Symposium - 2019 - Immigration

Detention as Deterrence

Emily Ryo *


Does immigration detention deter unauthorized migration? This is a pressing question with critical policy implications given that the U.S. government has detained tens of thousands of migrants in reliance on this deterrence rationale. Briefly described, the federal government has argued that “one particular individual may be civilly detained for the sake of sending a message” to others “who may be considering immigration.” 1 R.I.L–R v. Johnson, 80 F. Supp. 3d 164, 188-89 (D.D.C. 2015). In recent times, the potential migrants to whom the federal government has sought to send such a message are, by and large, from Mexico and Central America. Emerging empirical research, however, provides little to no evidence that detention has had the type and level of deterrent effect desired by the federal government. 2 See, e.g., Adam Cox & Ryan Goodman, Detention of Migrant Families as “Deterrence”: Ethical Flaws and Empirical Doubts, Just Security (June 22, 2018),; Tom K. Wong, Ctr. for Am. Progress, Do Family Separation and Detention Deter Immigration? 1-5 (2018), Why might this be so? This Essay addresses this question by examining three key “deterrence hurdles” that present challenges to detention as deterrence.

Part I of the Essay provides the policy context and theoretical framework for thinking about immigration detention as deterrence. I describe how and why U.S. policymakers have tried to use immigration detention to deter unauthorized migration. I then introduce the three deterrence hurdles that Paul Robinson and John Darley explore in their study of substantive criminal law. 3 Paul H. Robinson & John M. Darley, Does Criminal Law Deter? A Behavioural Science Investigation, 24 Oxford J. Legal Stud. 173 (2004); see also Paul H. Robinson, Distributive Principles of Criminal Law: Who Should be Punished How Much? 21-71 (2008). These hurdles or prerequisites to deterrence are the legal knowledge hurdle, the rational choice hurdle, and the perceived net cost hurdle. Part II considers how each of these hurdles likely operates in the immigration law context and perhaps in even more complex and unexpected ways than in the criminal law context. Part III maps an agenda for future research by highlighting some of the key inquiries that can shed light on these complexities and their implications for policymaking.

I. Background

A. Policy Context

Immigration detention in the U.S. is civil confinement for which the officially stated purpose is to facilitate the removal of individuals who do not have permission to remain in the country. With important exceptions, individuals whose immigration proceedings are pending in immigration courts may be released so that they can wait for the adjudication of their cases on the outside. Such a release requires that the individual not pose a danger to the public. In addition, if the individual is deemed to pose a flight risk, the posting of a cash bond and/or compliance with certain supervision conditions (such as reporting requirements or the wearing of an ankle monitor) may be required to mitigate that risk.

The use of immigration detention to deter unauthorized migration can be traced back to the U.S. detention policy targeting Cubans and Haitians arriving by boat in the early 1980s. Some observers argued that the growing influx of these migrants was due to U.S. policies that paroled the migrants into the U.S. with work authorization pending adjudication of their immigration proceedings. There was growing perception that these policies incentivized unauthorized migration. In response, a federal task force considered alternative policies and recommended that the government “detain as a matter of course all arriving immigrants who could not establish a prima facie claim for admission to this country.” 4 Michele R. Pistone, Justice Delayed Is Justice Denied: A Proposal for Ending the Unnecessary Detention of Asylum Seekers, 12 Harv. Hum. Rts. J. 197, 226 (1999) (quoting Louis v. Nelson 544 F. Supp. 973, 979-80 (S.D. Fla. 1982)). The Reagan Administration adopted this recommendation, reasoning that the new detention policy would “more effectively deter illegal immigration to the United States.” 5 Id. at 227 n.166 (quoting Administration’s Proposals on Immigration and Refugee Policy: Joint Hearing Before the Subcomm. on Immigration, Refugees, and Int’l Law of the H. Comm. on the Judiciary, 97th Cong. 6 (1981) (statement of William French Smith, Att’y Gen. of the United States)). Notably, the policy was adopted despite the Attorney General’s caution that “[d]etention could create an appearance of ‘concentration camps’ filled largely by blacks.” 6 Louis v. Nelson, 544 F. Supp. 973, 980 n.19 (S.D. Fla. 1982) (quoting Plaintiffs’ Exhibit 4 at 8), aff’d in part & rev’d in part, Jean v. Nelson, 711 F.2d 1455 (11th Cir. 1983).

In more recent times, detention as deterrence has been used to justify long-term family detention. In response to the increasing number of Central American families and children arriving at the southwest border seeking asylum in 2014, Jeh Johnson, the Secretary of Homeland Security during the Obama administration, announced a “No-Release Policy.” Under this policy, the Department of Homeland Security (DHS) detained the families and generally refused to consider them for release on bond, recognizance, or other conditions of release. This was true even for families who were found to have a “credible fear” of persecution, which meant that there was a “significant possibility” of their receiving asylum. 7 Karen Musalo & Eunice Lee, Seeking a Rational Approach to a Regional Refugee Crisis: Lessons from the Summer 2014 “Surge” of Central American Women and Children at the US-Mexico Border, 5 J. on Migration & Hum. Security 137, 142, 145 (2017); see also 8 U.S.C. § 1225(b)(1)(B)(v) (“‘credible fear of persecution’ means that there is a significant possibility . . . that the alien could establish eligibility for asylum”).

In announcing this policy, Secretary Johnson declared: “Frankly, we want to send a message that our border is not open to illegal migration, and if you come here, you should not expect to simply be released.” 8 Julia Preston, Detention Center Presented as Deterrent to Border Crossings, N.Y. Times (Dec. 15, 2014), This policy was challenged in a federal district court, which issued a preliminary injunction against the No-Release Policy. 9 R.I.L–R v. Johnson, 80 F. Supp. 3d 164, 170, 191 (D.D.C. 2015). Further, the court affirmed that general deterrence was an impermissible rationale in a civil detention context. The court also noted that the government had failed to present empirical evidence that the detention policy achieved its desired effect of deterring potential migrants from Central America.

Nonetheless, detention has increased and the idea that detention can deter migration has gained new currency under the current administration. The Trump Administration has expanded immigration detention by subjecting a greater number of individuals to detention and by making it more likely that detention will be prolonged. The Trump Administration’s determination to expand detention is largely borne out of the belief that releasing immigrants into the interior while their immigration cases are pending before immigration courts (a practice that President Donald Trump has decried as “catch and release”) encourages others to attempt to enter the U.S. For example, as part of its “zero-tolerance” policy implemented in April 2018, the Trump Administration began detaining and criminally prosecuting all migrants apprehended at the U.S.-Mexico border.

This policy resulted in the separation of thousands of migrant children from their parents, since children cannot be held in criminal detention. The ensuing media storm and public outcry, however, forced the Trump Administration to retract this policy and instead propose detaining families together in immigration detention facilities throughout the pendency of criminal and immigration proceedings. In a court filing seeking to implement this family detention policy in circumvention of its obligations under the Flores settlement agreement to limit the length of detention for minors, the government argued: “The more constrained DHS’s ability to detain families together . . . the more likely it is that families will attempt illegal border crossing.” 10 Defendants’ Ex Parte Application for Limited Relief from Settlement Agreement at 14, Flores v. Sessions, No. 85-4544-DMG (AGRx) (C.D. Cal. June 21, 2018). For the district court’s decision denying the government’s request for relief, see Flores v. Sessions, No. 85-4544-DMG (AGRx), 2018 WL 4945000 (C.D. Cal. July 9, 2018). In the aftermath of the district court’s rejection of this argument, the Trump Administration has proposed new regulations that would allow indefinite family detention. 11 Apprehension, Processing, Care, and Custody of Alien Minors and Unaccompanied Alien Children, 83 Fed. Reg. 45,486 (proposed Sept. 7, 2018) (to be codified at 8 C.F.R. pt. 212 & 236, 45 C.F.R. pt. 410).

President Trump himself has stated that an increasing number of Central American families are traveling with their children to seek entry into the U.S. because U.S. law prevents the detention of children for lengthy periods. 12 David Nakamura & Nick Miroff, Trump Announces Plan to Block Some Migrants from Seeking Asylum at the U.S.-Mexico Border, Offers Few Details, Wash. Post (Nov. 1, 2018), Applying the same logic, President Trump has also threatened to build tent cities for Central Americans arriving in caravans, declaring: “We’re putting up massive cities of tents . . . . You don’t have to release.” 13 Id. Of note, the migrants who are the targets of this deterrence strategy include non-asylum seeking and asylum-seeking migrants. 14 For a detailed overview of laws and policies on immigration detention, see Emily Ryo & Ian Peacock, A National Study of Immigration Detention in the United States, 92 S. Cal. L. Rev. 1, 10-14 (2018).

B. Theoretical Framework

Paul Robinson and John Darley consider whether criminal law deters crime. They present a variety of social-psychological and behavioral economics evidence to show that (a) there are prerequisites or special conditions that must be satisfied before criminal law can deter crime, and (b) these prerequisites are rarely ever satisfied in the real world. In their own words, the argument is as follows:

We do not deny that having a criminal justice system that administers punishment can have a deterrent effect. It is even possible that changes in police procedures or allocation of resources can have an effect on crime rates. But we are profoundly skeptical that the formulation of criminal law rules or even sentencing policies or practices can have the deterrent effect that common wisdom assumes it has. 15 Robinson & Darley, supra note 3, at 197.

Robinson and Darley’s skepticism is rooted in their recognition of the myriad challenges to deterrence presented by the legal knowledge hurdle, the rational choice hurdle, and the perceived cost hurdle. The legal knowledge and rational choice hurdles are critiques of the basic assumptions of the rational choice framework underpinning the deterrence theory. The perceived cost hurdle, on the other hand, is a critique from within the rational choice framework that suggests that even if the basic assumptions of the rational choice framework are satisfied, noncompliance may still follow. In the next Part, I briefly explain how each of these hurdles have been conceptualized in the criminal law context. I then analyze how these hurdles might operate in the immigration law context in ways that cast doubt on the assumption of policymakers that immigration detention will have a substantial deterrent effect.

Before proceeding, it is important to note that although my analysis draws heavily from research on criminal law and criminal offender decision-making, immigration detention is civil confinement under the law. For purposes of my analysis in this Essay, however, it is important to note that immigration detention takes place in jail or jail-like facilities, confinement conditions are often worse than criminal incarceration, and immigrant detainees themselves experience their detention as a form of punishment. 16 See Emily Ryo, Fostering Legal Cynicism Through Immigration Detention, 90 S. Cal. L. Rev. 999, 1024-31 (2018) (describing conditions of confinement in immigration detention that are indistinguishable from, or even worse than, criminal incarceration, from the perspective of detainees). This feature of detention is precisely why some policymakers believe prospective migrants will fear—and thus will be deterred by—detention.

II. Deterrence Hurdles in the Immigration Law Context

A. The Legal Knowledge Hurdle

The first prerequisite to deterrence is that people know the law. The legal knowledge that is relevant here is not technical knowledge about some esoteric regulations that only practicing lawyers would be expected to possess in their respective specialty areas. Rather, legal knowledge for the purposes of this discussion refers to a basic understanding about whether certain conduct is permitted or prohibited under the criminal law, and the legal sanctions associated with violating that law.

Studies show that people are not particularly knowledgeable about the criminal law rules and the magnitude of legal penalties associated with various crimes. 17 See Robert Apel & Daniel S. Nagin, Perceptual Deterrence, in The Oxford Handbook of Offender Decision Making 121, 125-28 (Wim Bernasco et al. eds., 2017); Robert Apel, Sanctions, Perceptions, and Crime: Implications for Criminal Deterrence, 29 J. Quantitative Criminology 67, 73-81 (2013). While some studies show that individuals with criminal and arrest experience tend to have relatively more accurate perceptions of criminal penalties compared to the noninstitutionalized population, the absolute level of knowledge even among offenders is surprisingly low. For example, in David Anderson’s study of male inmates at two medium-security state prisons and a county jail, the inmates were asked: “When you committed the crime, did you know what the likely punishment would be if you were caught?” Only 22% of the inmates reported, “I knew exactly what the punishment would be” for the crime (note that this does not mean they were correct in their knowledge); 18% reported, “I had no idea, or thought I knew but was wrong”; and another 35% reported, “I didn’t think about it.” 18 David Anderson, The Deterrence Hypothesis and Picking Pockets at the Pickpocket’s Hanging, 4 Am. L. & Econ. Rev. 295, 303 tbl.1 (2002).

What do potential migrants know about U.S. immigration laws in general and immigration detention policies in particular? For example, what do they know about laws that mandate their detention, how long they could be detained under the law, and under what circumstances they could be released from detention? To my knowledge, no studies have directly addressed these questions. But there are good reasons to think that the legal knowledge hurdle presents an even greater and more complex set of challenges in the immigration law context than in the criminal law context.

First, in criminal law, the relevant target populations of the law are residents of the U.S. In contrast, in immigration law, the target population of concern (for the purposes of the present discussion) are individuals who reside outside of our territorial boundaries and under the jurisdiction of foreign governments. In the current era of globalization, cross-country knowledge dissemination may be rapid and relatively seamless in certain domains and for certain populations. However, with respect to foreign laws and policies, the potential for gaps, lags, and distortions in knowledge transmission may be substantial. Some of the challenges to knowledge transmission in this context may include language divides, cross-country differences in legal systems and legal cultures, and the lack of access to modern communication technology such as the Internet for individuals residing in rural or impoverished areas. These transmission challenges are likely to be magnified in a rapidly evolving policy landscape shaped not only by federal statutes but also executive orders, court decisions, and state laws related to enforcement.

There are a number of other factors relating to the nature of the information sources that may also impact the efficacy and efficiency of cross-border legal knowledge transmission. The origin country governments play an important role in elevating legal knowledge in communities that send migrants abroad. For example, the U.S. government has collaborated with the governments of Mexico, Honduras, El Salvador, and Guatemala for years to launch “awareness campaigns” to warn their citizens about the dangers of illegal border crossing, and to spread the message that illegal border crossers will be deported. 19 See CBP Addresses Humanitarian Challenges of Unaccompanied Child Migrants, U.S. Customs & Border Protection (Sept. 1, 2014), (accessed by searching for “” in the Internet Archive Wayback Machine); see also U.S. Customs & Border Protection, Fiscal Year 2015: CBP Border Security Report 1 (2015), No studies have assessed the impact of these campaigns. But high levels of governmental distrust in these countries suggest that many individuals might discount or reject the messages contained in these campaigns. Exacerbating this situation are competing counter-messages disseminated by human smugglers who may provide inaccurate or misleading legal information designed to incentivize migration. 20 See, e.g., Nick Miroff, Faced with Migration “Crisis,” U.S. Border Chief Finds No Easy Fix in Central America, Wash. Post (Sept. 30, 2018),; Ron Nixon, U.S. Campaign Against Migration Goes Unheard, or Unheeded, in Guatemala, N.Y. Times (Oct 7, 2018), It is also possible that these types of awareness campaigns can increase rather than decrease migration, consistent with tax noncompliance studies that show that publicizing the “tax gap” makes salient the prevalence of noncompliance, provoking a higher incidence of tax cheating. 21 Dan M. Kahan, Trust, Collective Action, and Law, 81 B.U. L. Rev. 333, 342 (2001) (“When government engages in dramatic gestures to make individuals aware that the penalties for tax evasion are being increased, it also causes individuals to infer that more taxpayers than they thought are choosing to evade.”).

Another key source of legal knowledge for prospective migrants is likely to be past and current migrants in the U.S. Here too, there are difficult challenges that prospective migrants face in obtaining reliable and useful legal knowledge. First, past and current migrants themselves may not have accurate information about the law. In my study of long-term immigrant detainees, I found that many detainees lacked even the most basic knowledge about procedural and substantive aspects of U.S. immigration law, as well as crucial facts about their own legal cases. 22 See Ryo, supra note 16, at 1034-36. Second, insofar as detention and deportation is stigmatized in some communities abroad or in the U.S., 23 See, e.g., Bryan Roberts et al., Voluntary and Involuntary Return Migration, in Deportation and Return in a Border-Restricted World: Experiences in Mexico, El Salvador, Guatemala, and Honduras 3, 5-6 (Roberts et al. eds., 2017) (discussing stigma of deportation); Liza Schuster & Nassim Majidi, Deportation Stigma and Re-migration, 41 J. Ethnic & Migration Stud. 635, 635-36 (2015) (same). past and current migrants are unlikely to be forthcoming with family and friends about their personal experiences with, and knowledge about, the U.S. legal system.

Finally, research on criminal law suggests that in the absence of concrete knowledge about the law, people commonly assume that the law is what they think it should be based on their own intuitions of justice. 24 See John M. Darley et al., The Ex Ante Function of the Criminal Law, 35 L. & Soc’y Rev. 165, 179-81 (2001). Insofar as this finding extends to areas of law beyond criminal law, the implications are substantial for the legal knowledge hurdle in immigration law. In my study of unauthorized migrants from Mexico, I found that the migrants viewed their decision to cross the border illegally as the only moral choice given their commitment to work and their families. 25 Emily Ryo, Less Enforcement, More Compliance: Rethinking Unauthorized Migration, 62 UCLA L. Rev. 622, 628, 652-53 (2015). Importantly, the migrants drew a sharp distinction between immigration law and other laws that prohibited self-evidently harmful acts—that is, laws that prohibit mala in se. 26 Id. at 654-55. In this context, it would be unsurprising to find that when prospective migrants lack knowledge about U.S. immigration laws in general and detention policies in particular, they assume that the law cannot be so unjust as to confine them in jail or jail-like facilities upon apprehension.

B. The Rational Choice Hurdle

The second prerequisite for deterrence is that individuals are able to conduct relevant cost-benefit analyses to arrive at utility-maximizing decisions. Yet a long line of psychological research and behavioral economic studies contradict this conception of human rationality. 27 See generally Gerd Gigerenzer & Daniel G. Goldstein, Reasoning the Fast and Frugal Way: Models of Bounded Rationality, 103 Psychol. Rev. 650 (1996) (reviewing concepts and studies); Daniel Kahneman, Maps of Bounded Rationality: Psychology for Behavioral Economics, 93 Am. Econ. Rev. 1449 (2003) (same). In the criminal law context, Robinson and Darley highlight a variety of factors that can distort potential offenders’ reasoning and interfere with their motivations and ability to make the proper calculations required for deterrence. For example, Robinson and Darley discuss issues such as the potential offenders’ states of mind (for instance, desires for revenge or retaliation, and rage or anger) that can induce flawed reasoning at the moment of contemplating a crime. Robinson and Darley also discuss contextual factors such as group decision-making (for instance, crimes committed by gang members) that can promote underestimation of risks. 28 Robinson & Darley, supra note 3, at 178-82.

In the rest of this Subpart, I discuss how these insights can help us to understand the nature of the rational-choice hurdle facing many prospective migrants from countries such as Mexico and Central America. First, given that “migration decisions are inherently risky,” 29 William A. V. Clark & William Lisowski, Prospect Theory and the Decision To Move or Stay, 114 Proc. Nat’l Acad. Sci. U.S. E7432, E7432 (2017). there is likely a differential selection into migration based on individuals’ risk tolerance. Research on internal and international migration provides empirical evidence of this selection process, indicating that migrants tend to be more risk-tolerant than non-migrants. 30 See, e.g., Vladimir Balaz & Allan M. Williams, Risk Attitudes and Migration Experience, 14 J. Risk Res. 583, 585 (2011); Clark & Lisowski, supra note 29, at E7437-39; David A. Jaeger et al., Direct Evidence on Risk Attitudes and Migration, 92 Rev. Econ. & Stat. 684, 686-88 (2010). These findings are important because studies on risk-taking suggest that risk-tolerant or risk-seeking individuals are more likely to underestimate actual risks involved in an activity. 31 See Robinson & Darley, supra note 3, at 179 n.15 (discussing studies on risk perceptions among criminal offenders). In addition, migrants who have been previously detained may suffer from the perceptual “resetting” phenomenon (a variant of the gambler’s fallacy) that leads them to erroneously underestimate the odds of being detained again. 32 Gambler’s fallacy refers to an erroneous belief that a certain random event is less or more likely given the occurrence of a previous event.

Second, there are many contextual factors that may complicate the decision-making process of prospective migrants. Although dangers of the journey to the U.S. may be generally well known (for instance, dehydration, heat, cold, assault, death), prospective migrants may be unable to fully appreciate these dangers ex ante under conditions of acute or chronic stress to which they are subject in their countries of origin. Individuals fleeing their countries due to poverty or violence are likely under substantial cognitive, physiological, or psychological stress. Research on the effects of stress and emotions on decision-making find that stress and certain types of emotions can deteriorate overall decision-making performance through a variety of mechanisms. 33 For helpful review, see Katrin Starcke & Matthias Brand, Effects of Stress on Decisions Under Uncertainty: A Meta-Analysis, 142 Psychol. Bull. 909, 921-26 (2016).

Third, although deterrence theory assumes individual decision-making, research shows that migration is typically a household decision. For example, migrants do not seek to maximize individual income as much as they seek to manage risks on the part of their households to deal with missing markets for capital, credit, and insurance. 34 Douglas S. Massey et al., Theories of International Migration: A Review and Appraisal, 19 Population & Dev. Rev. 431, 436-38 (1993). Recognizing that migration decisions are collective household decisions is important for at least two reasons. First, insofar as a given migrant is seeking to maximize her family’s welfare instead of only her individual welfare, the risks that such a migrant is willing to assume may be much greater than risks that a migrant who is thinking only of her own welfare. Second, studies show that individual decision-making, including offender decision-making, is altered in many ways by the presence of others and by small group dynamics. For example, decisions made by groups are generally more extreme than the decisions made by individual members (a phenomenon known as “group polarization”). 35 Andrew Hochstetler, Co-Offending and Offender Decision-Making, in Encyclopedia of Criminology & Criminal Justice 570, 571-72 (Gerben Bruinsma & David Weisburd eds., 2014).

It is also possible that certain social dynamics may allow individuals to largely bypass altogether the decision-making process required by deterrence theory. For example, in some communities with high rates of outmigration, migration may be considered socially normative or taken for granted. In such contexts, individualized cost-benefit analysis may be rendered peripheral or nonmaterial. As one migrant explained in Lawrence DeLuca and colleagues’ study of how migrants understand the risks of U.S.-Mexico border crossing: “[W]e follow the foot prints left by others, you follow them and that is it.” 36 Lawrence A. DeLuca et al., United States–Mexico Border Crossing: Experiences and Risk Perceptions of Undocumented Male Immigrants, 12 J. Immigrant Minority Health 113, 117 (2010). Migration may also become associated with socially desirable traits such as manhood or courage. 37 Chad Broughton, Migration as Engendered Practice: Mexican Men, Masculinity, and Northward Migration, 22 Gender & Soc’y 568, 569 (2008). Such a dynamic can also minimize the salience of individualized cost-benefit analysis. This situation is powerfully illustrated in this account shared by another migrant in Lawrence DeLuca and colleagues’ study:

There are people that talk about it and they tell us . . . listen, I crossed but you have to go through this and that and a lot of walking. They convince us to try, they say give it a try, you can do it. You are a man and you can withstand it. So I thought, well why not, I would give it a try to see if I could make it. 38 DeLuca et al., supra note 36, at 118 (emphasis added).

My argument is not that migrants are any less “rational” than non-migrants. I have discussed elsewhere the importance to policymaking of understanding unauthorized migrants as complex moral agents. 39 Ryo, supra note 25, at 664. Likewise, it is important to recognize migrants as individuals capable of making reasoned and logical decisions given the information to which they have access. My argument here instead is that migrants—just like anybody else—likely do not engage in the type of cost-benefit analysis demanded by deterrence theory. To be clear, deterrence theory does not require full and perfect “rationality”—only that individual behavior is responsive, in some way on the margin, to threat of sanctions. However, in light of many situational or contextual factors that can easily complicate or render irrelevant the type of individual utility maximization assumed by deterrence theory, the rational choice hurdle presents an important set of challenges for the detention-as-deterrence policy.

C. The Perceived Net Cost Hurdle

The third deterrence hurdle that Robinson and Darley examine is the perceived net cost hurdle. The central issue underlying this hurdle is whether the potential offender will decide that the perceived costs of legal sanction outweigh the perceived benefits of noncompliance. Robinson and Darley discuss a variety of difficulties involved in establishing a punishment probability and severity that would be meaningful to potential offenders. For example, Robinson and Darley point out that lawmakers assume that lengthy prison sentences will generate a greater deterrent effect by increasing the perceived cost of offending. Yet research suggests that this aspect of the cost-benefit calculus is much more complicated than commonly assumed. 40 See Robinson & Darley, supra note 3, at 190.

In the immigration law context, the net cost hurdle requires us to ask whether prospective migrants will choose to stay home even if they understand the law and can properly bring their cost-benefit analysis to bear on their decision-making. For many unauthorized migrants and asylum seekers, the benefits of migration will often outweigh any perceived costs, including harsh enforcement measures. The following perspective of a fifteen-year-old boy in Honduras who is planning his journey to the U.S. is instructive: “Here we live in fear . . . I’ve thought of it a lot. I will go.” 41 UNICEF, Broken Dreams: Central American Children’s Dangerous Journey to the United States 2 (2016), This resolve is to be expected given his life circumstances. His sister was killed by unknown assailants, five teenagers were gunned down outside the youth outreach center where he used to hang out, and the grocer in his neighborhood was shot.

Research suggests that such calculations and conclusions by prospective migrants from Central America and Mexico are common. Jonathan Hiskey and colleagues’ quantitative study showed that the majority of survey respondents in Honduras were well aware of the heightened risks of making the journey to the U.S. and the greater chance of deportation migrants faced compared to the previous year. However, the study also found that this awareness had no effect on their intentions to migrate to the U.S. By contrast, the respondents’ experiences of crime victimization were significant predictors of emigration intentions. The authors concluded: “Hondurans are far more driven by a desire to ‘leave the devil they know’ than they are dissuaded to leave by the possible risks that may await them.” 42 Jonathan T. Hiskey et al., Leaving the Devil You Know: Crime Victimization, US Deterrence Policy, and the Emigration Decision in Central America, 53 Latin Am. Res. Rev. 429, 442 (2018).

The accounts of migrants at the border in the aftermath of the Trump Administration’s announcement to institute indefinite family detention are also instructive. Take the case of a 26-year-old mother from Guatemala with two young children who was interviewed at the border in July 2018. She had heard about the Trump Administration’s plan to expand family detention to replace the family separation policy: “It worried her, but it did not change her mind. She wants her children to grow up without the constant fear they would live with in Guatemala, she said.” 43 Kate Morrissey, Asylum Seekers in Tijuana Not Deterred by Immigration Policy Changes, San Diego Union Trib. (July 1, 2018),

From a deterrence standpoint, the foregoing discussion suggests that the perceived costs of migration have not yet reached a point where they offset the potential benefits that migrants associate with living and working in the U.S. This is precisely the position of some observers who call for even stricter enforcement measures and more punitive legal sanctions. Such a position, however, raises serious questions that are fundamental to any deterrence policy—namely, how much enforcement or punishment is enough or possible and at what cost to society?

Immigration detention is costly. In fiscal year 2018, the federal government is projected to spend over $3 billion, or $8.43 million per day, on immigration detention. 44 Laurence Benenson, The Math of Immigration Detention, 2018 Update: Costs Continue to Multiply, Nat’l Immigr. F. (May 9, 2018), Apart from economic costs, there are many other negative social and political consequences of detention—not only to detainees themselves but to U.S. citizens and American society as a whole. 45 Emily Ryo, Understanding Immigration Detention: Causes, Conditions, and Consequences, 15 Ann. Rev. L. & Soc. Sci. (forthcoming 2019) (discussing civic and democratic costs of detention in the form of intergenerational and international diffusion of distrust about the U.S. legal system and the rule of law). There are also what Adam Cox and Ryan Goodman refer to as “ethical redlines” to deterrence policies. 46 Cox & Goodman, supra note 2. One can imagine extreme enforcement strategies designed to inflict maximum human suffering that will most certainly ensure behavioral changes, but such strategies are bound to violate fundamental moral principles, not to mention legal standards. In short, there are practical, legal, and moral limits to increasing the costs of migration.

III. Directions for Future Research

Thus far, I have drawn upon what we know from the existing literature in psychology, behavioral economics, and criminology to suggest that detention as deterrence is unlikely to operate in the way that some policymakers might expect or desire. This is not to say that detention will not deter anyone. On the contrary, detention may indeed serve as an effective deterrent for some individuals.

For example, it is possible that repeat border crossers may be more deterrable than first-time crossers, given that repeat crossers may have a higher level of legal knowledge than first-time crossers. Alternatively, repeat crossers may be less deterrable than first-time crossers given that repeat border crossers are more likely to have family members in the U.S. 47 A recent study shows that individuals with family members in the U.S. are more likely to intend to return to the U.S. after deportation. See Jeremy Slack et al., In Harm’s Way: Family Separation, Immigration Enforcement Programs and Security on the US-Mexico Border, 3 J. on Migration & Hum. Security 109, 114-15 (2015). Individuals migrating in search of work may be more deterrable than asylum seekers and refugees. On the other hand, “economic migrants” may experience circumstances that they deem just as dire as those fleeing political persecution. The foregoing discussion underscores the need for research on the varying “height” of deterrence hurdles for different types of migrants. In the remainder of this Part, I highlight other key questions that detention-as-deterrence policy raises that require careful investigation.

The first set of questions relates to how the deterrence hurdles that I have discussed above actually operate on the ground. Take the legal knowledge hurdle, for instance. Conspicuously absent in the existing literature is research about how legal knowledge is obtained and disseminated throughout migrant communities. What do prospective migrants know about immigration detention, how do they know it, and how do they transmit that knowledge? What role do governmental and nonprofit organizations play in disseminating legal knowledge? What is the relationship between the migrants’ knowledge and their trust in government—both their own governments and the U.S. government? Do migrants’ views about the morality of their actions shape their perceptions of not only what they believe the law ought to be, but also what the law is in this area?

The second set of questions that the detention-as-deterrence policy raises is whether the policy serves other goals even if it does not actually deter migrants. Margaret Taylor argues that detention serves an important symbolic function of “convinc[ing] the general public that something is being done about a particular problem.” 48 Margaret H. Taylor, Symbolic Detention, 20 Def. Alien 153, 155 (1997). Relatedly, detention could “bring about an overall atmosphere of deterrence and compliance within the country.” 49 Id. at 157 (quoting Doris Meissner, Commissioner, U.S. Immigr. & Naturalization Serv., Address Before the Carnegie Endowment for International Peace (Sept. 10, 1996), in Fed. News Serv., Sept. 1996) (emphasis added). I refer to this use of detention as symbolic deterrence. There are important empirical questions that arise from recognizing detention as symbolic deterrence. Which groups of migrants have been targeted for symbolic deterrence? What are the actual messages received by different segments of the public, and how do they integrate those messages to form attitudes toward the government and immigrants? To what extent does symbolic deterrence activate distrust and legal cynicism, rather than trust and confidence, in the U.S. government—both domestically and internationally?

The third set of questions that detention as deterrence raises relates to the possibility that detention might be deterring legal behavior. Specifically, detention likely deters some detainees from pursuing claims of relief from removal to which they are legally entitled to pursue. For example, one attorney whose client had fled Honduras after her son had been killed by a gang member recounted how her client had passed the initial asylum screening but continued to be detained by immigration officials. Eventually, the client decided to give up her legal case: “She just couldn’t stomach those conditions . . . . She knew she had many more months in there, and ICE clearly wasn’t going to release her. She just couldn’t bear it.” 50 Michael Barajas, What the Rise of Immigrant Detention Means for Some of the Most Vulnerable People Seeking Refuge on Our Border, San Antonio Current (Nov. 1, 2016, 6:00 AM), How prevalent is this form of deterrence of legal behavior? Which groups of detainees are most vulnerable to its operation? Under what confinement conditions and stages in detention do detainees waive their legal rights?


In many areas of law—including in immigration law—government officials often assume, without sound theoretical or empirical basis, that legal or policy changes can change behavior. Yet criminal deterrence literature suggests that people generally do not know the law, are bad at rational decision-making, and even if they can make rational decisions, will choose to commit the crime because the perceived benefits often outweigh the perceived costs. Drawing on these insights, this Essay has sketched in broad strokes some of the hurdles to deterrence in the immigration law context. To be sure, these hurdles will take on different forms and varying degrees of salience for different migrants depending on their legal experiences and motivations for migrating. On the whole, however, the hurdles that this Essay has analyzed constitute important challenges to the operation of detention as deterrence. Importantly, the complexity of human cognition, behavior, and social dynamics that led Robinson and Darley to question the deterrence effects of criminal law may be even more magnified in the immigration law context.

From a research standpoint, my analysis should serve as a call and a road map for researchers to develop a rigorous field of study devoted to understanding how prospective migrants around the world actually make decisions and choices under uncertainty, and how they respond to legal and policy changes intended to alter their behavior. From a policy standpoint, this Essay offers a new framework for rethinking the use of immigration detention to achieve deterrence goals.

* JD, PhD; Associate Professor of Law & Sociology, University of Southern California Gould School of Law. This research was supported by the Carnegie Corporation of New York. I am grateful to Adam Cox, Jon Hiskey, Paul Robinson, Daria Roithmayr, and Margaret Taylor for their insightful comments on earlier drafts of this essay. The statements made and views expressed are solely the responsibility of the author. Direct all correspondence to Emily Ryo, USC Gould School of Law, 699 Exposition Blvd Los Angeles, CA 90089. Email: