Migration and Cooperative Economic Development. GONZALEZ, MERCEDES, AND AGUSTIN ESCOBAR 1990 The Impact of IRCA on the Migration Patterns of a Community in Los Altos, Jalisco, Mexico. Commission for the Study of International Migration and Cooperative Economic Development, Working Paper no. 41. Washington, D.C.: Commission for the Study of International Migration and Cooperative Economic Development. GRINDLE, MERILEE S. 1988 Searching for Rural Development: Labor Migration and Employment in Mexico. Ithaca, N.Y.: Cornell University Press. GUZMAN, RALPH C. 1979 "La repatriacion forzosa como solucion politica concluyente al problema de la emigraci6n ilegal: una perspectiva hist6rica." In CENTRO DE ESTUDIOS INTERNACIONALES 1979, 137-63. HANCOCK, RICHARD 1959 The Role of the Bracero in the Economic and Cultural Dynamics of Mexico: A Case Study of Chihuahua. Stanford, Calif.: Hispanic-American Society. 38 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. HEER, DAVID M., AND JEFFREY S. PASSEL 1985 "Comparison of Two Different Methods for Computing the Number of Undocu- mented Mexican Adults in the Los Angeles PMSA." Paper presented to the Population Association of America, 21-24 Apr., Boston, Massachusetts. HOFFMAN, ABRAHAM 1974 Unwanted Mexican Americans in the Great Depression: Repatriation Pressures, 19291939. Tucson: University of Arizona Press. IMF (INTERNATIONAL MONETARY FUND) 1987 International Financial Statistics: 1987 Yearbook. Washington, D.C.: International Monetary Fund. JASSO, GUILLERMINA, AND MARK R. ROSENZWEIG 1990 The New Chosen People: Immigrants in the United States. New York: Russell Sage. JONES, RICHARD C., ED. 1984 Patterns of Undocumented Migration: Mexico and the United States. Totowa, N.J.: Rowman and Allanheld. KEELY, CHARLES B., AND BAO NGA TRAN 1989 "Remittances from Labor Migration: Evaluations, Performance, and Implications." International Migration Review 23:500-525. KOSSOUDJI, SHERRIE A. 1990 "Playing Cat and Mouse at the Border: Does the INS Alien Apprehension Strat- egy Alter the Aggregate Supply of Illegal Labor?" Paper presented at the Annual Meetings of the Population Association of America, 3-5 May, Toronto, Ontario. KOSSOUDJI, SHERRIE A., AND SUSAN I. RANNEY 1984 "The Labor Market Experience of Female Migrants: The Case of Temporary Mexican Migration to the U.S." International Migration Review 18:1120-43. 1986 Legal Status as Union Membership: Legal and Illegal Wage Rates of Mexican Immigrants. Population Studies Center Research Report no. 86-103. Ann Arbor: Population Studies Center, University of Michigan. LESKO ASSOCIATES 1975 Final Report: Basic Data and Guidance Required to Implement a Major Illegal Alien Study during Fiscal Year 1976. Washington, D.C.: U.S. Immigration and Naturalization Service. LOPEZ, GUSTAVO 1986a La casa dividida: un estudio de caso sobre migraci6n a Estados Unidos en un pueblo michoacano. Zamora: Colegio de Michoacan. 1986b "Migraci6n temporal a Estados Unidos en una comunidad mexicana." In Se fue a volver: Seminario sobre Migraciones Temporales en America Latina. Mexico City: Pispal/Ciudad/CENEP. 1988 "La migraci6n a Estados Unidos en G6mez Farias, MichoacAn." In LOPEZ, ED., 1988, 125-34. LOPEZ, GUSTAVO, ED. 1988 Migraci6n en el occidente de Mexico. Zamora: Colegio de Michoacan. LOPEZ, GUSTAVO, AND SERGIO ZENDEJAS 1988 "Migraci6n internacional por regiones en Michoacan." In CALVO AND LOPEZ 1988, 81-102. LOWELL, B. LINDSAY, FRANK D. BEAN, AND RODOLFO 0. DE LA GARZA 1986 'The Dilemmas of Undocumented Immigration: An Analysis of the 1984 SimpsonMazzoli Vote." Social Science Quarterly 67:118-27. MARTIN, PHILIP L. 1990 "Harvest of Confusion: Immigration Reform and California Agriculture." International Migration Review 24:69-95. MASSEY, DOUGLAS S. 1981 "The New Immigration to the United States and the Prospects for Assimilation." Annual Review of Sociology 7:57-85. 1987 "Understanding Mexican Migration to the United States." American Journal of Sociology 92:1372-1403. 39 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review MASSEY, DOUGLAS S., RAFAEL ALARCON, JORGE DURAND, AND HUMBERTO GONZALES 1987 Return to Aztlan: The Social Process of International Migration from Western Mexico. Berkeley and Los Angeles: University of California Press. MASSEY, DOUGLAS S., AND LAURENCE BASEM n.d. "Determinants of Savings, Remittances, and Spending Patterns among Mexican Migrants to the United States." Sociological Inquiry, forthcoming. MASSEY, DOUGLAS S., KATHARINE M. DONATO, AND ZAI LIANG 1990 "Effects of the Immigration Reform and Control Act of 1986: Preliminary Data from Mexico." In BEAN, EDMONSTON, AND PASSEL 1990, 183-210. MASSEY, DOUGLAS S., AND FELIPE GARCIA ESPANA 1987 "The Social Process of International Migration." Science, no. 237:733-38. MINES, RICHARD 1981 Developing a Community Tradition of Migration: A Field Study in Rural Zacatecas, Mexico, and California Settlement Areas. Monographs in U.S.-Mexican Studies no. 3. La Jolla: Center for U. S.-Mexican Studies, University of California at San Diego. 1984 "Network Migration and Mexican Rural Development: A Case Study." In JONES 1984, 136-55. MINES, RICHARD, AND ALAIN DE JANVRY 1982 "Migration to the United States and Mexican Rural Development: A Case Study" American Journal of Agricultural Economics 64:444-54. MINES, RICHARD, AND DOUGLAS S. MASSEY 1985 "Patterns of Migration to the United States from Two Mexican Communities." LARR 20, no. 2:104-23. MORALES, PATRICIA 1981 Indocumentados Mexicanos. Mexico City: Grijalbo. NORTH, DAVID S. n.d. "Cinco mitos en torno a la inmigraci6n ilegal hacia Estados Unidos." Revista de la Universidad de Guadalajara 9, no. 2:11-20. NORTH, DAVID S., AND MARION F. HOUSTOUN 1976 The Characteristics and Role of Illegal Aliens in the U. S. Labor Market: An Exploratory Study. Washington, D.C.: Linton. OROZCO, JUAN LUIS 1990 "El negocio de los ilegales: Zganancias para quien?" Ph.D. diss., Estudios de America Latina, Paris. PASSEL, JEFFREY S. 1985 "Undocumented Immigrants: How Many?" Proceedings of the Social Statistics Section, American Statistical Association, 1985:65-72. 1986 "Undocumented Immigration." Annals, American Academy of Political and Social Science 487:181-200. PASSEL, JEFFREY S., AND KAREN A. WOODROW 1984 "Geographic Distribution of Undocumented Immigrants: Estimates of Undocumented Aliens Counted in the 1980 Census by State." International Migration Review 18:642-71. 1987 "Change in the Undocumented Alien Population in the United States, 1979-83." International Migration Review 21:1304-23. PLEHN, CARL C., IRA B. CROSS, AND MELVIN M. KNIGHT, EDS. 1929- University of California Publications in Economics, Berkeley: University of California 1932 Press. PORTES, ALEJANDRO, AND RUBEN C. RUMBAUT 1990 Immigrant America: A Portrait. Berkeley and Los Angeles: University of California Press. RANNEY, SUSAN I., AND SHERRIE A. KOSSOUDJI 1983 "Profiles of Temporary Mexican Labor Migrants to the United States." Population and Development Review 9:475-92. REICHERT, JOSHUA S. 1979 "The Migrant Syndrome: An Analysis of U.S. Migration and Its Impact on a Rural Mexican Town." Ph.D. diss., Princeton University. 40 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �MEXICAN MIGRATION TO THE U.S. 1981 "The Migrant Syndrome: Seasonal U.S. Wage Labor and Rural Development in Central Mexico." Human Organization 40:56-66. 1982 "A Town Divided: Economic Stratification and Social Relations in a Mexican Migrant Community." Social Problems 29:411-23. REICHERT, JOSHUA S., AND DOUGLAS S. MASSEY 1979 "Patterns of U.S. Migration from a Mexican Sending Community: A Comparison of Legal and Illegal Migrants." International Migration Review 13:599-623. 1980 "History and Trends in U.S.-Bound Migration from a Mexican Town." International Migration Review 14:475-91. REUBENS, EDWIN P. 1980 "Immigration Problems, Limited-Visa Programs, and Other Options." Appendix F in U. S. Immigration Policy and the National Interest: Report of the Select Commission on Immigration and Refugee Policy. Washington, D.C.: U.S. Government Printing Office. RIONDA, LUIS MIGUEL 1986 "Continuidad y accesi6n social en Copandaro." In Estudios Michoacanos II, edited by Carlos Herrej6n, 245-56. Zamora: Colegio de Michoacan. 1990 "Y jalaron pal norte." M.A. thesis, Colegio de Michoacan. ROBERTS, KENNETH D. 1982 'Agrarian Structure and Labor Mobility in Rural Mexico." Population and Development Review 8:299-322. 1984 'Agricultural Development and Labor Mobility: A Study of Four Mexican Subregions." In JONES 1984, 74-92. ROUSE, ROGER C. 1989 "Mexican Migration to the United States: Family Relations in the Development of a Transnational Migrant Circuit." Ph.D. diss., Stanford University. SAMORA, JULIAN 1971 Los Mojados: The Wetback Story. Notre Dame, Ind.: University of Not SANTIBANEZ, ENRIQUE 1930 Ensayo acerca de la inmigraci6n mexicana en los Estados Unidos. San Antonio, Tex.: Cleeg. SELBY, HENRY A., AND ARTHUR D. MURPHY 1984 The Mexican Urban Household and the Decision to Migrate to the United States. Occasional Papers in Social Change, no. 4. Philadelphia, Pa.: Institute for the Study of Human Issues. SHADOW, ROBERT D. 1979 "Differential Out-Migration: A Comparison of Internal and International Migration from Villa Guerrero, Jalisco (Mexico)." In CAMARA AND VAN KEMPER 1979, 67-84. STARK, ODED, AND J. EDWARD TAYLOR 1989 "Relative Deprivation and International Migration." Demography 26:1-14. STUART, JAMES, AND MICHAEL KEARNEY 1981 Causes and Effects of Agricultural Labor Migration from the M California. La Jolla: Center for U.S.-Mexican Studies, University of California, San Diego. TAYLOR, J. EDWARD 1986 "Differential Migration, Networks, Information, and Risk." In Research in Human Capital and Development: Migration, Human Capital, and Development, edited by Oded Stark, 4:147-71. Greenwich, Conn.: Jai. 1987 "Undocumented Mexico-U.S. Migration and the Returns to Households in Rural Mexico." American Journal of Agricultural Economics 69:626-37. 1988 "Life in a Mexican Village: A SAM Perspective." Journal of Development Studies 25:5-24. TAYLOR, PAUL S. 1929 "Mexican Labor in the United States: Migration Statistics." In PLEHN, CROSS, AND KNIGHT 1929, 237-55. 1930 "Mexican Labor in the United States: Dimmit County, Winter Garden District, South Texas." In PLEHN, CROSS, AND KNIGHT 1930, 393-464. 41 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms �Latin American Research Review 1931 "Mexican Labor in the United States: Bethlehem, Pennsylvania." In PLEHN, CROSS, AND KNIGHT 1931, 1-24. 1932 "Mexican Labor in the United States: Chicago and the Calumet Region." In PLEHN, CROSS, AND KNIGHT 1932, 25-284. 1933 'A Spanish-Mexican Peasant Community: Arandas in Jalisco, Mexico." IberoAmericano 4:1-80 (published by the University of California Press). TRIGUEROS, PAZ, AND JAVIER RODRIGUEZ 1988 "Migraci6n y vida familiar en Michoacan." In LOPEZ, ED., 1988, 201-21. U.S. INS (IMMIGRATION AND NATURALIZATION SERVICE) 1989 1988 Statistical Yearbook of the Immigration and Naturalization Service. Washington, D.C.: U.S. Government Printing Office. VERDUZCO, GUSTAVO 1987 "Los falsos supuestos de la Ley Simpson-Rodino." Foro Internacional 27:168-70. WARREN, ROBERT, AND JEFFREY S. PASSEL 1987 'A Count of the Uncountable: Estimates of Undocumented Aliens Counted in the 1980 United States Census." Demography 24:375-93. WIEST, RAYMOND 1973 "Wage-Labor Migration and the Household in a Mexican Town." Journal of Anthropological Research 29:108-209. 1979 "Implications of International Labor Migration for Mexican Rural Development." In CAMARA AND VAN KEMPER 1979, 85-97. 1984 "External Dependency and the Perpetuation of Temporary Migration to the United States." In JONES 1984, 110-35. YATES, P. LAMARTINE 1981 Mexico's Agricultural Dilemma. Tucson: University of Arizona Press. 42 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:44:45 UTC All use subject to https://about.jstor.org/terms � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Mexican Migration to the United States: A Critical Review Creator An entity primarily responsible for making the resource Durand, Jorge Massey, Douglas, S Subject The topic of the resource Immigration & society Mexican Immigration Latin America Description An account of the resource An examination of “two particular salient issues — The number of Mexican migrants to the United States and the quantity of their monetary remittances to Mexico” Publisher An entity responsible for making the resource available Latin American Research Review Date A point or period of time associated with an event in the lifecycle of the resource 1992 Type The nature or genre of the resource Text Review Article Format The file format, physical medium, or dimensions of the resource PDF Identifier An unambiguous reference to the resource within a given context https://www.jstor.org/stable/2503748 Source A related resource from which the described resource is derived JSTOR Language A language of the resource English https://dh.scu.edu/exhibits/files/original/90/2936/Rise_and_Fall_of_U.S._Low-Skilled_Immigration.pdf 0c70f398eec83a099fd255374418dbf8 PDF Text Text The Rise and Fall of U.S. Low-Skilled Immigration Author(s): GORDON HANSON, CHEN LIU and CRAIG McINTOSH Source: Brookings Papers on Economic Activity , (SPRING 2017), pp. 83-151 Published by: Brookings Institution Press Stable URL: https://www.jstor.org/stable/10.2307/90013169 REFERENCES Linked references are available on JSTOR for this article: https://www.jstor.org/stable/10.2307/90013169?seq=1&cid=pdfreference#references_tab_contents You may need to log in to JSTOR to access the linked references. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms Brookings Institution Press is collaborating with JSTOR to digitize, preserve and extend access to Brookings Papers on Economic Activity This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON University of California, San Diego CHEN LIU University of California, San Diego CRAIG M c INTOSH University of California, San Diego The Rise and Fall of U.S. Low-Skilled Immigration ABSTRACT    From the 1970s to the early 2000s, the United States experienced an epochal wave of low-skilled immigration. Since the Great Recession, however, U.S. borders have become a far less active place when it comes to the net arrival of foreign workers. The number of undocumented immigrants has declined in absolute terms, while the overall population of low-skilled, foreign-born workers has remained stable. We examine how the scale and composition of low-skilled immigration in the United States have evolved over time, and how relative income growth and demographic shifts in the Western Hemisphere have contributed to the recent immigration slowdown. Because major source countries for U.S. immigration are now seeing and will continue to see weak growth of the labor supply relative to the United States, future immigration rates of young, low-skilled workers appear unlikely to rebound, whether or not U.S. immigration policies tighten further. I mmigration is a divisive issue in public discourse about U.S. economic policy. At the center of the debate is how to address inflows of undocumented immigrants. During previous decades, inflows of illegal aliens were substantial. The Pew Research Center estimates that between 1990 Conflict of Interest Disclosure: The authors received financial support for this work from the Center on Global Transformation at the University of California, San Diego. With the exception of the aforementioned affiliations, the authors did not receive financial support from any firm or person for this paper or from any firm or person with a financial or political interest in this paper. They are currently not officers, directors, or board members of any organization with an interest in this paper. 83 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �84 Brookings Papers on Economic Activity, Spring 2017 Figure 1. U.S. Foreign-Born Population, Age 18–64, 1970–2015 Millionsa 30 All 20 High school or less 10 Less than high school 5 2 1980 1990 Year 2000 2010 Sources: U.S. Census Bureau, decennial census, American Community Survey; authors’ calculations. a. Note that this axis is on a log scale. and 2007, the U.S. population of undocumented residents, which as of 2013 accounted for nearly two-thirds of the U.S. foreign-born adult population with 12 or fewer years of schooling, grew on net by an annual average of 510,000 individuals (Borjas 2016; Passel and Cohn 2016). These inflows contributed to a sizable increase in the U.S. supply of low-skilled, foreign-born workers (figure 1). During the 1990–2007 period, the number of working-age immigrants with 12 or fewer years of schooling more than doubled, rising from 8.5 million to 17.8 million. Since the Great Recession, however, U.S. borders have become a less active place when it comes to net inflows of low-skilled labor from abroad. The undocumented population declined in absolute terms between 2007 and 2014, falling on net by an annual average of 160,000 individuals, while the overall population of low-skilled immigrants of working age remained stable. Viewed through the lens of the U.S. business cycle, the recent slowdown in low-skilled immigration hardly comes as a surprise (Villarreal 2014). Construction is the second-largest sector of employment for undocumented labor and the third-largest among all low-skilled immigrants (Passel and Cohn 2016). Because the collapse in the U.S. housing market helped precipitate the Great Recession (Mian and Sufi 2014), it follows logically that This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 85 the downturn in home building after 2006 would have triggered a drop in new arrivals of low-skilled, foreign-born workers. Yet, there are good reasons to believe that the Great Recession may have merely advanced forward in time an inevitable reduction in low-skilled immigration. Today, about half of low-skilled immigrants are from Mexico and another onequarter are from elsewhere in the Latin American and the Caribbean countries. Because these countries had marked declines in fertility after the late 1970s, they started to see slower growth in the size of cohorts coming of working age in the 2000s, thereby weakening a key demographic push factor for emigration (Hanson and McIntosh 2010, 2012). Just as relatively strong growth in U.S. GDP and Latin American labor supplies a generation ago helped initiate the great U.S. immigration wave of the late 20th century, the reversal of these conditions may be launching the United States into an era of far more modest low-skilled labor inflows (Hanson and McIntosh 2009, 2016). The policy dilemma facing the United States is thus not so much how to arrest massive increases in the supply of foreign labor, but rather how to prepare for a lower-immigration future. The pertinent issues for economists to address include how the scale and composition of low-skilled labor inflows have changed over time, whether the drop in inflows is primarily a cyclical phenomenon or represents a secular decline, and how the U.S. economy would adjust to an environment with modest numbers of low-skilled, foreign-born workers entering the labor force each year. These questions guide the analysis in this paper. We begin by summarizing trends in low-skilled immigration over the last several decades. As is well known, supplies of less-educated, foreignborn labor increased sharply after 1970, while their predominant national origins shifted from Europe to Latin America. Perhaps less appreciated, the demographic structure of this population has also changed, moving from younger, recent arrivals toward an older, more settled population. Which types of individuals select into immigration also appear to have changed, a pattern we examine in detail for the case of Mexico, given its outsize importance as a source country for U.S. immigrants. In 1990, those having recently migrated from Mexico to the United States—as captured by the population censuses of the two countries—were drawn more heavily from just above versus just below the mean of potential labor market earnings in Mexico (Chiquiar and Hanson 2005). This mild positive selection weakened during the 1990s and the 2000s, such that by 2010 the population of recent Mexican immigrants was close to a random draw of working-age individuals from Mexico, with a slight overrepresentation of individuals This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �86 Brookings Papers on Economic Activity, Spring 2017 from the middle of the skill distribution. Although immigrant selection captured in census data may be subject to measurement error associated with undercounts of undocumented immigrants (Fernández-Huertas Moraga 2011), selection patterns in these data are similar to those in the Mexican Family Life Survey, which appears less subject to missing information on undocumented migrants (Kaestner and Malamud 2014). The largely neutral selection of immigrants from Mexico in terms of skill implies that any future shock to Mexican immigration—such as a further tightening of U.S. borders—would target middle-income earners in Mexico, while affecting low-wage earners in the United States. Recent changes in low-skilled immigration have occurred in a tumultuous environment for the U.S. labor market. Even before the economic turbulence that occurred after 2006, there were adverse changes in the demand for low-skilled labor associated with automation and increased import competition from low-wage countries (Autor and Dorn 2013; Autor, Dorn, and Hanson 2013; Pierce and Schott 2016; Cortes, Jaimovich, and Siu 2016). At the higher end of the labor market, demand for young, college-educated labor has also weakened (Beaudry, Green, and Sand 2016). Together, these changes have combined to create a period low wage growth since 2000 for all but the highest-earning U.S. workers (Valletta forthcoming). To put recent changes in U.S. labor market conditions in a global context, we compare the level and volatility of U.S. income with that in major sending countries for low-skilled immigrants. The gap between the 25th percentile of the income distribution in the United States and the 50th percentile of the income distribution in Mexico—which approximates the expected gains in earnings for the typical Mexican migrant—was stable during the 1990s and early 2000s but shrank noticeably after 2007. Relative volatility in income growth has also changed. The Great Moderation heralded a period of steady U.S. GDP growth from the early 1980s to the mid-2000s (Bernanke 2004), a calm that was brought to an end by the Great Recession. In Mexico and other migrant-sending nations of the Western Hemisphere, the pattern is roughly the opposite. The 1980s and early 1990s were periods of high macroeconomic volatility, whereas the 2000s were a period of steady if not spectacular economic growth. Shrinking income gaps and reduced income volatility between the United States and major migrant-sending nations have eased pressures for net labor flows into the United States. Another factor contributing to the decline in low-skilled immigration is changes in U.S. enforcement against illegal labor inflows (Roberts, Alden, and Whitley 2013). Between 2000 and 2010, the number of U.S. Border This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG McINTOSH 87 Patrol agents policing the U.S.–Mexico border doubled, from 8,600 to 17,500 officers, and has since remained at historically high levels. Concurrently, the U.S. government intensified immigration enforcement in the interior of the country, which led to an increase in deportations of noncriminal aliens—many of whom are apprehended through traffic stops or other routine law enforcement operations—from 116,000 individuals in 2001 to an average of 226,000 a year from 2007 to 2015.1 Increases in border enforcement, which deter potential migrants from choosing to enter the United States (Gathmann 2008; Angelucci 2012), and in interior enforcement, which reduces the existing population of undocumented immigrants and may also deter future immigration, appear likely to continue under the administration of Donald Trump (Meckler 2017; Kulish and others 2017). Looking toward the future of U.S. low-skilled immigration, forces are at work that are likely to soften pressures for labor inflows and that will remain in place for the next several decades. By the mid-1970s, the size of U.S. cohorts coming of working age was growing much more slowly than in Mexico and the rest of Latin America, creating steady pressure for migration to the United States. However, by the mid-2000s this demographic push factor had largely disappeared. Because the United States’ neighbors to the south are today experiencing much slower growth in the labor supply, the future immigration of young low-skilled labor looks set to decline rapidly, whether or not more draconian policies to control U.S. immigration are implemented. If changes in global macroeconomic conditions and U.S. enforcement policy have combined to weaken recent growth in the U.S. supply of lowskilled, foreign-born labor, what are the implications for U.S. labor markets? As a way of answering this question, we examine the net impact of immigration-induced changes in the labor supply on U.S. labor-market tightness. To perform this analysis, we apply the approach of Lawrence Katz and Kevin Murphy (1992) to data from the U.S. Current Population Survey (CPS), which involves modeling the relative hourly earnings of more- and less-skilled labor as a function of their relative supplies and a flexible time trend, meant to capture the evolution of labor demand. We estimate the model using earnings and employment data for the period 1976–2007 and then project relative earnings through 2015, using either 1. Noncitizens (including legal immigrants) convicted of an aggravated felony, a drug crime, or multiple crimes involving moral turpitude are subject to deportation upon or before completion of their prison sentence. Deportations of criminal aliens also increased in the 2000s, from 73,000 in 2001 to an average of 156,000 per year over 2007 to 2015. See Gonzalez-Barrera and Lopez (2016). This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �88 Brookings Papers on Economic Activity, Spring 2017 actual labor supplies or labor supplies under counterfactual assumptions about low-skilled immigration. If, counterfactually, the low-skilled, foreign-born labor supply had grown at the same rate during the period 2008–15 as it did from 1994 to 2007, our simple model implies that the wage gap between more-skilled and less-skilled labor would have been 6 to 9 percentage points higher in 2015. This finding, though not a general equilibrium assessment of the wage effects of U.S. immigration, illustrates the magnitude of the immigration slowdown in terms of U.S. wage pressures. To the extent that slowing low-skilled immigration puts downward pressure on the skill premium, we would expect firms to invest more in automation and other changes in production techniques that reduce reliance on low-skilled labor (Card and Lewis 2007; Lewis 2011), effects that are likely to register most strongly in immigrant-intensive industries such as agriculture, construction, eating and drinking establishments, and nondurable manufacturing. Our work complements the existing literature on immigration, much of which takes national changes in low-skilled foreign labor supply as given and examines its impact on the earnings of U.S. native-born workers.2 As is well known, estimates of the wage effects of immigration vary widely across studies (Blau and Mackie 2016). Results depend on how one defines the geographic scope of labor markets, skill groups within these labor markets, and the interchangeability of native- and foreign-born workers on the job (Borjas 2003, 2013; Card 2001, 2009; Ottaviano and Peri 2012; Dustmann, Frattini, and Preston 2013). To explain instability in the wage effects of immigration, the literature has studied factors that may confound empirical analysis, including offsetting migration by nativeborn workers (Borjas 2006), the location choices of immigrant workers (Cadena and Kovak 2016), firm-level changes in technology (Lewis 2011), occupational downgrading by immigrant workers (Peri and Sparber 2009; Dustmann, Frattini, and Preston 2013), and measurement error in labor market earnings (Aydemir and Borjas 2011). Relative to existing work, we offer the inverse perspective of how and why the low-skilled immigrant labor supply has changed. Given the abundance of research on how 2. Other literature on the effects of low-skilled immigration in the United States examines its consequences for local consumer prices (Cortes 2008), the labor supply of highskilled, native-born women (Cortes and Tessada 2011), local housing prices (Saiz 2007), state GDP growth (Edwards and Ortega 2016), cultural diversity (Ottaviano and Peri 2005), and occupational employment and wages of native-born workers in local labor markets (Burstein and others 2017). This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG McINTOSH 89 immigration affects U.S. wages, the factors that govern the magnitude of low-skilled immigration are understudied. Our work helps address this gap in knowledge. I. �The Presence of Low-Skilled Immigrants in the U.S. Labor Force We begin our analysis with an overview of the characteristics of low-skilled immigrants in the United States and then examine how selection into U.S. migration among individuals from Mexico has changed over time. For the analysis in this section and the next, we focus on individuals of working age, defined as those 18 to 64 years of age. We utilize data from the U.S. population censuses, the American Community Survey (ACS), and the CPS; and from Mexico’s population census, available from the Integrated Public Use Microdata Series. I.A. Characteristics of Low-Skilled Immigrants A preliminary issue we must address is how to define low-skilled labor. When it comes to the analysis of immigration, the literature alternatively defines low-skilled workers as those with less than a high school education (Borjas 2003) or those with a high school education or less (Card 2001).3 This difference matters, because those completing less than 12 years of schooling are an ever-smaller share of the U.S. native-born population but continue to account for a majority of adults in low- and middle-income countries. In the nations that send large numbers of low-skilled migrants to the United States—including Colombia, Cuba, the Dominican Republic, Ecuador, El Salvador, Guatemala, and Mexico—compulsory schooling is through grade 8 or 9, as opposed to being through age 16 in most U.S. states. The median worker in many sending countries thus has well less than the equivalent of a U.S. high school education (Clemens, Montenegro, and Pritchett 2008). Cross-national differences in compulsory education are manifest in the distribution of years of schooling among less-educated foreign- and native-born adults in the United States. In 1970, those not completing high school accounted for just over half of U.S. native-born adults with a high school education or less, a share that declined to 3. We define high school education to mean completing 12 years of school, whether or not a degree is granted, a convention we adopt because the meaning of a high school degree varies across countries. Throughout the paper, we use high school education and 12 years of schooling interchangeably. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �90 Brookings Papers on Economic Activity, Spring 2017 29.4 percent in 1990 and to 16.6 percent in 2015 (table 1). Among the U.S. foreign-born adult population with a high school education or less, the share with less than 12 years of schooling has also fallen but from a much higher base, beginning at 65.2 percent in 1970 and dropping to 55.0 percent in 1990 and to 44.7 percent in 2015. To ensure that our analysis is robust to the definition of skill, we utilize both education-based definitions of low-skilled labor.4 When viewed over the sweep of the last half century, the U.S. lowskilled, foreign-born population has transformed not just in terms of its size but also in its demographic structure. These evolutions are evident in tables 1 and 2, which present summary statistics on U.S. low-skilled foreign- and native-born individuals going back to 1970 using data from the U.S. census and ACS. In 1970, when the presence of the foreign born in the U.S. population was at a historic low, low-skilled immigrants, in comparison with the native born, were relatively old and likely to be female. This population came in its majority (52.9 percent) from Europe, was dominated by individuals who had arrived in the United States in 1960 or earlier (66.1 percent), and had a near majority (45.6 percent) with eight or fewer years of schooling. As the incipient immigration wave gained momentum, the composition of low-skilled immigrants became younger, more likely to have come from Latin America, and more educated. These changes were most dramatic between 1970 and 1990. During this period, the fraction of foreignborn people age 18–33 rose from 28.6 to 43.2 percent, the fraction of males rose from 41.8 to 48.8 percent, and the fraction completing 12 years of education rose from 34.8 to 45.0 percent. In terms of origin countries, among immigrants with a high school education or less, the fraction born in Mexico rose from 11.6 to 34.0 percent, the fraction born elsewhere in Latin America (and the Caribbean) rose from 13.2 to 23.7 percent, and the fraction born in Asia rose from 5.7 to 16.2 percent.5 The 1970 to 1990 increase in the shares of immigrants coming from Mexico and the rest of Latin 4. Education is, of course, an imperfect definition of skill. Language barriers and occupational licensing present obstacles to foreign-born workers in integrating themselves into the U.S. labor force, which may induce some immigrants to downgrade occupationally by taking jobs for which, based on their observable skills, they would appear overqualified (Lazear 1999, 2007; Dustmann, Frattini, and Preston 2013). 5. Half of the 1970–90 increase in low-skilled immigration from Asia (55.1 percent) is from Southeast Asia, with much of this inflow associated with a substantial but temporary rise in U.S. refugee admissions from the region that occurred following the end of the Vietnam War. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms 38.1 33.5 28.5 23.5 27.2 49.2 3.7 6.8 2.9 10.0 4.8 67.2 4.6 Age 18–33 34–49 50–64 Years of schooling 0–8 9–11 12 Industry, share of labor force Agriculture Construction Eating and drinking establishments Nondurable manufacturing Personal services Other industries Unemployment rate 3.2 6.1 5.8 14.5 7.0 58.7 4.6 45.6 19.6 34.8 28.6 34.9 36.5 41.8 75.8 Foreignborn 2.9 6.8 3.8 8.9 3.2 66.6 7.8 14.3 24.5 61.2 43.4 28.1 28.5 46.6 63.9 Nativeborn 4.0 5.7 6.8 13.0 5.8 56.8 7.8 39.2 19.0 41.7 40.2 33.0 26.8 44.4 69.1 Foreignborn 1980 2.8 8.0 4.8 7.6 3.3 65.2 8.4 8.6 20.8 70.6 41.6 31.7 26.6 48.6 59.1 Nativeborn 5.7 7.8 8.7 9.5 6.9 51.6 9.8 38.2 16.8 45.0 43.2 34.5 22.3 48.8 64.1 Foreignborn 1990 2.1 9.1 5.3 5.5 2.9 67.6 7.5 4.9 15.8 79.3 36.4 37.9 25.6 50.3 56.8 Nativeborn Sources: U.S. Census Bureau, decennial census, American Community Survey. a. All values are expressed as percentages. The sample is restricted to individuals age 18–64 with 12 years of education or less. 46.4 71.6 Sex Male Female Nativeborn 1970 5.5 10.8 9.2 7.5 6.3 52.1 8.6 33.7 16.5 49.8 42.2 38.2 19.6 51.2 63.9 Foreignborn 2000 2.3 9.2 7.6 4.3 2.9 63.9 9.7 4.0 12.6 83.4 38.2 27.9 33.8 54.3 46.2 Nativeborn Table 1. Characteristics of Native-Born and Foreign-Born U.S. Working-Age Population with High School Education or Less, 1970–2015a 6.9 14.8 11.3 5.8 7.2 47.9 6.1 29.6 15.1 55.3 27.3 42.3 30.5 51.2 60.5 Foreignborn 2015 �92 Brookings Papers on Economic Activity, Spring 2017 Table 2. Summary Statistics for Foreign-Born U.S. Working-Age Population with High School Education or Less, 1970–2015a 1970 1980 1990 2000 2005 2015 Years of residence in the United States 0–5 19.2 23.2 6–10 14.8 18.4 11+ 66.1 58.4 24.2 21.6 54.2 22.8 19.1 58.1 21.8 18.4 59.9 11.9 13.0 75.1 Age of arrival in the United States 0–14 22.1 13.6 15–25 28.0 34.8 26+ 44.0 51.4 14.7 42.4 43.0 19.1 47.5 33.4 18.1 47.3 34.3 20.2 45.0 34.4 Country or region of origin, less than high school education Mexico 15.4 33.3 47.5 60.6 Central America 1.6 3.5 8.7 10.8 Caribbean 8.4 10.7 9.2 7.2 South America 2.6 3.6 3.3 3.4 Southeast Asia 1.7 4.6 6.6 5.8 Other Asia 3.3 5.4 5.6 4.7 Africa 0.4 0.5 0.4 0.7 Middle East 1.6 1.5 1.2 0.9 Europe 51.6 25.9 11.4 5.0 Other 13.5 11.0 6.2 0.9 64.0 12.4 5.7 3.2 4.7 4.0 1.0 0.8 3.5 0.7 59.3 15.9 6.0 3.0 4.8 6.0 1.6 0.9 2.0 0.5 Country or region of origin, high school education or less Mexico 11.6 23.2 34.0 Central America 1.6 3.3 7.7 Caribbean 8.6 11.1 10.8 South America 3.0 4.4 5.3 Southeast Asia 1.7 4.9 7.5 Other Asia 4.0 7.1 8.7 Africa 0.5 0.8 0.9 Middle East 1.8 1.8 1.6 Europe 52.9 30.9 16.5 Other 14.4 12.5 7.2 48.1 10.1 8.8 6.1 6.9 7.0 2.2 1.4 8.0 1.6 45.1 12.2 9.4 5.7 6.9 8.9 3.0 1.6 5.8 1.3 44.4 9.3 9.9 5.8 7.8 7.7 1.7 1.4 10.2 1.8 Sources: U.S. Census Bureau, decennial census, American Community Survey. a. All values are expressed as percentages. The sample is restricted to individuals age 18–64 with 12 years of education or less. America is even larger among those with less than a high school education, rising from 15.4 to 47.5 percent and from 12.6 to 21.2 percent, respectively. By 1990, nearly 7 in 10 (68.7 percent) of the least-skilled U.S. immigrants of working age came from other nations in the Western Hemisphere. Jorge Durand, Douglas Massey, and Rene Zenteno (2001) describe this era of U.S. immigration as one marked by a preponderance of itinerant workers who came to the United States to take seasonal jobs, especially This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG McINTOSH 93 on farms in the Southwest, and often returned home during periods when labor demand was slack. During the two decades after 1970, the share of low-skilled immigrant workers employed in agriculture did rise, from 3.2 to 5.7 percent (as compared with a decline of 3.9 to 3.0 percent among the low-skilled, native-born workers of working age), and the fraction with 10 or fewer years of residence in the United States grew from 34.0 to 45.8 percent.6 However, throughout the sample period, farm workers accounted for only a small share of low-skilled immigrant employment. During the first decades of the late-20th-century immigration wave, lowskilled immigrants spread themselves across a wide range of jobs, while concentrating more heavily, when compared with their native-born counterparts, in agriculture, construction, eating and drinking establishments, nondurable manufacturing, and personal services. In subsequent decades, the low-skilled immigrant population has become more mature and more settled, at least when measured in terms of length of U.S. residence. By 2015, three-quarters (75.1 percent) of lowskilled immigrants had resided in the United States for 11 or more years, while the share of the population age 18–33 had dropped to 27.2 percent. Since 1990, the fraction of low-skilled immigrants from Mexico and the rest of Latin America has continued to rise, reaching 45.1 and 27.3 percent, respectively, in 2015. Among immigrants with less than a high school education, these shares are 59.3 and 24.9 percent, respectively, meaning that today, nearly 9 in 10 (85.2 percent) of the least-skilled, foreign-born workers are from Latin America and the Caribbean. It is particularly important that Mexico’s dominance as a source country for low-skilled immigrants peaks in 2005, at 48.1 percent of those with a high school education or less and 64.0 percent of those with less than a high school education. The 4.7 percentage point drop in Mexico’s share of the least-skilled immigrant population from 2005 to 2015 is largely offset by Central America’s jump during the same period of 3.5 percentage points. As we discuss in section III, demographic push factors help account for Mexico’s recent relative decline and Central America’s continuing relative gain as source regions. After Latin America, Asia remains the next most important source region of low-skilled immigration, in 2015 accounting for 15.8 percent of all low-skilled immigrants and 10.8 percent of those with less than 12 years of schooling. 6. The question for length of U.S. residence in the ACS reads, “When did this person come to live in the United States?” with the instruction, “If this person came to live in the United States more than once, print latest year.” This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �94 Brookings Papers on Economic Activity, Spring 2017 Over time, low-skilled immigrants have become more specialized in particular lines of work. The share employed in immigrant-intensive sectors in 2015 reached 14.8 percent in construction (up from 7.8 percent in 1990), 11.3 percent in eating and drinking establishments (up from 8.7 percent in 1990), 7.2 percent in personal services (up from 6.9 percent in 1990), and 6.9 percent in agriculture (up from 5.7 percent in 1990). The one immigrant-intensive sector registering a decline in its share of low-skilled immigrant employment is nondurable manufacturing—which includes apparel, footwear, furniture, and textiles—industries whose overall employment in the United States has fallen sharply in recent decades due to technological change and competition from China and other lowwage countries (Autor, Dorn, and Hanson 2013). The transition of the U.S. low-skilled immigrant population from sojourners to settlers, first noted by Wayne Cornelius (1992) nearly three decades ago, today appears to be largely complete. Part of this shift is the natural result of a dynamic process of immigration in which early arrivals initially dominate the population, only to decline in importance as the existing stock grows and matures (Piore 1980). However, the shift is also the result of the pronounced slowdown in low-skilled immigration since the mid-2000s, as seen in figure 1. Because the immigration levels portrayed in table 2 reflect changes in net immigration, they are uninformative about whether this slowdown is the result of reduced inflows of new immigrants, larger outflows of existing immigrants returning to their home countries, or some combination of the two. We next summarize evidence on changing inflows and outflows of immigrants over time. Figure 2 gives counts of immigrants by current age, age of arrival in the United States (inferred from years of U.S. residence), and census year for three source regions— Mexico, other countries in Latin America and the Caribbean, and Southeast Asia—which together account for the large majority of low-skilled immigration in the United States. To avoid concerns about tracking individuals who educate themselves out of the low-skilled category over time, we include all immigrants from these source regions, regardless of schooling. Several patterns are apparent in the data. First, for most current-age groups and in most census years, the largest cohorts are those arriving between 15 and 24 years of age. That is, for a given current-age group, if we compare bars that have the same shading (thus comparing different arrival-age cohorts in the same census year for the same current-age group), those in the 15–24 arrival-age category are the largest in nearly This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �95 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Figure 2. Number of Working-Age Immigrants by Arrival Age, Current Age, and Year of Arrival, 1980–2010 Mexico Thousands 1,500 1980 1990 2000 2010 1,000 500 5–14 15–24 5–14 15–24 15–24 25–34 5–14 25–34 15–24 25–34 35–44 35–44 Arrival age Current age Latin America, excluding Mexico Thousands 600 1980 1990 2000 2010 400 200 5–14 15–24 5–14 15–24 15–24 25–34 5–14 25–34 15–24 25–34 35–44 35–44 Arrival age Current age Southeast Asia Thousands 200 1980 1990 2000 2010 100 5–14 15–24 15–24 5–14 15–24 25–34 25–34 5–14 15–24 25–34 35–44 35–44 Arrival age Current age Sources: U.S. Census Bureau, decennial census, American Community Survey; authors’ calculations. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �96 Brookings Papers on Economic Activity, Spring 2017 all cases. Second, between 2000 and 2010, there are substantial declines in the sizes of given arrival-age or birth-year cohorts. For individuals from Mexico arriving in the United States between age 5 and 14, the number who are age 15–24 in 2000 is much larger than those who are age 25–34 in 2010. We see similar declines in the number of Mexican immigrants who are age 25–34 in 2000 and the number who are age 35–44 in 2010, both for the cohort arriving between age 5 and 14 and the cohort arriving between age 15 and 24. Similar patterns hold for immigrants from other countries in Latin America and from Southeast Asia. Declines in cohort size, as measured in the census, may result from mortality, return migration, or changes over time in the fraction of individuals in a cohort who are enumerated in the census. Given the youth of the cohorts considered, mortality seems unlikely to explain this decline. Moreover, given that we expect enumeration rates to increase with length of residence in the United States, declines in enumeration seem an unlikely explanation, which leaves return migration as the most likely cause for the decline in measured immigrant cohort sizes between 2000 and 2010. The net impact of these changes is that the size of immigrant cohorts in 2010 is skewed heavily toward individuals who have more than 10 years of residence in the United States. For immigrants from Mexico in 2010 (indicated by the darkest shaded bars), those with fewer than 10 years of U.S. residence are the smallest cohort among all current age groups, a pattern that holds for other countries in Latin America and for Southeast Asia as well. I.B. The Presence of Low-Skilled Immigrants in the U.S. Labor Force To consider how the presence of low-skilled immigrants in the U.S. labor force has changed in recent years, we focus on movements at annual frequencies using data from the CPS. Because the CPS only begins asking questions about nativity in 1994, our use of these data is for that year forward. We use two measures of the working-age population: raw data on body counts; and these values expressed in terms of productivity-equivalent units (PEUs), following the weighting procedure used by David Autor, Lawrence Katz, and Melissa Kearney (2008). Consistent with the post-1970 rise in low-skilled immigration seen in figure 1, figure 3 shows that the presence of low-skilled, foreign-born workers in the U.S. working-age population rose steadily from 1994 to 2007 but has been stable since. The left panel of figure 3 plots four measures of low-skilled immigration. The top line gives the share of the This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �97 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Figure 3. Percentage of Low-Skilled, Foreign-Born Workers in the U.S. Working-Age Population, 1993–2014 Percent weighted by productivityequivalent units Percent Raw percent Percent Foreign-born, high school or less 8 8 Mexican-born, high school or less 6 6 Foreign-born, less than high school 4 2 4 2 Mexican-born, less than high school 1995 2000 2005 Year 2010 1995 2000 2005 Year 2010 Sources: U.S. Census Bureau, Current Population Survey; authors’ calculations. foreign born with a high school education or less among all working-age individuals in the United States. This fraction rose from 6.5 percent in 1994 to 9.1 percent in 2007, before stabilizing in subsequent years, settling at 8.8 percent in 2015. Just under half these foreign-born individuals were born in Mexico (43.1 percent in 1994; 47.3 percent in 2015). When, alternatively, we define low-skilled immigrants more narrowly as those with less than 12 years of schooling, we also see a growing immigrant presence in the U.S. working-age population, rising from 3.6 percent in 1994 to 4.5 percent in 2007, and showing little change thereafter. Individuals born in Mexico account for a high fraction of the foreign-born population with less than a high school education (61.1 percent in 1993, 64.4 percent in 2014). Body counts of low-skilled immigrants overstate their presence in the U.S. labor force to the extent that these individuals have low labor productivity relative to the average U.S. person of working age. To measure the population in terms of PEUs, we apply the approach taken by Autor, Katz, and Kearney (2008), which involves reweighting individuals by their This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �98 Brookings Papers on Economic Activity, Spring 2017 projected relative earnings.7 Specifically, the weight attached to an individual is the ratio of the average weekly wage among full-time, full-year workers in his or her race, gender, education, and labor market experience cell to the average weekly wage for white, male, high school graduates with 8 to 12 years of potential work experience.8 Population shares expressed in terms of PEUs appear in the right panel of figure 3. These shares are naturally smaller than in the left panel, owing to the fact that low-skilled immigrant workers have low earnings relative to other U.S. workers. Using the productivity-adjusted measure, foreign-born individuals with 12 or fewer years of schooling reach 6.5 percent of the U.S. working-age population in 2007, a share that declines slightly to 6.3 percent in 2015.9 Low-skilled immigrants tend to have high rates of labor force participation and employment when compared with similarly skilled native-born workers (Borjas 2016). Consequently, the population shares shown in figure 3 may give an incomplete characterization of the presence of the lowskilled, foreign-born workers in the effective U.S. labor supply. Figure 4 reports the shares of low-skilled immigrants in total hours worked, both using raw hours (left panel) and productivity-adjusted hours (right panel). The share of total hours worked by immigrants with 12 or fewer years of schooling rose from 5.2 percent in 1994 to 8.4 percent in 2007, before falling modestly to 8.0 percent in 2015. When expressed in PEUs, these shares are 3.6, 5.8, and 5.5 percent, respectively. 7. When applying wage-based weights to the entire population, we assume that nonworking individuals would earn the same average wage as full-time workers who share their age, gender, race, education, and nativity profile. Because employment rates increase with potential earnings, this assumption may lead our productivity-adjusted shares of the lowskilled immigrant population to overstate the shares one would calculate based on “true” wage weights. This problem is partially ameliorated when we examine the share of lowskilled immigrants in total hours worked, as we do in figure 4. 8. We construct these weights as follows. First, we divide workers into labor market groups broken down by gender, two education categories (less than 12 years of education, exactly 12 years of education), and eight experience categories (0–4, 5–9, 10–14, 15–19, 20–20, 25–29, 30–34, and 35–39 years of potential labor market experience). Then, for each gender-education-experience group, we calculate the weight as average weekly earnings in each year (for full-time, full-year workers, defined to be those working at least 35 hours per week and 40 weeks a year) divided by average weekly earnings for white, male, high school graduates with 8 to 12 years of labor market experience. 9. The number of Mexican-born workers in the United States increased by more than 350,000 per year over the 20 years from 1980 to 2000. Negative net migration of 160,000 per year subsequent to 2007 therefore represents a drop of half a million people per year relative to the prior trend, enough to constitute a noticeable change in the foreign-born population when cumulated over a decade of low migration. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �99 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Figure 4. Percentage of Low-Skilled, Foreign-Born Workers in Total Hours Worked, 1993–2014 Percent weighted by productivityequivalent units Percent Raw percent Percent 8 Foreign-born, high school or less Mexican-born, high school or less 6 4 8 Foreign-born, less than high school 2 4 Mexican-born, less than high school 1995 2000 2005 Year 6 2010 2 1995 2000 2005 Year 2010 Sources: U.S. Census Bureau, Current Population Survey; authors’ calculations. Because undocumented immigrants account for a large share of the lowskilled immigrant population, and because the large majority of these individuals come from Mexico and Central America, low-skilled, foreign-born labor accounts for a relatively high fraction of employment in the states along the U.S. border with Mexico. Figure 5 plots the share of low-skilled immigrants in hours worked for the four U.S. border states (Arizona, California, New Mexico, and Texas), again in terms of both raw hours and productivity-adjusted hours. Among these border states, the share of foreign-born workers with 12 or fewer years of education in total hours worked rose from 11.9 percent in 1994 to 16.2 percent in 2005 and then dropped to 14.1 percent in 2015. Given the propensity of low-skilled immigrants to concentrate in particular sectors, it is not surprising that, in selected industries, they have come to account for a substantial fraction of total employment. As seen in table 3, in 2015 immigrants with 12 or fewer years of schooling account for 29.3 percent of total hours worked in agriculture (up from 3.9 percent in 1970), 21.8 percent in personal services (up from 6.4 percent in 1970), This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �100 Brookings Papers on Economic Activity, Spring 2017 Figure 5. Percentage of Low-Skilled, Foreign-Born Workers in Total Hours Worked for the U.S. Border States with Mexico, 1993–2014a Percent weighted by productivityequivalent units Percent Raw percent Percent 15 Foreign-born, high school or less 15 Mexican-born, high school or less 10 5 10 Mexican-born, less than high school 1995 2000 5 Foreign-born, less than high school 2005 Year 1995 2010 2000 2005 Year 2010 Sources: U.S. Census Bureau, Current Population Survey; authors’ calculations. a. The U.S. border states are Arizona, California, New Mexico, and Texas. Table 3. Percentage of Foreign-Born Workers with a High School Education or Less, by Industry, 1970–2015a Industry Agriculture Construction Eating and drinking establishments Nondurable manufacturing Personal services Other industries 1970 1980 1990 2000 2015 3.9 3.9 8.3 5.9 6.4 3.0 7.0 4.4 8.5 7.1 8.9 3.3 13.5 6.9 11.4 7.9 12.5 3.7 21.4 12.0 15.6 11.2 17.7 5.1 29.3 20.3 16.8 13.5 21.8 5.8 Sources: U.S. Census Bureau, decennial census, American Community Survey. a. All values are expressed as percentages. The sample is restricted to individuals with 12 years of education or less. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 101 20.3 percent in construction (up from 3.9 percent in 1970), 16.8 percent in eating and drinking establishments (up from 8.3 percent in 1970), and 13.5 percent in nondurable manufacturing (up from 5.9 percent in 1970), as compared with just 5.8 percent of employment in all other industries. For these immigrant-intensive industries, future changes in low-skilled immigration matter immensely. I.C. Who Chooses to Migrate to the United States? Is the increase in low-skilled immigration in the United States the result of increasing immigration from countries that are relatively abundant in low-skilled labor or the result of low-skilled labor being relatively likely to select into international migration? One cannot answer this question by examining U.S. data alone. Differences in educational attainment across countries would make the average worker from, say, Mexico appear to be low skilled in the U.S. labor market, whereas at home he or she would fall in the middle of the earnings distribution. In seminal research, George Borjas (1987) derived the conditions under which immigrants are negatively or positively selected in terms of skills. Conditions favoring negative selection—meaning that immigrants are drawn disproportionately from the bottom half of the skill distribution— are high returns to skills in the sending country relative to the receiving country, and migration costs that are proportional to worker productivity (for example, costs that have an iceberg form), which combine to give lessskilled workers a relatively strong incentive to migrate. Migration costs that are fixed in nature and a marginal utility of income that is not strongly decreasing favor positive selection of immigrants in terms of skills (Grogger and Hanson 2011), in which case immigrants are drawn more heavily from the top half of the skills distribution. Whether immigrants are negatively or positively selected in terms of skills matters for how labor movements affect the distribution of income in sending and receiving countries and for the ease with which immigrants from low-income countries integrate themselves into the labor markets of high-income countries. If, for example, immigrants from Mexico are negatively selected in terms of skills, shocks that contribute to a positive net flow of labor from Mexico to the United States would tend to decrease Mexican wage inequality—by reducing Mexico’s relative supply of lowwage workers—and to increase U.S. wage inequality—by expanding the U.S. relative supply of low-wage workers. Further, immigrants who are negatively selected in terms of skills may face greater challenges in This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �102 Brookings Papers on Economic Activity, Spring 2017 assimilating economically in the U.S. labor market and may be more likely to be a net drain on public resources (Borjas 2016). To examine the composition of low-skilled immigration in the United States from the sending country’s perspective, we focus on the case of Mexico, which is by far the largest source country for U.S. labor inflows, accounting for nearly half of all U.S. low-skilled immigrants and nearly two-thirds of those with less than 12 years of schooling. We extend forward in time the analysis of Daniel Chiquiar and Gordon Hanson (2005), which utilizes the methodology of John DiNardo, Nicole Fortin, and Thomas Lemieux (1996) for constructing counterfactual wage distributions.10 To examine differences in the distribution of skills between Mexican residents (that is, nonmigrants in Mexico) and Mexican immigrants, we compare the actual wage density in Mexico for Mexican residents with the counterfactual wage density that Mexican immigrants in the United States would obtain if they were paid according to Mexico’s prevailing wage structure. This comparison reveals from where in Mexico’s wage distribution migrants to the United States are drawn. Because this analysis projects U.S. immigrants onto Mexico’s wage distribution based on workers’ observable skills, it ignores the role of unobserved characteristics in migration and earnings. And because it takes Mexico’s current wage distribution as given, the analysis is silent about the equilibrium impact of immigration on U.S. or Mexican wages. Let f i(w x) be the density of wages w in country i, conditional on observed characteristics x, h(x i = MX) be the density of observed characteristics among workers in Mexico, and h(x i = US) be the density of observed characteristics among Mexican immigrants in the United States. The density of wages that would prevail for Mexican immigrants in the United States if they were to be paid according to the price of skills in Mexico is given by (1) MX ( w ) = ∫ f MX ( w x ) h ( x i = US ) dx . gUS This quantity corresponds to the counterfactual distribution of wages that arises from projecting the skill distribution of Mexican immigrants in the 10. Also on the selection of immigrants from Mexico in terms of observable skill, see Feliciano (2001), Orrenius and Zavodny (2005), Mckenzie and Rapoport (2007, 2011), and Akee (2010). This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 103 United States onto the current wage structure of Mexico. Although this distribution is unobserved, we can rewrite it as ( 2) MX ( w ) = ∫ θ f MX ( w x ) h ( x i = MX ) dx , g US where Mexico’s conditional wage distribution f MX(w x) and the skill distribution of its resident population h(x i = MX) are observed, and where ( 3) θ= h ( x i = US ) . h ( x i = MX ) Hence, we can obtain the counterfactual wage density that we desire in equation 1 simply by applying the appropriate weight q to the existing dis­ tribution of wages in Mexico. To compute this weight, we use Bayes’s theorem to write (4) h( x ) = h ( x i = US ) Pr ( i = US ) Pr ( i = US x ) and ( 5) h( x ) = h ( x i = MX ) Pr ( i = MX ) . Pr ( i = MX x ) Combining equations 4 and 5, we obtain an expression for q that is the ratio of the conditional probability that a Mexican-born worker resides in the United States, Pr(i = US x)/Pr(i = MX x), to the unconditional probability that a Mexican-born worker resides in the United States, Pr(i = US)/ Pr(i = MX). We estimate these probabilities via a logit model, use the estimates to calculate q, and apply the q weights to estimate the counterfactual wage density in equation 2.11 11. This method for constructing weights ignores differences in labor force participation rates in the two countries. Whereas labor force participation among male residents of Mexico and male Mexican immigrants in the United States are similar, labor force participation is higher among immigrant Mexican women than among nonmigrant Mexican women. Not accounting for these differences would tend to overstate negative selection among immigrants. See Chiquiar and Hanson (2005) for details and for methods to account for crossnational differences in labor force participation. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �104 Brookings Papers on Economic Activity, Spring 2017 We construct actual and counterfactual wage densities for males and females, separately, in the years 1990, 2000, and 2010. Earnings are annual labor income for individuals age 18–64. We estimate the logit regressions used to predict whether an individual born in Mexico resides in the United States separately for men and women as a function of education (seven categories, based on years of schooling: 0–4, 5–8, 9, 10–11, 12, 13–15, 16+) and age (46 categories, one for each year in the range 18–64). The population is all working-age individuals in Mexico and Mexican immigrants in the United States who have resided in the country for 10 or fewer years. Results are similar when we expand the analysis to include immigrants with 20 or fewer years of U.S. residence, who constitute the large majority of working-age Mexican immigrants in the United States. Wage densities are plotted relative to mean log earnings for workers in Mexico of a given gender in a given year, such that actual wage densities are centered on zero. Figure 6 presents the results, where in each plot the dashed line is the actual wage density for Mexico and the solid line is the counterfactual wage density in Mexico for current Mexican immigrants. For the case of males, shown in the left-side panels of figure 6, we see that in each year, the actual and counterfactual densities are very similar to each other, suggesting that the observable skills of Mexican immigrants match closely those of individuals who have not migrated abroad. In 1990, the counterfactual wage density lies slightly to the right of the actual wage density, indicating that Mexican immigrants are mildly positively selected in terms of observable skills. This difference is more defined in figure 7, which plots the difference between counterfactual and actual wage densities. In 1990, this difference, as seen in the top-left panel, has a negative mass just below zero and a positive mass just above zero, indicating that male immigrants are underrepresented among those who would earn slightly less than mean earnings in Mexico and overrepresented among those who would earn slightly more than mean earnings in Mexico. The slight rightward shift in the counterfactual relative to the actual wage density is also present in 2000 and 2010. However, the difference between actual and counterfactual densities becomes less pronounced over time, such that in the top-left panel of figure 7, the negative hump below zero and the positive hump above zero are smaller in 2000 than in 1990 and smaller still in 2010 relative to 1990. These changes are also seen in the top-right panel of figure 7, which reports the double difference in densities: counterfactual relative to actual wage densities in 2010 relative to this difference in either 1990 or 2000. The double difference using 2010 and 1990 is larger than that for 2010 and 2000, indicating a lessening of positive selection This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �105 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Figure 6. Actual Wage Density in Mexico and Counterfactual Wage Density for Mexican Immigrants in the United States Males, 1990 Females, 1990 Density 0.5 0.4 Density Mexican residents in Mexico Mexican immigrants in the United States 0.5 0.4 0.3 0.3 0.2 0.2 0.1 0.1 –0.4 –0.2 0 0.2 Log wage 0.4 –0.4 Males, 2000 Density 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 –0.2 0 0.2 Log wage 0 0.2 Log wage 0.4 Females, 2000 Density –0.4 –0.2 0.4 –0.4 Males, 2010 –0.2 0 0.2 Log wage 0.4 Females, 2010 Density Density 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 –0.4 –0.2 0 0.2 Log wage 0.4 –0.4 –0.2 0 0.2 Log wage 0.4 Sources: U.S. Census Bureau, decennial census, American Community Survey; Instituto Nacional de Estadística y Geografía, Mexican decennial census; authors’ calculations. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �106 Brookings Papers on Economic Activity, Spring 2017 Figure 7. Selection of Immigrants from Mexico in Terms of Observable Skills Migration selection evaluated at 1990 wage, males Density Double difference, males Density 0.1 0.1 2000 0.05 2010 minus 1990 0.05 2010 0 0 –0.05 –0.05 1990 –0.1 –4 –2 0 2 Log wage 4 –4 Migration selection evaluated at 1990 wage, females Density 0.1 0.1 0.05 0.05 0 0 –0.05 –0.05 –0.1 –0.1 –2 0 2 Log wage 4 –2 0 2 Log wage 4 Double difference, females Density –4 2010 minus 2000 –0.1 –4 –2 0 2 Log wage 4 Sources: U.S. Census Bureau, decennial census, American Community Survey; Instituto Nacional de Estadística y Geografía, Mexican decennial census; authors’ calculations. over time. By the time that we arrive in 2010, working-age Mexican immigrants who reside in the United States appear to be close to a random draw on the population of working-age individuals in Mexico. The three right-side panels of figure 6 repeat the analysis for women. Among women, we see evidence of stronger positive selection in 1990 and 2000 when compared with men. In each year, the rightward shift of the counterfactual wage density relative to the actual wage density is more pronounced than the corresponding density difference for males. As with males, the strength of positive selection diminishes over time, such that by 2010 the counterfactual and actual wage densities for women are very This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 107 similar. We conclude that by 2010, the selection of immigrants from Mexico is close to neutral in terms of observable skills. As mentioned above, these results are silent about the pattern of immigrant selection in terms of unobservables. One concern about the results shown in figures 6 and 7 is that we use census data to evaluate immigrant selection. Any undercount of the Mexicoborn population in either Mexico or the United States that depends systematically on an individual’s age or education could result in biased estimates either of the wage density for Mexico or of the counterfactual wage density that we construct for Mexican immigrants in the United States. There is a long-standing belief among demographers that the U.S. census undercounts undocumented immigrants in the United States (Warren and Passel 1987). To address this undercount issue, some studies evaluate immigrant selection using data exclusively from Mexico (Ibarraran and Lubotsky 2007). In noteworthy work, Jesús Fernández-Huertas Moraga (2011) uses data from Mexico’s national employment survey (Encuesta Nacional de Empleo, ENE), which follows households for five consecutive quarters and includes in the survey questions about whether household members have migrated to the United States during the period since the last survey was conducted. Distinct from Chiquiar and Hanson (2005), Fernández-Huertas Moraga (2011) finds that Mexican immigrants are negatively selected in terms of skills, as captured by residuals from Mincerian wage regressions. The ENE, however, has measurement problems of its own. It suffers from high rates of attrition by households from the sample within the five-quarter survey window, which a recent study by the National Academies of Science, Engineering, and Medicine concludes makes it problematic as a data source for evaluating Mexican migration to the United States (Carriquiry and Majmundar 2013). Fortunately, there is a source that provides longitudinal data on households in Mexico and that tracks information on individuals who migrate to the United States. The Mexican Family Life Survey (MxFLS) has been conducted in three waves—2002, 2006, and 2009—with a recontact rate of respondents between each wave of 90 percent. Of particular importance, the survey follows household members who migrate to the United States between waves. Robert Kaestner and Ofer Malamud (2014) use data from the first two MxFLS waves to analyze the selection of immigrants according to various measures of skills. Similar to what one sees in census data, migrants to the United States in the MxFLS are more likely to be young. In terms of education, both male and female migrants are more likely to have middle levels of schooling (4–9 years for men, 4–12 years for women) than This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �108 Brookings Papers on Economic Activity, Spring 2017 to have low levels of schooling (0–3 years). For men, but not for women, migrants are less likely to have very high levels of schooling (more than 12 years) than to have very low levels (0–3 years). The MxFLS also provides a measure of cognitive ability in the form of a Raven’s Progressive Matrices test score (Raven, Raven, and Court 2000). Although cognitive ability is a frequently discussed source of skill in the analysis of earnings (Heckman and Vytlacil 2001), few data sources provide evidence of how cognitive skills relate to migration decisions. Among both men and women, Kaestner and Malamud (2014) report no difference between migrants and nonmigrants in terms of their Raven scores, suggesting that the two populations have a similar distribution of observable cognitive abilities. Following Fernández-Huertas Moraga (2011), Kaestner and Malamud (2014) also examine migrant selection in terms of observable and unobservable characteristics using Mincerian wage regressions. Their analysis shows that workers with the highest predicted earnings or the highest residual earnings in the first MxFLS wave—meaning those among the top quintile of predicted or residual wage earners—are less likely to migrate but that there is no pattern of selection among lower-wage individuals. Despite problems with possible undercounts of undocumented migrants in census data, they provide a characterization of immigration selection that is comparable to that based on high-quality longitudinal micro data. Immigrants from Mexico to the United States are overrepresented among individuals whose skills place them in the middle of Mexico’s wage distribution and mildly underrepresented among individuals who would be very low-wage or very high-wage earners in their home country. When we examine U.S. immigration from other source countries, evidence of positive selection in terms of observable skills such as education is even more pronounced (Grogger and Hanson 2011). In nearly all source countries for U.S. labor inflows, immigrants are relatively likely to come from among the more educated.12 I.D. Summary The U.S. population of low-skilled immigrants has gone through an epochal half century of growth, transforming from a small cadre of older immigrants from Europe to a large population of immigrants from Latin America and Asia who are nearing middle age and who have now lived in 12. One exception to this pattern is Puerto Rico, which as an unincorporated territory of the United States is not subject to the same barriers to U.S. immigration as foreign nations (Borjas 2008). This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 109 the United States for an extended period of time. Immigrants from Mexico, who account for one-half to three-fifths of the low-skilled foreign born population, depending on the definition of skill, are preponderantly individuals who would be middle-income earners in their birth country. As the United States looks forward to an era of weakened incentives for lowskilled immigration due to changing labor demand and labor supply conditions at home and abroad, it will be shocks to middle-wage workers in migrant-sending countries that matter disproportionately for who migrates. Efforts to reduce the existing population of low-skilled immigrants, such as through increased deportations of undocumented immigrants, would target a population that appears to have a long tenure of residence in the United States. II. Labor Demand, Labor Supply, and Low-Skilled Immigration In this section and the next, we examine factors affecting the net flow of low-skilled immigrants into the United States. We begin in this section by describing recent changes in conditions surrounding low-skilled immigration, including income differences between the United States and major migrant-sending countries, U.S. immigration policy, and relative labor supply growth in the United States and major sending countries. We then analyze for the case of Mexico the contribution of labor demand and labor supply shocks to migration to the United States. II.A. Income Differences between Countries Perhaps the simplest manner in which to evaluate the incentive for immigration is to compare income between countries. Beginning with Larry Sjaastad (1962), economists have modeled immigration as an investment decision, in which the upfront cost of migration yields an income flow over time equal to the difference in earnings between the home and foreign economies. There may be considerable heterogeneity in the time horizon over which individuals consider migration (Dustmann 2003). Seasonal workers may focus on income differences between countries no more than a few months in advance, other individuals may be uncertain about their desire to relocate permanently and so put weight on the income differences they expect to be sustained over the next several years, and still others may treat migration as a long-term decision and therefore evaluate the expected discounted difference in income streams over their full working lives. To examine high-frequency changes in the incentive for immigration, we abstract away from such heterogeneity and consider point-in-time This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �110 Brookings Papers on Economic Activity, Spring 2017 income differences between the United States and migrant-sending countries, an approach taken in the large body of literature that uses the gravity model to analyze bilateral migration flows (Karemera, Oguledo, and Davis 2000; Clark, Hatton, and Williamson 2007; Bertoli and Fernández-Huertas Moraga 2013). Even in making point-in-time income comparisons, one faces many choices for how to measure income. One approach is to evaluate earnings for individuals with similar observable skills who were born in the same country and now live in different countries. Using data from the U.S. and Mexican population censuses, Hanson (2006) reports that in 2000 the average hourly wage for a 28- to 32-year-old male with 9 to 11 years of education is $2.40 in Mexico and $8.70 among recent Mexican immigrants in the United States (these income values, like those we report below, are adjusted for purchasing power parity in terms of 2000 dollars). At a labor supply of 35 hours per week and 48 weeks per year, this would yield an annual income gain of $10,600. Combining data from Mexico’s national survey of income and expenditures with data from the U.S. census, Michael Clemens, Claudio Montenegro, and Lant Pritchett (2008) obtain similar results, estimating that in 2000 the annual income gain to migration for a 35-year-old Mexican male with 9 to 12 years of education is $9,200. Comparing migrants with nonmigrants is problematic if there are unobserved characteristics that affect both the migration decision and an individual’s income-earning ability. An alternative approach is to use longitudinal data for the same individual, which allows comparisons of income before and after migration. Mark Rosenzweig (2007) uses data from the New Immigrant Survey to estimate the change in income for new U.S. permanent legal immigrants in 2003. He checks their current U.S. earnings against earnings in the last job they held in their country of origin. For a legal immigrant from Mexico with 9 to 12 years of education, the average gain in income is $15,900 (at 35 hours a week and 48 weeks a year). Comparing the same individual in two countries corrects for selection into the migration associated with unobserved, time-invariant individual characteristics but may introduce other complications. If, in preparing to migrate, individuals reduce their labor supply in a manner that diminishes income (or if negative shocks to income precipitate migration), this approach may overstate the income gains to migration.13 13. Since Rosenzweig (2007) examines legal immigrants, his figures are not directly comparable to those of Hanson (2006) or Clemens, Montenegro, and Pritchett (2008), whose samples include all immigrants. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 111 Evaluating how the incentive to migrate to the United States has changed across countries and over time is complicated by the fact that few countries produce annual household survey data, and census data are amassed infrequently. Our approach is to construct income differences between countries by combining annual data on average income from national accounts with data on the variance in income as inferred from summary statistics on income inequality. Although statistics on income inequality, such as the Gini coefficient, are often constructed at less than an annual frequency, they tend to change slowly from one year to the next (Solt 2016), which permits interpolation of their values to create an annual series. Under the assumption that income is log-normally distributed across households, which is approximately consistent with data for many countries (Pinkovskiy and Sala-i-Martin 2009), one can use the Gini coefficient to calculate the variance of income across individuals and then combine this value with average income to construct income at different percentiles of the distribution (Grogger and Hanson 2011).14 Given the neutral selection of immigrants from Mexico in terms of observable skills, the 50th percentile (equal to $8,800 in 2000) is a natural choice for the reference income of a prospective Mexican migrant. To select the reference income in the United States for a typical immigrant from Mexico, we choose the percentile of the U.S. income distribution that yields an income gain to migration in 2000 that is approximately equal to the average income gain for migrants as described by Hanson (2006); Clemens, Montenegro, and Pritchett (2008); and Rosenzweig (2007). The 25th percentile of the U.S. income distribution ($20,100 in 2000) serves this purpose. In the left panel of figure 8, we report the ratio of the 50th percentile of the Mexican income distribution to the 25th percentile of the U.S. income distribution, where we construct these values using Gini coefficients from UNU-WIDER’s World Income Inequality Database, and per capita GDP, adjusted for purchasing power parity, from the World Bank’s World Development Indicators.15 This ratio is stable in the 1990s and early 2000s, averaging 0.44 between 1990 and 2007. During this period, 2 14. Suppose log income is normally distributed with mean µ and variance – s . Given an estimate of the Gini coefficient G, the standard deviation of log income is s = √2F-1([G + 1]/2). The value of log income at the a quantile is then µe(sza-s2/2), where za is the ath percentile of N (0,1). 15. Because Gini coefficients are not available in all years, we interpolate values for missing years. The series on Gini coefficients ends in 2012 in some countries and in 2013 in others. We assume that Gini coefficients in later years equal those in the last year for which data are available. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �112 Brookings Papers on Economic Activity, Spring 2017 Figure 8. The Ratio of the 50th Percentile of Income in the Sending Country to the 25th Percentile of Income in the United States, 1990–2015 Mexico Latin America, excluding Mexicoa 0.55 0.35 0.5 0.3 0.25 0.45 0.2 1995 2000 2005 Year 2010 1995 2000 2005 Year 2010 Sources: UNU-WIDER, World Income Inequality Database; World Bank, World Development Indicators; authors’ calculations. a. The Latin American countries considered are Colombia, the Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, and Jamaica. a middle-income earner in Mexico who chooses to become a low-income earner in the United States would see his or her real earnings increase by a factor of 2.3. After the Great Recession, the U.S.–Mexico income difference compresses, with the ratio of the 50th percentile of Mexican income to the 25th percentile of U.S. income rising to an average of 0.53 between 2008 and 2015 and to 0.58 during the later period of 2011 to 2015. In the right panel of figure 8, we report the corresponding ratio of the 50th percentile of the sending country’s income to 25th percentile of U.S. income for a composite of other countries in Latin America and the Caribbean. We choose the next-largest sending countries for which data on Gini coefficients are available—Colombia, the Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, and Jamaica—and we weight each country’s income by its relative share of working-age, low-skilled immigrants in the United States in 2000.16 The time path of relative income is similar 16. In 2000, the share of U.S. low-skilled, working-age immigrants accounted for by these countries is 15.0 percent (4.1 percent for El Salvador, 2.8 percent for the Dominican Republic, 2.3 percent for Guatemala, 1.7 percent for Jamaica, 1.6 percent for Colombia, 1.3 percent for Honduras, and 1.1 percent for Ecuador). Gini coefficients are unavailable for Cuba and Haiti (2.5 and 1.5 percent of U.S. low-skilled immigrants in 2000, respectively), leading us to leave them out of figures 8 and 9. Significant sending nations for U.S. low-skilled immigrants outside the Western Hemisphere (and their shares of this population in 2000) include Vietnam (4.1 percent), the Philippines (2.2 percent), China (2.1 percent), South Korea (1.5 percent), Germany (1.4 percent), Italy (1.2 percent), Canada (1.2 percent), India (1.2 percent), and Poland (1.1). This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �113 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Figure 9. Volatility of GDP Growth, 1993:Q1–2016:Q1a Mexico Standard deviations of GDP growth Latin America, excluding Mexicob Standard deviations of GDP growth 12 5 9 6 Mexico United States 3 Latin America 4 3 2 1 1995 2000 2005 2010 2015 Year United States 2004 2008 2012 2016 Year Sources: World Bank, World Development Indicators; authors’ calculations. a. GDP growth volatility is shown for rolling eight-quarter windows. Tick marks on the horizontal axes indicate quarter 1. b. The Latin American countries considered are Colombia, the Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, and Jamaica. to that for the U.S.–Mexico comparison, though the absolute income gap is larger. The income ratio is stable from 1990 to 2007, averaging 0.22, and then rises after the onset of the Great Recession, averaging 0.30 from 2008 to 2015. Since 2007, relatively slow U.S. income growth and rapid growth in neighboring countries have compressed the income gap between the United States and migrant-sending nations, presumably weakening incentives for immigration. In forming expectations about future income differences between countries, prospective migrants are likely to consider not just the level of income but also its variance. Over short time horizons, higher perceived variance in income in the sending country relative to the receiving country may add to the incentive for migration. At a monthly frequency, changes in attempted undocumented migration from Mexico to the United States, as captured by apprehensions at the U.S.–Mexico border, are strongly sensitive to changes in the dollar–peso real exchange rate, with attempted entry surging during periods following currency crises in Mexico (Hanson and Spilimbergo 1999; Monras 2015). When expanding data to include countries throughout the Western Hemisphere, emigration rates to the United States are larger for cohorts subject to a higher incidence of financial crises in their home country (Hanson and McIntosh 2012). To characterize changes in U.S. income volatility relative to that in migrant-sending economies, the left panel of figure 9 reports the standard This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �114 Brookings Papers on Economic Activity, Spring 2017 deviation in quarterly real GDP growth in Mexico and the United States for rolling eight-quarter windows covering the period 1991:Q1–2016:Q1. Throughout this time span, volatility in GDP growth is higher in Mexico (average eight-quarter standard deviation of 3.7 percent) than in the United States (average eight-quarter standard deviation of 2.0 percent). However, there are evident changes in relative volatility over time. After the 1995 peso crisis, volatility spiked in Mexico while remaining low in the United States. During the ensuing 10 years, volatility remained uniformly higher in Mexico, though well below the elevated levels of crisis periods. With the onset of the global financial crisis of 2008–10, volatility jumps in both economies, declining thereafter to roughly equal levels. Reduced differences between Mexico and the United States in income volatility reflect the improved execution of monetary and fiscal policies in Mexico—which, as in much of Latin America, has helped lower inflation, reduce government debt, and stabilize GDP growth (Edwards 2009). In the right panel of figure 9, we compare volatility in quarter-to-quarter GDP growth in the United States with the same migrant-sending countries examined with regard to relative GDP levels in figure 8. Here again, we see that volatility in GDP growth in migrant-sending nations has decreased relative to the United States, which has presumably dampened pressures for cross-border labor flows. II.B. U.S. Immigration Policy Low-skilled immigrants enter the United States through three channels: with permanent legal residence visas (green cards), with temporary work visas, and as undocumented entrants. During the 2000s, the U.S. government engaged in a massive buildup in border enforcement efforts, with most newly committed resources allocated to the U.S. border with Mexico. To understand how changes in immigration policy may have affected incentives for low-skilled immigration, we review recent adjustments in U.S. policy mechanisms. LEGAL IMMIGRATION The vast majority of low-skilled immigrants who obtain green cards do so through family sponsorship, for which visa eligibility derives from having a relative who is a U.S. citizen or legal resident, or as refugees or asylees (Rosenzweig 2007). The number of U.S. green cards and the policies governing their allocation have been stable since 1990. In that year, the Immigration Act set the annual number of familysponsored visas at 480,000, the annual number of employer-sponsored visas (which go primarily to skilled workers) at 140,000, and the annual number of diversity visas (allocated via lottery to countries that have historically low migration to the United States) at 55,000. Visas available to immediate This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 115 relatives of U.S. citizens are uncapped, though applications for these visas may be subject to processing delays. The number of green cards given to refugees and asylees, though having no set cap, shows no trend over time, having fallen from 114,000 per year in the 1990s to 83,000 per year in the 2000s, before rising to 109,000 per year during the 2010–15 period.17 Any increase in low-skilled immigration via permanent legal visas thus cannot have occurred through expanded quotas for green cards. It must instead have occurred through increases in the number of low-skilled immigrants qualifying for, applying for, and receiving visas from the annual allocation of visas.18 Qualifying for a green card under family sponsorship requires having an immediate relative who is a U.S. citizen—which gives one access to visas that are not subject to numerical limit—or a more distant relative who is a U.S. citizen or legal resident—which allows one to apply for the fixed annual allocation of green cards. Because the number of new applications exceeds the annual cap on green cards, and has for many years, there is often a substantial lag between the time of application and the time of visa receipt, with waiting times of several years in length being common. The waiting time depends in part on one’s visa preference category, which is a function of how closely related one is to a U.S. resident, and in part on the number of green card applicants from an individual’s country of birth who are higher up in the visa queue. Family sponsorship for green cards makes immigration a self-reinforcing process. As the number of permanent legal immigrants from a sending country increases, so too does the number of residents of that country who are eligible for a green card. For example, permanent visas awarded to residents of Mexico rose from 64,000 per year in the 1970s to 166,000 per year in the 1980s and to 225,000 per year in the 1990s, before dropping to 169,000 per year since 2000.19 Visa growth from the 1970s to the 17. A refugee is a foreign resident who is unable or unwilling to remain in his or her country of nationality because of fear of persecution based on race, religion, social group, or political opinion; an asylee is a foreign national who meets the conditions of a refugee and is already in the United States. A refugee is eligible to apply for a green card after one year of U.S. residence. At the beginning of each fiscal year, the president, in consultation with Congress, sets a worldwide ceiling on refugee admissions. 18. All data on legal immigration are from the U.S. Department of Homeland Security’s Yearbook of Immigration Statistics (https://www.dhs.gov/immigration-statistics). 19. Regarding admission of permanent legal residents from nations in the rest of Latin America and the Caribbean, green cards issued have risen from 50,000 per year in the 1970s, to 180,000 per year in the 1980s, to 205,000 per year in the 1990s, and to 250,000 per year since 2000. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �116 Brookings Papers on Economic Activity, Spring 2017 2000s partly reflects the growing population of Mexican residents who have family members who are legal U.S. residents, which has expanded the pool of eligible green card applicants. However, idiosyncratic changes in immigration policy are also at work. The 1990s blip in green cards awarded to residents of Mexico was partly a result of the legalization of undocumented immigrants under the Immigration Reform and Control Act of 1986, which delivered green cards to undocumented residents who met eligibility requirements on a one-time basis. The recent slowdown in low-skilled immigration is evident in green card allocations. Green cards awarded to Mexican residents declined from 175,000 per year during the 2001–05 period to 140,000 per year during the 2011–15 period. Because more residents of Mexico were eligible for green cards in 2010 than in 2000, the slowdown in green cards issued must be due to a decrease in demand for U.S. visas, which may be due to improved economic conditions in Mexico relative to the United States. Another form of legal immigration available to low-skilled, foreignborn workers is a temporary work visa. These visas permit a non-U.S. resident to work in the United States for a period of less than one year. The H-2A program provides work permits to agricultural workers, while the H-2B program gives work permits to nonagricultural workers, often for seasonal jobs in construction or tourism. The number of H-2 visas has risen over time—H-2A visas from 46,000 in 2006 to 284,000 in 2015, and H-2B visas from 97,000 in 2006 to 120,000 in 2015. However, because these visas permit stays of less than one year and are nonrenewable, they account for no more than a small share (less than 3 percent) of the over 17 million low-skilled, working-age immigrants who resided in the United States as of the mid-2010s.20 UNDOCUMENTED IMMIGRATION The most significant recent changes in the United States’ policy governing low-skilled immigration pertain to how the country monitors and enforces its borders and ports of entry. Undocumented immigrants gain entry to the United States either by overstaying their legal immigration visas or by crossing a U.S. border or entry point illegally.21 The United States has substantially expanded the resources it devotes to 20. That is, if a current H-2 visa holder desires to return on an H-2 visa in the following year, he or she must return to his or her country of residence and seek admission out of the following year’s visa allocation. 21. As of the mid-2000s, approximately 45 percent of undocumented immigrants in the United States appeared to be visa overstayers (many of whom do not remain in the United States in the longer term). See Pew Research Center (2006) and U.S. Department of Homeland Security (2016b) for estimates of annual overstay rates by country. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �117 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Figure 10. Number of U.S. Border Patrol Agents, 1992–2016 Nationwide 20,000 Southwestern states 15,000 10,000 5,000 1996 2000 2004 Year 2008 2012 Source: U.S. Department of Homeland Security. preventing undocumented labor inflows (Roberts, Alden, and Whitley 2013). Figure 10 plots the number of U.S. Border Patrol agents stationed at the U.S.–Mexico border and other entry points. In 2016, 85.9 percent of agents were stationed in the Southwest, a share similar to that in 1992. The expansion in personnel at the border—which increased by a factor of 4.8 by 2016—encompasses only part of the buildup. There have also been substantial investments in infrastructure at the border and changes in how those caught attempting undocumented entry are treated. To comprehend the dimensions of these changes, consider how the environment along the San Diego–Tijuana segment of the U.S.–Mexico border today compares with that in 1992, before the modern enforcement buildup began. In 1992, there were 1,009 Border Patrol agents assigned to the San Diego region, which stretches from the Pacific Ocean for about 60 miles east, among the 3,555 agents stationed along the entire U.S.–Mexico border. Barriers at the border itself were insubstantial, consisting in many areas, including those adjacent to the heart of urban Tijuana, of no more than a chain-link fence, in which large holes were frequently cut. In 1992, the Border Patrol apprehended 545,000 individuals in the San Diego sector, representing 542 apprehensions per agent. Across the entire U.S.–Mexico border, there were 1,134,000 This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �118 Brookings Papers on Economic Activity, Spring 2017 apprehensions, representing 319 apprehensions per agent. Agents spent much of their time chasing down migrants as they attempted to run into the United States and find cover in San Diego neighborhoods. More than 95 percent of those apprehended were Mexican nationals, and nearly all were subject to “voluntary removal,” under which they face no legal sanction for being apprehended. After capture, most were bused across the nearest border crossing, leaving them free to attempt entry again soon thereafter (Hanson 2007). Thus, as of the early 1990s, the U.S.–Mexico border was porous, the enforcement presence was unsophisticated and lightly resourced, and sanctions against migrants attempting illegal entry were weak. Today, the San Diego–Tijuana border, as with much of the U.S.–Mexico border, is a much different place. The number of Border Patrol officers in San Diego has grown to 2,325, among the 17,026 stationed along the entire border. San Diego and Tijuana are now separated by multiple layers of border barriers, which include rows of closely spaced, vertically mounted steel beams that reach 18 feet in height. These barriers constitute part of the 650 miles of fencing along the U.S.–Mexico border, 600 miles of which were constructed between 2006 and 2010 (Roberts, Alden, and Whitley 2013), which cover nearly all the U.S–Mexico border that does not coincide with the Rio Grande, a river that spans the near entirety of Texas’s border with Mexico. The San Diego–Tijuana border is patrolled by Border Patrol agents in SUVs, who traverse groomed roads constructed between each layer of border fencing, with manned and unmanned aircraft surveilling from above. Night-vision-capable video cameras posted every few hundred yards provide a continuous feed to Border Patrol stations nearby. In 2015, apprehensions in the San Diego sector were down to 26,000 (11 apprehensions per agent) and down to 337,000 for the U.S.–Mexico border as a whole (29 apprehensions per agent). Whereas, in the past, the Border Patrol spent much of its time physically apprehending migrants, today its job is to serve as a deterrent force against those who would consider illegal entry. In fiscal year 2017, the U.S. Department of Homeland Security spent an estimated $7 billion on salaries and benefits for Border Patrol agents and Customs and Border Protection officers (whose employment numbers are roughly equal); $3.6 billion on Coast Guard efforts to maintain the security of U.S. ports, waterways, and coastal areas; $2.9 billion on the detention and removal of deportable aliens; and $410 million to maintain infrastructure and purchase communications equipment related to border security (U.S. Department of Homeland Security 2016a). This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 119 Sanctions against undocumented immigration have also changed. The era of voluntary removal is over, replaced by a Consequence Delivery System (Argueta 2016). The disposition of those apprehended is conditional on their previous crossing activity and other circumstances. Since 2000, nearly all those apprehended at the border (meaning within 100 miles of a border and 14 days of entering the United States) have been fingerprinted and recorded in a digital database. Consequences depend on whether the apprehension is the first ever or a repeat event. Since the early 2010s, the large majority of those apprehended (more than 85 percent) have been subject at a minimum to “expedited removal” (or “reinstatement of removal,” if they have been removed before), which is a formal and immediate removal order that carries the considerable penalty of making the individual ineligible for legal U.S. immigration during the subsequent 10 years (enforceable via an individual’s fingerprint record). Those with multiple prior apprehensions may be subject to a “warrant of arrest” and misdemeanor prosecution. Roughly one-third of those deported are now repatriated to a port of entry far from their attempted crossing point, which disrupts smuggling operations in which individuals pay smugglers for multiple attempts to cross the border (as a hedge against the risk of apprehension).22 Since the enactment of the Consequence Delivery System, recidivism rates have dropped. During the 2005–07 period, 25 to 30 percent of those apprehended were caught within the same year. Recidivism began to decline in 2009, when the consequence program was rolled out, and in 2015 it stood at 15 percent. The intensity of immigration enforcement has also increased in the U.S. interior.23 U.S. Immigration and Customs Enforcement (ICE) is the government agency tasked with locating and removing “deportable aliens” in the U.S. interior, meaning all immigrants whose criminal activities— whether related to immigration or nonimmigration infractions—warrant deportation. By working more closely with local law enforcement agencies, ICE agents have expanded the deportations of individuals accused of minor infractions, including those driving without a license or driving 22. The Alien Transfer Exit Program repatriates Mexican nationals through geographic areas different from their attempted point of entry (Argueta 2016). 23. Changes in interior enforcement are important, in light of the fact that about twofifths of undocumented immigrants may have entered the country on legal visas, which they subsequently overstayed (Passel and Cohn 2016). By increasing border and interior enforcement simultaneously, the Department of Homeland Security may reduce incentives for border crossers to become visa overstayers. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �120 Brookings Papers on Economic Activity, Spring 2017 under the influence of alcohol or other substances (Thompson and Cohen 2014). These changes in part account for the increase in the deportations of noncriminal aliens (that is, those whose nonimmigration crimes alone do not warrant deportation) from 112,000 per year during the first three years of the George W. Bush administration (2001–03) to 223,000 per year during the first three years of the Barack Obama administration (2009–11). Deportations of criminal aliens—those whose nonimmigration crimes do warrant deportation—has also increased, from 77,000 per year in 2001–03 to 164,000 per year in 2009–11, which may reflect a combination of an expanding population of criminal aliens and increased efforts by ICE to locate and remove these individuals when they finish their prison terms. The Department of Homeland Security’s ICE budget in fiscal year 2017 was $6.2 billion (U.S. Department of Homeland Security 2016a). What have these changes in policy meant for undocumented immigration? One indication of the impact of border enforcement is movements in the price for smuggling services. Most of those attempting to cross the U.S.–Mexico border hire a smuggler, known as a coyote, to serve as a guide through the desert and mountain regions of Arizona and Texas, where most undocumented immigrants now attempt to cross the border. Measures of coyote prices are available from the Border Patrol, which asks a subset of those apprehended whether they hired a coyote and the price paid; from the Mexican Migration Project (MMP), which surveys individuals in Mexico about their previous border-crossing experiences; and from the Survey of Migration on the Northern Border of Mexico (Encuesta sobre Migracíon en la Frontera Norte de México, EMIF Norte), which surveys migrants returning from the United States at bus stations and other transportation points in Mexico.24 None of these sources are free of measurement problems. Border Patrol data are only available from those apprehended and questioned about their behavior, while the MMP and EMIF Norte are based on the selected sample of migrants who have returned to Mexico. Although average coyote prices differ across these sources, their time trends are similar (Roberts and others 2010). Border Patrol data show smuggler prices rising from $1,000 in 1999 to $1,600 in 2008 (in 2007 dollars). Using data from the MMP, Christina Gathmann (2008) estimates that a 24. For the methodology of the Mexican Migration Project, see http://mmp.opr.princeton. edu/research/design-en.aspx; and for EMIF Norte, see http://www.colef.mx/emif/eng/bases_ metodologicas.php. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 121 10 percent borderwide increase in enforcement (measured in man-hours) increases the average smuggler price by 4.9 percent. Using this elasticity, the increase in Border Patrol manpower on the U.S.–Mexico border from 2007 to 2015 of 27.6 percent would have increased smuggler prices by 13.6 percent. The cumulative effect of the U.S. enforcement buildup is hundreds of miles of new fencing, the rollout of technologically sophisticated border surveillance, a near quintupling of Border Patrol agents since the early 1990s, and the criminalization of illegal border crossings since the late 2000s. These changes combine with the recent compression in income differences between the United States and major sending nations to weaken incentives for low-skilled labor inflows. II.C. Demographic Pressures for U.S. Immigration In the 1970s, 1980s, and 1990s, macroeconomic shocks and relatively low incomes in Mexico and the rest of Latin America helped trigger labor flows to the United States. What sustained these flows over time was rapid growth in the relative labor supplies of these countries (Hanson and McIntosh 2012). Whereas the U.S. baby boom came to a halt in the early 1960s, Latin America’s baby boom did not abate until two decades later. Differences in the timing of the U.S. and Latin American demographic transitions mean that though the sizes of U.S. cohorts coming of working age began to slow in the early 1980s, they kept growing in Latin America until the 2000s, fueling pressures for emigration. The relationship between labor supply growth and changes in U.S. migration appears in figure 11, which charts, for countries in Latin America and the Caribbean, the percent change in migration rates to the United States from 1980 to 2015 against the percent change in national birth cohort sizes over this interval. A strong positive relationship is evident, with the R2 on the population-weighted linear fit equal to 0.45. As noted, the most important origin countries in terms of absolute number of current migrants residing in the United States are Mexico, El Salvador, Guatemala, the Dominican Republic, and Honduras. Mexico stands out in this group, with roughly 10 times the number of U.S. immigrants in the 15–40 age group as the next largest origin country in Latin America (figure 12). Mexico also stands out in terms of having the largest drop in migration between 2010 and 2015. Although most countries in Latin America see some decrease in the number of immigrants in this age group between 2010 and 2015, the number of Mexican-born individuals age This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �122 Brookings Papers on Economic Activity, Spring 2017 Figure 11. U.S. Labor Supply Growth versus U.S. Migration Changes from Latin America and the Caribbean, 1980–2015 Percent change in migration rate Honduras 1,250 Guatemala 1,000 El Salvador 750 500 Venezuela Dominican Antigua and Peru Mexico Trinidad and Republic Barbuda Haiti Tobago Nicaragua 250 Grenada Bolivia Paraguay Ecuador Guyana Bahamas Jamaica Colombia Belize Costa Rica Chile 0 Barbados Panama Uruguay –25 0 25 50 75 100 Percent change in birth cohort size 125 150 Sources: U.S. Census Bureau, American Community Survey; United Nations, World Population Prospects. Figure 12. Migration Rates and Migration Counts for Age 15–40, 1980–2015 Migration rates Migration counts Percent Million El Salvador 6 5 15 Mexico 10 4 3 Guatemala 5 Honduras 1990 2000 Year Dominican Republic 2 1 2010 1990 Source: U.S. Census Bureau, American Community Survey. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms 2000 Year 2010 �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 123 15–40 residing in the United States fell by more than 1.1 million over these five years.25 The durability of the decrease in migration from Mexico to the United States depends strongly on the reason it occurred. If the slowdown is due primarily to earlier shifts in population growth (Hatton and Williamson 2011), then given projected near constant U.S.–Mexico labor supply ratios, we may expect large-scale Mexican emigration to be a thing of the past. If the recent border enforcement buildup plays a significant role in the immigration slowdown (Gathmann 2008; Amuedo-Dorantes, Puttitanun, and Martinez-Donate 2013), then the pace of immigration from Mexico may be effectively in the control of the U.S. government.26 Alternatively, if labor demand shocks are primarily responsible for the immigration slowdown (Villarreal 2014), the hiatus in high levels of immigration may end once the U.S. economy recuperates more fully. To understand the causes of the decline in Mexican migration to the United States, we turn to data from the Mexican population census. We exploit variation in labor supply and per capita GDP across Mexican states to explain emigration. This analysis updates the work of Hanson and Craig McIntosh (2010) to the 2000–10 period.27 We count the base size of Mexican state-age-gender birth cohorts when they are first seen in the data, and use successive censuses to count the number of individuals remaining in Mexico each year. Because more than 95 percent of emigrants from Mexico go to the United States (Passel and Cohn 2009), and because we study young cohorts in which mortality is low, these numbers provide a 25. Our data indicate that the total number of Mexican-born individuals in the United States of all ages fell by 272,000 from 2010 to 2015, roughly in line with the estimate from Gonzalez-Barrera (2015) using data from the Mexican National Survey of Demographic Dynamics (Encuesta Nacional de la Dinámica Demográfica, ENADID) that the United States lost 141,000 Mexican-born individuals from 2009 to 2014. This muted change relative to the population age 15–40 indicates both that the young are more sensitive to changes in conditions than the old, and foreshadows the results later in this section that the population of older Mexicans will continue to grow even once the number of younger individuals starts to fall (Giorguli-Saucedo, García-Guerrero, and Masferrer 2016). Due to how mortality increases with age, our measure of net migration (differences between migrant numbers and birth cohort size) becomes less reliable as cohorts become older. 26. Other factors driving migration from Mexico to the United States in recent decades include Mexican policy reforms in the 1990s that privatized land rights, which allowed rural residents to sell their land, or to leave for urban areas without fear of relinquishing their claim to communal land (de Janvry and others 2015). 27. Although the 2015 Mexican micro census Conteo de Población y Vivienda should permit a similar exercise to be conducted, we found the resulting population estimates to be too noisy to use. Hence, the analysis uses data only through 2010. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �124 Brookings Papers on Economic Activity, Spring 2017 usable estimate of net emigration to the United States from each Mexican state. To form our dependent variable, we aggregate individuals into threeyear birth cohorts, and then calculate decadal changes in the percentage of the cohort that has emigrated, as measured by the change in its size. To investigate the role of labor supply in emigration, we include as a regressor the log ratio of the Mexican state birth cohort size to the U.S. birth cohort size. We restrict the analysis to individuals age 15–40, which is the age range during which most migration occurs. The regression specification is (6) N   GDPic16  dmicgt = β 0 + γ 1 log  icgt  + γ 2 log  16   GDPUSc   NUSgt   GDPitC  + γ 3 log  + α i + µ g + ηc + ρt + ε icgt ,  GDPUStC  where dmicgt is the change in the emigration rate for Mexican state i, birth cohort c, gender g, and census year t. To focus on low-skilled immigration, we take the log of the ratio of the Mexican state birth cohort size Nicgt to the current U.S. native-born population with less than a high school education NUSgt. Because we may be concerned that education decisions among U.S. natives are endogenous to Mexican immigration rates, we instrument for this ratio using the log ratio of the Mexican birth cohort to the entire respective U.S. birth cohort. Labor demand shocks are captured by the log ratio of GDP per capita in a Mexican state to the United States in the year that a cohort was age 16, GDP16ic /GDP16 USc, as well as the log ratio of contemporary GDP per capita in a Mexican state to the United States, GDPitC / GDPCUSt. We select age 16 because it is a common year for entry into the labor market in Mexico; relative income in this year indicates prevailing economic conditions at a time when individuals first make choices about labor supply. To express labor demand factors in terms of deviations from trend changes in economic activity, we use residuals from a regression of the log ratio of Mexican state GDP per capita to U.S. GDP per capita on state-specific intercept and slope terms. To control for confounding shocks on migration, we include fixed effects for the Mexican state (ai), gender (µg), birth cohort (hc), and census wave (rt). The regression specification in equation 6 is motivated by the migration model of Borjas (2006), in which differences in relative labor supply combine with shocks to labor demand to create the incentive for the movement of labor between economies. Adjustment costs prevent This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �125 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Table 4. Analysis of Net Migration at the Mexican State Levela Decadal change in net migrationb Independent variablec Labor supply with less than a high school educationd Innovations to GDP per capita at age 16 Innovations to GDP per capita in census year No. of observations R2 (1) (2) (3) (4) Pooled Excluding 2010 Men Women 0.1441*** (0.010) -0.0197* (0.010) 0.022 (0.020) 3,328 0.122 0.1643*** (0.011) -0.0735*** (0.016) -0.0472 (0.039) 2,432 0.175 0.1733*** (0.015) -0.0262 (0.017) 0.0234 (0.034) 1,664 0.108 0.1190*** (0.013) -0.0134 (0.012) 0.0187 (0.024) 1,664 0.189 Sources: U.S. Census Bureau, decennial census; Instituto Nacional de Estadística y Geografía, Mexican decennial census; authors’ calculations. a. The units of analysis are three-year birth cohorts. The log of the ratio for Mexican state birth cohort size to the U.S. birth cohort size is used as an instrument for the log of the ratio for Mexican state birth cohort size to the U.S. labor supply with less than a high school education. Robust standard errors in parentheses are clustered by birth cohort and weighted by birth cohort size. Statistical significance is indicated at the ***1 percent, **5 percent, and *10 percent levels. b. The mean decadal change in net migration for the pooled sample is 0.0619, holding the Mexican state cohort size fixed at its initial value. c. Each independent variable is the log of the ratio for the Mexican state to the United States. d. The native U.S. labor supply with less than a high school education is measured contemporaneously. migration from quickly equilibrating wages between regions, meaning that initial differences in labor supply—in particular, at the time a cohort enters the labor market—contribute to continuing pressure for migration in subsequent periods as a cohort ages. Hence, the migration rate is a function of initial labor supply and labor demand conditions, as well as subsequent innovations to labor demand. The results of this analysis, presented in table 4, confirm a prominent role for labor supply in driving migration from Mexico to the United States. Column 1 shows the pooled results using all available census waves including 2010, column 2 excludes 2010, column 3 shows the results for males only, and column 4 shows the results for females only. In all cases, the log Mexico–U.S. labor supply ratio is positive and strongly statistically significant. The coefficient in the first column implies that a 10 percent increase in relative labor supply would translate into a 1.4 percentage point increase the decadal flow of net migration. The relationship is slightly weaker among women, with a coefficient approximately 70 percent as large as for men, but still very precisely estimated (t value of 9.1). The coefficient of 0.1441 in column 1 of table 4 combined with the doubling of This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �126 Brookings Papers on Economic Activity, Spring 2017 the labor supply ratio between Mexico and the United States from 1970 to 2000 can more than explain the rise in the decadal net migration rate from 2.5 to 8.3 percent during this period. Now that U.S. cohorts are growing more rapidly than their Mexican counterparts, our results suggest that the drop in the average log labor supply ratio for a Mexican state to the United States, from -3.82 in 2000 to -3.73 in 2010, is responsible for more than four-fifths of the observed decrease in the decadal average net migration rate to the United States during that time interval (from 8.3 to 6.6 percent). The effect of labor demand, as measured by the log ratio of GDP per capita for a Mexican state to the United States, is weaker and less stable. Contemporaneous GDP ratios are never significant and alternate in sign across specifications, while the GDP ratio at age 16 is consistently negative but significant only in the specifications that pool men and women. Surprisingly, it appears that this relationship becomes less pronounced during the Great Recession; when we exclude 2010, the fit becomes significant at the 1 percent level and the coefficient is almost four times as large in absolute value.28 The relatively large swings in Mexican GDP during this period, as well as the fact that the recession occurred close to the end of the decade, may have dampened the sensitivity of migration to shocks during this interval. Nevertheless, positive income shocks to Mexican states (or negative income shocks to the United States) clearly have the overall effect of slowing migration. How much of the migration slowdown can be attributed to the Great Recession? In trying to understand the labor demand effects of the Great Recession shock on migration, we conduct a simple simulation exercise. We use the marginal effect from the model estimated in column 2 of table 4— for the period preceding the Great Recession—to ask the out-of-sample question as to what would have happened to Mexico–U.S. migration if the United States had not experienced the Great Recession. We simulate the counterfactual log GDP ratios that would have occurred if the United States had remained on its long-term trend of GDP per capita. Because it is the GDP-at-age-16 variable through which income changes primarily affect migration, GDP shocks operate by altering the initial labor supply choices of individuals when they first enter the labor force: to seek work in Mexico or to move to the United States. The left panel of figure 13 shows the time 28. Villarreal (2014) also finds declining trend migration and a weak discontinuous effect of Mexican and U.S. GDP on migration during the Great Recession. His analysis suggests that migration tracks U.S. employment rates quite closely during this time period, but due to endogeneity concerns we do not pursue this control for economic conditions. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �127 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Figure 13. Actual and Simulated GDP per Capita in the Absence of the Great Recession, 1960–2015a Log GDP per capita ratio, Mexico to United States GDP per capita Dollars 60,000 Ratio Quadratic fit 50,000 20,000 Simulated ratio –1.1 40,000 30,000 –1.05 –1.15 U.S. GDP per capita Actual ratio –1.2 –1.25 10,000 –1.3 1970 1980 1990 2000 2010 Year 2002 2004 2006 2008 Year Sources: U.S. Census Bureau, American Community Survey; United Nations, World Population Prospects; International Monetary Fund; authors’ calculations. a. Data for the years preceding the Great Recession are actual; data for the Great Recession years are simulated by projecting the long-term trend. series projection of U.S. GDP in the absence of the recession, and the right panel shows the actual and counterfactual log labor demand ratios that would have resulted. The right panel shows that the difference in log GDP ratios from actual versus predicted in 2010 opens up to about 0.06 = 1.25 - 1.19 log points, or 6 percentage points. This indicates that the total predicted effect on the migration rate for cohorts turning 16 after 2007 is 0.0042 = 0.06 × 0.07 log points, or roughly a half percentage point decrease in the decadal migration rate. Using the 2010 age cohort sizes and migration rates, there were 22 million Mexican-born individuals between the age of 15 and 25, and we would have expected 1.9 million of them to migrate to the United States. Adjusting the decadal migration rate downward by 0.5 percent for the decade that transpired between the Great Recession and the 2015 ACS, we would have expected a decrease in the total stock of migrants of 109,000 arising from the labor demand shock to those exposed to the Great Recession shock when 16 or younger. Given the results in the previous section illustrating that the stock of migrants in this age group fell by more than 1.1 million between 2010 and 2015, it would appear that labor demand shocks as captured by GDP per capita can explain only a modest portion of the reversal in migration flows. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �128 Brookings Papers on Economic Activity, Spring 2017 To what extent are other factors, such as the ramp-up in enforcement at the border, responsible for the decreases in Mexico–U.S. migration between 2007 and 2015? We attempted to investigate the role of border enforcement using an instrumentation approach that first calculated the share of migrants from each Mexican state apprehended in each Border Patrol sector in the earliest available year, 1999. We then multiplied this sector- or state-specific enforcement incidence by an index of overall Border Patrol effort, the number of Border Patrol “linewatch” hours per year. The resulting instrument proved to be strongly positively correlated with Mexican state emigration rates, indicating that despite our effort to exogenize enforcement, it may be so strongly endogenous to migration that one cannot estimate a credible long-term impact of border enforcement on successful crossings. Our compromise, reduced-form approach described above omits direct measures of enforcement, whose effect on migration may therefore be absorbed by other covariates. How may the exclusion of enforcement affect our results? One possibility is that that enforcement responds endogenously to relative GDP ratios, in which case it is captured by the reduced-form relationship between GDP per capita and migration (Hanson and Spilimbergo 2001). A second possibility is that enforcement is orthogonal to our core explanatory variables, and hence remains in the residual. A third possibility is that omitted enforcement variation is incidentally correlated with changes in labor supply, in which case its effects load onto this variable. In any case, it appears clear that the push factors driving migration from Mexico to the United States have abated sharply during the past decade, and hence the marginal effectiveness of border enforcement spending in terms of prevented crossings is falling. Taking our results at face value, the analysis suggests that labor supply shocks play a major role in driving low-skilled immigration flows in the United States. This fact, combined with the relatively predictable nature of future population growth, provides an opening for predictive analysis. We therefore turn next to a forecasting exercise using data from all the Latin American sending countries to assess U.S. immigration pressures decades into the future. II.D. Summary From the early 1980s to the mid-2000s, there were robust pressures for low-skilled immigration in the United States. U.S. incomes for lowskilled workers far exceeded those in migrant-sending nations; the U.S. macroeconomy was considerably more stable than Latin America’s; and enforcement against illegal entry, though not entirely lax, permitted large This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 129 inflows to occur. These conditions changed abruptly after the Great Recession. Gaps in the level and volatility of income between the United States and migrant-sending nations have compressed, while there has been an extensive buildup in U.S. immigration enforcement. Despite these recent changes in the pattern of relative income growth, it appears that changes in relative labor supply growth have mattered the most for current trends in U.S. immigration. III. Low-Skilled Immigration in the Long Run It may be tempting to view the period since the Great Recession as a temporary pause in the U.S. immigration wave that began in the 1970s. After all, U.S. incomes for low-wage labor are still roughly twice those in Mexico, and are even larger when compared with those of other sending nations in Latin America. Should not high levels of immigration resume once the U.S. economy returns to a period of normal growth? Such a perspective downplays both the recent increase in U.S. enforcement and the demographic determinant of recent U.S. labor inflows from the Western Hemisphere. Looking forward, demographic pressures for U.S. immigration are set to weaken significantly. In this section, we construct a model of long-run changes in U.S. immigration, which we use to project U.S. labor inflows in coming decades. We then characterize how changes in low-skilled immigration may affect labor market conditions in the United States. III.A. �A Predictive Analysis of Future Migration from Latin America to the United States To evaluate the effects of future demographic change on U.S. immigration, we turn to the national level and incorporate the long-term population growth forecasts for countries in the Western Hemisphere provided by the United Nations’ World Population Prospects.29 We examine how immigration from Latin American countries may have changing effects on U.S. labor inflows in coming decades. Recognizing that future migration episodes will also be driven by unanticipated economic and political shocks, differential demographic growth estimates for the next 15 years provide one of the clearest lenses on the future that is available. Labor 29. To avoid circularity in studying how future population growth will drive migration from countries whose population will in turn be determined by migration, we use the UN’s “no migration” population forecast, which ignores as-yet-unobserved future migration in its projected population estimates. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �130 Brookings Papers on Economic Activity, Spring 2017 Figure 14. Relative Labor Supply Ratios by Origin Country for Age 15–40, 1980–2050 Change in labor supply ratio, origin country to United States, 1980 = 1 3 Actual Projected Guatemala 2.5 Honduras 2 Mexico 1.5 El Salvador 1990 2000 2010 2020 2030 2040 Year Sources: United Nations, World Population Prospects; author’s calculations. supply provides a uniquely forecastable component of migration pressures: The cohorts that will enter the labor force in the next 16 to 20 years have already been born, and changes in cohort sizes for the next two decades can be predicted relatively accurately using current trends in fertility.30 A simple visual perspective on the issue is given in figure 14, which plots observed and projected population relative to the United States for major sending countries from 1980 to 2050, where we normalize 1980 values to 1. The most striking demographic transition is in El Salvador; having reached a peak in relative labor supply in 2010, its population relative to the United States is projected to decline rapidly in the future, reaching 1980 levels again by 2050.31 Mexico follows a similar temporal pattern, but with a slower future decline. Guatemala and Honduras have seen similarly steep increases in relative labor supply, reaching roughly 250 percent of their 30. Many migration episodes, such as the recent surge of Syrians into Europe and the arrival of Vietnamese immigrants in the United States in the late 1970s, were driven by shocks other than labor supply. Some shifts in labor supply, such as the rapid fertility decreases in Catholic Southern Europe and Latin America, were not forecasted. Nonetheless, given our lack of ability to anticipate future income shocks, demographic differentials remain an attractive way of predicting medium-term migration trends. 31. Because the U.S. population is growing, declines in these relative population size ratios do not imply absolute declines in the populations of origin countries. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 131 1980 values by 2015. The future experiences of these two Central American countries diverge strongly, with Guatemala’s relative labor supply continuing to grow almost linearly for another 25 years, whereas Honduras’s supply begins to decline immediately. Continuing robust labor supply growth in Guatemala may allow it to partially replace diminished U.S. migrant inflows from other Latin American countries. However, given Guatemala’s small size—its 2015 population was 16 million, compared with Mexico’s population of 127 million—its migrant-sending capacity is limited. To use population forecasts as the basis for a predictive model of future U.S. immigration, we first calculate five-year birth cohort ratios in the historical (census) and future (United Nations) data in the same manner, and project future GDP based on growth forecasts from the International Monetary Fund so as to be able to take log GDP per capita ratios. Using data for the 25 available countries in Latin America and the Caribbean, we estimate the model using observed migration rates from 1980 to 2015, and use the resulting parameters to project age-gender-country–specific migration rates through 2040. In order to capture the dramatic shift in immigration enforcement that occurred between 1990 and 2000, we estimate a trend-break model that fits piecewise linear time trends to the “low enforcement” era preceding 2000 and the “high enforcement” era starting in 2000.32 Because we want to estimate effects for the full age distribution of immigrants in the United States, we use all available age cohorts and not just those of young workers. The estimating equation is ( 7)  N   GDPit  micgt = β 0 + γ 1 log  icg  + γ 2 log  + α i + µ g + ηc  GDPUSt   NUScg  + ∑ α i δ 2 t + ∑ ϕ1t α i t1 + ∑ ϕ 2 i α i t2 + ε icgt . i i i Here, micgt is the net migration rate to the United States for source country i, age cohort c, and gender g, at the time of census wave t. Regressors include the log ratio of the origin-country birth cohort to the U.S. birth cohort (Nicg/NUScg) and the contemporaneous log GDP ratio of the origin-country’s GDP per capita to U.S. GDP per capita (GDPit /GDPUSt). We include fixed 32. These trend breaks (which we allow to be country-specific) are consistent with the inclusion of time period fixed effects in the first-differenced model in equation 6. We move from first differences in equation 6 to levels in equation 7 to accommodate forecasting immigrant stocks. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �132 Brookings Papers on Economic Activity, Spring 2017 Table 5. Results of the Prediction Regressiona Independent variable Log birth cohort ratio for origin country to United States Log birth cohort ratio × under 40 years old Log GDP ratio for origin county to United States Log GDP ratio × under 40 years old Female Female × under 40 years old Covariatesb No. of observations R2 Net migration rate 0.7716 (0.760) 3.9292*** (0.734) 1.9581*** (0.518) -1.4201* (0.798) 0.3558*** (0.116) -0.7988*** (0.116) Yes 3,310 0.842 Sources: U.S. Census Bureau, American Community Survey; United Nations, World Population Prospects; International Monetary Fund; authors’ calculations. a. The units of analysis are five-year birth cohorts. The analysis is conducted on 25 Latin American countries’ net migration rates to the United States over the years 1980–2015. Robust standard errors in parentheses are clustered by birth cohort and weighted by birth cohort size. Statistical significance is indicated at the ***1 percent, **5 percent, and *10 percent levels. b. Covariates included but not reported are: (i) five-year age cohort dummies; (ii) country fixed effects; (iii) a country dummy for the year 2000 and later; (iv) a country-specific linear time trend for the lowenforcement era of 1980–90; (v) a country-specific time trend for the high-enforcement era of 2000–15; and (vi) the interactions between all these covariates and being under 40 years old. effects for the origin country (ai), the age cohort (hc), and gender (µg). We let time trends (and trend breaks) be specific to the origin country by allowing each country to have separate intercept and time trend terms within the low-enforcement and high-enforcement eras. First, we interact the country fixed effect with a dummy variable indicating the year 2000 and later (d2t), then we interact the fixed effect with a 1980–90 trend (t1), and finally with a second term picking up the high-enforcement 2000–15 time trend (t2). We predict over the interval 2020–40, meaning that we forecast for the same number of periods as are used to estimate the postenforcement trend. In using a broad set of age cohorts (age 2–67), we partition the regression to estimate separate coefficients for the young (age 2–37) and the old (age 42–67), given the divergent migration trajectories of these two groups. The parameter estimates for the central parameters in the cross-country regression for Latin America and the Caribbean are given in table 5, where we suppress the large number of country-level intercept and slope This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �133 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Figure 15. Migration Rates by Origin Country for Age 15–40, 1980–2040 Percent Actual Projected El Salvador 15 10 Honduras Guatemala 5 Mexico 1990 2000 2010 Year 2020 2030 Sources: U.S. Census Bureau, American Community Survey; United Nations, World Population Prospects; International Monetary Fund; authors’ calculations. interactions. The regression results contain several noteworthy features. First, the strong push factor of large cohort size is experienced entirely before age 40, after which there is no significant relationship between the log birth cohort ratio and net migration rates. Second, the effect of GDP ratios flips across ages; for the young cohorts, high income in the origin country has no effect on migration, whereas for the older cohorts, high origin-country income accelerates migration. This latter result is consistent with gravity model estimates of bilateral migration (Clark, Hatton, and Williamson 2007). It could be explained by migration costs being a greater constraint to migration across Latin America as a whole than in Mexico, with positive income shocks enabling migration by credit-constrained individuals (Mckenzie and Rapoport 2007). Having estimated the model on decadal changes in the migration rate, we then predict the net migration rate for future decades, using the International Monetary Fund’s GDP forecasts and UN population growth forecasts for future decades. We maintain the time trends estimated on data for 2000 and later, such that we presume the strong immigration enforcement regime stays in effect. The results are shown in figure 15. With this predicted future migration rate in hand, we multiply these by the UN-projected future birth cohort sizes to calculate predicted migrant counts for each cohort. These This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �134 Brookings Papers on Economic Activity, Spring 2017 Figure 16. Stock of Foreign-Born Migrants by Origin Country for Age 15–40, 1980–2040 Millions 6 5 4 Mexico 3 Actual Projected 2 1 Guatemala El Salvador 1990 2000 2010 Year 2020 Honduras 2030 Sources: U.S. Census Bureau, American Community Survey; United Nations, World Population Prospects; International Monetary Fund; authors’ calculations. values are then summed up to obtain migration totals for each country in each year. As seen in figure 15, this empirical structure predicts declining net migration rates from Mexico over the coming decades, with other major Latin American destinations seeing roughly constant rates. Figure 16 shows that Mexican-born migrant stocks in the key age 15–40 group are predicted to drop to less than half their current level by 2040. Rather than showing that the Great Recession has caused a temporary pause in an ongoing wave of immigration from Mexico, these long-term trends suggest that the 1990– 2007 housing boom may have caused a temporary surge in migration, arresting a demographically driven long-run slowing.33 Given the strong role that demographic factors play in our estimation model, convergence in fertility rates across the Americas removes a powerful factor pushing workers across borders. We find little support for the idea that Latin American immigration will surge again as the U.S. economy recovers. Our focus on the declining population of migration-age individuals overlooks an important role that Mexican-born individuals will play 33. Despite using a much longer panel and a different estimation structure than that used by Hanson and McIntosh (2016), these results confirm these previous predictions that new inflows of working-age Mexicans will drop substantially by the middle of the 21st century. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �135 GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH Figure 17. Age Frequencies of Mexican-Born Immigrants in the United States, by Year Thousands per five-year age cohort 2040, projected 2,000 1,500 2015 1,000 500 1980 10 20 30 40 50 Midpoint of five-year age cohort 60 Sources: U.S. Census Bureau, American Community Survey; United Nations, World Population Prospects; International Monetary Fund; authors’ calculations. in American demographics. In a manner even more pronounced than for Mexico itself—which has recently undergone a rapid demographic transition (Tuiran and others 2002)—the U.S. Mexican-born population will age very quickly. We can draw frequencies of Mexican-born individuals in our data, starting with the observed numbers in 1980 and 2015, and then plot the predicted values in 2040, as shown in figure 17. Whereas the modal Mexican-born resident in the United States was 20 years old in 1980 and 40 years old in 2015, he or she will be almost 60 years old by 2040. Rapid aging arises from the confluence of declining fertility in Mexico and the demographic amplifier of emigration, which pushes a larger share of larger cohorts into the United States and therefore accentuates the implications of Mexico’s demographic transition for the age structure of the Mexico-born population on the U.S. side of the border. A large elderly population of undocumented immigrants is a policy challenge that the United States has hitherto not faced. Stepping back to examine the age distribution of all Latin American immigrants, we see broad evidence of an aging population in the United States. Table 6 illustrates that the total Latin American–born population under 40 in the United States by 2040 will be only 56 percent of its current size (4.9 million versus 8.8 million), while the foreign-born This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �136 Brookings Papers on Economic Activity, Spring 2017 Table 6. Thousands of Foreign-Born Residents in the United States 1980 2015 2040 projecteda Country Younger than 40 Older than 40 Younger than 40 Older than 40 Younger than 40 Older than 40 Antigua and Barbuda Bahamas Barbados Belize Bolivia Chile Colombia Costa Rica Dominican Republic Ecuador El Salvador Grenada Guatemala Guyana Haiti Honduras Jamaica Mexico Nicaragua Panama Paraguay Peru Trinidad and Tobago Uruguay Total 2.8 7.8 16.5 10.0 9.6 24.5 101.5 101.5 115.1 61.3 74.5 4.6 49.3 37.0 63.0 25.3 123.3 1,574.7 28.6 39.4 2.2 38.8 46.9 8.8 2,485.5 0.8 4.0 8.3 3.6 3.5 10.3 40.2 40.2 43.8 24.0 17.8 2.2 12.6 11.6 25.2 10.1 59.9 452.3 12.1 16.6 0.7 15.6 17.1 4.7 805.1 4.4 13.7 8.9 15.3 30.3 32.2 253.0 37.3 439.0 186.7 650.2 5.5 557.1 81.1 256.0 369.0 227.0 5,299.0 95.0 25.5 7.0 157.4 73.8 13.7 8,838.1 10.9 18.5 33.1 25.4 36.6 54.4 374.6 40.0 536.9 222.3 655.7 19.1 346.1 164.7 349.9 211.7 391.8 5,663.8 138.0 60.2 7.0 248.5 131.6 23.3 9,764.1 6.8 4.5 1.1 3.7 -61.4 -29.4 -83.7 -38.0 402.1 22.8 594.1 0.2 813.5 55.9 205.0 488.0 151.0 2,707.0 -69.7 -35.4 -167.8 -81.0 22.8 -4.5 4,907.7 8.9 47.1 28.1 33.5 86.7 73.4 782.5 50.3 1,104.4 281.0 1,601.7 19.5 1,102.3 186.2 967.7 698.1 539.8 14,146.0 336.2 50.3 45.3 536.7 150.1 54.3 22,930.1 Sources: U.S. Census Bureau, American Community Survey; United Nations, World Population Prospects; International Monetary Fund; authors’ calculations. a. Negative values are an artifact of the linear model used to forecast future flows, and are not possible in practice. See the text for interpretation, specifically note 34. population over 40 will be 235 percent of its current size (22.9 million versus 9.8 million). The model predicts a negative net migration rate for some counties of origin in 2040 (negative values are not possible in practice; but in our linear predictive model, they indicate a phenomenon that could be interpreted as net migration pressure out of the United States).34 The main 34. Negative net migration rates are an artifact of the linear model used to forecast future flows. Our model predicts only 14 percent of the dyads to have negative net migration in 2020, but future decreases in population growth in sending countries drive the share of predicted negative dyads to 24 percent by 2040. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 137 policy question posed by first-generation immigrants from Latin America and the Caribbean thus appears likely to shift from one of the labor market effects of large-scale labor inflows to one of the cost of social programs and health care for an elderly immigrant population with relatively low incomes (and relatively low rates of naturalization when compared with high-skilled immigrants). III.B. �Changes in Low-Skilled Immigration and U.S. Labor Market Tightness We have seen that incentives for low-skilled immigration in the United States have changed markedly since the early 2000s, and have already manifested demographic pressures that are likely to compress migration inflows in coming decades. How will these developments affect the U.S. economy? A reduction in the relative supply of low-skilled labor, by putting upward pressure on wages for these workers, may operate directly by causing changes in the U.S. wage structure (Borjas 2003). Alternatively, wage pressures may induce firms to alter their production techniques in a manner that mitigates the wage effects of shocks to the relative labor supply by generating endogenous changes in labor demand (Lewis 2011). Whichever form labor market adjustments take, the magnitudes of these adjustments are likely to be determined by the implicit pressure of changes in immigration inflows on U.S. wages, which we analyze next. As a final exercise, we consider how the United States’ demand for and supply of labor have evolved over time and how the supply of low-skilled, foreign-born workers meshes with these changes. Our approach employs the methodology of Katz and Murphy (1992), as applied by Autor, Katz, and Kearney (2008), to examine the relative earnings of more- and lessskilled workers. The exercise we perform allows us to translate the recent slowdown in low-skilled immigration into implied pressures on the wage premium enjoyed by skilled workers. Consider a production function with constant elasticity of substitution that takes as its arguments the employment of low-skilled and high-skilled workers, where within each skill group we treat native-born and foreignborn workers as perfect substitutes, as is consistent with recent evidence (Borjas, Grogger, and Hanson 2012).35 From the first-order conditions for 35. The fact that the CPS does not include measures of nativity until 1994 makes this assumption a necessity if we are to estimate equation 6 based on time series variation. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �138 Brookings Papers on Economic Activity, Spring 2017 firm profit maximization, we obtain an expression for the relative wages of high-skilled and low-skilled labor, (8 ) log ( wht ) − log ( wlt ) = γ 0 + γ 1 [ log ( N ht ) − log ( N lt )] + γ 2 X t + ε t , where log(wht) - log(wlt) is the log U.S. wage for high-skilled workers relative to the log U.S. wage for low-skilled workers, log(Nht) - log(Nlt) is the log U.S. supply of high-skilled workers relative to the log U.S. supply of low-skilled workers, and Xt is a vector of controls that capture labor demand shocks.36 Each skill group is made up of a combination of native-born and foreign-born labor. By taking the difference in earnings between skill groups in equation 8, we remove from the specification labor demand shocks that are common to high-skilled and low-skilled workers (for example, aggregate changes in labor demand associated with recessions, and growth in total factor productivity). Following Autor, Katz, and Kearney (2008), within each skill group we measure wages as average weekly earnings—holding constant the age, gender, and racial composition of workers—and we measure employment in terms of labor supplies expressed in PEUs. High-skilled workers are those with at least a college education, whereas low-skilled workers are those with a high school education or less. We estimate equation 6 with annual data from the CPS for the period 1963–2007, and we use the results to predict relative earnings for the period 1963–2015, which includes the out-of-sample range 2008–15. In our baseline specification for equation 8, which includes a time trend as the only additional covariate, the coefficient estimate for g1 is -0.42 (with a t value of 9.9) when we define the low-skilled group to be workers with a high school education or less; and it is -0.18 (with a t value of 8.0) when we define the low-skilled group to be workers with strictly less than a high school education. Consistent with the theory underlying equation 8, increases in the relative supply of skilled labor drive down the wage premium for skill. These estimates change little when we expand the period from 1963 to 2015 or include the following additional covariates: a quadratic time trend, the aggregate unemployment rate, and the log real 36. This estimation approach makes the strong assumption that labor is freely mobile across occupations. See Burstein and others (2017) for an analysis that uses a Roy (1951) model in the analysis of how immigration affects labor market outcomes at the occupation level. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 139 Figure 18. Skill Premium for All Industries, 1963–2015 College and above versus less than high school Percent 1.1 0.9 Percent Counterfactual predicted wage gap Katz and Murphy (1992) predicted wage gap 1.1 0.9 0.7 0.7 0.5 College and above versus high school or less Observed wage gap 0.5 2007 1970 1980 1990 2000 2010 Year 1970 1980 1990 2000 2010 Year Sources: U.S. Census Bureau, Current Population Survey; authors’ calculations. federal minimum wage.37 The first coefficient—for the impact of relative labor supply on the wage gap between college- and high school–educated workers—compares in value with the estimates given by Autor, Katz, and Kearney (2008) of -0.40 to -0.62, depending on the covariates included, for the period 1963–2005, suggesting that our coefficient estimates are at the low end of those obtained in previous empirical work. Figure 18, which presents the results, shows three series for relative earnings: the actual skill premium for the period 1963–2015; the projected skill premium for 1963–2015 based on equation 6, using coefficients estimated on data for 1963–2007; and a counterfactual projection of the skill premium, where we again use the estimation results for equation 8, but now replace the relative labor supplies we feed into the projection with a counterfactual series in which we assume that the number of low-skilled immigrant workers grows at the same rate from 2008 to 2015 as it does for 1994 to 2007. For the out-of-sample period 2008–15, the first projection is based on observed changes in the relative supply of skilled labor (which embody 37. The time trend carries a positive coefficient, indicating a positive trend in the relative demand for skilled labor; the unemployment rate carries a negative sign but is imprecisely estimated; and the minimum wage enters negatively, indicating that a higher minimum wage compresses the skill premium. The assumption that the time trend for relative labor demand is linear over the course of several decades is of course quite strong. Nevertheless, allowing for a quadratic time trend has minimal impact on the estimate of g1. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �140 Brookings Papers on Economic Activity, Spring 2017 actual changes in both high-skilled and low-skilled immigration), whereas the second projection is based on the observed labor supply through 2007 and the counterfactual supply thereafter (which suppresses the slowdown in low-skilled immigration). In constructing the counterfactual projection, it is worth noting that inflows of both low-skilled and high-skilled immigrants slowed after 2007. Our counterfactual labor supply series, by imposing continued growth for low-skilled but not for high-skilled immigration, thus understates the post2007 growth in the relative supply of skilled labor. The resulting counterfactual projection of the skill premium therefore corresponds to an artificial setting, which we view as useful for describing the magnitude of the lowskilled immigration slowdown in terms of wage pressures but not for evaluating the impact of immigration on earnings during this period.38 For low-skilled workers defined to be those with less than a high school education (the left panel of figure 18), the actual skill premium is flat from the early 1960s to the late 1970s, rises steadily from the late 1970s to the mid-2000s, and is flat again thereafter. During the within-sample period 1963–2007, the predicted skill premium rises more slowly than the actual skill premium in the mid-1970s, suggesting that the relative demand for skill rises more slowly than the linear trend would indicate, and rises more rapidly than the actual skill premium from the late 1970s to the late 1990s, suggesting growth in the demand for skill that exceeds the linear trend during this interval. Turning to the out-of-sample period 2008–15, the post-2007 slowdown in immigration tempers growth in the supply of low-skilled labor, causing the predicted skill premium to rise more slowly after 2007 than before. Replacing actual relative labor supplies with the counterfactual series that assumes sustained growth in low-skilled immigration (that is, a 2007–15 annual growth rate equal to the 1994–2007 annual growth rate), the projected path of the skill premium naturally lies above the projected path based on observed data. If low-skilled immigration had not slowed after 2007, the relative supply of skilled labor would have grown more slowly, which in turn would have mandated a larger increase in relative earnings for skilled labor. The difference between the two projected wage series in 2015 is 8.6 log points, which indicates that the magnitude of the slowdown in low-skilled immigration—holding all else equal, including high-skilled 38. For analyses of the impact of immigration on earnings, see Card (2001); Borjas (2003); Ottaviano and Peri (2012); and Dustmann, Frattini, and Preston (2013). This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 141 immigration—is consistent with a decrease in the skill premium of 1.1 percent a year during the 2007–15 period. To put this magnitude in context, the observed increase in the wage premium for college-educated workers versus workers with less than a high school education during the 1980–2007 period of rapidly rising wage inequality is 1.6 percent a year. Again, we do not take this value to be the true change in wages due to the immigration slowdown (because we are not addressing changes in high-skilled immigration), but rather as an indication of the magnitude of the immigrationinduced change in the labor supply expressed in terms of wage pressures. When we instead define low-skilled workers to be those with a high school education or less (right panel of figure 18), the broad patterns for the actual skill premium are similar, though the flattening in the premium during the 2000s is less pronounced and the absolute premium is smaller. For the out-of-sample period 2008–15, it is again the case that the projected skill premium based on actual labor supplies lies above the observed skill premium, indicating an increase in the demand for skill that is less than the linear trend. Comparing this projected skill premium with that which obtains when using counterfactual labor supplies (involving no post-2007 slowdown in immigration), the latter exceeds the former by 6.1 log points in 2015, or a difference of 0.8 percent a year during the period 2007–15. To put this magnitude in context, the observed increase in the wage premium for college-educated workers versus workers with a high school education or less for 1980–2007 is 1.1 percent a year. Because low-skilled immigrants are a smaller share of the skill group with a high school education or less than of the skill group with strictly less than a high school education, the implied wage pressures of the immigration slowdown are weaker when we move to this more expansive definition of being low-skilled. Similar patterns are observed when we restrict the sample from all industries (figure 18) to low-skill industries (figure 19). III.C. Summary The U.S. immigration wave of the late 20th century was enabled to a substantial extent by the rapid growth of the labor supply in Latin America and the Caribbean relative to the United States. Because labor supply growth in migrant-sending nations is slowing and will continue to slow, the demographic push for U.S. immigration is abating. Abetting these demographic forces is the substantial increase in U.S. immigration enforcement, which thus far has been maintained. Absent economic or political crises in the Western Hemisphere that reignite international migration, standard migration models predict that migration rates from major U.S.-sending This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �142 Brookings Papers on Economic Activity, Spring 2017 Figure 19. Skill Premium for Low-Skill Industries, 1968–2015 College and above versus less than high school College and above versus high school or less Percent Percent Counterfactual predicted wage gap 1.1 1.1 Observed wage gap 0.9 0.9 0.7 0.7 0.5 Katz and Murphy (1992) predicted wage gap 2007 1970 1980 1990 Year 2000 0.5 2010 1970 1980 1990 Year 2000 2010 Sources: U.S. Census Bureau, Current Population Survey; authors’ calculations. nations will drop sharply in coming decades. Indeed, the weakening of these migration pressures began in the early 2000s, and may have been masked by the temporary labor demand boost provided by the U.S. housing boom. The resulting post-2007 slowdown in low-skilled immigration is of a magnitude consistent with a decrease in the wage gap between high-skilled and low-skilled U.S. labor of 6 to 9 percentage points. If, as predicted by demographic forces, low-skilled immigration continues to decline in future decades, U.S. firms—especially those located in U.S.–Mexico border states and in the immigrant-intensive industries of agriculture, construction, eating and drinking establishments, and nondurable manufacturing—are likely to face pressure to alter their production techniques in a manner that replaces low-skilled labor with other factors of production. IV. Looking Forward From the early 1970s to the early 2000s, the United States experienced an epochal wave of low-skilled immigration, which was the combined result of relatively high U.S. incomes, relatively stable U.S. GDP growth, relatively slow U.S. labor supply growth, and moderately permissive immigration enforcement. Since the mid-2000s, each of these drivers has attenuated. The U.S. macroeconomy is no longer so stable relative to migrant-sending countries; U.S. labor supply growth is now similar to that in much of the This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 143 Western Hemisphere; and the U.S. borders, having been heavily fortified, are much harder to cross without a visa. The future of low-skilled immigration thus appears to be less about streaming inflows of young workers from lower-income nations and more about the needs of an aging population of lower-income adults that is settled in the United States. Those within this group who are undocumented— somewhere between one-half and three-fifths—do not qualify for most federally funded welfare benefits, including Medicare and Medicaid. In recent decades, the primary fiscal effects of low-skilled immigration were the cost of primary and secondary education for the children of immigrants, and, to a lesser extent, for publicly funded health care for the subpopulation of this group that was born in the United States (Blau and Mackie 2016). Judith Treas and Zoya Gubernskaya (2015) document that 51 percent of the foreign-born population is covered by some form of public insurance, as compared with 35 percent of the native born, suggesting that the costs of caring for the foreign born are likely to fall disproportionately on publicly funded programs. Given our estimates of an increase of 13 million (134 percent) in the population of foreign-born immigrants over the age of 40 by the year 2040, there may be considerable growth in the demand for safety net programs as a result of past and future immigration. Under existing financing rules, U.S. states and localities would be the entities primarily responsible for shouldering these costs. In light of the changing demographics of migrant-sending nations, and the apparent effects of the existing immigration enforcement surge, the current emphasis of the U.S. government on further intensifying immigration enforcement is puzzling. One interpretation of the planned enforcement buildup is that it is driven by politics. Having lived through the great immigration wave of the last 35 years, some native-born voters may be upset by the laxity of past enforcement and willing to reward politicians who are seen as atoning for these transgressions. Supporting stronger enforcement may be a way for politicians to signal their disapproval of earlier policy choices. Such signaling would come at a substantial cost, however, given that the U.S. immigration enforcement budget now exceeds $20 billion a year. Another interpretation is that intensifying enforcement is an effort to forestall future claims on public resources. The aging of the low-skilled, foreign-born population means that by increasing deportations today—when many low-skilled immigrants are approaching middle age— the United States may avoid demand for social spending in the future. If U.S. voters oppose providing public benefits to low-skilled immigrants— and if the U.S. government cannot credibly commit to deny benefits to This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �144 Brookings Papers on Economic Activity, Spring 2017 low-income, elderly, foreign-born residents down the road—then expanding current deportations may reduce the expected drain on U.S. public coffers in later decades.39 The cost of these extra deportations—beyond the incremental spending on enforcement—includes reducing the supply of workers who are in their prime earning years, who have accumulated substantial U.S. labor market experience, and who are well established in their communities. Changes in U.S. immigration policy affect not just the U.S. economy but also the economies of migrant-sending nations. To the extent that immigration enforcement has played a dominant role in the recent slowdown in migration, improvements in the welfare of U.S. workers come in part at the cost of potential migrants who forgo higher wages, and their nonmigrant compatriots, who now face more crowded labor markets at home. Because the modal immigrant from Mexico would be a middle-income earner at home, expanding deportations and tightening border security would tend to expand labor supplies and depress earnings in the middle of Mexico’s wage distribution (and from higher quantiles of the wage distributions in other sending countries for migrants). From the perspective of those born in sending countries, wages take a hit from immigration restrictions whether or not workers decide to migrate. Mexico, by virtue of its status as a transit country for undocumented immigrants, is doubly exposed to changes in U.S. immigration enforcement. Many Central Americans planning to enter the United States traverse Mexico illegally on their way north. Stronger U.S. enforcement may have the indirect consequence of increasing the supply of undocumented Central Americans seeking to live and work in Mexico. Given that Guatemala is the one major U.S. migrant-sending nation in the Western Hemisphere that will continue to experience high rates of labor supply growth in coming decades, Mexico faces the real possibility that continued tightness in U.S. immigration policy would increase its supply of low-skilled, foreign-born residents. Taking immigration controls as a means to improve the plight of lowskilled U.S.-born workers, how would we expect the incidence of benefits to stack up against alternative policies? An increase in minimum wages may benefit those with jobs that pay equilibrium wages above a higher wage 39. This characterization of political support for immigration enforcement is roughly consistent with the framework used by Alesina, Baqir, and Easterly (1999), in which enthusiasm for public spending is diminished by increased ethnic and racial diversity in a jurisdiction. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 145 floor, at a cost to consumers who purchase labor-intensive goods and services and to workers excluded from the labor market. Investments in skill development and retraining may target native-born workers more finely, but the effectiveness of such programs is in question (LaLonde 1995). The attractiveness of immigration restrictions relative to these alternative policies also depends on the extent to which we take the welfare of the foreignborn into account. Immigration controls therefore appear to be a prolabor instrument that comes at a high cost to consumers and foreign-born workers relative to alternative potential policies. It may be premature to declare that the most recent episode of high U.S. immigration is over. Many factors could cause the population of lowskilled immigrants in the United States to begin growing again. Prime among these is increased economic or political instability in the Western Hemisphere. Indeed, heightened insecurity in Central America, due in large part to violence associated with organized crime, appears to have increased recent labor outflows from the region. Mexico, for its part, has not had a financial crisis since 1995. The reform of the country’s electoral laws in 1997 created a political process that is competitive and free of the vote rigging that marred the Institutional Revolutionary Party’s 70-year rule during the 20th century. However, the openness of Mexico’s economy leaves it vulnerable to external shocks, in particular in the United States, which is the destination for more than 80 percent of its exports. At least in the short to medium runs, the U.S. government itself seems to be in a position to determine—whether through its trade or its immigration enforcement policies—the potential supply of low-skilled immigrants. ACKNOWLEDGMENTS    We thank Janice Eberly, Adriana Kugler, Edward Lazear, James Stock, and the participants in the Spring 2017 Brookings Panel on Economic Activity for helpful comments on a draft of this paper; Daniel Leff for excellent research assistance; and the Center on Global Transformation at the University of California, San Diego, for financial support. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �146 Brookings Papers on Economic Activity, Spring 2017 References Akee, Randall. 2010. “Who Leaves? Deciphering Immigrant Self-Selection from a Developing Country.” Economic Development and Cultural Change 58, no. 2: 323–44. Alesina, Alberto, Reza Baqir, and William Easterly. 1999. “Public Goods and Ethnic Divisions.” Quarterly Journal of Economics 114, no. 4: 1243–84. Amuedo-Dorantes, Catalina, Thitima Puttitanun, and Ana P. Martinez-Donate. 2013. “How Do Tougher Immigration Measures Affect Unauthorized Immigrants?” Demography 50, no. 3: 1067–91. Angelucci, Manuela. 2012. “US Border Enforcement and the Net Flow of Mexican Illegal Migration.” Economic Development and Cultural Change 60, no. 2: 311–57. Argueta, Carla N. 2016. “Border Security: Immigration Enforcement Between Ports of Entry.” Report no. R42138. Washington: Congressional Research Service. Autor, David H., and David Dorn. 2013. “The Growth of Low-Skill Service Jobs and the Polarization of the US Labor Market.” American Economic Review 103, no. 5: 1553–97. Autor, David H., David Dorn, and Gordon H. Hanson. 2013. “The China Syndrome: Local Labor Market Effects of Import Competition in the United States.” American Economic Review 103, no. 6: 2121–68. Autor, David H., Lawrence F. Katz, and Melissa S. Kearney. 2008. “Trends in U.S. Wage Inequality: Revising the Revisionists.” Review of Economics and Statistics 90, no. 2: 300–23. Aydemir, Abdurrahman, and George J. Borjas. 2011. “Attenuation Bias in Measuring the Wage Impact of Immigration.” Journal of Labor Economics 29, no. 1: 69–112. Beaudry, Paul, David A. Green, and Benjamin M. Sand. 2016. “The Great Reversal in the Demand for Skill and Cognitive Tasks.” Journal of Labor Economics 34, suppl. 1, pt. 2: S199–S247. Bernanke, Ben S. 2004. “Fedspeak.” Speech given at the Meetings of the American Economic Association, San Diego, January 3. Bertoli, Simone, and Jesús Fernández-Huertas Moraga. 2013. “Multilateral Resistance to Migration.” Journal of Development Economics 102: 79–100. Blau, Francine D., and Christopher Mackie, editors. 2016. The Economic and Fiscal Consequences of Immigration. Washington: National Academies Press. Borjas, George J. 1987. “Self-Selection and the Earnings of Immigrants.” American Economic Review 77, no. 4: 531–53. ———. 2003. “The Labor Demand Curve Is Downward Sloping: Reexamining the Impact of Immigration on the Labor Market.” Quarterly Journal of Economics 118, no. 4: 1335–74. ———. 2006. “Native Internal Migration and the Labor Market Impact of Immigration.” Journal of Human Resources 41, no. 2: 221–58. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 147 ———. 2008. “Labor Outflows and Labor Inflows in Puerto Rico.” Journal of Human Capital 2, no. 1: 32–68. ———. 2013. “The Analytics of the Wage Effect of Immigration.” IZA Journal of Migration 2: article 22. ———. 2016. “The Labor Supply of Undocumented Immigrants.” Working Paper no. 22102. Cambridge, Mass.: National Bureau of Economic Research. Borjas, George J., Jeffrey Grogger, and Gordon H. Hanson. 2012. “On Estimating Elasticities of Substitution.” Journal of the European Economic Association 10, no. 1: 198–210. Burstein, Ariel, Gordon H. Hanson, Lin Tian, and Jonathan Vogel. 2017. “Tradability and the Labor-Market Impact of Immigration: Theory and Evidence from the U.S.” Working Paper no. 23330. Cambridge, Mass.: National Bureau of Economic Research. Cadena, Brian C., and Brian K. Kovak. 2016. “Immigrants Equilibrate Local Labor Markets: Evidence from the Great Recession.” American Economic Journal: Applied Economics 8, no. 1: 257–90. Card, David. 2001. “Immigrant Inflows, Native Outflows, and the Local Labor Market Impacts of Higher Immigration.” Journal of Labor Economics 19, no. 1: 22–64. ———. 2009. “Immigration and Inequality.” American Economic Review 99, no. 2: 1–21. Card, David, and Ethan G. Lewis. 2007. “The Diffusion of Mexican Immigrants during the 1990s: Explanations and Impacts.” In Mexican Immigration to the United States, edited by George J. Borjas. University of Chicago Press. Carriquiry, Alicia, and Malay Majmundar, editors. 2013. Options for Estimating Illegal Entries at the U.S.–Mexico Border. Washington: National Academies Press. Chiquiar, Daniel, and Gordon H. Hanson. 2005. “International Migration, SelfSelection, and the Distribution of Wages: Evidence from Mexico and the United States.” Journal of Political Economy 113, no. 2: 239–81. Clark, Ximena, Timothy J. Hatton, and Jeffrey G. Williamson. 2007. “Explaining U.S. Immigration, 1971–1998.” Review of Economics and Statistics 89, no. 2: 359–73. Clemens, Michael, Claudio E. Montenegro, and Lant Pritchett. 2008. “The Place Premium: Wage Differences for Identical Workers across the U.S. Border.” Working Paper no. 148. Washington: Center for Global Development. Cornelius, Wayne A. 1992. “From Sojourners to Settlers: The Changing Profile of Mexican Immigration to the United States.” In U.S.–Mexico Relations: Labor Market Interdependence, edited by Jorge A. Bustamante, Clarke W. Reynolds, and Raúl A. Hinojosa Ojeda. Stanford University Press. Cortes, Guido Matias, Nir Jaimovich, and Henry E. Siu. 2016. “Disappearing Routine Jobs: Who, How, and Why?” Working Paper no. 22918. Cambridge, Mass.: National Bureau of Economic Research. Cortes, Patricia. 2008. “The Effect of Low-Skilled Immigration on U.S. Prices: Evidence from CPI Data.” Journal of Political Economy 116, no. 3: 381–422. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �148 Brookings Papers on Economic Activity, Spring 2017 Cortes, Patricia, and José Tessada. 2011. “Low-Skilled Immigration and the Labor Supply of Highly Skilled Women.” American Economic Journal: Applied Economics 3, no. 3: 88–123. de Janvry, Alain, Kyle Emerick, Marco Gonzalez-Navarro, and Elisabeth Sadoulet. 2015. “Delinking Land Rights from Land Use: Certification and Migration in Mexico.” American Economic Review 105, no. 10: 3125–49. DiNardo, John, Nicole M. Fortin, and Thomas Lemieux. 1996. “Labor Market Institutions and the Distribution of Wages, 1973–1992: A Semiparametric Approach.” Econometrica 64, no. 5: 1001–44. Durand, Jorge, Douglas S. Massey, and Rene M. Zenteno. 2001. “Mexican Immigration in the United States.” Latin American Research Review 36, no. 1: 107–27. Dustmann, Christian. 2003. “Return Migration, Wage Differentials, and the Optimal Migration Duration.” European Economic Review 47, no. 2: 353–69. Dustmann, Christian, Tommaso Frattini, and Ian P. Preston. 2013. “The Effect of Immigration along the Distribution of Wages.” Review of Economic Studies 80, no. 1: 145–73. Edwards, Ryan, and Francesc Ortega. 2016. “The Economic Impacts of Removing Unauthorized Immigrant Workers: An Industry- and State-Level Analysis.” Washington: Center for American Progress. Edwards, Sebastian. 2009. “Forty Years of Latin America’s Economic Development: From the Alliance for Progress to the Washington Consensus.” Working Paper no. 15190. Cambridge, Mass.: National Bureau of Economic Research. Feliciano, Zadia M. 2001. “The Skill and Economic Performance of Mexican Immigrants from 1910 to 1990.” Explorations in Economic History 38, no. 3: 386–409. Fernández-Huertas Moraga, Jesús. 2011. “New Evidence on Emigrant Selection.” Review of Economics and Statistics 93, no. 1: 72–96. Gathmann, Christina. 2008. “Effects of Enforcement on Illegal Markets: Evidence from Migrant Smuggling along the Southwestern Border.” Journal of Public Economics 92, nos. 10–11: 1926–41. Giorguli-Saucedo, Silvia E., Víctor M. García-Guerrero, and Claudia Masferrer. 2016. “A Migration System in the Making: Demographic Dynamics and Migration Policies in North America and the Northern Triangle of Central-America.” Mexico City: El Colegio de México, Center for Demographic, Urban and Environmental Studies. Gonzalez-Barrera, Ana. 2015. “More Mexicans Leaving than Coming to the U.S.: Net Loss of 140,000 from 2009 to 2014; Family Reunification Top Reason for Return.” Washington: Pew Research Center. Gonzalez-Barrera, Ana, and Mark Hugo Lopez. 2016. “U.S. Immigrant Deportations Fall to Lowest Level Since 2007.” Washington: Pew Research Center. Grogger, Jeffrey, and Gordon H. Hanson. 2011. “Income Maximization and the Selection and Sorting of International Migrants.” Journal of Development Economics 95, no. 1: 42–57. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 149 Hanson, Gordon H. 2006. “Illegal Migration from Mexico to the United States.” Journal of Economic Literature 44, no. 4: 869–924. ———. 2007. “The Economic Logic of Illegal Immigration.” Special Report no. 26. New York: Council on Foreign Relations. Hanson, Gordon H., and Craig McIntosh. 2009. “The Demography of Mexican Migration to the United States.” American Economic Review 99, no. 2: 22–27. ———. 2010. “The Great Mexican Emigration.” Review of Economics and Statistics 92, no. 4: 798–810. ———. 2012. “Birth Rates and Border Crossings: Latin American Migration to the US, Canada, Spain, and the UK.” Economic Journal 122, no. 561: 707–26. ———. 2016. “Is the Mediterranean the New Rio Grande? US and EU Immigration Pressures in the Long Run.” Journal of Economic Perspectives 30, no. 4: 57–82. Hanson, Gordon H., and Antonio Spilimbergo. 1999. “Illegal Immigration, Border Enforcement, and Relative Wages: Evidence from Apprehensions at the U.S.– Mexico Border.” American Economic Review 89, no. 5: 1337–57. ———. 2001. “Political Economy, Sectoral Shocks, and Border Enforcement.” Canadian Journal of Economics 34, no. 3: 612–38. Hatton, Timothy J., and Jeffrey G. Williamson. 2011. “Are Third World Emigration Forces Abating?” World Development 39, no. 1: 20–32. Heckman, James, and Edward Vytlacil. 2001. “Identifying the Role of Cognitive Ability in Explaining the Level of and Change in the Return to Schooling.” Review of Economics and Statistics 83, no. 1: 1–12. Ibarraran, Pablo, and Darren Lubotsky. 2007. “Mexican Immigration and SelfSelection: New Evidence from the 2000 Mexican Census.” In Mexican Immigration to the United States, edited by George J. Borjas. University of Chicago Press. Kaestner, Robert, and Ofer Malamud. 2014. “Self-Selection and International Migration: New Evidence from Mexico.” Review of Economics and Statistics 96, no. 1: 78–91. Karemera, David, Victor Iwuagwu Oguledo, and Bobby Davis. 2000. “A Gravity Model Analysis of International Migration to North America.” Applied Economics 32, no. 13: 1745–55. Katz, Lawrence F., and Kevin M. Murphy. 1992. “Changes in Relative Wages, 1963–1987: Supply and Demand Factors.” Quarterly Journal of Economics 107, no. 1: 35–78. Kulish, Nicholas, Vivian Yee, Caitlin Dickerson, Liz Robbins, Fernanda Santos, and Jennifer Medina. 2017. “Trump’s Immigration Policies Explained.” New York Times, February 21. LaLonde, Robert J. 1995. “The Promise of Public Sector–Sponsored Training Programs.” Journal of Economic Perspectives 9, no. 2: 149–68. Lazear, Edward P. 1999. “Culture and Language.” Journal of Political Economy 107, suppl. 6: S95–S126. ———. 2007. “Mexican Assimilation in the United States.” In Mexican Immigration to the United States, edited by George J. Borjas. University of Chicago Press. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �150 Brookings Papers on Economic Activity, Spring 2017 Lewis, Ethan. 2011. “Immigration, Skill Mix, and Capital Skill Complementarity.” Quarterly Journal of Economics 126, no. 2: 1029–69. Mckenzie, David, and Hillel Rapoport. 2007. “Network Effects and the Dynamics of Migration and Inequality: Theory and Evidence from Mexico.” Journal of Development Economics 84, no. 1: 1–24. ———. 2011. “Can Migration Reduce Educational Attainment? Evidence from Mexico.” Journal of Population Economics 24, no. 4: 1331–58. Meckler, Laura. 2017. “Trump Orders Wall at Mexican Border.” Wall Street Journal, January 25. Mian, Atif, and Amir Sufi. 2014. “What Explains the 2007–2009 Drop in Employment?” Econometrica 82, no. 6: 2197–223. Monras, Joan. 2015. “Immigration and Wage Dynamics: Evidence from the Mexican Peso Crisis.” Discussion Paper no. 8924. Bonn: Institute for the Study of Labor (IZA). Orrenius, Pia M., and Madeline Zavodny. 2005. “Self-Selection among Undocumented Immigrants from Mexico.” Journal of Development Economics 78, no. 1: 215–40. Ottaviano, Gianmarco I. P., and Giovanni Peri. 2005. “Cities and Cultures.” Journal of Urban Economics 58, no. 2: 304–37. ———. 2012. “The Effects of Immigration on US Wages and Rents: A General Equilibrium Approach.” In Migration Impact Assessment: New Horizons, edited by Peter Nijkamp, Jacques Poot, and Mediha Sahin. Cheltenham, U.K.: Edward Elgar. Passel, Jeffrey S., and D’Vera Cohn. 2009. “Mexican Immigrants: How Many Come? How Many Leave?” Washington: Pew Research Center. ———. 2016. “Overall Number of U.S. Unauthorized Immigrants Holds Steady since 2009: Decline in Share from Mexico Mostly Offset by Growth from Asia, Central America and Sub-Saharan Africa.” Washington: Pew Research Center. Peri, Giovanni, and Chad Sparber. 2009. “Task Specialization, Immigration, and Wages.” American Economic Journal: Applied Economics 1, no. 3: 135–69. Pew Research Center. 2006. “Modes of Entry for the Unauthorized Migrant Population.” Fact Sheet. Washington. http://www.pewhispanic.org/2006/05/22/ modes-of-entry-for-the-unauthorized-migrant-population/ Pierce, Justin R., and Peter K. Schott. 2016. “The Surprisingly Swift Decline of US Manufacturing Employment.” American Economic Review 106, no. 7: 1632–62. Pinkovskiy, Maxim, and Xavier Sala-i-Martin. 2009. “Parametric Estimations of the World Distribution of Income.” Working Paper no. 15433. Cambridge, Mass.: National Bureau of Economic Research. Piore, Michael J. 1980. “The Economic Role of Migrants in the U.S. Labor Market.” In Sourcebook on the New Immigration, edited by Roy Simon. New Brunswick, N.J.: Transaction Books. Raven, John E., John C. Raven, and John H. Court. 2000. Manual for Raven’s Progressive Matrices and Vocabulary Scales. Section 3: Standard Progressive Matrices. San Antonio: Psychological Corporation. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms �GORDON HANSON, CHEN LIU, and CRAIG M c INTOSH 151 Roberts, Bryan, Edward Alden, and John Whitley. 2013. “Managing Illegal Immigration to the United States: How Effective Is Enforcement?” Washington: Council on Foreign Relations. Roberts, Bryan, Gordon H. Hanson, Derekh Cornwell, and Scott Borger. 2010. “An Analysis of Migrant Smuggling Costs along the Southwest Border.” Working Paper. Washington: U.S. Department of Homeland Security, Office of Immigration Statistics. https://www.dhs.gov/xlibrary/assets/statistics/publications/ ois-smuggling-wp.pdf Rosenzweig, Mark. 2007. “Education and Migration: A Global Perspective.” Presentation given at the Workshop on Equity and Growth, Commission on Growth and Development, New Haven, Conn., September 26–27. http://siteresources. worldbank.org/EXTPREMNET/Resources/489960-1338997241035/Growth_ Commission_Workshops_Equity_Rosenzweig_Presentation.pdf Roy, A. D. 1951. “Some Thoughts on the Distribution of Earnings.” Oxford Economic Papers 3, no. 2: 135–46. Saiz, Albert. 2007. “Immigration and Housing Rents in American Cities.” Journal of Urban Economics 61, no. 2: 345–71. Sjaastad, Larry A. 1962. “The Costs and Returns of Human Migration.” Journal of Political Economy 70, no. 5, pt. 2: 80–93. Solt, Frederick. 2016. “The Standardized World Income Inequality Database.” Social Science Quarterly 97, no. 5: 1267–81. Thompson, Ginger, and Sarah Cohen. 2014. “More Deportations Follow Minor Crimes, Records Show.” New York Times, April 6. Treas, Judith, and Zoya Gubernskaya. 2015. “Policy Contradictions and Immigrant Families.” Public Policy & Aging Report 25, no. 3: 107–12. Tuiran, Rodolfo, Virgilio Partida, Octavio Mojarro, and Elena Zúñiga. 2002. “Fertility in Mexico: Trends and Forecast.” In Population Bulletin of the United Nations: Completing the Fertility Transition. Special Issue nos. 48–49. New York: United Nations. U.S. Department of Homeland Security. 2016a. “Budget-in-Brief: Fiscal Year 2017.” Washington. https://www.dhs.gov/sites/default/files/publications/FY2017_BIBMASTER.pdf ———. 2016b. “Entry/Exit Overstay Report: Fiscal Year 2015.” Washington. https://www.dhs.gov/publication/entryexit-overstay-report Valletta, Robert G. Forthcoming. “Recent Flattening in the Higher Education Wage Premium: Polarization, Skill Downgrading, or Both?” In Education, Skills, and Technical Change: Implications for Future U.S. GDP Growth, edited by Charles Hulten and Valerie Ramey. University of Chicago Press. Villarreal, Andrés. 2014. “Explaining the Decline in Mexico–U.S. Migration: The Effect of the Great Recession.” Demography 51, no. 6: 2203–28. Warren, Robert, and Jeffrey S. Passel. 1987. “A Count of the Uncountable: Estimates of Undocumented Aliens Counted in the 1980 United States Census.” Demography 24, no. 3: 375–93. This content downloaded from 24.4.24.213 on Fri, 04 Mar 2022 00:14:55 UTC All use subject to https://about.jstor.org/terms � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource The Rise and Fall of U.S. Low-Skilled Immigration Creator An entity primarily responsible for making the resource GORDON HANSON, CHEN LIU and CRAIG McINTOSH Subject The topic of the resource Brookings Papers on Economic Activity Description An account of the resource From the 1970s the the early 2000s, the United States experienced an epochal wave of low-skilled immigration. Since the Great Recession, however, U.S. borders have become a far less active place when it comes to the net arrival of foreign workers. The number of undocumented immigrants has declined in absolute terms, while the overall population of low-skilled, foreign-born workers has remained stable. We examine how the scale and composition of low-skilled immigration in the United States have evolved over time, and how relative income growth and demographic shifts in the Western Hemisphere have contributed to the recent immigration slowdown. Because major source countries for U.S. immigration are now seeing and will continue to see weak growth of the labour supply relative to the United States, future immigration rates of young, low-skilled workers appear unlikely to rebound, whether or not U.S. immigration policies tighten further. Publisher An entity responsible for making the resource available Brookings Institution Press Date A point or period of time associated with an event in the lifecycle of the resource 2017 Type The nature or genre of the resource Journal Article Format The file format, physical medium, or dimensions of the resource PDF Identifier An unambiguous reference to the resource within a given context https://shibbolethsp.jstor.org/start?entityID=https%3A%2F%2Fidp.scu.edu%2Fopenathens&dest=https://www.jstor.org/stable/90013169&site=jstor Language A language of the resource English