The Yale Law Journal


Abolish ICE . . . and Then What?

07 Nov 2019

abstract. In recent years, activists and then politicians began calling for the abolition of the United States’s interior immigration-enforcement agency: U.S. Immigration and Customs Enforcement (ICE). Many people have misinterpreted the call to “Abolish ICE” as merely a spontaneous rhetorical device used to express outrage at the current Administration’s brutal immigration policies. In fact, abolishing ICE is the natural extension of years of thoughtful organizing by a loose coalition of grassroots immigrant-rights groups. These organizations are serious, not only about their literal goal to eliminate the agency, but also about not replacing it with another dedicated agency of immigration police. Accordingly, the proposal to eliminate ICE necessarily raises the question of how, in a post-ICE world, the United States would enforce its immigration laws. Missing from the public discourse, however, is an affirmative vision for the mechanics of a just and humane immigration-enforcement system that could follow the abolition of ICE. Drawing on lessons from our own and other nations’ past immigration-enforcement schemes, enforcement mechanisms employed by other federal agencies, and interviews with leaders of the “Abolish ICE” movement, I seek to begin to fill this void. This Essay suggests a paradigm shift in immigration enforcement toward the creation of an enforcement scheme that does not rely on detention, mass deportation, or any dedicated agency of immigration police but is nevertheless realistic and effective at increasing compliance with immigration law. The new immigration enforcement principles set forth herein are intended as a starting point for the immigrant-rights movement and for policymakers to use, critique, and improve upon.


The movement to abolish the U.S. Immigration and Customs Enforcement agency (ICE) burst into the national consciousness in the summer of 2018.1 The idea was the natural extension of years of thoughtful organizing by a loose coalition of grassroots immigrant-rights groups.2 They had become convinced that efforts to reform ICE were futile and that a more radical approach was needed. The Trump Administration’s policy of separating and detaining parents and children and the upset victory of U.S. Representative Alexandria Ocasio-Cortez, who had built her campaign in large part around a call to abolish ICE, created a tipping point for the movement.3 Prominent national politicians quickly lined up behind the idea.4 But that initial rush of support has waned in the past year, as many have struggled to explain how we would enforce immigration laws without ICE.

ICE’s brutality, lawlessness, and ineffectiveness have been well documented in academic literature and, more powerfully, by immigrant-community leaders, who have proposed specific and thoughtful changes to our immigration-enforcement paradigm.5 These changes, however, focus almost exclusively on a negative vision of what we need to eliminate in our current enforcement scheme: detention, deportations,6 and the entanglement of our criminal-justice and immigration systems, among other flaws.7 Lacking in the public discourse is a clear affirmative vision for a practical immigration-enforcement system8 that is not dependent on mass detention and deportation.9 This Essay seeks to offer such a vision for a new immigration-enforcement paradigm10 that is humane and effective but that does not rely on the existence of ICE, or any immigration police agency.

The proposal builds on the work of others who have examined immigration-enforcement strategies within the context of the current agency system,11 as well as upon the literature discussing administrative enforcement in other contexts.12 In addition, a critical and primary source for this Essay is a series of interviews conducted with immigrant-community leaders who generously shared their time and insight.13

This Essay proceeds in two Parts. Part I lays out the case for the abolition of ICE. Part II proposes four key pillars of a humane and effective immigration-enforcement system that could follow ICE’s dissolution.

I. the case against ice

ICE was created in the aftermath of the September 11 attacks14 and was fundamentally ill-conceived from its inception. It has since amassed a notorious record of abuse, illegality, waste, and ineffectiveness. Its predecessor agency, the Immigration and Naturalization Service (INS), had been housed in the Department of Justice (DOJ) and had a combined services and enforcement mission.15 However, in creating ICE, Congress excised the services mission, placing those functions in another agency, U.S. Citizenship and Immigration Services (USCIS). All immigration-related agencies then came under the purview of the new Department of Homeland Security (DHS).16 The message was clear: immigration policy would be a component of the nation’s counterterrorism efforts, and ICE would be singularly focused on punitive enforcement. While ICE’s civil-enforcement efforts, by its own assessment, are unrelated to DHS’s national security mission,17 that mission has powerfully influenced ICE’s heavy-handed tactics.

ICE’s abusive tactics are well documented. The barbarism of separating toddlers from their parents, holding them in cages, and forcing them to defend themselves in complex legal proceedings against trained government lawyers is a daily reality.18 In addition, there is an established and ongoing pattern of “egregious medical neglect in [immigration] detention facilities across the country.”19 Not surprisingly, ICE’s mistreatment of detainees has led to a startling pattern of deaths in ICE custody.20 In addition, ICE has amassed a well-earned reputation as a dishonest,21 racist,22 and rogue agency that regularly flouts legal limits.23

The one area where ICE has demonstrated startling success has been in garnering resources for itself. Its funding has risen from $3.3 billion in 2003, the year after its creation, to $7.5 billion in 2018—an increase of approximately 130%.24 The United States now spends more on immigration enforcement than on all other federal criminal law enforcement combined.25 And what has been gained from this enormous spending increase? Certainly, the number of deportations has skyrocketed.26 But if the goal is to increase compliance with the law, the surge in funding has failed miserably. In 2000, just before the creation of ICE, the country’s undocumented population stood at 7 million.27 DHS’s most recent estimate of the undocumented population is 12 million.28 Thus, while ICE’s resources have more than doubled, the undocumented population has grown by over 70%. Waste and mismanagement are part of the problem,29 but at base, these numbers demonstrate that ICE’s ever-increasing investment in detention and deportation has simply failed to increase compliance with U.S. immigration law.

Demonstrating the failure of ICE’s enforcement paradigm, however, merely begs the question whether there is a better way. Documenting ICE’s irredeemable defects is the beginning, not the end, of the inquiry. The Abolish ICE movement can only succeed if it can plot a path forward with an affirmative vision for a more humane enforcement scheme that is both effective and realistic.

II. pillars of a humane and effective immigration enforcement system

Any successful proposal for a post-ICE immigration-enforcement scheme must achieve two distinct goals. First, the new system must forswear and commit to cure the inhumanity and brutality that characterize immigration enforcement under ICE—most notably by eliminating immigration detention and mass deportation. Second, the new system must be more effective at promoting compliance with immigration law. Some critics of the Abolish ICE movement have suggested that the two goals are mutually inconsistent, but they need not be so.30 Below, I propose four pillars of a new immigration-enforcement paradigm—one that would radically reshape immigration enforcement and create a humane system that could be both more effective at encouraging compliance with U.S. immigration laws and dramatically less expensive. The pillars are foundational elements of such a system that would need to be fleshed out in significant detail in future legislation. I offer them now in broad strokes to help policy-makers and advocates conceive of, and articulate, the basic contours of a new immigration-enforcement paradigm. In doing so, I establish a target to work toward now and to refine into winnable legislation when the political climate allows.31 The proposal draws lessons from our own and other nations’ past immigration-enforcement schemes, from mechanisms employed by other federal agencies, and from the expertise of the Abolish ICE movement leaders interviewed for this Essay.

A. Pillar One: Optimal Enforcement Scaling

DHS was founded with the mission of achieving a “100% removal rate.”32 Accordingly, ICE’s goal from the outset was, and remains, to deport every person potentially subject to deportation.33 This approach, which has driven the extraordinary increase in ICE’s funding, is both wildly impractical34 and dramatically out of step with historical norms. During the twentieth century, when the United States also had significant levels of noncompliance with immigration laws, the nation deported under 25,000 people per year on average.35 Thus far in the twenty-first century, the United States has deported over 300,000 people per year on average—a startling 1,200% increase.36 Accordingly, the first inquiry is to determine whether the current extraordinary levels of punitive enforcement are justified.

Any well-conceived enforcement scheme must identify its optimal scale by balancing the societal costs of punitive enforcement against the marginal compliance such enforcement can achieve and the societal benefits associated with that additional compliance. For example, extremely high levels of enforcement are justified by the Nuclear Regulatory Commission because even low levels of noncompliance risk grave societal harm. In other areas, such as the regulation of marijuana, sex work, or quality-of-life crimes, there is a growing consensus that the cost and collateral harms of high levels of enforcement, the low deterrent value of heavy-handed enforcement, and the relatively minor societal injuries associated with noncompliance, militate in favor of low enforcement levels.

In the administrative arena, the approach of the Internal Revenue Service (IRS) stands out as an example of an intentional low-level punitive enforcement strategy. Initially, the IRS was required to audit all tax returns.37 Over time, this enforcement approach came to be seen as overzealous.38 In response, Congress implemented reforms that caused punitive enforcement activity to decline significantly.39 The IRS now employs a notably low level of punitive enforcement, annually auditing under one percent of returns and prosecuting only a couple hundred people for failure to file.40 This shift away from punitive enforcement and toward compliance assistance has not undermined the IRS’s enforcement of tax laws. Indeed, the United States now enjoys one of the world’s highest tax-compliance rates.41

In the immigration arena, the societal costs of the current high levels of punitive enforcement are profound. The destruction of family units, deaths, mistreatment, and the traumatization of entire communities are well documented. Further, the fiscal costs are extraordinary.42 On the other side of the equation, the harms associated with noncompliance are hotly contested. Immigration restrictionists point principally to three categories of alleged harm: criminality of undocumented immigrants, the cost of providing public benefits and services to undocumented immigrants, and the harms that may flow from labor competition with undocumented workers.43 However, the social-science data establish that these alleged harms are either nonexistent or minor.

Study after study has concluded that undocumented immigrants pose no heightened risk of criminality.44 Undocumented immigrants are also ineligible for virtually all federal benefits programs.45 Indeed, many undocumented workers pay taxes and pay into Social Security and other benefits systems—notwithstanding their ineligibility to receive such benefits—leading some to argue that they are a net benefit to the public coffer.46 Evidence on whether undocumented workers create increased labor competition in low-wage fields is more contested,47 but the overwhelming weight of the evidence demonstrates that undocumented workers are a critical net benefit to the U.S. economy.48 Indeed, much of the literature demonstrates that the greatest harms of immigration noncompliance are suffered by the undocumented population itself.49 Moreover, even if we were to assume some significant level of harm from noncompliance, high levels of enforcement are only justified if they actually reduce noncompliance. The weight of the evidence, however, suggests that ICE’s heavy-handed tactics are of limited value in reducing noncompliance.50

Accordingly, the first pillar requires policy-makers to scrutinize the level of punitive enforcement employed by ICE and to identify the optimal scale. In doing so, they will likely see that the current high level of punitive immigration enforcement is of limited utility in increasing compliance and carries significant fiscal and human costs, which outweigh the relatively minor societal harms associated with the deterrable noncompliance. A humane and effective immigration-enforcement scheme thus requires a radical reduction in the scale of punitive enforcement. This first pillar could be largely implemented through executive direction and without any change in law.

B. Pillar Two: Mandatory Preferences for Compliance Assistance

There are large categories of undocumented immigrants who are both potentially subject to deportation and also eligible to obtain lawful status. For example, someone who came to the United States lawfully, overstayed their visa, but married a citizen may either be deported or allowed to obtain legal permanent residence. Immigration authorities have a choice between two enforcement pathways: punish the noncompliance through deportation, or allow the individual to come into compliance by applying for permanent residence.

Administrative-law scholars have widely praised the trend toward “cooperative enforcement” where administrative agencies outside the immigration context increasingly favor efforts to help entities come into compliance over punitive measures.51 When the regulated parties are corporations rather than immigrants, the government seems comfortable with this approach. It is the approach now favored by federal agencies like the Occupational Safety and Health Administration, the Food and Drug Administration, the Environmental Protection Agency, and the Securities and Exchange Commission.52 Indeed, it is the approach that the IRS has favored in modern times, which has helped it successfully transition away from heavy-handed punitive enforcement.53

But when it comes to immigration, the government has done the opposite.54 Instead of diverting people out of the punitive-enforcement stream—ICE’s deportation machine—into compliance-assistance mechanisms—USCIS application processes—it does the reverse. When an eligible-but-deportable individual applies for an immigration benefit, however, USCIS frequently declines to grant their application and diverts them into the deportation system.55 There are large categories of undocumented individuals eligible to obtain lawful status right now.56 There are other significant categories of people being pushed through the deportation pipeline who will, in time, become eligible for lawful status.57 In addition, there are many more people eligible to obtain or maintain status through mechanisms that are currently only available defensively in deportation proceedings, such as cancelation or withholding of removal.58

Prosecutorial discretion is the mechanism that has generally been used to determine which enforcement pathway to pursue, and it has failed to deliver a reliable preference for compliance assistance. Accordingly, the second pillar of a humane and effective immigration-enforcement scheme is to enact mandatory rules that give people the right to affirmatively pursue pathways to lawful status before they can be subject to any punitive enforcement action. This mandatory preference could be implemented through regulation59 and, like the first pillar, would not require any congressional action.60 In addition, Congress should consider making compliance mechanisms that are presently only available defensively, instead available on an affirmative basis. Most dramatically, implementing this pillar could address the majority of noncompliance if Congress created new affirmative pathways to citizenship for undocumented individuals, as it has done in the past.61 With a significantly reduced investment in punitive enforcement, the United States could increase compliance by redirecting resources to USCIS (or some successor agency responsible for delivering immigration benefits) to speed up processing times, expand legal assistance, and reduce application fees for immigration benefits.

C. Pillar Three: A System of Proportional Consequences

Even if the United States dramatically reduces the scale of punitive enforcement and relies principally on compliance assistance, there will still be situations where individuals cannot be brought into compliance and where enforcement is warranted. Accordingly, there must be some consequences that can be triggered by noncompliance. The current regime relies only on a single penalty—deportation—which is grossly disproportionate to the overwhelming majority of immigration offenses. A binary choice between no penalty and the harshest possible penalty is not how other effective enforcement systems work. Imagine the injustice that would follow if we had a criminal-justice system where the only two choices were no penalty or the death penalty. Accordingly, implementing a new system of scalable penalties is the third pillar of a humane and effective immigration-enforcement scheme.

To create such a system, the United States should implement a set of scalable penalties.62 Fines are the most obvious option and are widely used across the administrative state. As recently as 2000, fines were also a critical part of the immigration-enforcement system. Certain individuals who entered unlawfully were permitted to obtain lawful permanent residency if they paid a $1,000 penalty.63 Any such fines would need to be of sufficient magnitude to be significant to the regulated population but also within their ability to pay. Other scalable penalties could include, for example, delayed access to immigration or other public benefits, lengthened pathways to citizenship, or mandated community-service obligations. As with the former $1,000 penalties noted above, any such scalable penalties would then unlock a previously unavailable pathway to status. Any such new penalties and the ability to access lawful immigration status thereafter would require an act of Congress. As with other administrative schemes, the penalties could be offered as a negotiated resolution of noncompliance by USCIS, or the successor services agency, and if no agreement could be reached, such penalties and any defenses thereto could be litigated through the immigration courts. ICE currently acts as the prosecutor in immigration court. However, as with most other federal administrative schemes, we would not need a separate agency to play such a role, as USCIS could serve that function itself.

D. Pillar Four: Minimizing the Use of Physically Coercive State Power

Even with the reforms outlined above, there will be outlier cases where the agency is unsuccessful in bringing individuals into compliance or where an individual fails to comply with the imposed scalable penalty, and where thereafter, the proportional punishment is deemed to be deportation.64 In such cases, the United States has a legal and moral obligation to find mechanisms to ensure people appear in court and comply with deportation orders without unnecessarily severe deprivations of liberty.

The first step in that effort must be to eliminate the immigration-detention system. Virtually every other federal agency has found a way to enforce its civil administrative scheme without putting people in cages. There is no reason why deportation proceedings or even the deportation process itself must begin with handcuffs. For most of U.S. history, those processes began with notices.65 The question then becomes, how can we more humanely ensure people’s appearance in immigration court and ultimately, for some, their compliance with deportation orders?

The answer begins with due process. The data demonstrate that the most important thing we can do to improve appearance rates in immigration court is to provide lawyers. The most recent publicly available data show that virtually every family who was released from immigration detention and had a lawyer showed up for all of their immigration court hearings (99%).66 Those without lawyers were significantly less likely to consistently appear (76%).67 Lawyers help ensure that individuals have accurate information about the time and place of hearings. Lawyers also remove the terror of walking into an unfamiliar courtroom alone and litigating in one of the most complex arenas of American law, against trained government prosecutors, without any legal training, and often in an unfamiliar language. In addition, there is a wealth of empirical research about how the perception of procedural fairness enhances compliance,68 which also helps explain the impact that lawyers have on appearance rates. Moreover, the feasibility of an assigned-immigration-counsel program has been robustly demonstrated in the City and State of New York, which have had universal appointed counsel programs dating back to 2013.69 To be sure, implementing such a program on a national scale would be costly, but the massive scale-down in punitive enforcement contemplated in pillar one would more than offset any such costs. Providing lawyers is thus the first step toward ensuring that individuals know and comply with their obligations in immigration court.70

In addition, other countries such as Canada have engaged in promising experiments with inducements to improve compliance with deportation orders.71 Canada used financial inducements of up to $2,000 to encourage voluntary compliance with deportation orders. With the United States spending an average of $12,000 per deportation, the cost-saving opportunities of such a program are significant.72 The incentivizing force of money is self-evident but could be particularly powerful in this context, where low-income individuals are often extremely scared about returning to countries where they may lack any means of economic survival. Other inducements for those who promptly comply with deportation orders—such as reduced wait times for readmission and continued access to earned domestic benefits like Social Security—could also be powerful tools to promote compliance.73

Finally, the immigration system can draw on lessons from the criminal-justice reentry movement. It has become accepted wisdom that if we want individuals leaving prison to successfully reintegrate into society, we need to invest in reentry services. Those services include, for example, job training, housing assistance, healthcare planning, and mental-health treatment. We provide these services to people coming out of prison because it is inhumane to deposit individuals on the street without the skills to survive and thrive. But we also deliver these services because they are critical to ensuring that individuals do not reoffend. Reentry services for individuals being deported could serve these same purposes. They could be powerful tools to help individuals reintegrate into their countries of origin, thereby reducing the brutality of deportation and easing the terror that leads some to resist compliance with deportation orders. Successful integration, in turn, is critical to reducing the chances of unlawful return to the United States. As with the financial incentives discussed above, care would need to be taken to ensure that such services themselves do not become an incentive for noncompliance.

Collectively, by providing due process, inducements for compliance, and reentry services, a new enforcement regime could increase compliance, reduce costs, and all but eliminate the use of physically coercive state power.74


ICE has failed by every measure. It has brutalized communities, wasted billions of dollars, and failed to increase compliance with immigration laws. The four pillars set forth above provide a blueprint for how a radically new vision for immigration enforcement could increase compliance while simultaneously reducing the human and fiscal costs of enforcement. This blueprint would allow us to abolish ICE and to create an effective and humane enforcement system without the need for any dedicated agency of immigration police. In place of ICE, the United States would need to increase public investment in an immigration agency, like USCIS, that delivers services to immigrants and that would be responsible for facilitating compliance-assistance programs. Such an agency, however, should not be housed inside DHS. Instead, the immigration agency should be housed in a department, like Health and Human Services, that has a mission consistent with appropriate immigration policy goals: to look after the well-being and economic vitality of immigrants and of the nation as a whole.

The ability to theorize a workable and humane immigration-enforcement system is a necessary but insufficient precursor to realizing the reforms that our nation’s immigration system desperately needs. Substantive immigration reform is a separate and perhaps even more critical need, because it is impossible to construct a just enforcement system if the laws being enforced are themselves unjust. To date, much effort has gone into theorizing needed substantive reforms. However, comparatively little work has been done to think through the structures of a just enforcement system. This Essay seeks to help move our national conversation about immigration enforcement forward by providing a starting point for the immigrant-rights movement and for policy-makers to use, critique, and improve upon.

Professor of Law, Benjamin N. Cardozo School of Law. I am grateful for the feedback and insights shared by the accomplished advocates and immigrant community leaders who generously participated in interviews for this project. In that regard, I would especially like to thank Silky Shah from Detention Watch Network; Mizue Aizeki and Marie Mark from the Immigrant Defense Project; Paromita Shah from Just Futures Law; Javier Valdez, Natalia Aristizabal, and Luba Cortes from Make the Road New York; and Jacinta Gonzalez from Mijente. In addition, I would like to acknowledge and thank Ahilan Arulanantham, Christopher Buccafusco, Michael Herz, Alan Hyde, Tom Jawetz, Amy Taylor, Michael Wishnie, and the participants of the Cardozo School of Law faculty workshop, for their contributions to the research and thought underlying this Essay. Finally, I am indebted to Agnes Baik, Alberto Casadevall, Kara Nowakowski, and Kathleen Wahl for their superb research assistance.