increased linewatch hours reduced apprehensions initially, the effect dissipated as migrants adapted their behavior.20

By using estimates or reports of unauthorized migration among people who reportedly managed to cross into the United States, these studies have looked at the net effect of enforcement on successful crossings, not on attempted ones. These studies thus assessed the combined effectiveness of border enforcement in deterring attempts and in catching migrants attempting to cross without inspection. Assuming, however, that these studies are mostly indicating the role of border enforcement on deterring an initial attempt to cross the border, the probability of apprehension would still need to be analyzed separately.

The probability of apprehension in the 1980s and early 1990s ranged between 30 percent and 40 percent (Massey and Singer, 1995). The Mexican Migration Project (MMP), which is described in Chapter 3, also publishes updated estimates of apprehension probabilities.21 While the probabilities estimated by the MMP are somewhat lower than those presented by Massey and Singer—around 20-25 percent for the 1980s, dipping to 15-20 percent in the early 1990s—the MMP estimates indicate a sizable subsequent rise in the probability of apprehension to levels between 30 and 40 percent during the 2000s. It is difficult to determine if the increase is related to the changing composition of communities in the MMP data or to real changes at the border. If one assumes that the trend is a result of the latter,22 then the probability of apprehension during the last decade was around one in three, a relatively low number.

It is also worth noting that increased enforcement during the past decade in the U.S. interior has worsened the labor market and living conditions of undocumented immigrants in the United States (Orrenius and Zavodny, 2009) and may have a deterrent effect on attempted crossings (Wein et al., 2009). However, scholarship on the subject has mostly been devoted to understanding how interior enforcement affects immigration to particular localities. For instance, Parrado (2012) finds that the 287(g) program23 has only been effective in deterring immigration at a few large


20 Enforcement can still affect the selection of people who cross, regardless of whether it affects the total number of people who cross successfully (Lozano and Lopez, 2010; Orrenius and Zavodny, 2005).

21 See (April 13, 2012).

22 The step decline in the probability of apprehension between 2010 and 2011 is most likely due to a drop in the number of MMP communities available for these analyses.

23 The Section 287(g) program was established in 1996 but implemented primarily after 2005. Under this program, state and local law enforcement agents receive DHS training and supervision to check the immigration status of arrestees in jails and prisons and to apprehend suspected unauthorized immigrants through traffic stops or other community interventions (National Research Council, 2011:42-43).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement