There Is No Evidence That TPS Designations Increase Irregular Migration to the United States
The figures in this column contain a correction.
Under the Immigration and Nationality Act, the U.S. secretary of homeland security is authorized to designate countries for Temporary Protected Status (TPS) when ongoing armed conflict, natural disasters, or other extraordinary and temporary circumstances exist that render the countries unable to adequately or safely handle the return of their nationals from abroad. Following such a designation, certain nationals from those countries who are already in the United States may qualify to receive temporary protection from deportation and work authorization. In addition to preventing such individuals from being returned to dangerous situations in their home countries, this also provides designated countries with the time and space needed to regain stability without having to deal with the added strain of safely and effectively reintegrating returnees into what are often chaotic and challenging situations. As of October 2020, there were approximately 411,000 TPS holders in the United States from 10 TPS-designated countries: El Salvador, Haiti, Honduras, Nepal, Nicaragua, Somalia, South Sudan, Sudan, Syria, and Yemen. TPS holders have made significant contributions to the U.S. economy, work in vital industries, and have become an integral part of their communities.
As a result of hurricanes Eta and Iota—two historic hurricanes that swept through Central America in November 2020—and the ongoing coronavirus pandemic and pervasive food insecurity, the Center for American Progress has called on the U.S. Department of Homeland Security (DHS), in consultation with the Department of State, to redesignate El Salvador, Honduras, and Nicaragua for TPS and to issue a new initial designation for Guatemala. Similar calls have been made by members of Congress in both the U.S. Senate and the U.S. House of Representatives, as well as by the U.S. Conference of Catholic Bishops and Catholic Relief Services, both of which work in affected areas and provide assistance to communities in need. So far, the Biden administration has provided no indication that it is beginning work on such designations.
Dating back to when the provision was first enacted more than 30 years ago, skeptics have expressed concerns that TPS might encourage more people to migrate to the United States. However, there is no evidence that TPS designations have acted as a magnet for future irregular migration—and no particular reason to think that they even would. For one, in order to qualify for TPS, a person must have already been in the United States prior to the date of the designation; people who arrive in the United States later are ineligible for protection unless and until the country is subsequently redesignated for TPS. In 2018, a study out of the University of Virginia found strong statistical evidence that TPS designations actually decrease the demand for both regular and irregular migration to the United States. In 2019, Georgetown Law School professor Andrew Schoenholtz analyzed TPS enrollment data in connection with 12 initial country designations that were redesignated in subsequent years: Angola, Burundi, Haiti, Kosovo Province, Liberia I, Liberia II, Sierra Leone, Somalia, South Sudan, Sudan, Syria, and Yemen. The small number of individuals who qualified for protection during those redesignation periods—sometimes higher than in the initial designation but sometimes lower—provides strong evidence that the initial designation did not attract a significant number of new entrants to the country. Schoenholtz additionally analyzed port-of-entry inadmissibility determinations and apprehensions at the border and throughout the interior of the country of Haitian nationals in the years preceding and following Haiti’s TPS designation. He found no evidence that TPS led to increased irregular migration from Haiti to the United States.
Building on these earlier studies, this column explores whether the granting of TPS to nationals from the Central American countries of Nicaragua and Honduras in 1999 and El Salvador in 2001 created a magnet effect for increased irregular immigration to the United States from these countries. Using publicly available data from DHS on the annual number of persons apprehended by immigration enforcement officials—which distinguish between the country of nationality of those apprehended and serve as a proxy for undocumented immigration—two important findings emerge. First, there is no immediate post-TPS increase in the number of Nicaraguans, Hondurans, and Salvadorans apprehended by immigration enforcement officials in the years immediately following the granting of TPS for any of these three countries. Second, trends in apprehensions of Nicaraguans, Hondurans, and Salvadorans generally do not statistically significantly differ from trends in the apprehensions of nationals from countries outside of Central America over the approximately 10-year period after the granting of TPS. In other words, there is no evidence of a post-TPS magnet effect over time.
Using apprehensions to evaluate whether TPS created a magnet effect
Between fiscal years 1993 and 2010, U.S. immigration enforcement officials apprehended approximately 22.33 million individuals. As the data show, the overwhelming majority of those apprehended—93.1 percent—were Mexican nationals. However, since 2000, the percentage of Mexican nationals apprehended steadily decreased, as did the total number of apprehensions. Consequently, the percentage of non-Mexican nationals apprehended increased during this same period. (see Figure 1)
Because apprehensions serve as a proxy for undocumented immigration, these data can be used to evaluate whether TPS has created a magnet effect. Figure 1 also shows apprehensions of Nicaraguans, Hondurans, and Salvadorans as a percentage of total apprehensions. These data show several important trends. First, the time trend for Nicaraguans is statistically indistinguishable from zero. In other words, the granting of TPS for Nicaraguans has no measurable impact on the percentage of Nicaraguans apprehended by immigration enforcement officials—neither immediately after the granting of TPS nor in the decade after, from FY 2000 to FY 2010. With respect to Honduras, its trend line from FY 2000 to FY 2010 mirrors the trend line for all other countries, which casts doubt on longer-term magnet effects. Indeed, to the extent that the post-TPS trend line for Honduras mirrors the trend line for all other countries, this suggests that the apprehension of Hondurans has less to do with TPS and is likely better explained by broader trends in interior immigration enforcement. Data on El Salvador generally mirror those of Honduras, which casts further doubt on a TPS magnet effect.
To analyze the data more rigorously, post-TPS trends for Nicaraguans, Hondurans, and Salvadorans are analyzed using the synthetic control method. The synthetic control method is a quasi-experimental research design that allows for the estimation of causal treatment effects by comparing treated units with algorithmically constructed synthetic control units. Nicaragua, Honduras, and El Salvador represent the treated units. Excluding Mexico, all other countries are used to construct synthetic control units. In other words, excluding Mexico, all other countries are weighted to approximate the pre-TPS trends for Nicaragua, Honduras, and El Salvador, respectively, which then allows for the comparison of post-TPS trends. If TPS created a magnet effect, then post-TPS trends in apprehensions of Nicaraguans, Hondurans, and Salvadorans should positively and statistically significantly differ from trends in apprehensions of nationals from countries that were not granted TPS.
Figure 2 confirms that granting TPS in 1999 to Nicaraguans following Hurricane Mitch had no magnet effect. As the figure shows, after 1999, the trend line for Nicaragua is consistently below the trend line for the synthetic control. In other words, whereas undocumented immigration measured by apprehensions increased in the post-TPS period in the synthetic control, actual apprehensions of undocumented Nicaraguans remained largely flat.
The trend line for Honduras, which was also designated for TPS in 1999, is mostly below the trend line for the synthetic control throughout the post-TPS period, with the exception of FY 2005. (see Figure 3) In other words, the data show that actual apprehensions of Hondurans were lower than expected in the post-TPS period when compared with the synthetic control. If granting TPS increased undocumented immigration from Honduras, the trend line for Honduras would be statistically significantly higher than the trend line for the synthetic control. This is not the case. Importantly, the trend lines are not statistically significantly different (p = .487), which suggests that granting TPS to Hondurans following Hurricane Mitch did not create a magnet effect.
Lastly, Figure 4 shows that the trend line for El Salvador, designated in 2001, is mostly below the trend line for the synthetic control throughout the post-TPS period, with the exception of FY 2006. In other words, actual apprehensions of undocumented Salvadorans were lower than expected when compared with the synthetic control. Again, if granting TPS created a magnet effect, the trend line for El Salvador would be statistically significantly higher than the trend line for the synthetic control. This is not the case—in fact, the trend line for El Salvador is statistically significantly lower (p = .008) than the trend line for the synthetic control.
Has the granting of TPS to Central Americans from Nicaragua, Honduras, and El Salvador created a magnet effect, increasing undocumented immigration to the United States from these countries? The data analyzed here provide no evidence to suggest that TPS has created a magnet effect. As concerns over whether granting TPS incentivizes undocumented immigration are not supported by the available empirical evidence, policymakers should continue to use TPS, in accordance with the statute, to provide humanitarian protections where warranted.
Tom K. Wong is an associate professor of political science and founding director of the U.S. Immigration Policy Center at the University of California, San Diego and a senior fellow at the Center for American Progress. Tom Jawetz is the vice president for Immigration Policy at the Center. Silva Mathema is an associate director for policy on the Immigration Policy team at the Center.
The synthetic control method is a quasi-experimental research design that allows for the estimation of causal treatment effects by comparing treated units with algorithmically constructed synthetic control units. El Salvador, Honduras, and Nicaragua represent the treated units in this analysis. Excluding Mexico as an outlier, all other countries are used to construct synthetic control units. The logic underpinning the synthetic control method centers on the impossibility of observing counterfactual outcomes. Because it is impossible to wind back time in order to observe trends in the apprehension of undocumented persons from El Salvador, Honduras, and Nicaragua if TPS status had never been granted, synthetic control units that approximate the pre-TPS characteristics of these countries are used instead to estimate counterfactual outcomes. Causal treatment effects can then be obtained by comparing treated units with synthetic control units. Post-TPS trends for Salvadorans, Hondurans, and Nicaraguans are compared to synthetic control units using the synth package, which implements the synthetic control algorithm.
The positions of American Progress, and our policy experts, are independent, and the findings and conclusions presented are those of American Progress alone. A full list of supporters is available here. American Progress would like to acknowledge the many generous supporters who make our work possible.