Enforcement Overdrive: A Comprehensive Assessment of ICE’s Criminal Alien Program

Executive Summary

The Criminal Alien Program (CAP) is a massive enforcement program administered by U.S. Immigration and Customs Enforcement (ICE) and has become the primary channel through which interior immigration enforcement takes place. Between two-thirds and three-quarters of individuals removed from the interior of the United States are removed through CAP. Each year, Congress allocates hundreds of millions of dollars to fund this program. Until now, however, little has been known about how CAP works, whom CAP deports, and whether CAP has been effective in meeting its goals.

Based on government data and documents obtained through the Freedom of Information Act (FOIA), this report examines CAP’s evolution, operations, and outcomes between fiscal years 2010 and 2013. That data shows that through CAP’s enormous web, ICE has encountered millions and removed hundreds of thousands of people. Yet, CAP is not narrowly tailored to focus enforcement efforts on the most serious security or safety threats—in part because CAP uses criminal arrest as a proxy for dangerousness and because the agency’s own priorities have been drawn more broadly than those threats.

As a result, the program removed mainly people with no criminal convictions, and people who have not been convicted of violent crimes or crimes the Federal Bureau of Investigation (FBI) classifies as serious. CAP also has resulted in several anomalies, including that it appears biased against Mexican and Central American nationals. Moreover, the number of CAP removals differs significantly from state to state.

ICE’s reliance on CAP to achieve its goals will likely continue as ICE further narrows its focus on removing noncitizens with criminal convictions and continues to seek partnerships with state and local law enforcement to find them. This examination of CAP’s outcomes from fiscal years 2010 to 2013 offers important insights into CAP’s operations over time and its potential impact on communities moving forward. In particular, it raises questions about the ability of a broad “jail check” program to effectively remove serious public safety threats without resulting in serious unintended consequences, such as those described in this report.

Background: CAP’s Expansion over Time

• While CAP was originally conceived as a “jail check” program narrowly tailored to remove noncitizens incarcerated for serious criminal convictions, it has become a massive enforcement web—indeed, the primary mechanism through which ICE removes people from the U.S. interior.
• When the former U.S. Immigration and Naturalization Service (INS) established CAP’s predecessor programs in 1988, only a limited number of crimes rendered a person removable. These crimes were commonly considered “serious” or “violent.” Subsequently, between 1990 and 1996, Congress expanded the criminal grounds for removal to potentially include minor crimes such as drug possession, simple assault, shoplifting, turnstile jumping, and disorderly conduct.
• The scope of CAP ballooned when Congress dramatically increased its funding nearly thirty-fold between fiscal years 2004 and 2008, from $6.6 million to $180 million. In fiscal year (FY) 2013, Congress began to dedicate the funding that Secure Communities had received towards CAP, resulting in a 64 percent increase in CAP funding, from $196.7 million in FY 2012 to $322.4 million in FY 2015.

CAP’s Outcomes: Large Web, Few Removals of Immigrants with Serious or Violent Convictions 

• Between FY 2010 and FY 2013, ICE encountered over 2.6 million persons through CAP, but removed fewer than 508,000. Of that number, 87,426 individuals had a conviction for a violent crime or a crime the FBI classifies as serious (a mere 3 percent of the total number of encounters).
• Out of more than half a million CAP removals that took place between FY 2010 and FY 2013, ICE classified the largest share (27.4 percent) as not “definite criminals”—i.e., ICE recorded no criminal conviction. The second- and third-most prevalent categories of CAP removals were of individuals whose “most serious” criminal conviction involved a “traffic offense” (20 percent) and “dangerous drugs” (18 percent), followed by “assault” (6 percent) and “immigration” (5 percent).
• During this time period, over 4 out of 5 CAP removals were of individuals with either no conviction or of individuals who have not been convicted of a violent crime or one that the FBI classifies as serious.
• Conversely, a very low percentage of individuals removed through CAP were convicted of extremely serious crimes. Specifically, 1.7 percent were convicted of burglary, 1.6 percent were convicted of robbery, 1.5 percent were convicted of sexual assault (including rape), 0.5 percent were convicted of homicide, 0.4 percent were convicted of kidnapping, and 0.1 percent were convicted of arson.

Do CAP Removals Match ICE’s Removal Priorities? 

• ICE has justified its interior removal efforts by stating that its intention is to remove “those convicted of the most serious crimes,” ranking immigrants with convictions as “Level 1, 2, or 3” offenders (in order of decreasing seriousness), and publicly emphasizing that an increasing share of its removals are of Level 1 or 2 offenders. But, in recent years, ICE’s criminal removal priorities have still encompassed many offenders besides those who had committed violent crimes or crimes the FBI classifies as serious.
• While the largest proportion (33 percent) of CAP removals falls into Level 1, almost two-thirds of all Level 1 removals were not associated with convictions for a violent crime or a crime that the FBI classifies as serious. Of all Level 2 removals, 76 percent involved individuals who have not been convicted of a violent crime or a crime classified as serious by the FBI. And among Level 3 removals, crimes other than those categorized as violent or serious accounted for 95 percent of removals.

CAP Appears to Be Biased against Mexican and Central American Nationals 

• Mexican and Central American nationals are overrepresented in CAP removals compared to the demographic profiles of those populations in the United States. People from Mexico and the Northern Triangle (Guatemala, Honduras, and El Salvador) accounted for 92.5 percent of all CAP removals between FY 2010 and FY 2013, even though, collectively, nationals of said countries account for 48 percent of the noncitizen population in the United States. Nationals of those countries, however, are not markedly more likely to be convicted of violent crimes or crimes the FBI classifies as serious.

The Geography of Removals: State-by-State Results 

• The number of CAP removals differs significantly from state to state. And this discrepancy is not associated with the size of the states’ noncitizen populations.  
• The states with the highest rates of removals include several with smaller immigrant populations (i.e., Mississippi, Wyoming, and West Virginia), as well as two with large immigrant populations (Texas and Arizona).
• Although further research is needed to explain these discrepancies in CAP removal rates, it is plausible that the state-by-state outcomes of CAP are related to local cooperation with ICE; ICE’s local capacity and presence in those states; and the availability of public transportation and driver’s licenses for undocumented immigrants.

Introduction

The Criminal Alien Program (CAP) has become the primary program through which Immigration and Customs Enforcement (ICE), a component of the U.S. Department of Homeland Security (DHS), conducts immigration enforcement in the interior of the United States. Between two-thirds and three-quarters of individuals ICE removes from the interior of the United States are removed through CAP—the vast majority in cooperation with state and local law enforcement. Each year, Congress allocates hundreds of millions of dollars to fund CAP. And ICE’s reliance on CAP to achieve its goals will likely continue as ICE further narrows its focus on removing noncitizens with criminal convictions and continues to seek partnerships with state and local law enforcement to find them.

There are reasons to question CAP’s efficacy and efficiency, and to look closely at unintended consequences. As this report discloses, CAP is not narrowly tailored to focus enforcement efforts on the most serious security or safety threats—in part because CAP uses criminal arrest as a proxy for dangerousness and because the agency’s own priorities have been drawn more broadly than those threats. As a result, the program is overbroad and arguably inefficient. CAP also has resulted in several anomalies, including an apparent bias against Mexican and Central American nationals, and the number of CAP removals differs significantly from state to state.

Prior to this report, little was known about how CAP works, whom CAP deports, and whether CAP has been effective in meeting its goals. As the debate grows over whether state and local cooperation with ICE helps or hinders public safety, this report provides crucial information regarding CAP’s evolution, operations, and outcomes between fiscal year (FY) 2010 and August 17, 2013, based on never-before-released government data[5] and documents obtained by the American Immigration Council and the American Immigration Lawyers Association’s Connecticut Chapter through the Freedom of Information Act (FOIA). Yet, as this report highlights, much is still unknown about CAP, and the government’s lack of transparency and poor data collection and sharing means that a full assessment of CAP remains elusive.

What the data does show, however, is that from FY 2010 to August 17, 2013, ICE officers encountered over 2.6 million persons, and removed nearly 508,000 through CAP. As an initial matter, it is unclear why, in a program designed to target removable noncitizens in jails, ICE spends so much time, money, and resources encountering those whom ICE does not even arrest, let alone remove. Moreover, despite this massive enforcement web, only 3 percent of the total number of CAP encounters resulted in removal of an individual with a conviction for a violent crime or a crime the Federal Bureau of Investigation (FBI) classifies as serious. Indeed, strikingly, CAP removed mainly people with no criminal convictions and people who have not been convicted of violent crimes or crimes the FBI classifies as serious. Over four out of five CAP removals from FY 2010 to August 17, 2013—83 percent—fell into these categories. Among those noncitizens whom ICE removed through CAP, the most common “most serious” conviction that ICE recorded was none (27 percent).

Of those removed with a conviction, a sizable proportion had not been convicted of a violent crime or a crime the FBI classifies as serious. The most prevalent “most serious” convictions that ICE recorded were traffic offenses (20 percent) and drug offenses (18 percent). And among those persons removed with traffic and drug convictions, many had committed lesser offenses. For example, 30 percent of those whose most serious conviction was a traffic offense were convicted of non-DUI offenses, and 39 percent of those whose most serious convictions were drug offenses were convicted of possession, with another 23.2 percent not specified. Among all removed individuals whose most serious convictions were marijuana offenses, 53.6 percent were convicted of possession. This data complements other reports showing that DHS’ removals of those with criminal convictions have largely been removals of those who committed drug or lesser offenses—in some cases, with harsh impact compared to their equities in the United States. And even though this report shows that ICE, over time, began better tailoring its CAP removals to its own priorities, still, most removed through CAP did not commit violent crimes or crimes the FBI classifies as serious.

In addition to removing large numbers of minor offenders, ICE has removed through CAP disproportionately high numbers of Mexicans, Guatemalans, and Hondurans compared to the composition by national origin of both the foreign-born and the undocumented populations in the United States. Interestingly, the data also shows that nationals of those countries are not markedly more likely to be convicted of violent crimes or crimes the FBI classifies as serious.

Finally, as states and localities have adopted differing policies regarding cooperation with ICE, state-by-state disparities in CAP removals have emerged. CAP led to the removal of a disproportionately high number of noncitizens in certain states in FY 2013. In particular, states including Mississippi, Wyoming, West Virginia, Kentucky, South Dakota, Nebraska, Texas, and Arizona experienced the highest rates of CAP removals per 1,000 noncitizens. Meanwhile, some states with substantial immigrant populations, including Massachusetts, New York, Maryland, New Jersey, Washington D.C., Connecticut, Florida, and Illinois, experienced comparatively lower rates of CAP removals per 1,000 noncitizens. A possible explanation is the emergence of “community trust” policies in FY 2013, such as the introduction of anti-detainer policies by urban localities, which reduced local cooperation with ICE regarding non-serious offenders, as well as those charged but not convicted. Possible explanations also include ICE’s detention capacity in different states, and the necessity for immigrants to drive in rural states, which increases the chances of contact with local law enforcement or ICE.

Looking forward, as Congress, states, and localities debate the impact of state and local cooperation with ICE on public safety, and ICE attempts to further narrow its enforcement priorities, excluding from removal those with no convictions or only minor convictions, the data in this report is crucial to understand. ICE asserts that it is protecting public safety in the interior of the United States by removing the “worst criminals.” Most noncitizens removed through CAP, however, were not the worst criminals, and many were not criminals at the time of removal. The data calls into question whether CAP is designed to effectively and efficiently achieve the government’s stated policy goals. Given the comparatively few CAP removals involving individuals who committed violent crimes or crimes the FBI classifies as serious, relative to the financial and human cost of deporting minor offenders, the value of current CAP funding levels is questionable.

Background

While CAP was originally conceived as a “jail check” program narrowly tailored to remove noncitizens incarcerated for serious criminal convictions, it has become a massive enforcement web—indeed, the primary mechanism through which ICE removes people from the U.S. interior. ICE has both expanded the target population of CAP’s core “jail check” activities to include those charged with any crime, whether convicted or not, and expanded CAP’s mission outside jails into immigrant communities. Meanwhile, Congress has consistently fueled those expansions with dramatic funding increases.

Rather than operating CAP as a freestanding program with specialized CAP officers, ICE merely designates its CAP work separately for budgetary purposes. In other words, ordinary ICE officers record their work as CAP work when performing CAP duties. As CAP’s former unit chief stated, “[a]ny [ICE] ERO [Enforcement and Removal Operations] officer at any time can conduct a Criminal Alien Program duty,” if (s)he is pursuing removal of a noncitizen with a criminal history. Thus, in FY 2014, CAP was funded for 1,495 full-time employees nationwide, even though, as of 2013, CAP only operationally employed a unit chief and 10 staff officers who oversaw the program from ICE headquarters in Washington, D.C.

CAP’s Traditional “Jail Check” Role

In CAP’s traditional “jail check” function, ICE identifies removable noncitizens who are incarcerated in jails or prisons and initiates removal proceedings against them. To this end, ICE collaborates with over 4,000 federal, state, and local facilities that provide ICE information about the foreign nationals in their custody. Federal facilities are required to report all self-identified foreign inmates to ICE, under an agreement between ICE and the federal Bureau of Prisons. Additionally, certain state and local facilities voluntarily provide ICE with lists of foreign nationals in custody, targeted lists of suspected noncitizens, or access to detainees or records, depending on the level of cooperation. State and local law-enforcement agencies have provided the bulk of CAP’s work. From FY 2010 to FY 2013, only 6.4 percent of CAP encounters were referred from federal facilities, compared to 91.8 percent referred from state or local law enforcement.

Once receiving information, the ICE officer ascertains a noncitizen’s removability either by screening records or traveling to a jail or prison to interview the noncitizen. Before November 2014, if the ICE officer found a noncitizen to be removable, the officer would typically lodge a “detainer” request for state or local law enforcement to hold the noncitizen, interview the noncitizen if necessary, and subsequently initiate removal proceedings if appropriate. Under new November 2014 guidance ending Secure Communities and announcing the Priority Enforcement Program (PEP), ICE officers are directed to request state or local law enforcement to notify ICE of the noncitizen’s release, rather than issue a detainer request (absent special circumstances).

CAP’s Reach into Immigrant Communities

Though CAP primarily functions in jails and prisons nationwide, ICE also has extended the program’s reach into communities through what the agency terms “at-large” activities, including pursuing individuals on criminal probation or parole, and working with other ICE enforcement initiatives, such as the National Fugitive Operations Program. In FY 2013, 97 percent of CAP removals appear to have been executed through CAP’s traditional jail check functions. It is generally unclear, though, when ICE arrests and removes noncitizens in the community, whether ICE counts those arrests as “CAP removals.”

CAP’s Dramatic Expansion in Mission, Scale, and Funding

CAP’s expansion of its “jail check” targets, as well as its activities in communities, evolved out of several events: (1) Congress broadening the legal grounds for criminal deportability, (2) subsequent dramatic funding increases for CAP, (3) ICE establishing the controversial “Secure Communities” program, with its mission to remove every noncitizen deportable for a criminal offense, (4) ICE prioritizing the removal of noncitizens with criminal convictions, and (5) ICE folding Secure Communities’ technology and operations into CAP, and apparently Secure Communities’ broad mission, too.

U.S. immigration-enforcement agencies have prioritized the removal of noncitizens with criminal convictions since at least 1986. When the former U.S. Immigration and Naturalization Service (INS) established CAP’s predecessor programs in 1988, the crimes that resulted in removal were a limited number commonly considered “serious” or “violent”—i.e., murder, gun trafficking, and drug trafficking, which had been designated by Congress as “aggravated felonies.” Subsequently though, between 1990 and 1996, Congress expanded the criminal grounds for removal to potentially include less serious crimes such as drug possession, simple assault, shoplifting, turnstile jumping, and disorderly conduct.

The scope of CAP then ballooned when Congress dramatically increased CAP funding nearly thirty-fold from fiscal years 2004 through 2008, from $6.6 million to $180 million; and in FY 2009 began to allocate to ICE $1 billion per year (and more in later years) through CAP and other programs to identify and remove “criminal aliens” (albeit prioritized “by the severity of [their] crime”) (Table 1).

Table 1: Appropriations for ICE, CAP, and Secure Communities, FY 2004-2016 (in millions)

One of those other programs was “Secure Communities,” through which the fingerprints of those arrested by local law enforcement were shared with ICE. ICE established Secure Communities in 2008 following Congress’ directive to “identify and remove every deportable criminal alien”—not just those convicted of serious or violent crimes. Secure Communities’ mission also included targeting noncitizens in the community, not just those incarcerated in jails. Both these functions were later integrated into CAP, expanding its massive reach.

In June 2010, ICE Director John Morton published a memorandum setting new civil enforcement priorities, with the goal of focusing removals on public safety and other threats. The priorities set out a three-tiered framework, with noncitizens who have criminal convictions as Priority 1—and, within Priority 1, individuals with convictions ranked by the severity of their crimes into three levels (Levels 1, 2, and 3). Subsequently, though, when ICE management set numeric goals for removals with convictions, overall goals were set rather than goals by level. This may have resulted in ICE removing more offenders at lesser levels. For example, ICE management for FY 2012 set an aggressive goal of 225,000 removals with convictions, stated that that goal was “[t]he only performance measure that will count,” and directed officers to “reallocate all available resources” to meet it. In response, an ICE Assistant Field Director in Atlanta offered to “process more petty offenses,” among other proposals.

CAP’s unit chief also testified in 2013 that CAP had targeted “any alien convicted of any crime” since 2010. Additionally, ICE’s internal “Criminal Alien Program Handbook” (May 2013) made clear that CAP targets included individuals “regardless of the status of conviction” (emphasis in original), i.e., even where the criminal charge was pending. CAP’s unit chief confirmed this practice. This guidance exists even though ICE has defined “criminality” in terms of a “recorded criminal conviction,” following Morton’s 2010 memorandum.

On January 22, 2013, ICE completed nationwide deployment of Secure Communities in all local jurisdictions. ICE then stated that it would transfer “full responsibility” of the day-to-day management of Secure Communities to CAP, and began to realign Secure Communities funding towards CAP. In FY 2013, Congress began to direct the funding that Secure Communities had received towards CAP, resulting in a 64 percent increase in CAP funding from $196.7 million in FY 2012 to $322.4 million in FY 2015. By February 2014, CAP had assumed operational responsibility for Secure Communities. In November 2014, DHS Secretary Jeh Johnson then stated that DHS was discontinuing the “Secure Communities program, as we know it,” although it replaced it with PEP.

As CAP subsumed Secure Communities operationally, it appears that CAP also assumed its broad mission. Today, CAP’s stated mission is essentially coextensive with ICE’s general priorities—to remove noncitizens that in ICE’s view are public safety threats, by using criminal history as a proxy for danger. CAP also targets noncitizens not only in jails, but also in the community.

CAP’s Outcomes: Large Web, Few Removals of Immigrants with Serious or Violent Convictions

Between October 1, 2009, and August 17, 2013, ICE encountered 2.6 million individuals under CAP. Such encounters consisted of an interview and/or screening of a person or his or her records to determine citizenship status, nationality, lawful presence, and legal right to remain in the United States. During the same period, ICE arrested over three-quarters of a million individuals through CAP, and removed more than half a million.

Over time, while the number of encounters through CAP has remained relatively high, the number of CAP arrests and removals has gone down. For instance, in FY 2010 there were roughly 329 arrests and 256 removals per 1,000 encounters. By 2013, that rate had decreased to 250 arrests and 139 removals per 1,000 encounters. In other words, CAP’s web has remained large while its arrests—and particularly removals—have decreased (Table 2).

Table 2. CAP Encounters, Arrests, and Removals by Fiscal Year

*Data provided for FY 2013 covers the period of October 1, 2012, through August 17, 2013.

It is unclear why, in a program designed to target removable noncitizens in jails, ICE spends so much time, money, and resources encountering those whom ICE does not even arrest, let alone remove. Indeed, CAP’s massive enforcement web even has ensnared U.S. citizens. In 2012, CAP and ICE management identified 278 U.S. citizens upon whom ICE had placed detainers.

That said, these overall numbers tell us little about the population that CAP has targeted and removed. Examining the details underlying the general numbers tells us more about CAP’s effectiveness and efficiency in identifying, apprehending, and removing genuine public safety threats.

Most interestingly, of the over half-million CAP removals that took place between FY 2010 and FY 2013, ICE classified the largest percentage (27.4%) as not “definite criminals”—i.e., ICE recorded no criminal conviction in its ENFORCE database. The second- and third-most prevalent categories of CAP removals were of individuals whose “most serious” criminal conviction, according to ICE, involved a “traffic offense” (20 percent) and “dangerous drugs” (18 percent), followed by “assault” (6 percent) and “immigration” (5 percent). The FBI’s National Crime Information Center (NCIC) crime coding scheme, and ICE’s internal coding based upon it, classify all drug offenses under the rubric of “dangerous drugs,” without any “non-dangerous” category. Collectively, people with no recorded conviction, or a drug, traffic, or immigration conviction, constituted 70.5 percent of all removals ICE attributed to CAP between FY 2010 and FY 2013 (Figure 1).

Figure 1: Most Serious Criminal Offense Convictions, FY 2010-2013

To more precisely examine CAP’s efforts to remove public safety threats, we classified removed individuals into three broad categories, following the Migration Policy Institute (MPI) and the FBI classification schemes:

1. Individuals with no conviction: I.e., “definite criminal” no, according to ICE.
   
   
2. Serious or violent: Individuals whose “most serious” conviction was a serious or violent conviction according to independent classification schemes. These include (a) crimes that the FBI classifies as Part I; (b) a subset of those the FBI classifies as Part II (i.e., Part II-violent); and (c) domestic abuse crimes. According to the FBI’s decades-old crime classification system, “Part I” includes eight types of crimes, which criminal scholars have commonly understood to represent the most serious crimes. Those eight categories are murder and non-negligent homicide, forcible rape, robbery, aggravated assault, burglary, motor vehicle theft, larceny-theft, and arson. Part II, on the other hand, includes a broad spectrum of crimes that range from more to less serious. In an effort to further disaggregate the broad universe of FBI-Part II crimes, MPI classifies them as violent or nonviolent. Part II-violent includes crimes such as assault, battery, kidnapping, hit and run, weapons offenses, and sex offenses. Following MPI’s approach, we included these crimes (Part II-violent), as well as domestic abuse crimes, in our “serious or violent” category.
   
   
3. Other: Individuals whose “most serious” conviction was for (a) a FBI Part 2 crime that is not violent according to MPI; (b) drug possession; (c) drug sale, distribution, or transportation; (d) an immigration crime (e.g., illegal entry or re-entry); (e) a nuisance crime; (f) driving under the influence (DUI); or (g) traffic offenses other than DUI.

Between FY 2010 and 2013, more individuals were removed through CAP without a criminal conviction (27.4 percent) than with a serious or violent conviction (17.2 percent). Put another way, based on this categorization, over 4 out of 5 CAP removals between FY 2010 and 2013 involved individuals with either no conviction or individuals who were not convicted of a crime classified as violent (following MPI’s definition) or serious, according to the FBI (Table 3).

Table 3: Removals by Broad Type of Crime, FY 2010-2013

Moreover, although over time CAP removed fewer individuals overall and a higher percentage of them with convictions, it essentially replaced removals of people who had not been convicted of a crime with removals of people who had not been convicted of a violent crime or of a crime that the FBI classifies as serious. Specifically, while the share of those removed under CAP without a conviction decreased 62.3 percent over time—from 37.9 percent in FY 2010 to 14.3 percent in FY 2013—the share of those removed under CAP who had not been convicted of a serious or violent crime increased 44.6 percent over time—from 45.8 percent in FY 2010 to 66.2 percent in FY 2013.

Furthermore, the proportion of individuals removed through CAP whose most serious criminal conviction related to traffic, drug, or immigration offenses increased from FY 2010 to FY 2013. The proportion of individuals removed with a traffic offense conviction increased each year, from 14.4 percent to 24.0 percent; the proportion of individuals removed with a drug conviction grew from 17.2 percent to 20.5 percent; and the proportion of individuals removed with an immigration-related criminal conviction grew from 4.4 percent to 6.5 percent (Appendix 1).

Additionally, breaking down those most prevalent convictions by specific offense types reinforces that a sizable proportion of the individuals ICE removed through CAP, if convicted, were convicted of less serious crimes. For example, 30 percent of all removed individuals whose most serious conviction was a traffic offense were convicted of non-DUI traffic offenses. Among “dangerous drug” offenders, 38.8 percent were convicted of possession and 23.2 percent did not have a “specified” conviction. Only 33.1 percent of the convictions were associated with selling, smuggling, or distributing drugs. Further, among all marijuana offenders, 53.6 percent were convicted for possession compared to 37 percent convicted of smuggling or selling drugs.

Lastly, a very low percentage of individuals removed through CAP were convicted of extremely serious crimes. Specifically, 1.7 percent were convicted of burglary, 1.6 percent of robbery, 1.5 percent of sexual assault (including rape), 0.5 percent of homicide, 0.4 percent of kidnapping, and 0.1 percent of arson (see Appendix 1).

In sum, given that the overwhelming majority of those removed through CAP do not fall within the most serious categories of crimes, as identified by the FBI, the data casts doubt on ICE officials’ assertions that ICE removals had narrowed to the worst public safety threats.

Do CAP Removals Match ICE’s Removal Priorities Targeting "Serious Criminals" or Public Safety Threats?

Although CAP removals have increasingly matched ICE’s removal priorities over time, ICE’s removal priorities did not necessarily correlate to the most serious criminal convictions or public safety threats.  

When ICE Director Morton’s June 2010 memorandum classified all noncitizens with criminal convictions as “Priority 1,” it sub-classified offenders into three Criminal Offense Levels (“COLs”), with Level 1 and 2 receiving principal attention:

• Level 1 offenders: noncitizens convicted of “aggravated felonies,” as defined in §101(a)(43) of the Immigration and Nationality Act, or two or more crimes each punishable by more than one year, commonly referred to as “felonies”;
• Level 2 offenders: noncitizens convicted of any felony or three or more crimes each punishable by less than one year, commonly referred to as “misdemeanors”; and
• Level 3 offenders: noncitizens convicted of crimes punishable by less than one year.

The data indicates that CAP removal patterns increasingly reflected Morton’s prioritization. Specifically, between FY 2011 and FY 2013 the share of ICE Level 1 and Level 2 offenders among all CAP removals increased—while the share of individuals with no conviction declined (Table 4).

Table 4. Removal Criminal Offense Level by Fiscal Year

Table 5. Removals by Broad Type of Crime and Removal Criminal Offense Level, 2011-2013

Note: Percentages may not sum to 100 due to rounding

This disconnect largely existed because ICE classified many common FBI Part II-nonviolent offenses or drug offenses as “Level 1.” More than half (52.5 percent) of Level 1 offenders removed through CAP were convicted of FBI Part II-nonviolent offenses or drug offenses. In particular, drug offenders removed through CAP were overwhelmingly classified as “Level 1.” Namely, 38.7 percent of drug possession offenders, and 96.1 percent of drug sale, distribution, or transportation offenders, were classified as “Level 1”—even though drug “distribution” can encompass minor conduct such as sharing without money changing hands, as a Human Rights Watch report points out. ICE still places a high priority on removing drug offenders, even as states have moved to decriminalize drugs such as marijuana.

Thus, while CAP has achieved some success in prioritizing removals according to ICE Levels, it has been less successful in narrowing enforcement removals to those who have been convicted of a violent crime or a crime that the FBI classifies as serious. Thus, CAP—and, in a broader sense, ICE’s prioritizations—may have facilitated to some extent ICE’s goal of removing the most "serious criminals" or public safety threats. However, by using criminal convictions as a proxy for danger, CAP also has resulted in the removal of a large number of individuals who did not meet those criteria.

Mexican and Central American Nationals are Overrepresented in CAP Removals

Additionally, the data shows that Mexican and Central American nationals are overrepresented in CAP removals compared to the demographic profiles of those populations in the United States. Collectively, people from Mexico and the Northern Triangle (Guatemala, Honduras, and El Salvador) accounted for 92.5 percent of all CAP removals between FY 2010 and FY 2013 (Table 6).

Table 6. Top Seven Countries Represented Among Persons Removed Through CAP, FY 2010-2013

Moreover, regardless of nationality, the proportion of individuals removed through CAP with convictions for violent crimes or for crimes that the FBI classifies as serious is relatively low, even across nationalities. For example, the share of Mexicans, Guatemalans, Dominicans, and Colombians among all people removed through CAP with a “violent or serious” conviction is slightly smaller than the share of nationals of the same countries among all CAP removals. Conversely, the proportion of nationals of Honduras, El Salvador, Jamaica, and “all other countries combined” among all people removed through CAP with a “violent or serious” conviction is slightly greater than the share of nationals of the same countries among all CAP removals (Table 8).

Table 8. Top Seven Countries by Broad Type of Crime, FY 2011-2013

All in all, these data suggest that, considering their demographic profiles, nationals of Mexico, Guatemala, and Honduras seem to be overrepresented among overall CAP removals, while the proportion of those nationals in each broad category of crime is not very different from the representation of their nationality among overall CAP removals.

Further research is needed to establish why nationals of certain countries tend to be overrepresented among CAP removals. For example, reports have questioned whether local law-enforcement officers have racially profiled minority residents or conducted pre-textual arrests so that ICE’s officers would check residents’ immigration status, once residents’ information was shared with ICE through CAP. A 2009 study of arrest data in Irving, Texas, found “strong evidence to support claims that Irving police engaged in racial profiling of Hispanics in order to filter them through the CAP screening system.” Our data raises concerns about whether racial disparities among arrests that lead to immigration enforcement may be more generalized practices.

The Geography of Removals: State-by-State Results

To identify the regions of the country in which CAP has been most active, we analyzed the location where individuals were apprehended prior to arrest and removal through CAP in FY 2013. In order to account for the fact that a higher proportion of noncitizens are living in certain states, we estimated CAP arrest and removal rates standardized to 1,000 noncitizens in each state, utilizing 2013 ACS estimates of the noncitizen population (Table 9).

Table 9. DHS Criminal Alien Program Removals By U.S. State, FY 2013

Sources: DHS Criminal Alien Removals, 2013; American Community Survey, 2009-2013.
*Results should be interpreted with caution due to small cell sizes for ACS noncitizen estimates.

States with similar noncitizen population sizes do not necessarily present similar rates of CAP removals. In fact, a Pearson correlation test shows no statistically significant correlation between the two variables. In other words, whether the foreign-born population living in a particular state is large or small does not explain whether the state would have a higher or lower CAP removal rate.

The states with the highest rates of removal per 1,000 noncitizens include several with smaller immigrant populations—e.g., Mississippi (17.4 per 1,000 noncitizens), Wyoming (10.1 per 1,000), West Virginia (9.3 per 1,000), Kentucky (7.9 per 1,000), South Dakota (7.4 per 1,000), and Nebraska (7.2 per 1,000)—as well as two with large immigrant populations, Texas (6.6 per 1,000) and Arizona (6.6 per 1,000). The states with the lowest rates of removals per 1,000 noncitizens include several with comparatively small immigrant populations—e.g., Rhode Island (0.2 per 1,000), Alaska (0.3 per 1,000), Vermont (0.3 per 1,000), Maine (0.6 per 1,000), and Hawaii (0.7 per 1,000)—as well as several with comparatively larger immigrant populations—Nevada (0.9 per 1,000), Massachusetts (1.0 per 1,000), New York (1.1 per 1,000), Maryland (1.2 per 1,000), New Jersey (1.3 per 1,000), Washington D.C. (1.4 per 1,000), and Connecticut (1.4 per 1,000). 

Several state-specific factors may have contributed to these varying arrest and removal patterns in FY 2013. Most relevant may be the extent of local cooperation with ICE generally, through programs such as 287(g) that deputize local officers to enforce immigration law, and CAP specifically, through “jail check” agreements and policies regarding ICE detainer requests. For example, from 2011 to 2013, several large U.S. cities limited ICE’s access to their jails, and stopped honoring ICE detainer requests for some or all noncitizens—particularly those with criminal charges, but not convictions, or those with lesser charges or convictions. Other state-specific factors may include availability of public transportation and ICE’s detention capacity. Although these are plausible explanatory factors, further research is needed to understand the varying impact of CAP at the state level.

Conclusion and Looking Forward

This examination of CAP’s outcomes from fiscal years 2010 to 2013 offers important insights into CAP’s operations over time and its potential impact on communities moving forward. Understanding CAP is critical to understanding ICE’s overall interior enforcement activities. This understanding takes on added importance as ICE begins implementing its 2014 enforcement priorities and engages state and local law enforcement through PEP.

Since the years for which this report examined data, CAP has remained the primary program through which ICE conducts enforcement in the interior of the United States. In FY 2014, CAP was responsible for 524,522 encounters, 122,826 arrests, and 68,244 removals, according to ICE—compared to ICE’s 102,224 overall “interior removals” (in other words, removals of individuals apprehended in the interior). With ICE still focusing on removing noncitizens with criminal convictions by leveraging state and local law-enforcement arrests, and with Secure Communities (and now PEP) fingerprint-sharing from state and local law enforcement subsumed into CAP’s operations, CAP will likely remain a key (if not the) linchpin of ICE’s efforts to achieve interior removals.

That said, CAP’s target population and removal processes will likely change. On November 20, 2014, DHS Secretary Johnson, among other actions, (1) narrowed ICE’s removal priorities from any noncitizen convicted of any crime to those “convicted of specifically enumerated crimes,” and (2) changed CAP’s enforcement processes after CAP receives information from local law enforcement, by replacing ICE “detainer” requests (i.e., to hold a noncitizen) with requests for notification of release, absent special circumstances.

Johnson’s first change, the narrowing of ICE’s priorities, could significantly reduce the numbers of noncitizens targeted through CAP in two key groups: (1) those without convictions (e.g., those only arrested or charged), which constituted 27.4 percent of those CAP removed from FY 2010 to 2013, and (2) those convicted of certain lesser offenses that no longer meet ICE’s priorities, potentially as many as 34.4 percent of those CAP removed from FY 2010 to 2013. These latter individuals are those whose most serious conviction is for a Part II-nonviolent offense, a nuisance offense, drug possession, an immigration-related criminal offense, or a non-DUI traffic offense—but who do not remain an ICE priority because they have not committed a criminal felony, an “aggravated felony” under immigration laws, three misdemeanor offenses arising on separate occasions, or a misdemeanor with a 90-day sentence or longer.

The likelihood of a decline in the number of noncitizens targeted by CAP, however, depends on ICE’s effective implementation of the 2014 Johnson Memo priorities. ICE would have to adjust CAP’s daily work so as to not target and remove some of those individuals whom CAP has been removing in large numbers. For example, it is unclear whether the May 2013 CAP guidance that targets those without convictions remains in place following the 2014 Johnson Memo. Additionally, ICE would have to adjust its daily work based on criminal offense level classifications (i.e., Level 1-2-3), since a significant amount of Level 1 and Level 2 offenders will no longer meet ICE’s new priorities—namely, 36.1 percent of Level 1 offenders and 55.5 percent of Level 2 offenders that CAP removed from FY 2011 through FY 2013.

But tellingly, this report shows that CAP encounter numbers have remained outsized relative to CAP removals—in FY 2014, nearly eight times as large, suggesting that the program may not be narrowly designed to target those who fall within the government’s priorities. Logically, if CAP is targeting fewer individuals, encounters should drop. Whether reductions in encounters occur or not, since large numbers of encounters appear either unnecessary or inefficient, Congress may want to more closely examine current levels of CAP funding.

As to Johnson’s second change—ICE’s replacement of detainer requests with notification requests—the impact upon CAP will likely be state-specific, depending on the willingness of state and local authorities to cooperate with ICE. Many local law-enforcement leaders, and the President’s Task Force on 21st Century Policing, note that cooperation with ICE inhibits community trust and thus local immigrant reporting of crimes. States and localities have, to date, implemented a range of protocols regarding ICE cooperation. Indeed, the state-by-state CAP arrest and removal data in this report may reflect differing levels of cooperation with ICE.

If ICE’s goal is to remove serious public safety threats, a broad “jail check” program such as CAP may not be an efficient way to do it. Jail check or arrest check programs, by their nature, funnel into immigration enforcement those noncitizens apprehended by local law enforcement—and those apprehended by local law enforcement tend to overwhelmingly be apprehended for lesser offenses. Congress should more closely examine CAP funding, given these dynamics. The government also should report on a regular basis statistics on encounters, arrests, and removals by specific types of criminal convictions. Finally, there must be more oversight over DHS’s steps to forestall racial profiling, given the disproportionate numbers of Mexicans and Central Americans removed through CAP.

Glossary

Encounters: “An ICE encounter of a person is defined as the interview, screening, and determination of his/her citizenship, nationality, and lawful presence (i.e., whether or not the alien is present in the United States after the expiration of the period of stay authorized by the Secretary or is present in the United States without being admitted or paroled), and legal right to remain in the United States of America. An encounter, detainer or charging documents issued by ICE does not necessarily result in the individual being placed into ICE custody.”

Arrests: “The arrest date refers to the date in which an individual was either booked into ICE custody or processed prior to being booked into ICE custody.”

Removals: “FY2010-2013 Removals include Returns, which include Voluntary Returns, Voluntary Departures and Withdrawals under Docket Control.”

Lead Type: “The Lead Type is a category of event, incident, or anything of interest to immigration authorities. The Lead Source is a category of the origin of information regarding an occurrence.”

ICE Criminal Offense level: Since FY 2011, ICE has defined criminality as whether or not an alien has an ICE Threat Level (convicted criminal) or not (non-criminal immigration violator). For purposes of prioritizing the removal of aliens convicted of crimes, ICE personnel refer to the following offense levels: Level 1, Level 2, and Level 3 offenders. Level 1 offenders are those aliens convicted of “aggravated felonies,” as defined in § 101(a)(43) of the Immigration and Nationality Act, or two (2) or more crimes each punishable by more than 1 year, commonly referred to as “felonies.” Level 2 offenders are aliens convicted of any other felony or three (3) or more crimes each punishable by less than 1 year, commonly referred to as “misdemeanors.” Level 3 offenders are aliens convicted of “misdemeanor” crime(s) punishable by less than 1 year.

“Definite Criminal”: “The “Definite Criminal Yes No” column is populated based on whether a criminal conviction is recorded in ENFORCE's Crime Entry Screen. Aliens may have criminal convictions that have not been recorded in the Crime Entry Screen.”

“Most Serious Criminal Charge”: Reflects “the most serious convicted criminal charge on record at the time of removal.” These categories are drawn from the NCIC criminal charge code that ICE reported to us within its removals data. We used a list describing ICE’s internal “business rules” for using NCIC codes, publicly available and posted on TRAC’s website. ICE’s data stated that this NCIC code reflects “the most serious convicted criminal charge on record at the time of removal.”

NCIC code: Classification of criminal offenses based on codes from the Department of Justice’s National Crime Information Center (NCIC).

Aggravated Felony: An “aggravated felony” is a term of art, defined by an immigration statute, used to describe a category of criminal offenses carrying particularly harsh immigration consequences for noncitizens convicted of such crimes. Regardless of their immigration status, noncitizens who have been convicted of an “aggravated felony” are prohibited from receiving most forms of relief that would spare them from deportation, including asylum, and from being readmitted to the United States at any time in the future. An “aggravated felony” need not be aggravated nor a felony under criminal laws. 

Reference Guide

CAP Initiatives

During its existence, ICE’s Criminal Alien Program has engaged in several initiatives that complement its traditional “jail check” activities, expand its “at-large” activities in the community, or both. These initiatives include:

• CAP Surges: A “CAP Surge” represents ICE’s effort to focus resources to intensify and augment CAP’s traditional “jail check” removals within a short time period. ICE does this within a particular geographical Field Office through a “Removal Surge Operation” (CAPRSO), and, since FY 2014, has added ICE officers from across the country to a Field Office through the CAP “Surge Enforcement Team” (CAPSET). The goal of a CAP Surge is to ensure that “100 percent of all priority criminal aliens booked into targeted facilities that are amenable to immediate removal obtain a removal order” (CAPRSO), or to “[i]ncrease the number of priority criminal aliens identified and fully processed prior to their release from custody” (CAPSET).

• Threats to the Community (TC): CAP’s new “Threats to the Community” (TC) initiative—an “at-large” initiative—consists of “targeted at-large enforcement operations led by CAP to investigate, arrest, and remove criminal aliens that pose the greatest threats to the community.” The initiative “targets criminal aliens utilizing ICE priorities, focusing on at-large criminal aliens identified through interoperability” [i.e. Secure Communities fingerprint-sharing], “as well as those jurisdictions which limit or refuse ICE access to their facilities.” ICE first publicly disclosed the initiative in February 2015, in its FY 2016 budget request.

• Joint Criminal Alien Removal Taskforce (JCART): In ICE’s words, “ICE’s Joint Criminal Alien Removal Taskforce (JCART) focuses on locating and arresting at-large criminal aliens with convictions for drug trafficking, violent crimes and sex offenses. Working closely with other agencies, JCART conducts special operations including criminal aliens who have been released from federal, state, or local custody.” In February 2013, ICE stated that JCART was then currently “active in the Los Angeles and New York Field Offices.” It appears that ICE may have since replaced JCART with the “Threats to the Community” (TC) initiative, although ICE’s website still lists JCART as an active initiative. 

• Violent Criminal Alien Section (VCAS): CAP’s Violent Criminal Alien Section (VCAS) initiative exists to facilitate federal criminal prosecutions of criminal violations that are discovered through ICE enforcement activities. VCAS focuses on “recidivist” noncitizens with convictions (i.e. those have offended more than once), so as to “enhance public safety” and deter future recidivism—either by deterring crime through prosecution, or simply incarcerating offenders. VCAS screens for recidivists encountered not only through CAP, but through local law enforcement or Fugitive Operations raids. VCAS then refers cases to federal prosecutors (i.e. the United States’ Attorneys’ Offices) for prosecution.

• Rapid Removal of Eligible Parolees Accepted for Transfer (Rapid REPAT Program): In ICE’s words, the “Rapid Removal of Eligible Parolees Accepted for Transfer (REPAT) Program is a joint partnership with state correctional/parole agencies designed to expedite the process of identifying and removing criminal aliens from the U.S. by allowing selected non-violent criminal aliens incarcerated in U.S. prisons and jails to accept early release in exchange for voluntarily returning to their country of origin.”

• Detention Enforcement and Processing Offenders by Remote Technology (DEPORT): In ICE’s words: “Approximately 27 percent of inmates in Federal Bureau of Prisons (BOP) custody are non-U.S. citizens. ERO created the Detention Enforcement and Processing Offenders by Remote Technology (DEPORT) Center in Chicago to process this population through CAP. ERO officers and agents assigned to the DEPORT Center conduct interviews of BOP inmates nationwide using video teleconference equipment. Through the combined effort of the DEPORT Center and local ERO resources, criminal aliens from all federal detention facilities are taken into ERO custody upon completion of their sentences."

• Phoenix Law Enforcement Area Response (LEAR): This initiative, specific to ICE’s Phoenix, Arizona field office, “provides a method for state and local law enforcement agencies (LEAs) in Arizona to directly contact ICE regarding suspected immigration violators.” ICE’s LEAR staff would respond to local agencies 24/7; ICE officers would determine nationality, immigration status, and removability, place detainers, and process noncitizens for removal. It is unclear if LEAR is still in operation. Since February 2013, ICE has not listed it in its annual budget requests.

APPENDIX 1: Most Serious Criminal Charge Category, by Fiscal Year

APPENDIX 2: “Most Serious Criminal Charge Category” by “Removal Criminal Offense Level”, FY2011-2013

(Excludes Cases that are “Not Definite Criminal”)