Crimmigration at the Local Level: Criminal Justice Processes in the Shadow of Deportation

Abstract

In recent decades, authorities have adopted a number of programs that tether the criminal and immigration enforcement apparatuses in novel ways. This mixed methods case study assesses the impact of such programs on local criminal justice processes and outcomes in King County, Washington. Although the empirical research on the effects of such programs is scant, the emerging literature on legal hybridity suggests that the enmeshment of the criminal and immigration systems is likely to enhance the state's power to detain and punish. The quantitative results support this hypothesis: non-citizens flagged by immigration authorities stay in jail significantly longer than their similarly situated counterparts. Qualitative focus group interviews with prosecuting and defense attorneys identify four key mechanisms by which Immigration Customs and Enforcement detainers alter the incentive structure, impact decisionmaking, and extend jail stays for non-citizens. Together, these findings suggest that immigration law and the threat of deportation now cast a long shadow over local as well as federal criminal proceedings, and enhance penal pain for non-citizens. Implications of these findings for the “crimmigration” literature and research on the effect of citizenship status on criminal justice outcomes are discussed.

Over the past decade, immigration policy has prioritized the removal of “criminal aliens,” that is, non-citizens who have been convicted of a crime.¹ Toward this end, federal immigration authorities have adopted a number of new programs that mobilize state and local criminal justice resources. For example, the Criminal Alien Program (CAP) and Secure Communities enable immigration authorities to identify non-citizens in local jails so that they may be transferred to federal immigration custody on their release from jail. These programs were implemented in the context of heightened immigration enforcement, and coexist with policies that dramatically expand the number of criminal offenses that disqualify people seeking to obtain or retain legal permanent status (Chacón 2012; Kanstroom 2007; Menjívar 2012)—as well as policies and practices that have led the United States to host the world's largest prison population (Western 2006). Together, these developments have expanded immigration enforcement's scope, and tethered federal immigration enforcement and local criminal justice systems in novel ways.

This study analyzes how these developments affect local criminal justice processes and outcomes. The empirical findings fill an important lacuna: as Eagly notes, “the existing scholarship has not adequately explored how immigration operates in the criminal sphere—namely, how the rights, procedures and systems traditionally associated with the criminal justice system have themselves been affected by interaction with the civil system of immigration” (2010: 1284). Although a large and growing literature explores the consequences of “crimmigration”—especially the intermingling of the federal criminal and civil immigration systems—little is known about how the institutional enmeshment of federal immigration enforcement and state/local justice institutions affects practices and outcomes in the latter (but see Eagly 2013).

Although prior research on the impact of current immigration enforcement practices on local justice outcomes is scant, the emerging literature on legal hybridity provides reason to suspect that programs that tether federal immigration enforcement to local justice institutions will enhance the state's power to punish and exclude. For example, in recent years, bureaucratic and legal actors at the local level created a variety of legally hybrid control tools after the courts invalidated the vagrancy and loitering laws that had historically enabled local authorities to regulate the movement of the socially marginal (Beckett and Herbert 2010a, 2010b; Beckett and Murakawa 2012). Examples include of new, legally hybrid control tools include gang injunctions, no contact orders, and various applications of trespass law that enable officials to regulate the movement of individuals perceived as disorderly from urban spaces for extended periods of time. These legally hybrid techniques blend elements of civil and criminal law; they also shift the burden of proof and restrict rights in ways that enhance the power of the state to control and banish the socially marginal. Similarly, legislation that authorized the transfer of some juveniles to the adult criminal system notably enhanced the power of prosecutors in juvenile courts (Harris 2007).

In these cases, then, the fusion of civil and criminal law worked to enhance state power. Studies suggest that a similar process has occurred in the federal immigration context, where authorities have incorporated many of the enhanced enforcement powers—but not the rights protections—associated with criminal law into the civil immigration system (Chacón 2012; Eagly 2010). As Legomsky explains, “immigration law has been absorbing the theories, methods, perceptions, and priorities associated with criminal enforcement while explicitly rejecting the procedural ingredients of criminal adjudication” (2007: 472). For example, in the federal system, non-citizen defendants are often held for extended periods of time on civil charges, only to be criminally charged weeks after their arrest; by holding these defendants on civil charges, authorities circumnavigate the right to criminal bond (Eagly 2010: 1306). Conversely, a majority of federal immigration criminal defendants are arrested by (civil) DHS enforcement officers, who typically interview arrestees without providing Miranda warnings before transferring arrestees to federal criminal authorities. In this context, then, the blurring of civil and criminal processes effectively denies federal non-citizen defendants the rights that are at least theoretically protected in the criminal system (Eagly 2010: 1309).

In short, recent studies of the development and application of legally hybrid control mechanisms provide reason to suspect that the tethering of federal immigration enforcement practices to local justice institutions is likely to enhance the state's capacity to detain and punish. Evidence that this is the case would have a number of important implications. Before describing these, we first provide some background regarding the programs that truss the federal immigration enforcement apparatus to local justice systems across the country.

Debating Secure Communities

Programs such as CAP and Secure Communities involve Immigration Customs and Enforcement (ICE) detainers: administrative requests from federal immigration authorities that jail administrators hold non-citizens whose release from jail has been ordered by a criminal court judge for up to 48 additional hours (plus weekends and holidays) so that immigration authorities may transfer non-citizen detainees to federal immigration custody.² Detainers, then, are both a proxy for non-citizen status and, where they are honored, an administrative tool with important material consequences.

In the absence of a robust empirical literature regarding the consequences of these programs, proponents and opponents make very different claims about them. On the one hand, ICE contends that programs that involve local criminal justice agencies in immigration enforcement via ICE detainers are an efficient means of locating suspected immigration law violators, especially those who pose grave threats to public safety; they also suggest that these programs have little impact on local criminal justice institutions beyond the addition of 48 hours to some jail stays.³ By contrast, critics maintain that these programs have a number of adverse consequences for immigrants and their families—as well as for local justice institutions (see Eagly 2013; Kholi, Markowitz, and Chavez 2011; National Immigration Forum 2011; Shahani 2010). In particular, opponents allege that these programs extend jail stays, even for people who are arrested for relatively minor offenses and who lack significant criminal histories (Greene 2012; National Immigration Forum 2011; Shahani 2010).

Despite the controversial nature of these programs, we were unable to find any published studies that analyze how CAP and Secure Communities impact criminal justice processes and outcomes, although a few unpublished reports have assessed the impact of these programs on jail stays and costs. These reports suggest that ICE detainers significantly extend jail stays. For example, a 2011 study found that Los Angeles jail inmates subject to ICE detainer requests spent an average of 20.6 more days in jail than other inmates (Greene 2012). Another study found that in Travis County, Texas, arrestees subject to ICE detainer requests spent, on average, three times longer (65–76 days) than others (22–26 days) (National Immigration Forum 2011). Similarly, a 2010 study found that New York City jail inmates flagged by ICE spent an average of 73 more days in jail than others (after controlling for race and offense level) (Shahani 2010).

Although suggestive, these reports suffer from two important limitations. First, they do not take into account many of the legal factors that might, along with ICE detainers, influence the amount of time arrestees spend in jail. By controlling for a wide range of legal factors that influence criminal justice outcomes, the present study assesses the impact of ICE detainers over and above other case characteristics. Second, although these studies suggest that ICE detainers prolong jail stays, it is unclear why this may occur. Our qualitative data allow us to identify the mechanisms by which these programs extend jail stays in King County, Washington. These findings have important implications for socio-legal studies of “crimmigration” and the impact of citizenship status on criminal justice outcomes, as described below.

Crimmigration

Scholars use the term “crimmigration” to refer to the growing enmeshment of the immigration and criminal legal systems (Stumpf 2006). This entanglement is rooted in recent legislation, including the 1986 Immigration Reform and Control Act (IRCA) and the Illegal Immigration Reform and Responsibility Act (IIRIRA) of 1996. These policy developments set into motion three dynamics that have led the immigration and criminal enforcement apparatuses to become increasingly intertwined. First, legislation adopted in the 1980s and 1990s dramatically increased funding for immigration control (Meissner et al. 2013). These funding increases enabled federal authorities to criminally prosecute a large and growing number of immigrants for “illegal entry” and “illegal reentry” prior to their removal proceedings (Chacón 2012; Cruz 2012; Eagly 2010; Inda and Dowling 2013; Stumpf 2013). Second, federal actors increasingly emphasize the need to prioritize the removal of non-citizens who have committed a crime (Eagly 2013). This focus on the “criminal alien” has largely replaced the prior emphasis on “illegal aliens” (Eagly 2013). Finally, Congress and the courts have notably expanded the number of crimes that disqualify people seeking to obtain or retain permanent legal status (Coutin 2011; Cruz 2012), thereby rendering “legal permanent status” potentially nonpermanent for millions of U.S. residents (Menjívar 2006, 2012).

As a result of these developments, the immigration and criminal justice systems are increasingly entangled. Researchers have used the term “crimmigration” to highlight different dimensions of the enmeshment that the term connotes. One body of scholarship focuses on crimmigration's ideological underpinnings, arguing that the same social and cultural forces that encourage punitive anticrime policies also depict immigrants as dangerous outsiders who threaten the safety and well-being of American citizens; in both cases, a dangerous mixture of fear, insecurity, racism, and resentment enhance receptivity to these images (Chavez 2008; Inda and Dowling 2013; Sklansky 2012; Tyler and Boeckmann 1997; Welch 2002, 2012).

Another body of literature calls attention to increased blurring of the institutional boundaries between the immigration and local criminal enforcement apparatuses. Some of these studies focus on the increased involvement of local authorities in the enforcement of federal immigration laws (Adler 2006; Provine et al. 2012; Varsanyi et al. 2012). Others highlight the increasingly punitive tenor of both civil and criminal immigration enforcement tactics, including the dramatic increase in the use of detention for suspected immigration law violators (Welch 2002, 2012) and the increased prosecution of “criminal aliens” in federal courts whose crimes consist mainly of entering the country without authorization (Hagan, Rodriguez, and Castro 2011; Stumpf 2013). These studies show that the processing of these criminal cases in the federal criminal courts is characterized by diminished procedural protections, “fast-tracked” plea bargaining, and mass legal representation, leading researchers to conclude that recent policy developments have fueled the creation of a two-tiered system of justice in the federal criminal courts, one for citizens and another for non-citizens (Camayd-Freixas 2009; Chacón 2012; Eagly 2010).

The crimmigration literature thus sheds important light on new cultural and institutional linkages between the criminal and immigration legal systems. Our analysis extends this body of research by showing that the institutional enmeshment of the criminal and immigration systems not only impacts federal criminal and immigration enforcement institutions, but also local criminal processes, institutions, and outcomes (see also Eagly 2013). Insofar as significantly more people are processed through local state and local justice institutions than by the federal courts, our findings suggest that the scope and effects of “crimmigration” are far broader than previously recognized. These findings also bolster Menjívar's (2012) argument that recent immigration policies have intensified the pain and suffering associated with possession of an uncertain legal status. Specifically, our study shows that recent policies significantly extend jail stays of non-citizens where ICE detainers are honored—regardless of whether non-citizen arrestee are charged or convicted of a crime. Because incarceration has a variety of adverse psychological, social, and physical effects, we interpret our findings as evidence that the institutional blurring of immigration and criminal law enforcement enhance penal pain for non-citizens.

Ethnicity, Citizenship Status, and Criminal Justice Outcomes

The present study also has important implications for studies assessing the impact of citizenship status on criminal justice outcomes. Many studies have found that ethnicity has a significant impact on criminal case outcomes (for an overview, see Light Massoglia, and King, 2014). Recently, however, researchers have explored the possibility that the apparent impact of ethnicity may be an artifact of citizenship status (Hartley and Armendariz 2011; Light 2014; Light Massoglia, and King 2014; Wolfe, Pyrooz, and Spohn 2011; Wu and Delone 2012).⁴ The results of the most recent, comprehensive and longitudinal analyses of the impact of citizenship status on federal sentencing outcomes support this hypothesis: citizenship status has a significant impact on case outcomes across a range of offense categories, and accounting for it substantially attenuates the impact of ethnicity (Light 2014; Light, Massoglia, and King, forthcoming). Moreover, the impact of citizenship status on penal outcomes has increased notably over time; its effect is especially pronounced for undocumented immigrants and in districts with large immigration populations (Light 2014).

A variety of theoretical perspectives suggest that the impact of citizenship status on criminal case outcomes may reflect the stigmatization and stereotyping of immigrant populations. Yet as Light, Massoglia and King note, “additional case processing information would help shed light on the nuanced and cumulative effect of citizenship throughout the criminal justice system” (2014: 18). Indeed, our study suggests the need to consider the institutional mechanisms by which citizenship status may impact case outcomes. That is, our study suggests that the effect of citizenship status on sentencing outcomes may be a function of institutional arrangements that tether the immigration and criminal legal systems rather than of the stigmatization of immigrants in the criminal justice system.

The present study, then, has important implications for a number of socio-legal literatures. In what follows, we describe the data and methods we use to answer our research question: does the presence of ICE detainers impact the length of jail stays, and if so, how? We then present our findings and situate our case study in comparative context. In the final section, we discuss the substantive and theoretical implications of our findings.

Data and Methods

Administrative Data

The original dataset provided by King County jail administrators included information regarding the 37,670 King County jail releases that took place in 2011. The individuals released from jail in 2011 may have been booked in years prior to or during 2011; they may have been transferred to prison, to federal authority or released to the community. Some of these individuals were charged with a felony, others with a misdemeanor, and some were never charged at all. Some individuals may have been booked and released from jail multiple times over the course of that year. Some of these people were convicted; others were not.

In our analysis, we excluded 2,619 releases associated with people who had pending administrative matters but no criminal charges and 133 cases with missing data. Diagnostic tests identified four cases as outliers, having both leverage and influence; these were also excluded from the analysis. After these exclusions, our sample included 33,587 separate jail bookings associated with a 2011 jail release date. The dataset includes information about a number of variables, including: booking and release date; reason for release; number of charges; seriousness ranking of each charge; type of each charge (i.e., felony/misdemeanor/investigation); offense category; and the race, ethnicity, and gender of the booked person. In several cases, these variables were recoded to facilitate analysis. “ICE detainer” was coded as a binary variable of 0 and 1. ICE issued a total of 821 ICE detainer requests for people released from a King County jail in 2011. In theory, ICE issues detainers only in cases involving non-citizen defendants. Although the data allow us to ascertain whether an ICE detainer was issued, they do not illuminate the precise nature of the target's citizenship status. For this reason, we are not able to compare the effects of ICE detainers for legal permanent residents versus undocumented or partially documented persons. The dependent variable in our quantitative analysis is the number of days each individual released from jail spent as in inmate of that institution. Our dependent variable includes only the number of days spent in jail, not prison or other confinement institutions.

Legal variables

Jail Days: Using the booking and release dates, we generated a total number of days spent in a King County jail and used this measure as the dependent variable in the regression model. Because the variable “jail days” is heavily skewed, this variable was logged for the regression analysis. As a result, the regression coefficients are interpreted as a percent change in the number of jail days attributable to each of the independent variables.

Charge Type, Seriousness Rank, Offense Type and Number of Charges: Charge type, included in the model as three dummy variables, designates whether the most serious charge associated with the booking was a felony, misdemeanor, or investigation. For approximately 9.2 percent of all bookings (3,081 cases), the most serious charge type was an “investigation,” meaning that a formal charge had not been filed by the prosecutor's office at the time of release; we refer to these cases at “not charged” in our analyses. We also control for the seriousness rank (as recorded by jail administrators) of the most serious charge, incorporating it into our model as variable ranging from 0 to 13 (cases not charged are listed as having a seriousness rank of 0).

In the original dataset, “number of charges” included investigations that did not result in formal charges and ICE detainers. We recoded the number of charges to include the number of charges for which a court cause number had been assigned, so that number of charges refers to the number of criminal charges filed. The number of charges ranges from 0 to 10. The original “number of charges” variable indicated that 46 cases had more than 10 charges. Because the data provided identify only the first 10 charges for each case, we use 10 as the maximum in the regression analysis. In the original dataset, 14 offense categories were identified; we collapsed these into six offense types: violent, property, drug, public order, sex, and other. These are incorporated into the model as six dummy variables. Ideally, we would have also been able to control for prior criminal convictions. Unfortunately, this information was not included in the jail data. However, we did obtain criminal history information for a subset of our sample.

Extra-legal variables

Race and ethnicity are included in these analyses as mutually exclusive categories in the form of six dummy variables: Hispanic/Latino, White, Black, Asian, Native American, Other/Unknown. We used Hispanic Surname Analysis to identify Latino/a inmates who were not identified as such by jail administrators. Unlike race and gender, which are ascribed and recorded by booking officers, jail administrators inquire about ethnicity (Hispanicity) in interviews with jail inmates within 72 hours of booking. However, some inmates are released before jail administrators are able to conduct this interview, leading to an undercount of Latinos in the jail data. We used therefore Hispanic Surname Analysis to estimate the proportion of defendants who identify as Latino. This program utilizes the U.S. Census Spanish Surname database and assigns a numeric value between 0 and 1 to all surnames in that database. The list used to identify defendants of Hispanic origin contained 12,497 different Spanish surnames that are classified by the Census Bureau as “Heavily Hispanic.” These numeric values represent the probability that a given surname corresponds to persons who identified themselves as Hispanic/Latino in the 1990 U.S. Census (Perkins 1993; Word and Perkins 1996).

Statistical Methods

We begin by providing descriptive statistics to illuminate the prevalence and distribution of ICE detainer requests, and to test for differences in means between the number of days people with and without ICE detainers spent in jail. Next, we present results from Ordinary Least Squares (OLS) statistical regression. Using logged jail days as the dependent variable, we examine the impact on jail stays of: ICE detainer; charge type (e.g., felony, misdemeanor, or investigation); the seriousness ranking of the most serious charge; the number of charges filed; and the offense type. We refer to these as “legal factors” or “case characteristics.” We also control for the impact of extra-legal factors—namely, the race, ethnicity and gender of the person who was booked—on the number of days spent in jail. The model is specified as:

$\begin{array}{c}Logged \left(Jail Days\right)=\alpha +{\beta }_1\left(ICE Detainer\right)+{\beta }_2(Number of Charges)\\ +{\beta }_3\left(Seriousness Rank\right)+{\beta }_4\left(Felony\right)+{\beta }_5\left(Investigation\right)\\ +{\beta }_6(Violent Offense)+{\beta }_7(Property Offense)+{\beta }_8(Drug Offense)\\ +{\beta }_9(Sex Offense)+{\beta }_{10}(Other Offense)+{\beta }_{11}\left(Male\right)+{\beta }_{12}\left(Black\right)\\ +{\beta }_{13}\left(Latino\right)+{\beta }_{14}(Native American)+{\beta }_{15}\left(Asian\right)+{\beta }_{16}\left(OthRace\right)\end{array}$

Focus Groups

To better understand how the presence of ICE detainers affects the criminal process, we conducted three focus groups with twenty defense and prosecuting attorneys. We included attorneys who mainly handle felony cases because our quantitative findings indicate that the impact of ICE detainers is far greater in cases involving felony rather than misdemeanor charges. Participants included eight prosecuting attorneys from distinct areas within the felony division of the King County Prosecuting Attorney's Office and twelve public defense attorneys from two different nonprofit defense firms that provide legal representation to indigent defendants. (Because public defenders were, at the time of the study, employed by four distinct nonprofit organizations in King County, we conducted two separate focus groups with defense attorneys from different offices to ensure that our findings did not reflect organization-specific practices).

Our aim was to bring the collective experiences and observations of attorneys to bear on our topic. We therefore elected to conduct group rather than individual interviews. The dynamic nature of focus groups allows participants to assist in the analysis of the data as the conversation unfolds, and provides researchers with the opportunity to field hypotheses and see if their subjects agree with the provisional conclusions they have drawn (Frey and Fontana 1991). Conversely, group discussions create the potential for more vocal members of the group to dominate the discussion. In this study, the facilitator made an effort to involve all group members in the conversation by soliciting feedback from those who were comparatively quiet. This technique was successful, perhaps because attorneys are used to speaking in group settings and tend not to be especially retiring. In the end, all participants offered substantial and important observations to the group discussions.

The focus group discussions lasted 60–90 minutes and were digitally recorded and transcribed for analyses. After reading and rereading the transcripts, we identified all the ways that attorneys indicated that ICE detainers and awareness of immigration consequences affect their decision-making and the criminal process. We then grouped these observations into categories. Once these categories were created, we developed memos on each contributing factor. Contrary or diverging findings were also noted and allow us to highlight potential variation in informants' experiences or understandings. We then identified representative excerpts from the interviews and use them below to illustrate the various mechanisms by which ICE detainers affect the criminal process and extend jail stays.

Although there was a high degree of consensus among prosecutors and defense attorneys about practice and policy in King County, we nonetheless sought confirmation of the inferences we drew from the focus group interviews from a number of practitioners with extensive experience in the criminal justice system. These included: prosecutors with the Seattle City Attorney's Office and King County Prosecuting Attorney's Office; two heads of defense organizations; and an attorney with the Washington Defender Association's Immigration Project (WDAIP).

Findings

We begin by describing the results of our quantitative analysis. Our sample included 790 bookings (2.4 percent of the total sample) in which defendants were subject to ICE detainer requests. Men were over-represented among people subject to detainer requests: 97 percent of those flagged by ICE, but 79.2 percent of other inmates, were male. More than three-fourths (76.3 percent) of inmates subject to ICE detainer requests were Hispanic/Latino, compared to only 10.2 percent of other inmates. Approximately one of six (16 percent) Latinos booked into jail was subject to an ICE detainer request. Nearly all (96.2 percent) of those with ICE detainers were released via a “transfer of custody,” meaning that nearly one in six Latinos were transferred to ICE on their release.

Charge Type

Each booking is associated with one of three charge types: felony, misdemeanor, or “not charged.” For over half (50.8 percent) of the individuals subject to ICE detainers, the most serious charge was a misdemeanor; just over one-third (36.3 percent) were charged with a felony. The remaining 13 percent were not charged with any crime prior to their release from jail. Although the results show that the majority of people flagged by ICE were not charged with a felony, it is nonetheless conceivable that people subject to ICE detainer requests have extensive criminal histories that include felony convictions or other serious crimes, as ICE avers.

To assess this possibility, we drew a random subsample of detainees from the larger sample. The subsample included 100 individuals who were released from a King County jail in 2011 and charged with a crime. We then compiled criminal history information for these individuals.⁵ The results of this comparison are shown in Table 1, and indicate that only 20 percent of the people subject to an ICE detainer request had one or more prior Washington State felony convictions (compared to 60 percent of people released from jail in 2011 without ICE detainer requests). Similarly, a comparatively small proportion of people with ICE holds had one or more prior Washington State convictions for a crime against persons (18 percent vs. 30 percent of people not subject to ICE detainers).⁶ On average, people subject to ICE detainers also had fewer prior felonies and prior convictions for crimes against persons in Washington State than others.

Table 1. Comparison of Criminal Records of People With and Without ICE Detainer Requests

Source: Authors' analysis of data compiled by the Northwest Defender's Association obtained from the Washington State Administrative Office of the Courts Judicial Information Services (JIS) Website. N=100.

In short, these findings indicate that four of five people flagged by ICE in 2011 had not been convicted of a serious crime in Washington State. Although this analysis does not include information about convictions from outside Washington State, these findings are consistent with the results of a recent analysis of national ICE data by the Transactional Records Access Clearinghouse (TRAC 2013) at Syracuse University. TRAC researchers analyzed information regarding nearly a million detainers issued between fiscal year 2008 and the start of 2012. Their results indicate that more than three-fourths (77.4 percent) of the individuals subject to ICE detainers had no criminal record either at the time the detainer was issued or afterward. Among the remaining 22.6 percent of people with a criminal record, only 8.6 percent had been convicted of a crime that ICE classified as a Level 1 (serious) offense. Taken together, these findings provide compelling evidence that ICE detainers in King County and elsewhere primarily target people who do not pose serious security risks to the community. In addition to their policy relevance, these findings suggest that any impact of ICE detainers on jails stays is not a function of the uneven distribution of prior criminal convictions.

ICE Detainer Requests and Type of Release

The jail data indicate that people subject and not subject to ICE detainer requests were released from jail in different ways. Nearly all (96.2 percent) of the individuals sought by ICE were subject to a “transfer of custody” (to ICE) on their release from jail. By contrast, the majority (57.8 percent) of people not subject to ICE detainers were released on bail, bond, or personal recognizance prior to adjudication of their criminal case. According to these data, then, people with ICE detainers are essentially ineligible for release prior to adjudication. In addition, nearly all (98.1 percent) of the people who left jail without having been charged with a crime but had an ICE detainer request were released to the custody of ICE (see Table 2). Thus, being booked into jail has very serious consequences for people flagged by ICE even if they are not subsequently charged with a crime.

Table 2. Reason for Release from Jail, Comparison of Those Charged and Not Charged

Source: Authors' analysis of data provided by the King County Department of Adult and Juvenile Detention, N=33,587.

ICE Detainers and Length of Jail Stay

The average jail stay for people with ICE detainer requests was significantly longer than for those without detainers. Specifically, people without ICE detainers spent an average of 17 days in jail, while those with an ICE detainer request spent an average of 46.3 days in jail. A t-test confirms that this difference is statistically significant at the alpha 0.000 level. Moreover, people with ICE detainer requests spend significantly more days in jail across all charge types (see Figure 1). For inmates with ICE detainer requests whose most serious charge is a felony, the mean number of days spent in jail is more than two times greater than the average number of days spent by people with felony charges but no ICE detainer request (105 vs. 44 days). For those charged with misdemeanors, the average number of jail days is nearly twice that of those without ICE detainer requests (15 vs. 8 days.) People who are not charged with a crime and have an ICE detainer request also spend twice as much time in jail as those who are not charged and do not have an ICE detainer request (4 vs. 2 days). These findings provide compelling evidence that ICE detainers significantly impact the amount of time inmates spend in jail. However, it is possible that these differences stem from case and individual characteristics rather than from ICE detainers themselves. We therefore used OLS regression techniques to isolate the unique impact of ICE detainers on jail stays.

Figure 1. Mean Jail Days by Charge Type, ICE Detainer vs. No ICE Detainer. Source: Authors' analysis of data provided by the King County Department of Adult and Juvenile Detention, N = 33,587. Differences shown are statistically significant (p = 0.000.)

Regression Results

Table 3 above summarizes the descriptive statistics for each variable included in the regression model. In our sample, the maximum time spent in jail was 993 days. Twenty-seven percent of bookings in our sample involved felony charges; 63 percent were charged with one or more misdemeanors; and 9 percent were released without being charged.

Table 3. Descriptive Statistics

The OLS regression results are shown in Table 4. Because the dependent variable is logged in this model, the coefficients have a semielastic rather than linear relationship with number of days in jail, and indicate a percent change in “jail days” for a one-unit change in each dependent variable. Using the standard formula for interpreting regression coefficients when the dependent variable is logged [100(e^β1 −1)], we have included the calculated, semielastic impact of each independent variable.

Table 4. OLS Regression Results of Jail Days

Note: Data shown are ordinary Least Squares (OLS) regression coefficients; standard errors are shown in parentheses. Reference categories: race=white; offense=public order offenses.

^* p < 0.05, ^**p < 0.01, and ^***p < 0.001 (two-tailed tests).

Table 5. Immigration-Related Practices and Policies in Four Urban Counties

^a The Phoenix Police Department has adopted a policy of direct delivery (Eagly 2013: 1183).

^b According to Eagly, Maricopa County prosecutors do recommend, and judges do sentence, non-citizen defendants to probation, but require that all probationers comply with federal immigration law (2013: 1189).

^c See note 2.

^d However, the Phoenix Police Department directs its officers to transport immigrants directly to immigration authorities rather than booking them in a county jail (Eagly 2013: 1182–3).

^e At the time the data were collected, King County jail administrators honored all ICE detainer requests. However, in December of 2013, the King County Council adopted ordinance 0285, which specifies that the county will only honor ICE detainers if the detainee has been convicted of a serious felony offense.

^f Under California state law, the manufacture or use of false documents to conceal one's immigration status is a criminal offense (Eagly 2013: 1167). However, this crime is rarely prosecuted in Los Angeles County (Eagly 2013: 1168).

The regression results indicate that all the legal factors included in the model—the number of charges, the seriousness rank of the most serious charge, and type of most serious charge, and type of offense—have a statistically significant impact on the number of days spent in jail at the alpha 0.001 level or smaller (see Table 4). Only the seriousness ranking of the most serious charge has a negative correlation with number of days in jail. This negative relationship likely reflects the fact that people booked on more serious charges are more likely to be transferred to prison to carry out their sentence. Importantly, the results indicate that ICE detainers more than double the expected number of days in jail, increasing the length of stay by 170 percent, after controlling for other legal factors. ICE detainers, then, have a unique impact on jail stays, significantly extending an individual's time in jail far beyond what one would predict based on case characteristics alone.

The Impact of ICE Detainers on the Criminal Process

While the regression results clearly show that arrestees who are flagged by ICE spend more time in jail than otherwise similar arrestees, they do not shed light on how or why this happens. The qualitative findings presented below identify the mechanisms by which ICE detainers extend jail stays. In our focus group interviews, we found a surprisingly high degree of consensus between prosecutors and defense attorneys regarding how ICE detainers alter the incentive structure and impact the criminal process in ways that extend jail stays for non-citizens. Interestingly, some of these dynamics involved unequal treatment of defendants based on their detainer status. In other cases, however, formally equal treatment also contributed to substantively unequal outcomes. Each of these mechanisms is described below.

Pre-trial release

Shortly after an arrestee is booked into jail, prosecutors decide whether to file charges and, if so, which charges to file. None of the attorneys with whom we spoke indicated that ICE detainers impacted prosecutors' filing or charging decisions. As one prosecutor explained,

In terms of actual charging, it was my practice and I don't think anyone in this room is going to disagree, that whether you have an ICE detainer or not doesn't impact what charges you file…

Although this prosecutor's account may seem self-serving, defense attorneys agreed that ICE detainers did not affect prosecutors' filing decisions. Prosecuting and defense attorneys also agreed that prosecutors were unlikely to dismiss a case simply because deportation was a likely outcome, as exemplified in this exchange between defense attorneys:

DEFENSE ATTORNEY 1: I think as a policy they don't do that [dismiss a case because of an ICE detainer], right?

DEFENSE ATTORNEY 2: Yeah. They want the conviction in case you come back, blah, blah, blah, blah.

DEFENSE ATTORNEY 1: That's right – right.

DEFENSE ATTORNEY 2: They want the conviction.

DEFENSE ATTORNEY 1: You know, in any number of cases we'll say to them, “This is stupid. It's a waste of time. It's a waste of money.” Especially if it's pretty clear the person's going to be deported. People might have success in that sometimes, but … dismissals on that basis alone are pretty rare.

While prosecutors and defense attorneys thus agreed that ICE detainers did not affect prosecutors' filing and charging decisions, prosecutors explained this in terms of their commitment to equal treatment rather than a desire to “get convictions”:

As a general practice, I typically try not to consider immigration consequences at all… just because my goal is to try and treat every defendant that comes in charged with a specific crime equally, fairly. And – and it's not fair to reduce a crime for a noncitizen that you wouldn't reduce for a citizen….

Prosecutors agreed that it was important that criminal cases run their course no matter what immigration matters might eventually impact defendants. For them, this meant that defendants with ICE detainers should not be released from jail prior to adjudication, because a transfer to ICE would disrupt the criminal process:

I would say… you kind of just put a nominal bail on there so that they're not going anywhere. So that they're not going get shipped to SeaTac [the federal detention center] and you're not going to be able to get them back.

Concern about “getting people back” was especially pronounced in cases involving victims:

And that would be our biggest concern: we may never get the person back. And to me that matters the most. Particularly if you have a victim in a case, we don't want to be sitting there explaining to a victim: “Well, some federal authority took this person and we can't pursue the case because of it.”

For prosecutors, then, it was important that the presence of an ICE detainer did not interrupt or halt a case that would otherwise move forward. In order for a case to proceed, the defendant could not be released from jail, because this would almost certainly lead to his or her transfer to federal custody. For this reason, prosecutors reported requesting bail rather than recommending release on personal recognizance for virtually all defendants subject to ICE detainers. But prosecutors emphasized that this did not mean that they were treating people with detainers unequally. Rather, ICE detainers were seen by prosecutors as evidence that the defendant was a “flight risk,” just as strong social ties to another locale would be indicative of “flight risk” in other cases:

PROSECUTING ATTORNEY 1: But if you ask for bail and you file a charge, then we'll keep them here under our jurisdiction.

PROSECUTING ATTORNEY 2: But the practical implication of what he is saying is, you would probably add a line in the bail paragraph letting the court know that there was a detainer.

PROSECUTING ATTORNEY 3: You would.

PROSECUTING ATTORNEY 4: Yeah.

PROSECUTING ATTORNEY 2: I mean, if you know about it, I think most people would probably put that in their bail paragraphs, right. I mean that's something the court's going to want to consider in either approving your bail request or approving your reduction. So if you know about it at the time, I think most of us would put it in there.

PROSECUTING ATTORNEY 3: Right, because you don't want the judge to unknowingly think that they're releasing the person, they're PRing [releasing on personal recognizance] this person. Because they're not getting out, they're just going to SeaTac. Which is just going to make it then potentially impossible for us to process the case…

PROSECUTING ATTORNEY 4: I mean it's the same way when – if you know that they've lived in California their whole life, their families are in California and they've told people they are going back down to California – you would put that in there too. I mean, it's the same type of flight risk [emphasis added].

Prosecutors thus emphasized that that they were not treating non-citizens differently by requesting bail whenever detainers were present; rather, starting from the premise that a detainer signals flight risk, they framed their decision to request bail in these cases as an indicator of their commitment to equal treatment of non-citizen and citizen defendants.

Prosecutors also noted that judges share their reluctance to release defendants with detainers from jail only to have them transferred to federal custody, in part because defense attorneys rarely object to them:

PROSECUTING ATTORNEY 1: This is not –we're not in there as prosecutors saying, “Holy crap, Judge, they've got an ICE detainer. Let's jack this bail up.” That does not happen. It's…

PROSECUTING ATTORNEY 2: ‘Cause that would go over so well.

PROSECUTING ATTORNEY 1: Oh yeah, the judges would laugh us out of the courtroom. At least our judges would – but there may be some jurisdictions where they wouldn't. But it is more by and large defense deferring or not wanting to address bail because of that accessibility issue.

PROSECUTING ATTORNEY 3: I mean, for them to be able to go over across the street versus having to go to SeaTac…. And then warrants get issued when they don't show up and they don't get transported… I mean, just the accessibility is – for any attorney, of course, it's easier to walk across the street to meet with your client than it is to go to another county…

Surprisingly, defense attorneys agreed that judges rarely release people with detainers from jail—and that they rarely sought to have this done:

DEFENSE ATTORNEY 1: If you ask, the court will say, “There's an ICE hold, counsel.” I mean, it's an unspoken rule that the judge is not going to release your client. That doesn't mean we don't try if the client wants us to try.

DEFENSE ATTORNEY 2: ‘Cause the court is going to say, “There's an ICE hold. I'm not going to release this person.”

Defense attorneys further noted that even if a defendant had the financial resources to post bail, they probably would not do so because they would almost certainly be unable to return to court and would therefore forfeit the money:

DEFENSE ATTORNEY 1: And another thing I think we should mention is that it's not just the judge. Because even if the judge grants our motion to release the person on a lower bail, they still can't bail out.

DEFENSE ATTORNEY 2: That's true. You're right.

DEFENSE ATTORNEY 1: They can't bail out because they would forfeit the bail.

In sum, prosecuting and defense attorneys agreed that ICE detainers did not impact prosecutors' filing and charging decisions, but did impact their recommendations to the court regarding pretrial release: prosecutors consistently recommend bail and argue against release on personal recognizance when detainers are present. Both groups further noted that defense attorneys rarely challenge prosecutors' bail recommendations because release on personal recognizance might very well mean transfer to the federal detention facility, where the client would be difficult for defense attorneys to access, and because it is typically in defense attorneys' clients' interests to receive credit for time served in jail. Both groups also agreed that judges nearly always follow prosecutor's bail recommendations in these cases, and that the few defendants who have the financial resources to post bail generally do not because they would likely forfeit their money once transferred to immigration custody. Although attorneys noted that there are rare exceptions to these general patterns, all agreed that pre-adjudication release was essentially not an option for people flagged by ICE.

By contrast, our quantitative results indicate that roughly half of all persons booked into jail without ICE detainers were released on personal recognizance prior to disposition. Non-citizens' de facto ineligibility for pretrial release is especially notable because a disproportionate share of people flagged by ICE face only drug charges, which, absent ICE detainers, do not normally preclude pretrial release.⁷ It is also noteworthy because the Bail Reform Act specifies that the right to bond applies regardless of citizenship status (Eagly 2010: 1305). De facto ineligibility for pretrial release is thus a significant denial of rights and an important reason why people with ICE detainers spend more time in jail than others.

Enhanced complexity of plea negotiations

Once charges have been filed, prosecutors and defense attorneys typically commence negotiations over the terms of a plea deal. Prosecutors and defense attorneys alike indicated that negotiations involving non-citizen defendants, and particularly green card holders, with detainers are often comparatively complex and protracted because so many criminal convictions now have immigration consequences, and because the presence of an ICE detainer calls attention to those consequences. One prosecutor explained it this way:

When I make an offer to somebody that the defense attorneys raised immigration issues on, it will slow things down because they will want to go and consult with immigration attorneys. And oftentimes that'll result in a couple of weeks delay as we go back and forth about – you know, they come talk to me and then they go talk to the immigration lawyer, and then they come back and talk to me and then they go back and talk to the immigration lawyer. I'd say more than a couple weeks – I've had things like that just go on until you are talking for a couple of months…

Defense attorneys also noted the greater complexity of these cases:

DEFENSE ATTORNEY 1: And a lot of times what will happen is the attorney is negotiating with the prosecutor, trying to get a plea that this will not result in an automatic deportation, or exclusion.

INTERVIEWER: Mm-hmm, okay.

DEFENSE ATTORNEY 2: So that also adds time, and makes some things more complicated.

INTERVIEWER: Uh-huh. So maybe it prolongs the plea negotiations?

DEFENSE ATTORNEY 1: Oh, yeah.

Prosecutors pointed out that consideration of immigration consequences often includes the whole family:

Sometimes the immigration consequences have consequences for the family. I mean, if it's an intra-familial offense, whether it's sex or DV, having someone deported can end up deporting the entire family. Or if not deporting the entire family, arguably impoverish the remaining members of the family.

This meant that prosecutors were also more likely to need to discuss matters with family members and victims:

I think a lot of it's just the – the protracted negotiation– that happens… Defense attorneys make special requests based on immigration issues that we need time to think about, and talk with the victims and make sure it's okay. It takes time.

Both prosecuting and defense attorneys also noted that some defendants were willing to serve more time in jail in exchange for receiving an “immigration-safe” charge.⁸ Such offers further complicate plea negotiations and present prosecutors with tricky questions about fairness. As one prosecutor explained,

I've had defense attorneys come in and say: “Okay, he's charged with a felony,” – a felony would have some adverse, uh, consequences as far as deportation. “He's willing to do more time if you give him misdemeanor.”

But it was not just the precise nature of the criminal charges that were the subject of intense negotiations in these cases: getting non-citizen defendants to stipulate that the facts presented in the probable cause certification are true and correct is also unusually complicated. The probable cause certification is a document produced by law enforcement that describes their basis for believing they had probable cause for the arrest. In cases without ICE detainers, most defendants agree to certify that these facts were correct as a matter of course. But as this defense attorney explained, negotiations over the “real facts” in cases involving ICE detainers were fraught, complex and time-consuming:

Ideally you don't want to let the state know that that [the defendant's immigration status] is an issue. But then if you have to work towards getting the state not to check the “real facts” box, or to amend the case to something that's not going to trigger deportation, now they know, and to a great extent they have you over the barrel because, you know, it's just another – it's a difficult position.

Prosecutors agreed that these negotiations made plea negotiations more difficult:

And so normally what we do is, “Okay, I'll give you an Assault IV [instead of an Assault II], just, you know, put in the statement: ‘plead guilty, I punched my girlfriend in the nose.’ You know, and so they stipulate to the real facts in a Probable Cause Certification, and so it's just all done really easy. And that way, you know, we can tell the victims, “Look, he's been held accountable for what he did. He admitted to punching you… But what I get a lot of times is defense attorneys saying like they can't stipulate to real facts in the Probable Cause Certification because that's going to have adverse immigration consequences, and that slows things down. Because then we have to meet with the victim and say: “Well, he's willing to take some responsibility, but not all… he doesn't want the court to know that your nose is broken. And they wouldn't say ‘He punched.’ He says that ‘I touched her offensively.’”

In short, defense and prosecuting attorneys alike emphasized that plea negotiations are generally more complex—and hence tend to take longer—in cases involving non-citizens. This is mainly a result of policy changes adopted in the 1990s that expanded the number and type of criminal convictions that have immigration consequences for legal permanent residents as well as the undocumented and partially documented. Cases involving non-citizens also raise complex ethical questions for prosecutors. The complexity of plea negotiations, combined with de facto ineligibility for pretrial release, means that defendants with ICE detainers stay in jail longer than similarly situated others.

Incentives to set for trial

The focus group data presented in the previous sections show that the presence of ICE detainers, combined with awareness of the immigration consequences of criminal convictions, render plea negotiations far more complex and time-consuming than they would otherwise be. These same forces also create incentives for non-citizen defendants to set for trial rather than accept a guilty plea, at least initially. In part, this is a consequence of prosecutors' reluctance to make deals for non-citizens that they would not also make for citizens:

So they're asking for things oftentimes that I am unwilling to do. And what I also find is that when I am telling folks, you know, “You're going to plead to what anyone else in these circumstances would plead to. And that might have immigration consequences.” But I think defendants are much more reluctant to accept the offer, which means oftentimes setting for trial.

Another prosecutor similarly noted that non-citizens often have more to lose and are therefore more willing to risk setting for trial:

If they have more to lose they're less likely to [plead guilty]. And frankly, I think there's plenty of times where they're not even making rational decisions. Right? They'll spend six months in jail when the recommendation was for thirty days, because they're afraid that they're going to be guaranteed to get kicked out, and they lose all their connections to whatever they have here.

However, attorneys also noted that some defendants, faced with an ICE detainer, attempted to expedite the process and “get it over with.” As one prosecutor explained:

I would also add that I think it goes the other direction too. For example, I had a guy last week who just wanted to go back home because his father was dying…. So he just wanted it done quick: done, plea, do it and get back. So I think it – and that happens more often than you would imagine, probably – that people do just want to get it over with and get done. And there are the other camp who, um, want to know what the immigration consequences are and go that route.

Defense attorneys agreed that faced with immediate transfer to federal custody, some defendants elected to plead guilty to “get it over with,” but that ICE detainers more commonly created an incentive for defendants to go to trial:

DEFENSE ATTORNEY 1: Well, some clients just want to – “Let's just get this over quickly so I can go back.” You've got that class of people. You've got the other class of people: “Well, I'm going to get deported anyway. So any plea offer I get isn't that helpful,” if they're truly undocumented and they think they might be deported but they want to fight it. So in some ways you're almost more likely to go to trial.

INTERVIEWER: Because they don't want that conviction on their record?

DEFENSE ATTORNEY 1: Or ‘cause – any benefit of the plea bargain is non-existent to them, so if they're going to…

DEFENSE ATTORNEY 2: Or if they have family here and they want to stay as long as possible.

DEFENSE ATTORNEY 3: Yeah, exactly.

In sum, attorneys observed that although some defendants flagged by ICE just want to “get it over with” and are, therefore, amenable to plea offers, it was more common for detainers to encourage people to set for trial in an attempt to avoid a conviction that would make it nearly impossible to make a successful claim to legal status in the immigration context. It is not clear whether such defendants actually take the risk of going to trial. Nevertheless, it is clear that the presence of ICE detainers changes the incentive structure and makes most defendants more reluctant to accept plea offers, at least initially. The incentive to set for trial, along with the ineligibility for pretrial release and prolonged plea negotiations, help explain why people flagged by ICE spend more than double the amount of time in jail than similarly situated others.

Alternatives to Confinement

Municipalities around the country are increasingly relying on alternative sanctions that do not involve confinement (Vuong et al. 2010). These alternatives include therapeutic courts, work release, electronic home monitoring, day reporting, and other “intermediate sanctions.” The adoption of these alternatives has been motivated by recognition of the high cost of confinement, and the by fact that incarceration is often counter-productive from both a public health and a public safety point of view (Vuong et al. 2010). Defense and prosecuting attorneys alike pointed out that people with ICE detainers are not eligible for such sentencing alternatives in King County. As these defense attorneys noted:

DEFENSE ATTORNEY 1: The other factor is, uh, it limits your options for sentencing alternatives, so you can't do drug court. You can't do mental health court.

DEFENSE ATTORNEY 2: You can't do work release.

Defense attorneys in the other office also made this point:

And what is really frustrating sometimes is that alternatives like drug court or work release are just not an option. CCAP [Community Center for Alternative Programs], not an option…

In summary, our qualitative findings suggest that King County prosecutors aspire to treat citizens and non-citizens alike, and to separate, as much as they are able, immigration matters from criminal justice. Yet the institutional enmeshment of immigration and criminal enforcement institutions manifested in ICE detainers often renders such “blindness” impossible, leaving prosecutors and defense attorneys to wrestle with difficult ethical and strategic dilemmas. Moreover, prosecutors' commitment to formal equality and immigration “blindness” often works to enhance the pain associated with the criminal process for non-citizens. For example, prosecutors construe their treatment of an ICE detainer as an indicator of “flight risk”; this understanding legitimates their claim that recommending bail in all cases involving an ICE detainer, but only in half of cases involving citizen-defendants, is consistent with their commitment to formal equality. But prosecutors' commitment to formal equality notwithstanding, de facto ineligibility for pretrial release and alternative sanctions, extended plea negotiations, and enhanced incentives to set for trial keep non-citizens in jail far longer than their citizen counterparts. The insertion of immigration enforcement in the criminal process thus dramatically alters that process, creating novel strategic and ethical dilemmas for court actors and enhancing penal pain for non-citizens.

Contextualizing King County

Our findings suggest that the institutional enmeshment of immigration enforcement and criminal justice is enormously consequential: ICE detainers profoundly altered the criminal process and notably extended jail stays when honored by King County jail administrators. However, it is unclear how broadly applicable these findings are, as the effects of detainers (where they are honored) are undoubtedly conditioned by a host of local circumstances and practices. Ingrid Eagly's recent (2013) comparison of immigration-related law, practice and procedure in three urban counties with large immigrant populations—Los Angeles, California, Harris County, Texas, and Maricopa County, Arizona —helps to contextualize our findings and offers insights regarding their generalizability.

Eagly's analysis shows that although immigration issues are central to the operation of criminal justice in all three locales, local immigration-related practices and policies vary significantly. According to Eagly, policy and practice in Los Angeles County comprise an “alienage-neutral model” in which “criminal justice actors endeavor to make decisions that limit the potential effects of immigration status and enforcement on criminal adjudication” (2013: 1157). For example, police officers, prosecutors, and judges do not inquire about defendants' immigration status. Similarly, prosecutors' bail recommendations derive from a schedule that ties bail amounts solely to criminal charges: “Prosecutors explain that to argue that alienage predetermines flight risk would be a local ‘faux pas’” (1161). Criminal justice actors in Los Angeles proactively alter their practices to counter the perceived unfairness of current immigration policy, and consider immigration sanctions to be “quasi-punishments” that should be taken into account (1163–4).

Policy and practice looks quite different in Harris County, Texas. Under the “Illegal-Alien-Punishment Model,” undocumented defendants are segregated and deliberately treated more harshly than citizen defendants. Eagly notes that this model is predicated on the view that undocumented persons are more likely to commit crimes and, when they do, are more criminally culpable than citizens. Once booked, “illegal aliens” are intentionally treated more harshly than others. For example, the bond schedule specifies that felony defendants who are undocumented or have been deported automatically receive a minimum bond of $35,000 (1174). In addition, Harris County District Attorney policy specifies that prosecutors may not offer plea deals that include probation to undocumented persons (1176). Eagly concludes that under this model, “defendants with questionable immigration statuses… are subject to a separate system of heightened bond, not offered plea bargains, and always incarcerated” (1180).

Practice and policy in Maricopa County, Arizona are different yet, and exemplify what Eagly calls the “Immigration-Enforcement Model.” In this model, policy and practice are designed to maximize the immigration enforcement potential of the local justice system by incorporating “the identification of civil immigration law violators into the standard mission of the criminal justice system” (1181). Maricopa County authorities seek primarily to identify and turn over non-citizens to federal authorities with the hope of expediting their deportation. Toward this end, police agencies investigate alienage status, systematically refer non-citizens to federal authorities, and detain unauthorized migrants without bond (1181–2). Court actors also make proactive efforts to determine the immigration status of all criminal defendants, and prosecutors seek to enhance the immigration consequences associated with conviction.

Table 1 provides a summary of the key practices and policies associated with these three locales. We also include information about practices and policies in King County, Washington. Interestingly, King County policies and practices do not neatly fit into any of Eagly's categories. Although King County is a relatively liberal jurisdiction with a comparatively small immigrant population, a few of its practices resemble those found in Harris County, Texas: in both places, non-citizens with ICE detainers receive higher bail amounts than citizens without detainers, rendering them largely ineligible for pretrial release; non-citizens are also, either de facto or de jure, ineligible for noncustodial alternative sanctions such as work release. However, many King County practices are similar to those found in Los Angeles. In both settings, law enforcement officers, prosecutors and judges generally refrain from inquiring about immigration status; prosecutors sometimes craft plea deals that avoid triggering deportation, and neither DOC or probation officers work with immigration authorities. Moreover, both Los Angeles and King County have immigration-blind bail schedules and take active steps to limit the presence of ICE in county courthouses. King County has gone even further than Los Angeles in one respect: since 1999, one of Washington State's nonprofit defense agencies has staffed the WDAIP, which provides written information about immigration consequences and consults with defense attorneys representing non-citizens.⁹

In short, King County resembles, the “alienage-neutral” model found in Los Angeles, although some of the informal practices described by respondents have consequences that resemble those achieved in Harris County through formal policy. The fact that detainers so profoundly affected the criminal process and extended jail stays in a relatively progressive county suggests that detainers may have even more dramatic effects in large portions of the country. Future case studies and comparative work may yield useful insights regarding this question.

Discussion and Conclusion

As studies of legally hybrid control mechanisms led us to expect, our findings indicate that programs such as Secure Communities have significant effects on criminal justice outcomes where ICE detainers are honored. Specifically, people subject to ICE detainers spend, on average, 170 percent more days in jail than others with similar case characteristics. Although we were unable to control for criminal history in our regression models, our analysis of the criminal records of a subsample of King County jail inmates, along with other national studies (TRAC 2013b; Thompson and Cohen 2014), suggest that these findings are not a function of an uneven distribution of criminal records. Indeed, our findings call into question the claim that immigration detainers primarily target people with comparatively serious criminal histories.

Importantly, our findings suggest that the extension of jail stays for non-citizens in King County does not occur because jail administrators hold detainees flagged by ICE for more than 48 hours after their release has been ordered by a criminal court judge. Rather, ICE detainers trigger four dynamics that profoundly alter the criminal process and extend jail stays for non-citizens. The first and arguably most important of these is the de facto absence of a pretrial release option for people subject to ICE detainers. The reluctance of courtroom actors to seek the pretrial release of people who would likely be transferred to ICE if released from jail is perfectly rational given their shared interest in processing criminal matters as efficiently as possible. Yet it also illustrates how “equal treatment” of defendants deemed to be flight risks can nonetheless produce systematically unequal outcomes. ICE detainers—together with widespread awareness of the looming specter of deportation—also make plea negotiations far more complex and time-consuming than they would otherwise be. Moreover, the fact that some convictions have immigration consequences creates incentives for non-citizens to set for trial or offer to do extra time in exchange for “immigration-safe” charges. Finally, ICE detainers render those they mark ineligible for alternative sentencing dispositions. Together, these mechanisms keep non-citizens in local jails far longer than their citizen counterparts in King County, Washington.

Our study has several limitations. Unfortunately, the jail release data did not allow us to compare the consequences of ICE detainers for legal permanent residents and undocumented persons. In addition, it is difficult to assess how generalizable our findings are, as the precise impact of ICE detainers in other locales likely depends in part on local practices and policies. Drawing on Eagly's (2013) comparative analysis of law and procedure in three large southwestern counties, we conclude that practices and policies in King County are relatively progressive. Certainly the County's provision of dedicated immigration experts to defense attorneys sets the county apart. Moreover, in King County, both prosecutors and judges aspire to “immigration-blindness,” that is, the formally equal treatment of citizens and non-citizens and the separation of immigration matters from criminal issues. Although defense and prosecuting attorneys agreed that non-citizens were not intentionally singled out for harsh treatment in King County (as they are in some locales), the effort to be “immigration blind” in a county where ICE detainers were (at the time of the study) honored by jail administrators prolonged non-citizen defendants' jail stays and denied them rights and opportunities that defendants without detainers enjoyed, including the right to bond and eligibility for non-confinement sanctions.

In short, although the effects of ICE detainers likely vary across locales, it is notable that ICE detainers significantly extended jail stays and led to the denial of rights even in liberal King County. Moreover, there is evidence that detainers also prolong jail stays in Los Angeles (Greene 2012), where, according to Eagly (2013), practitioners and policymakers alike attempt to counter the perceived unfair impact of immigration policy by purposefully taking immigration consequences into account and treating them as quasi-punishments. It thus appears that where ICE detainers are honored, the threat of deportation casts a long shadow over local criminal proceedings—even where authorities attempt to either ignore or mitigate that reality. Interestingly, the King County council recently voted to limit the circumstances under which ICE detainer requests will be honored by jail administrators, joining a number of other counties that have elected to limit their cooperation with federal government (Admur 2014). However, a majority of counties continue to fully participate in the Secure Communities program.

Our findings have important substantive implications. Feeley's (1992) classic account of the handling of misdemeanor cases in New Haven, Connecticut suggests that the criminal justice process is inherently punitive, so much so that misdemeanants endeavor mainly to disentangle themselves from the courts as quickly as possible. Where ICE detainers are honored, the conjoining of the immigration and criminal systems appears to prolong both criminal case processing and jail stays, thus enhancing the degree of penal pain experienced by non-citizen defendants. Incarceration, including short-term jail stays, has a number of adverse social, psychological and economic consequences (Clear 2007; Freudenberg 2002; Freudenberg et al. 2008). In particular, the denial of pretrial release not only prolongs incarceration but also increases the likelihood of conviction and receipt of a prison sentence (Devers 2011; Phillips 2007, 2008). De facto ineligibility for alternative sentences also increases the likelihood that non-citizens will be sentenced to jail or prison. Insofar as detainers extend non-citizen’ jail stays, prolong their contact with the criminal justice system, and increase the likelihood of conviction, they are an important means by which penal pain is differentially imposed on non-citizens.

Although our findings are based on a single case study, socio-legal scholars analyzing other venues and court systems have reached similar conclusions. For example, Cade argues that, “the deportation and misdemeanor prosecution systems interact to produce graver injustices than observers have previously understood” (2012: 1811). Similarly, Eagly notes that non-citizens in the federal system are “treated differently along alienage lines” (2010: 1317). For example, 75 percent of all citizen defendants, but only 14 percent of non-citizen defendants, facing a sentence of six months or less are sentenced to probation (2010: 1318). Similar gaps exist for people with longer recommended sentences (2010: 1318). Moreover, the Federal Bureau of Prisons designates all non-citizen defendants “deportable aliens” without an investigation into their legal status, flight risk or dangerousness; this classification, in turn, subjects non-citizen defendants to harsher treatment in prison, including assignment to facilities with higher security levels, more stringent recommendations, fewer recreational programs (2010: 1318).

There is, then, a significant body of evidence suggesting that the intermingling of the criminal and immigration systems notably disadvantages non-citizens caught up in the criminal justice system in various locales. These findings support and extend Menjívar's (2012) argument that recent immigration policies have destabilized the identities and lives of many immigrants, and intensified the suffering associated with possession of an uncertain legal status. Based on her interviews with Central American immigrants, Menjívar finds that recent policy changes constitute legally sanctioned violence; this violence, in turn, has enhanced the vulnerability and pain her respondents experience in three important social domains: work, school and family. Our analysis suggests that the enmeshment of criminal and immigration law also enhance non-citizens' pain and suffering in another realm as well: local criminal justice systems.

The findings presented here also have a number of conceptual and theoretical implications. Recent studies find that citizenship status has an important effect on sentencing outcomes in the federal courts (Hartley and Armendariz 2011; Light 2014; Light, Massoglia, and King forthcoming; Wolfe, Pyrooz, and Spohn 2011; Wu and Delone 2012). Although these studies show that non-citizens—including both legal permanent residents and especially undocumented immigrants—are more likely to be sentenced to prison and receive longer prison sentences than their similarly situated citizen-counterparts, this quantitative literature does not shed light on the mechanisms by which citizenship status affects criminal case outcomes. As these researchers note, a variety of theoretical literatures suggest that immigrants are subject to negative stereotypes and stigmatization; these cultural and social-psychological process may explain why non-citizens are disadvantaged in the federal courts. Although this interpretation is entirely plausible, our analysis underscores the need to also consider the possibility that the enmeshment of criminal and immigration institutions—and the institutional practices that facilitate this enmeshment—may also help to explain why non-citizens are sanctioned comparatively harshly. Future research comparing the impact of citizenship status in locales that do and do not honor ICE detainer requests may help clarify the relative importance of this institutional mechanism.

Our findings also contribute to the “crimmigration” literature by showing that the effects of the convergence between criminal and immigration law reach beyond the federal criminal and civil immigration systems. “Crimmigration” has led not only to the criminalization of immigration offenses and the creation of a “two-tiered” system of justice at the federal level, but, more broadly, appears to have transformed the criminal process for non-citizens in state and local justice systems in ways that enhance the pain associated with criminal punishment. Given the large number and growing number of non-citizen residents of the United States and the unprecedented magnitude of the U.S. criminal justice system, the impact of immigration law and enforcement on the criminal process can no longer be ignored.