Pressured into a Preference to Leave? A Study on the “Specific” Deterrent Effects and Perceived Legitimacy of Immigration Detention

Abstract

Immigration detention is formally not a punishment, but governments do use it to deter illegal residence. This study examines whether and how immigration detention affects detainees’ decision-making processes regarding departure, thereby possibly resulting in de facto “specific deterrence.” Semistructured face-to-face interviews were conducted in the Netherlands with 81 immigration detainees, and their case files were examined. Evidence is found for a limited, selective deterrence effect at the level of detainee's attitudes: most respondents considered immigration detention a painful and distressing experience, but only a minority—mostly labor migrants without family ties in the Netherlands—developed a preference to return to their country of citizenship in hopes of ending their exposure, including repeated exposure, to the detention. In line with defiance theory, we find that eventual deterrent effects mostly occurred among detainees who also attributed some measure of legitimacy to their detention. Among some detainees, the detention experience resulted in a preference to migrate to a neighboring European country.

Immigration detention is an administrative measure to ensure that migrants cannot abscond while preparations for deportation are being made (see Cornelisse 2010; Wilsher 2012). Two main types exist: (1) preadmission detention at the border, involving foreigners not admitted to the state's territory, and (2) pre-expulsion detention of foreigners whose stay in the territory is, has, or is likely to become unauthorized. Although formally not a punishment, governments do use immigration detention to deter unwanted immigrants from the territory (Campesi 2015, DeBono 2013; Hasselberg 2014; Kalhan 2010; Leerkes and Broeders 2010; Mainwaring 2012; Martin 2012; Pickering and Weber 2014). That claim rests on three main observations: (1) barring exceptions, detention occurs under regimes resembling criminal imprisonment; (2) immigration detainees tend to experience the detention as a kind of punishment, and (3) various policy makers have publicly stated that immigration detention is meant to pressure detainees into leaving. For example, in the Netherlands, the country on which the present analysis focuses, a former Minister of Immigration and Asylum argued that the purpose of immigration detention is to “incite to departure” (“prikkelen tot vertrek”) (Parliamentary Documents II 2010/11, 19 637, no. 1396).

It is illegal under international law to use immigration detention to deter future asylum seekers or to dissuade those who have commenced their claims from pursuing them (United Nations High Commissioner for Refugees 1999; see also Mainwaring 2012). It is less clear whether states can use it as a means to coerce rejected asylum seekers and other deportable migrants into a willingness to leave. Various nongovernmental organizations believe not, and have criticized the lengthy maximum duration of immigration detention as well as the prison-like conditions under which immigration detention is carried out (see for example Amnesty International 2009; Jesuit Refugee Service-Europe 2010). Social scientists, too, have pointed at the harmful effects of immigration law more generally for the well-being and social incorporation of those with weaker legal statuses (see Menjívar and Abrego 2012). We agree that immigration detention should, in accordance with international and national laws, only be used as a last resort in order to prevent the risk of absconding, and should be carried out under conditions adjusted to the administrative nature of the detention.

In addition to the legal-normative question of what uses of immigration detention are permissible and acceptable, there is also the empirical-theoretical question of whether and how immigration detention affects migration preferences, and, if so, whether its influences are a result of deterrence, or whether other mechanisms, such as the perceived (il)legitimacy of the detention, (also) play a role. The central question of this contribution is: How, and to what extent, does immigration detention—under conditions found in the Netherlands in 2011—affect the willingness of detainees to leave the territory of the detaining state, and are eventual changes in detainees’ migration preferences produced by deterrence? The analysis is based on unique data that was initially gathered in cooperation with the International Organization of Migration (IOM) (Kox 2011). Semistructured face-to-face interviews were conducted with 81 immigrants who were being held in pre-expulsion detention in 2011, and information was obtained about the administrative outcome of their detention. In this article, we analyze detainees’ attitudes as they were expressed during detention. Elsewhere, we analyzed the relationship between detainee attitudes and detention outcomes (Leerkes and Kox 2016), but some results are also reported here. Post-deportation attitudinal changes lie outside the scope of the study. The data were fully reanalyzed after the cooperation with IOM had ended.

It is conceivable that immigration detention indeed causes forms of deterrence that, in some ways, resemble what would be called “specific deterrence” in the context of criminal imprisonment.¹ Here, the wish to prevent additional detention—either continued or repeated detention—possibly produces “law-abiding behavior” among those who have actually been detained. Three varieties of such deterrence potentially exist. First, detainees may be pressured into cooperating with the deportation procedure by giving up their claims to legal status or by disclosing their nationality and/or identity. The detaining state has an interest in identifying and documenting (undocumented) detainees, as no country of origin accepts undocumented returnees (Broeders 2007; Ellermann 2008). Second, those who are released because of a failed deportation procedure may try to leave the country on their own so as to prevent repeated detention. Third, deportees may refrain from re-immigrating in order to prevent repeated detention. This study focuses on the first and second variety of specific deterrence, so understood.

There is only fragmentary and contradictory evidence about whether and how immigration detention impacts detainees’ migration preferences and behavior. To our knowledge, the present study is unique in addressing these questions on the basis of a substantial number of interviews conducted in detention centers. Hasselberg (2014: 481) conducted ethnographic research among 18 foreign-national offenders who were facing deportation from the United Kingdom, but had been granted bail from immigration detention. Most respondents indeed perceived the detention as being designed to pressure them into leaving, and some of them—especially those who had been detained repeatedly—felt that the detention “‘break[s] one down’ to the point of agreeing to deportation.” Kalhan (2010), too, mentions in passing that for the United States, there is a deterrent effect in the sense that detainees may give up their claims to legal status and comply with deportation in order to end the detention. Mainwaring (2012), however, argues that economic hardship and political persecution give African migrants little choice but to continue their migration project regardless of the risk of being detained in Malta for 18 months. Unfortunately, her analysis potentially suffers from selection regarding the dependent variable, as only unauthorized migrants were interviewed outside of detention, thereby possibly excluding individuals from the sample who were deterred by detention and no longer stayed on Malta. Van Kalmthout (2004) conducted extensive fieldwork in two Dutch detention centers, but the authors did not specifically investigate the eventual impact of detention on migration preferences. They do report that most respondents did not want to go to their country of citizenship in the short term, and that their preferences did not seem to change much.

The analysis focuses on deterrence, but we also pay ample attention to how migrants perceive the legitimacy of immigration detention. There are different conceptualizations of legitimacy (see Fallon 2005). As a legal concept, legitimacy refers to lawfulness and internal consistency of the law, such as a law being in accordance with the constitution. As a philosophical concept, legitimacy means that a law or institution is believed to be right and justifiable in the light of some philosophical principle, such as the principle of sovereignty. Here, we follow a sociological conceptualization of legal legitimacy, which pertains to the normative acceptance of the law in society, especially among those who are targeted by the law. A practice can be in accordance with the law, while not being perceived as legitimate by the ruled. Criminal justice research has shown that perceived legitimacy of the law is crucial for compliance. Usually, normative models of compliance are even found to have more explanatory power than economic (“rational choice”) models of compliance such as found in deterrence theory (see for example Tyler 2003; Tyler et al. 2010), and this also seems to hold for migration decisions that are reached outside of detention (Ryo 2013). There also is the possibility that deterrence and perceived (il)legitimacy interact in complex ways. Defiance theory, for example, argues that deterrent effects require that the targets of a sanction perceive the sanction as legitimate (Sherman 1993). While we contend that this claim is too strong, it seems probable that deterrent effects become more likely when the detainee attributes some measure of legitimacy to the detention. Paying attention to perceived (il)legitimacy also is a way to do justice to, and provide more information about, the contested nature of immigration detention, or what Bosworth (2011, 2014) calls its “legitimacy deficit,” thereby strengthening the detainee perspective in this debate.

In what follows, it is shown that immigration detention, though generally perceived as rather painful and distressing, produced limited specific deterrent effects at the attitudinal level, which were concentrated among respondents classified as labor migrants and, relatedly, among respondents who attributed some measure of legitimacy to their detention. It is argued that most migrants had interests that outweighed their interest in the detention being ended, and that the detention also triggered mechanisms—including resistance on normative grounds—that actually reinforced detainees’ preferences to stay in the country of the detaining state.

The next section provides some contextual information on the Dutch case. Subsequently, we elaborate the theoretical framework, describe how the fieldwork was carried out, and report our results. In the concluding section, we discuss the broader scientific implications of this contribution for the study of international migration and the study of legal compliance.

The Dutch Immigration Detention Regime

As there are international differences in how immigration detention is organized (see Wilsher 2012), it is useful to provide some contextual information on the Dutch case in 2011. According to the EU Return Directive, immigration detention may last for six months, unless the detainee is considered uncooperative in revealing his/her identity and/or has been convicted of certain crimes, in which case the detention may be extended to 18 months. Within this limit, there is considerable international variation in Europe. The Netherlands is in a cluster of countries also including Germany, Denmark, and Italy, where the maximum detention duration is 18 months. These countries in particular, seem to use the detention for deterrence purposes given the long maximum detention duration. In that respect, they resemble the United States, Australia, and the United Kingdom (which opted out of the EU Return Directive), which do not have a pre-set maximum detention duration and sometimes use it for years in a row. Other EU countries, such as France (45 days) and Belgium (five months), have considerably shorter maximum detention durations, or use the detention for relatively short periods (in 2008, the average detention duration in Sweden was 20.8 days).² In the Netherlands, about half of those detained in 2011 spent more than three months in detention, while a quarter were detained for six months or more. On average, pre-expulsion detention lasted 74 days (Van Schijndel and Van Gemmert 2012). In 2011, the total inflow in Dutch immigration detention amounted to 6,104 persons, and 5,844 persons entered pre-expulsion detention (Van Schijndel and Van Gemmert 2012).

Noncitizens can be detained under Dutch law if various conditions are met. These include the foreigner lacking a right to stay in the country (or being likely to lose that right) and there being a realistic probability of deportation. If there is insufficient view to deportation, or if the government has made insufficient progress toward deporting the individual, the detention has to be terminated. Released detainees are usually given a day train ticket and an order to leave the country. Those who stay can be re-detained, provided that there is a “new fact” that supposedly increases the probability of deportation, such as new information about the individual's identity, or when at least a year has passed since the previous detention period. Immigration judges may decide that a detention is unlawful, and immigration attorneys have significant influence over such decisions (see Ryo 2016), but such institutional safeguards have not prevented significant levels of repeated immigration detention. In 2010, 45 percent of the detainees were released because of a failed expulsion procedure, and 27 percent had been in Dutch immigration detention previously (Kox 2011). In the case of repeated detention, the cumulative detention duration may exceed 18 months.

At the time of writing, a governmental proposal to develop a specific detention regime for immigration detainees is being discussed in parliament,³ but in 2011, the pre-expulsion detention regime was modeled after the regime for crime suspects who are detained before their trials. Detainees were to stay in their cells—mostly designed to hold two people—between 5 pm and 8 am, and between 12 noon and 1 pm. When out of their cells, they were to spend time in a common recreation room with cooking facilities, television, table tennis, and parlor games. A program of 18 hours a week was on offer, including airing (one hour per day), sports (45 minutes twice a week), recreational activities (six hours a week), library visits (one hour per week), religious ceremonies (one hour per week), and visits (two hours per week). There were no “re-integration facilities” such as labor, educational programs, or training courses.

In the Netherlands, some detainees are given the option of leaving the country via “Assisted Voluntary Return from Detention” (AVRD), instead of being deported. In case of AVRD, detainees are released at the airport where they meet an IOM employee in the departure hall. This happens under a number of conditions, which includes the “removal” not being delayed.

An additional specificity of the European context is that released detainees have the option of (illegal) onward migration: migrating to other European countries is illegal, but not too difficult, since land borders are normally rarely policed within the Schengen Area, an area comprising 26 European states that have abolished all passport controls at their mutual borders. The Area includes the Netherlands and the two countries where it has land borders with (Germany and Belgium).⁴ In the past, rejected asylum seekers and (other) unauthorized migrants could obtain legal stay in a different European country, if only temporarily—i.e., during the asylum procedure—by submitting an asylum request there. However, under the Dublin III Regulation (EC 604/2013), a successor of the Dublin Convention that became in force in 1997, those reapplying for asylum in a different European country, or who apply for asylum while they could have done so sooner in a different European country, may be rejected without further consideration and be returned to the “responsible state” for examining the asylum claim, usually the state through which the migrant first entered the EU (see Thielemann and Armstrong 2013).

Compared to the United States, where an estimated 52 percent of all unauthorized immigrants in 2014 were Mexicans (Passel and Cohn 2016), the unauthorized population in the Netherlands is quite heterogeneous in terms of national origin. In 2011, more than 60 percent of the detainees were from countries that, on their own, represented less than 5 percent of the detainees (Van Schijndel and Van Gemmert 2012). The most prevalent nationality was Iraqi with 9.6 percent. It has been estimated that women represented almost one third (29 percent) of the unauthorized population in 2012 (Van der Heijden et al. 2015). Since women run significantly lower risks of apprehension than men (Leerkes et al. 2012), they only represented 13 percent of all immigration detainees in 2011 (Van Schijndel and Van Gemmert 2012).

Conditional Deterrence?

Immigration detention in the Netherlands has three main direct “migration outcomes”: (1) deportation or AVRD to a country where the detainee's admission is guaranteed, usually the detainee's country of citizenship, (2) a continued unauthorized stay in territory of the detaining state, and (3) illegal onward migration to another European country.⁵ Before being detained, each detainee will generally prefer continued stay in the territory of the detaining state, although this does not hold for all detainees.⁶ “Specific deterrence” at the attitudinal level can be said to have occurred when a preference for continued stay has changed into (1) a preference to return to the country of citizenship or (2) onward migration to a different European country—and that the change is due to the detention being experienced as frightening or discouraging.

Potentially, two aspects of the immigration detention experience could produce such attitudinal deterrent effects. The first involves the perceived severity of the immigration detention experience. Much like criminal prisoners, immigration detainees are subjected to various forms of deprivation, or what Sykes (1958) famously called the “pains of imprisonment.” A reduction of autonomy and a loss of liberty are inherent in detention and most detainees experience a reduced access to social relationships as visitors face travel costs and visiting hours, internet access and opportunities to make phone calls are limited, and potential visitors who are unauthorized themselves are unlikely to visit the centers. Additionally, detainees experience material deprivation, especially those who were ordinarily employed. Detainees typically share a cell with a stranger and cannot work but receive some pocket money to buy some goods in the detention shop. Deprivation of security may also be relevant, as, in Sykes's words, each immigration detainee is “thrown into prolonged intimacy” with other detainees. This can provoke anxiety, especially when other detainees have a criminal background and/or are mentally instable. Crewe (2011) recently argued that Sykes's pains of imprisonment, though still relevant today, represent only a part of the painfulness of imprisonment as detainees may also be subjected to what he calls the “pains of uncertainty and indeterminacy.” This certainly also holds for immigration detainees as they do not know in advance how long they will be detained (within the maximally allowed detention duration), while it is often unclear whether they will be deported (see Becket and Evans 2015).

The second aspect involves the perceived risk of repeated immigration detention. Being detained implies that one has been “caught,” and that one's unauthorized status is known to the authorities. People who have experienced a sanction tend to perceive the probability of similar sanctions in the future as higher than before the sanction was imposed (Matsueda et al. 2006; Piquero and Pogarsky 2002), and this may also hold for the sanction of immigration detention. Perceived probability of a sanction, in turn, tends to increase rule compliance (see, for example, Tyler 2003). In other words, we may find detainees reasoning that it is in their interest to leave the territory of the detaining state because, having been apprehended and detained, they believe that repeated detention is likely should the deportation procedure fail.

For two reasons, detention may be perceived as painful and repeated detention as probable, without continued stay becoming the least preferred detention outcome. First—taking the rational choice perspective that is implied in deterrence theory—there may be a kind of threshold effect before actual deterrence occurs (see Tyler 2003). After all, the “utility” of staying in the country of the detaining state, though lowered by the “costs” of the detention experience, may still be higher than the “utility” of being sent to one's country of citizenship or, should the deportation procedure fail, of going to a third country. Similarly, it is conceivable that detainees experience the detention as painful, while they do not see it as a major deterioration of their situation, such as when the person concerned normally leads a strongly marginalized existence in the territory of the detaining state when out of detention.

Second, the deterrent effects may be offset by resistance should the detention be perceived as illegitimate. Normative models of rule compliance maintain that rules are followed because people believe that doing so is right or, as the neo-institutionalists emphasize (Scott 2008), as “normal,” that is, “as the way we do these things.” A distinction is usually made between the normative evaluation of outcomes and procedures (Tyler 2003). Outcome legitimacy means that the law's content is considered fair; procedural legitimacy means that rules are enforced in ways that are perceived as just, by actors who are trusted. The first type of legitimacy encompasses a judgement about the acceptability of the official objective of (pre-expulsion) immigration detention, namely that the detainee leaves the territory over which the detaining state claims territorial sovereignty, usually by “returning” to his or her country of citizenship. The second type of legitimacy encompasses a normative judgement about the acceptability of how and by whom that outcome is realized. This both entails a judgement about the use of immigration detention in cases like the detainee's, and a judgement about how the detention is organized in terms of detention duration, the possibility of repeated detention, available programs, interactional styles of the guards, and so forth. Migrants who perceive their deportation and/or their deportation procedure as illegitimate can be expected to try to obstruct the deportation procedure.

Defiance theory has aimed to synthesize deterrence theory with legitimacy theories (Sherman 1993). It argues that the deterrent effects of a sanction require that those subjected to the sanction perceive it as legitimate. While this claim seems too strong—we can conceive deterrent effects in the absence of perceived legitimacy—it does suggest the possibility of a kind of “interaction effect” in the sense that people may be most likely to comply when the product of perceived severity, perceived certainty and perceived legitimacy reaches a kind of optimum. In this view, increases in perceived severity are likely to eventually undermine compliance by reducing perceived (process) legitimacy.

Data and Method

The data were initially collected as part of a research project for IOM Netherlands, which wanted to learn more about how detainees experienced the detention, and how they assessed IOM's presence in the centers in the context of AVRD. The project was directed by the second author who was hired for the project by IOM; the Sociology department of Erasmus University Rotterdam—under guidance of the first author—provided advice and assisted with the fieldwork and analysis.

Semistructured interviews with 81 immigration detainees were conducted in May, June, and July 2011 in the three main detention centers that were operational in the Netherlands at the time (Rotterdam, Zaandam, and Zeist). All respondents were randomly selected from the resident lists. IOM employees working in the detention centers tried to approach 232 detainees; of these, 144 persons could not be interviewed: (1) 62 persons were not in the centers because of departure, release, or appointments at courts or embassies, (2) 51 persons were unwilling to be interviewed, or anticipated that it would be too emotionally demanding, and (3) 30 candidates faced communication issues. Apart from the 81 interviews, seven interviews were terminated prematurely due to unexpected language barriers or because these interviewees found the interview too emotionally demanding; these were excluded from the analysis. The interviews took place in the detention centers’ consulting rooms and lasted between 45 minutes and 3 hours.

The interviews were conducted by the authors, by IOM employees who specialize in client contacts in the clients’ mother tongue, and by a university research assistant.⁷ All interviewers received interview training. While some respondents criticized the IOM, most did not feel that IOM put pressure on them to leave the Netherlands, and we have the impression that respondents felt comfortable telling their stories, also given the detailed information they provided, including during the interviews with the IOM employees. The interviewers recorded the interview or took notes, depending on the possibilities for recording and the preferences of interviewee and interviewer. All interviews were held in the respondent's first or second language, with ten different languages being used in total. A telephonic interpreter assisted with five interviews. The respondents’ case files held by IOM, the Dutch Immigration and Naturalization Service (IND) and the Repatriation and Departure Service (DT&V) were also studied in order to gather procedural information, including the respondents’ juridical position and history, possible previous detainments, and the administrative outcome of the detention in terms of deportation, AVRD, or release in the Netherlands. Respondents were guaranteed anonymity, but the Custodial Institutions Agency knew who participated, as respondents could only be interviewed by appointment in a detention setting.

Respondents were interviewed about their situation in their country of citizenship and the Netherlands, their stay in immigration detention, their thoughts on the future and their attitude with regard to staying in or leaving the Netherlands, and the possible consequences of their stay(s) in immigration detention for their migration preferences. We realized that the respondents possibly perceived an interest in creating an impression that they were “cooperating” with the removal procedure: detainees who are seen as uncooperative by immigration judges risk an extension of the detention period. In order to minimize that risk, relatively long and open interviews were designed, during which we attempted to create “rapport.” For example, before sensitive questions regarding “return” were asked, respondents were given the opportunity to explain in their own words why and when they had migrated, what their situation in the Netherlands had looked like, and what they thought of immigration detention. Additionally, respondents were not explicitly asked whether they were “cooperating” with the Dutch authorities, but—using open questions—were interviewed about eventual changes in their migration preferences. For example, it was asked “Have your return intentions changed during your stay in immigration detention? How? Why/why not?” Two questions measured the perceived legitimacy of the detention: (1) “Do you think the Dutch government has a right to detain you? (Why/Why not?)” and (2) “Do you think the detention conditions are appropriate given the reason for your detention? (Why/Why not?).” Both questions incited statements about perceived process legitimacy. The first question in particular also provided information about perceived outcome legitimacy.

After the project for IOM had ended, all interviews were reanalyzed. Both authors independently coded the respondents’ statements, including their migration preferences at the time of the interview and their (retrospectively reported) preferences before the detention period. For repeatedly detained respondents, the stated migration preference before the first detention period was also coded. There was insufficient information to systemically classify the respondents on perceived outcome and process legitimacy separately, but their statements did allow us to tentatively divide the respondent into a “low” and a “higher” perceived legitimacy group. The first group believed that the Dutch government had no right to detain and deport them whatsoever and also believed that the detention conditions were inappropriate in their case. Most respondents in the second group were also quite critical about at least some aspects of the detention, but either indicated that the government had some kind of right to detain people in cases such like theirs (be it for a shorter period, not repeatedly, and so forth) and/or found the detention conditions appropriate. If transcripts initially resulted in different codings, the final coding was decided on after discussion among the authors.⁸

While analyzing the interviews, we inductively found indications that the effects of detention, and the reasons that respondents had to try to resist deportation, depended on respondents’ migration projects. In order to examine such differences in a more systematic way, we classified the respondents into family migrants, asylum migrants, and labor migrants. This was done using the information that respondents provided during the interviews regarding their reasons for leaving their respective countries of origin. Family migrants were defined as respondents who mentioned that a desire to reunify with family or a partner in The Netherlands had been among the reasons for leaving their country of origin. Asylum migrants indicated that a desire to apply for asylum had been among the reasons for emigrating. Respondents who had come to Europe without reunifying with immediate family members, and without the intention to apply for asylum, were classified as labor migrants. Most of them had migrated to Europe and (eventually) the Netherlands to work; a few had come in hopes of being able to study in Europe or had left their country of origin without a clear plan for the future. It could be said that we followed a hierarchical classification. First, we determined whether the respondent left his or her country or origin in order to reunify with family (yes: “family migrant”; no: “asylum or labor migrant”); as a second step, we then determined for each “asylum or labor migrant” whether he or she had left the country or origin to apply for asylum (yes: “asylum migrant,” no: “labor migrant”).

The classification of respondents into these three groups—it may be better to speak of partially overlapping “clusters of respondents”—served as a heuristic tool that helped us to understand and describe the implications of what are meaningful differences in respondents’ migration projects. It should be emphasized that we do not mean to essentialize immigrant categories. Evidently, people may migrate for more than one reason in complex “configurations of motives,” which may also change during the migration project. Some respondents classified as asylum and labor migrants eventually obtained a partner in the Netherlands. A few respondents classified as labor migrants applied for asylum at a later stage. Various respondents classified as family or asylum migrants also mentioned economic reasons for leaving their country of origin. While all asylum migrants in our sample had migrated without reunifying with direct family members, some had remote relatives in the Netherlands.

We repeatedly discussed ethical aspects among each other, including the question of whether it is ethical to conduct research on the (in)effectiveness of immigration detention given the possibility that the results may be used in society to legitimize immigration detention, including the—ethically and legally dubious—use of lengthy immigration detention under prison-like conditions. Unfortunately, there is little scientific debate on such normative aspects (see Düvell et al. 2009), and we found that scientific codes of ethics provide limited, perhaps insufficient, guidance.⁹ We eventually concluded that it is justified and important to conduct social scientific research on this topic. We consider the research not a priori harmful in democratic societies, where different publics may or may not use research findings for different purposes, including purposes that researchers do not personally advocate. We hope that our findings contribute to a richer scientific and public debate on the functions, human rights performance, and (il)legitimacy of immigration detention.

Results: Pressured to Depart?

Quantitative Overview

The analysis focuses on how the respondents experienced the detention. However, given the relatively large sample size and the use of probability sampling, we decided to also present some quantitative findings based on our codings, so as to make optimal use of the data (see Namey et al. [2008] on the possibilities and limitations of quantifying qualitative data). Before describing how the respondents experienced the detention, we therefore start with a description of the main characteristics of the research group, followed by a quantitative overview of the reported migration preferences of the respondents.

Research Group

The respondents, 73 men and eight women, were from 36 different countries, mainly located in North and West Africa, South/Southwest Asia, the Middle East, and Latin America (mostly Surinam). Ages varied from 18 to 67 and averaged around 34. The detention frequency varied from one to no fewer than eight times, and 33 respondents had been detained previously. The average duration of the most recent detention at the time of the interview was 4.4 months, varying from a few days up to 15 months. Some respondents had only been apprehended because of illegal stay in the country, others also had police contacts due to suspicions regarding felonies and/or minor misdemeanors. All but five respondents were illegally residing in the Netherlands at the time of the interview: four had a precarious form of legal stay as part of a procedure to obtain a residence permit, and one risked losing his residence permit because he had been sentenced for a crime. Information by the Custodial Institutions Agency indicated that the sample was quite representative with regard to sex, age, and migration motive, with former asylum seekers making up about half of the detainee population. Detainees with longer detention durations were somewhat overrepresented, however, and some nationalities—including Somalian, Georgian, Algerian, Ghanaian, and Eritrean—were underrepresented.

As was mentioned in the previous section, respondents were classified into family migrants, asylum migrants, and labor migrants based on their initial migration motive. The three groups indeed differed on a number of characteristics that can be expected to be correlated with migration motive, including country of origin, number of former residence permit holders, and duration of stay in the Netherlands. The groups were quite similar in other respects, including age and sex composition, and there were no notable differences with respect to their detention duration.

The respondents classified as family migrants (N = 13) mostly originated from countries that have been source countries of immigration to the Netherlands for some time, like the former colony Surinam, and the former “guest worker” country Morocco, but others were from Tunisia, Pakistan, Egypt, and Syria. Family ties do not preclude an unauthorized status as the requirements for legal family migration include income and age requirements, while certain family relations, such as those between siblings or between parents and adult children, do not normally give a right to a residence permit (Leerkes and Kulu-Glasgow 2011). Most family migrants had immigrated without state permission, but four used to have a residence permit, which had been ended because of criminal offending. The duration of stay of the family migrants varied from a number of months up to 32 years, with an average of 14 years, making this the most “established” group with the longest average duration of stay. The longer duration of stay was also reflected in a somewhat larger share of respondents who had been detained in immigration detention before.

The asylum migrants (N = 39) originated from 25 different countries, mostly located in North or West Africa and Asia (including the Middle East), and, less so, Eastern Europe. A substantial number were from countries that were experiencing military conflicts at the time, or that had oppressive regimes, including Afghanistan, Iran, Iraq, Sudan, Ethiopia, Russia, and China. They had been in the country between 0 and 27 years, with an average of six years. Some seemed to have weak asylum claims; others had possibly encountered difficulties in proving a well-founded fear of persecution or fled unsafe conditions that are not recognized as grounds for asylum. States may also end protection when conditions in countries of origin have improved only to some extent (see Black and Gent 2006). Seven asylum migrants were former residence permit holders. Contrary to the family migrants, their permits had not been ended because of crimes, but because they were no longer deemed in need of protection.

The labor migrants (N = 29) originated from 13 different countries, mostly located in Africa and Asia, including Morocco, Nigeria, Egypt, China, India, and Turkey. Three were from South America, and one from Eastern Europe. In this group, the duration of stay in the Netherlands varied from some days up to 32 years, with an average of nine years. European countries have a rather strict admission policy for labor migrants, especially when it concerns “unskilled” labor. Residence permits are only granted when the employer can prove that the work cannot be done by a native worker or (other) EU citizen. Indeed, only two respondents in this group had ever obtained a residence permit.

Migration Preferences

Figure 1 provides an overview of respondents’ migration preferences at the time of the interview and—retrospectively reported—just before the detention period during which the respondents were interviewed. The figure pertains to all respondents (N = 81), regardless of whether they had been detained in immigration detention before. Figure 2 specifically pertains to the 33 respondents who had been detained in immigration detention repeatedly. It should be stressed that these figures are indicative, as they are based on our interpretation of the respondents’ statements in a semistructured interview.

Figure 1. Migration Preferences Before and During Detention (All Respondents).

Figure 2. Migration Preferences Before and During Detention (Repeatedly Detained Respondents).

Just before the detention, 10 of 81 respondents preferred to go to their country of citizenship; against 17 respondents (not necessarily the same individuals) at the time of the interview. The number of respondents expressing a preference for onward migration similarly increased somewhat from 12 to 16. Eleven respondents were still undecided, or were unable to make their preferences clear. All in all, the number of respondents with a preference for continued stay decreased from 53 to 37 (19 percent of 81 respondents) during the detention. Among the 33 repeatedly detained respondents, migration preferences just before the detention had sometimes been produced by previous detention periods (see Figure 2). When these previous detentions are also taken into account, the number of respondents preferring return migration had increased by 16 ([17–10] + [9–0]=16) during the detention, while the number of respondents preferring onward migration had increased by seven ([16–12] + [5–2]=7). These results suggest that immigration detention pressures a notable minority into a preference to leave the country.

All changes in migration preferences were reported to be detention-related, but repeatedly detained respondents in particular, also attributed eventual changes to factors other than detention, or indicated that they were a more indirect result of detention. For instance, some of them had changed their mind because of difficulties they had encountered after previous detentions, for example because they had lost their jobs because of the detention.

Labor migrants seemed to be more likely than family and asylum migrants to have changed their mind. Those classified as labor migrants represented 35 percent of the sample, but 63 percent of the additional respondents who eventually preferred “return” (10 of 16), and 57 percent of the additional respondents who preferred onward migration at the time of the interview (4 of 7). As expected, respondents in the higher legitimacy group were relatively likely to report a preference to “return”: while only 38 of the 81 respondents (44 percent) attributed at least some measure of legitimacy to their detention, this percentage was 82 percent (14 of 17) among those who, at the time of the interview, preferred to go to their country of citizenship. This was not the case for those expressing a preference for onward migration; only 25 percent (4 of 16) of them were in the higher legitimacy group.

Part of the overrepresentation of labor migrants among those reporting a preference to go to their country of citizenship seemed to be related to an association between perceived legitimacy and type of migration. Nineteen of 29 labor migrants (66 percent) were in the higher legitimacy group, against 13 of 39 asylum migrants (33 percent) and six of 13 family migrants (46 percent). These differences echo culturally dominant beliefs and established regulations with regards to immigration rights—while labor migration is generally not considered a right, most countries do recognize a conditional right to asylum and family migration (see Walzer 1983).

The gender composition of the research group reflected the underrepresentation of women among immigration detainees. Accordingly, the number of female respondents (N = 8) was too small to make assertions about eventual gender differences. When comparing men and women, we did observe more or less similar patterns, suggesting that immigration detention produces limited specific deterrent effects among men and women alike, which are concentrated among labor migrants and those attributing some measure of legitimacy to the detention, irrespective of gender. Three female respondents—all part of the labor migrant and higher legitimacy group—developed a return preference during the detention, five persisted in a preference to stay.

After the project for IOM had been completed, we asked the Dutch government to provide us with information about the administrative outcome of the respondents’ detention. We learned that more than two-thirds of them (59 of 81, 69 percent) had eventually been released in the Netherlands, including 10 of the 17 respondents (59 percent) who had expressed a preference to go to their country of citizenship.

In the next section, we describe how family migrants, asylum migrants, and labor migrants experienced the detention, and why immigration detention did or did not alter their migration preferences. The interview fragments also show that there was a notable degree of “blurring” between the three types of migrants, and that migration motives sometimes changed during an individual's migration project.

The Voice of the Detainee

Family Migrants

Most family migrants had a strong socio-cultural attachment to the territory of the detaining state due to their family ties and the length of their stay in the Netherlands. As a consequence, they generally tried to resist deportation and were reluctant to migrate to a different European country. They certainly did not consider detention a trivial matter, however. A Surinamese man, who had been reunified with his elderly mother as an adult, and had lived in the country without state permission for 13 years, was a case in point. Having been detained twice before—the last time had been for nine months—he found the detention quite stressful, yet he insisted on staying in the Netherlands.

“This is no life. I mean for illegal Dutchmen in general. It's not good for honest people. If you keep somebody here for nine months, you've stolen a year of his life … Last time I slept with four people in a cell [for nine months]. … My health is still fine but I am a bit stressed about what's going to happen. I talk in my sleep and I've been to a psychologist.”

Like most respondents, the man primarily complained about the loss of freedom, the perceived unfairness of the detention, the uncertainly about the detention outcome, and his lack of perspective more generally. Additionally, he seemed to suffer from a loss of status that comes with cell-sharing. Several other respondents similarly experienced status deprivation, but focused more on material aspects, or at least articulated their concerns in these terms. A Moroccan man, for instance, lamented the quality of the food:

“It's distasteful. We only get bread and milk, one liter of milk for four days, half a loaf of bread for two days, 100 grams of sugar for four days. I try to make my own food, but the [detention] shop asks one euro per tomato! Outside that buys you a kilo!”

Despite experiencing deprivation, family migrants were generally reluctant to comply with arrangements for departure. Their local family ties, and the associated ambition to settle in the Netherlands permanently, was not easily offset by the “pains of immigration detention,” not even repeated and prolonged detention. Additionally, family migrants were quite critical about both the outcome and process legitimacy of the detention. A different Surinamese man, who had spent 3.5 months in his second detention, was another case in point. Beginning in the late 1960s, his whole family had gradually migrated to the Netherlands: first his eldest sister, then all of his remaining siblings and mother. When his marriage ended in 2001, he was the last family member to emigrate, after which he remarried and became a father:

“My wife and child are here. I want my child to make that decision when she's old enough, to decide herself where to live and be. My home is here because I've been living here for so long. The Netherlands is not the land of milk and honey, and it isn't easy [being unauthorized], but I'm in my fifties and I don't want to give up my life here. (…) You may protect your borders and you don't have to allow everybody, but the sanction needs to make sense and they need to listen to each story and pick the rotten apples. I have a permanent address with my wife and child but I cannot officially register there [being unauthorized]. And now they say that I am at risk of absconding. I will not let myself be deported!”

Many family migrants also saw no viable alternatives but to stay in the Netherlands since they did not have any relatives, friends, or economic opportunities in their country of origin or third countries. The aforementioned Moroccan, for example, had lost one of his hands while working illegally for a butcher, without being entitled to disability insurance as an unauthorized immigrant. With the exception of his elderly mother, all of his family members lived in Europe, and even though he no longer had close contact with his family—which he attributed to his accident—he argued that he could not return to Morocco as a beggar:

“Here, I at least get some support from friends and a Moroccan foundation.”

As expected under deterrence theory, eventual changes in migration preferences sometimes also seemed to be related to increases in the perceived certainty of (repeated) immigration detention. A young man wanted to return to Surinam, since his sisters reasoned that he was likely to be re-caught—and perhaps become a burden to them—now that he had been apprehended for unauthorized residence:

“Well, look. When the detention wasn't there, it went well. But not anymore. Once they have you they will catch you again just like that. And I've made some money and stuff, and now I want to return. I can work with my brother-in-law [until I have arranged my papers for the Netherlands]. So my sisters told me to go back to Surinam.”

In some instances, detention actually seemed to reinforce preferences to stay in the territory of the detaining state. Group-wise detention sometimes seemed to enable migrants to exchange techniques to resist deportation:

“We talk a lot about how they can't really force us to return. And we advise each other regarding how it will be held against you in court if you do not sufficiently [give the impression that you] cooperate.”

Additionally, information on opportunities to deal with the challenges of being unauthorized is sometimes also shared:

“I can't work [being an unauthorized immigrant in the Netherlands], I can only deal drugs. A guy offered me to make some money [when I am outside] by receiving packages of drugs every other week. But I‘m not into that kind of thing.”¹⁰

Asylum Migrants

Most asylum migrants similarly considered the detention quite severe, and were aware of the possibility of repeated detention even though this group was to a lesser extent repeatedly detained in comparison with the other two groups. Even more so than family migrants, most of them nonetheless tried to resist deportation. In their case, this was not primarily because of a strong socio-cultural attachment to the Netherlands, but rather because they “feared return.” An Iranian woman, who claimed to have fled an abusive husband, explained why being deprived of liberty had not changed her mind:

“I am enclosed by four walls and it feels like I'm in prison. Outside you're free, you can call, go shopping. I've been here for four months…. The people here do their best but it's very difficult…… Perhaps when you look at it from a higher level, that you can understand the system. But I don't see it for myself. I came for safety, received these rejections, and now I am locked up. It's not fair when somebody comes for safety…. I have children, and I want nothing more than to be with them. But I cannot return home due to the problems with my husband… I face a death threat when I return…”

Additionally, asylum migrants were even more critical about the legitimacy of the detention than family migrants. This is, for example, what a rejected Afghan asylum seeker said when an interviewer inquired how he perceived the legitimacy of the detention:

“What do you think yourself? I feel just like you if you would sit here. Every day hundreds of people die in Afghanistan. There is injustice. Troops go there, but it only makes things worse. That's why I was forced to come here, and then I get locked up. If someone here [a European] dies because of the war in Afghanistan they organize a march and mourn for ten days; there, hundreds of people die each day and nothing is done. There are no humane people here.”

Some respondents indicated, perhaps paradoxically, that a residence permit would make it easier to leave, as it would enable them to re-migrate should conditions be or become as unsafe as they feared. A Liberian respondent complained about the boredom of detention, the difficulty of communicating with the outside world, and his stress and associated medical problems including migraines, stomach aches, and depression. Yet this is what he said after almost a year of detention:

“I can't go back. This is only possible if I have a residence permit so I can come back if there are problems there. It's not safe right now because of the political situation and crime. I know that from the internet, friends, and CNN.”

Asylum migrants tended to emphasize asylum-related obstacles to returning, and to some extent this may have been a way for them to justify and articulate their right to stay in Europe. Apart from safety issues, however, some of them also faced socioeconomic obstacles. Most of them had migrated to settle in Europe permanently, or for an indefinite period of time, and seemed to have left little to “return” to. One of our interviewers reported the following about a West African man, aged 30, who had been in Europe for 12 years:

“He said his parents died in atrocious conditions, that he was imprisoned, and saved by accident. … It's his second detention. He says that all days are the same, that there's nothing to do but to ‘sit, sit, sit’, and that he's only ‘half a man'. His health has worsened, he eats badly. He doesn't consider return, because he feels that he's entitled to a residence permit, and because his parents have died and his house was destroyed. He also says he is not a criminal and doesn't understand why he can't stay at a regular address.”

All detainees found immigration detention a painful experience, but a few valued at least some aspects of it. For example, one respondent, who had initially come to Europe to seek asylum, had been in the country for 27 years and was detained for the sixth time.

“I cannot return. I don't want to. My life's here. I'm detained, but it doesn't matter to me. Look, you're not free, but it's not bad.”

This relatively positive evaluation of detention was probably related to the respondent's frame of reference. He had become addicted to drugs, had been living on the streets, and was regularly ill. Other strongly marginalized detainees were similarly negative about the loss of liberty and autonomy, but also valued certain aspects of the detention to some extent, such as the medical care facilities (also see Leerkes 2016).

Compared to family migrants, asylum migrants were more likely to consider onward migration to another EU country or, in some cases, to Canada or the United States. Thus, despite the aforementioned Dublin Regulation, substituting a different European country for the Netherlands still turned out to be a way to continue migration projects, albeit irregularly. In fact, some respondents had actually lived elsewhere in Europe after their asylum claim had been rejected, or after previous deportation procedures had failed. Others still intended to go elsewhere in Europe should they be released. An Iraqi, for example, had gone to Belgium in hopes of obtaining a residence permit there. His statements show how a perceived severity of the detention without at least some measure of perceived legitimacy of the detention may pressure migrants into (illegal) onward migration rather than “return.”

“It's not good. I have to go to Belgium again. There it's better. Here they keep you seven, eight, ten, 18 months… In Belgium it's two, three, four, and you're free. And I can get a residence permit there [this is actually unlikely]. Interviewer: What specifically is so bad about this detention? Respondent: Well, you're imprisoned. You don't know when you'll get out, you don't sleep or eat well, with others in a cell, showering together, just dirty. I don't get it! I don't come for money or something, I come for safety. I never had any problems with the police, and in Europe they lock me up.”

Those among our respondents who had actually been “displaced” to a different European country had all been apprehended there, and had been transferred to the Dutch authorities to be re-detained. After having experienced that the Dublin regulation ties asylum seekers to the (first) detaining state, some asylum migrants no longer preferred onward migration.

A few asylum migrants wanted to go to their country of citizenship. This mostly occurred after several detentions in combination with other difficulties of being unauthorized. Such respondents also seemed to have weaker asylum claims—at least they expressed milder fears concerning return. Some of them were in the lower legitimacy group, suggesting that a perceived legitimacy of immigration detention is not, as Sherman's (1993) defiance theory would have it, a necessary requirement for a return preference to develop; some detainees were simply “tired” after repeated detention. A Turkish Kurd had lived in a foster home in Greece until the age of 14, when he migrated irregularly to Italy, where the police allegedly denied him the opportunity to apply for asylum. As there was little work, he went to France, Belgium, and, eventually, the Netherlands, where he applied for asylum claiming to be from Iraq because he had been told that would increase his chances of obtaining asylum.

“I've seen nothing outside [since arriving in Europe]. Just prison. The first time they caught me at the border and kept me for six months. They released me, I went to Norway, got sent back to the Netherlands, was locked up for 11 months. Then I went to Belgium, got sent back, was detained a year. And now it's 6 months already! I'm so angry. Every day somebody comes to talk. First a lawyer, then a woman, then a man, now you. The people [guards] are nice, but I'm kept for nothing, just for papers. They have no right to detain me. I came for asylum and I have done nothing wrong. But I'm really tired. I've never seen Iraq but I really want to go to Turkey. I can work there. I watch Turkish TV every day [Turkey was indeed experiencing an economic boom]. But I'm not registered there, and they keep trying to send me to Iraq.”

Labor Migrants

Deterrent effects at the attitudinal level were stronger among labor migrants than among family and asylum migrants, but the former did not perceive immigration detention as more severe, nor did they perceive repeated detention as more probable. Rather, the differential impact of the detention experience on respondents’ migration preferences seemed to be related to both the differential migration projects among the three groups and to different perceptions of the legitimacy of the detention. While most family and asylum migrants had a strong orientation towards a permanent stay, most labor migrants either intended to go to their country of citizenship after having earned a certain amount of money, or lacked a clear plan about how long they would be away but still intended to “return” eventually. Still others intended to establish themselves in Europe permanently, but were open to work in a different European country. In fact, a significant number of labor migrants had lived elsewhere in Europe. As was mentioned in the quantitative overview, labor migrants were also notably less negative about the legitimacy of immigration detention.

A Nigerian man, who had migrated to work and “because he had always dreamt about living in Amsterdam,” represented a case where latent return intentions became more manifest under the influence of detention. He came with the intention of being away for seven years, but had been apprehended after three. In the meantime, he had been doing relatively well sharing a room with a friend and saving money by way of informal jobs via an Amsterdam church. He would not mind continuing that life in Amsterdam were he to be released, but after nine months of detention he was prepared to go to Nigeria; he feared “going crazy” in detention where he was regularly having “weird dreams.” Like others who developed a return preference, he accounted for these changes in terms of deterrence, that is, in terms of perceived severity and risk. However, he also indicated that the government did have the right to detain him, although no more than a month, suggesting that the influence of legitimacy may largely operate on a more subconscious level.

In some cases, pre-existing plans to eventually return to one's country of citizenship were brought forward because of a more encompassing and gradual process of deterrence in which repeated detention was one element. A Moroccan man from a middle-class family had immigrated in the 1990s. After being homeless for some time, he had found informal work in horticulture, as well as housing that he shared with other unauthorized immigrants. It had become increasingly difficult for him to work in the greenhouses because of checks by the labor inspectorate. The Dutch policy of discouraging unauthorized residence—which was mainly developed in the 1990s and early 2000s—then seemed to have initiated a process of marginalization and criminalization, which eventually contributed to a still unfulfilled desire to go to Morocco:

“When I could no longer work in greenhouses, I started doing irregular jobs for acquaintances, like painting, carpentry work. But I didn't earn enough. I also became addicted to cocaine, and committed petty crimes. These were added up in court, and I got 33 months of imprisonment, and was declared an undesirable alien [continued residence as an undesired alien is punishable as a crime against the state]. When I was detained for so long and so frequently, I wanted to return. Since then I've been presented to the embassy six times. They know my nationality and identity but cannot do anything [he indicated that the Moroccan authorities do not want to take him back because of his record of offenses]. One time I tried to go back myself, but got arrested in Belgium [on my way to Morocco].”

The man was classified into the higher legitimacy group, yet was quite critical about repeated detention, a practice that clearly shows the informal “incapacitation” function of immigration detention (see Leerkes and Broeders 2010):

“I think that people [like me] should be detained until they can be expelled. And if that doesn't work out, they should find a different solution. People should be able to go themselves or stay here. I can't be deported. They know it, too, and yet they detain me again.”

For geographical reasons, many undocumented labor migrants initially migrate to Mediterranean EU countries, but a notable number eventually travel to Western Europe. For a number of such respondents, the perceived severity of immigration detention contributed to a desire to go back to their initial destination countries. A Palestinian stated:

“As a teenager I went to Libya. My parents had died, so nothing kept me in Gaza. I worked in Libya for ten years before going to Italy, both for work and the experience. After three years I could no longer find work when three friends went here. So I decided to also take the train up North. I got arrested after spending 1.5 weeks here. The detention conditions are okay, but it's terrible to be locked up. When I'm free again, I want to go back to Italy. The Netherlands is a beautiful country but the law is horrible….”

A few respondents developed a preference to leave, but hoped that they would be away only temporarily. A Moroccan man, who had worked in various informal jobs and had eventually entered into a relationship with a Dutch girl of Moroccan origin, thereby also illustrating how labor migrants may resemble family migrants over time. He found it difficult to be deprived of his liberty and income, and therefore preferred to go to Morocco to arrange his immigration documents there. Should that be impossible, he hoped to go from Morocco to Spain, and reunite with his girlfriend there.¹¹ He intended to go to Spain directly should the deportation procedure fail. Several of his friends had gone there after having being detained in the Netherlands, and claimed that Spain was preferable. He explained:

“I had work and we [my girlfriend and I] went to the beach each week. I was happy. But here we are in our cells from 5 pm to 8:15 am. That's almost 15 hours! It drives me crazy: I've done nothing, why am I here? And I don't get enough food and need to do my own shopping on [an allowance of] 10 euro per week, but that's not enough.”

Like most family and asylum migrants, several labor migrants nonetheless persisted in a preference to stay, even when confronted with prolonged and/or repeated detention. They usually mentioned that they were afraid to leave “empty handed”: they could not return without bringing a sufficient amount of savings from their earnings. In these cases, detention sometimes seemed to extend rather than shorten migration projects. For example, a Chinese man, who was in his third immigration detention period in eight years, had initially planned to stay in Europe for a few years only. Although his detention experiences had made him believe that “the Dutch government is really determined to arrest me”—thereby appearing to confirm theoretical assumptions about the effect of being caught on the perceived probability of being re-caught—this was insufficient for him to consider returning.

“Every time I'd saved some money, they put me in detention. Every time my life's on track, I get arrested. I've lost everything I had. My planning was not to stay, but now I have to.”

A Cameroonian man, who had stopped contacting his family when he found himself unable to send remittances, similarly explained:

“You're keeping me in jail. How can I consider leaving this country without money?”

Conclusion and Discussion

This study examines whether and how immigration detention—under conditions found in the Netherlands in 2011—affects the willingness on the part of detainees to leave the territory of the detaining state, and whether eventual influences can be regarded the product of deterrence. The analysis sheds light on two types of “specific deterrence”; first, the possibility that detainees are pressured into cooperating with the deportation procedure and, second, the possibility that detainees are pressured into leaving the country on their own, should they be released because of a failed deportation procedure. The present contribution focuses on stated preferences; it was not observed whether respondents were cooperating with the deportation procedure or whether they actually went to a neighboring country after a failed deportation procedure (various repeatedly detained individuals did mention that they had done so after earlier detention periods). It should be remarked that an eventual preference on the part of migrants to leave the country of the detaining state, does not automatically translate into “permanent departure,” as the administrative detention outcome also depends on supra-individual factors, including the willingness by receiving states to provide laissez passers (Ellermann 2008; Leerkes and Kox 2016). Furthermore, migrants who are pressured into a preference to “return” may start a new (illegal) migration project after having been repatriated (Brotherton and Barrios 2011; Cardoso et al. 2016), while those who manage go to a third country may eventually come back to the country where they were detained, such as when former detainees are returned to the “responsible country” under the Dublin Regulation after having been apprehended elsewhere in Europe.

A first main conclusion is that immigration detention seems to be producing limited selective deterrent effects at the attitudinal level. At the time of the interview, 17 out of 81 respondents expressed a preference to go to their country of citizenship—in some cases in hopes of starting a new (legal or illegal) migration project from there—and in 16 cases that preference was reported to have been coerced, at least in part, by detention, including repeated detention. Additionally, the number of respondents preferring onward migration increased by seven when previous detention periods are taken into consideration. These results are obtained for a detention regime that is relatively punitive comparatively, especially in the European context given the relatively long maximum and actual detention duration, and the common practice of repeated detention.

As expected, most respondents found the detention quite painful, reflecting both Sykes's classic “pains of imprisonment,” Crewe's (2011) “pains of uncertainty and indeterminacy,” and what could perhaps be called the “pain of perceived illegitimacy,” that is, the belief that one does not deserve immigration detention—either not at all, or not under conditions that characterized the Dutch detention regime in 2011. Much in line with deterrence theory, respondents appeared to be making a kind of rational decision by weighing the “costs” of immigration detention against the “benefits” of a continued stay in the Netherlands or Europe more generally. In most cases, however, the outcome of that comparison turned out to be a persistent preference for continued stay; respondents’ interests in continuing their stay in the Netherlands typically outweighed their interests in ending the immigration detention experience—thus indicating strong “threshold effects.” Deterrent effects were also found to be limited because the detention experience also triggers mechanisms that actually reinforce a preference to stay. Some respondents, who were afraid of the stigma of returning “empty handed,” reasoned that they actually needed to prolong their migration project in order to make up for income that was lost during the detention. Others received advice from other detainees on how to resist deportation and/or on how to survive without a resident permit more generally (see Campesi 2015). A low perceived legitimacy of the detention seemed to be reinforcing migrants’ preferences to stay as well; while the “pain of perceived illegitimacy” sometimes added to a desire to leave the territory of the detaining state, so as to end one's exposure to it, it also triggered resistance.

The nature of the interests offsetting the “costs” of immigration detention varied by migration motive. Respondents classified as family migrants often persisted in a preference to stay due to their family ties and socio-cultural attachment to Dutch society. Asylum migrants emphasized safety concerns and an interest to stay away from their country of citizenship. They were relatively open to going to a different European country but found themselves being tied to the country of the first asylum application because of the Dublin regulation. Respondents classified as labor migrants were more likely to emphasize financial obstacles to return.

A second conclusion is that the specific deterrent effects were concentrated among labor migrants. Some of them brought pre-existing return migration plans forward. Others hoped to be able to evade repeated detention by seeking employment elsewhere in Europe, should the deportation procedure fail. On average, migrants classified as labor migrants seemed to have a weaker commitment to continued stay than family and asylum migrants, but they also seemed to consider their detention somewhat more legitimate, which brings us to our third and final main conclusion: while all respondents who developed a preference to leave attributed these attitudinal changes to the perceived severity of the detention and/or the risk of repeated detention, those developing a preference to “return” to their country of citizenship were notably less negative about the outcome and/or process legitimacy of the detention. That observation, which is consistent with defiance theory (Sherman 1993), suggests that deterrent effects mostly occur if some degree of perceived legitimacy is also present, and that perhaps some degree of compliance would also have occurred had the government used a “lighter” measure than detention—or that such measures would actually have been more effective, as they can be expected to perform better on perceived process legitimacy. It also suggests that the distinction between forced return and voluntary return is blurred in reality; there may be a measure of (indirect) force in voluntary return, and a trace of voluntariness in forced return. To the extent that perceived severity and risk in the absence of some measure of perceived legitimacy were capable of changing migration preferences, they mostly pressured detainees into a preference for onward illegal migration to a different European country—thus turning them into a kind of “fugitives” who contribute to a spatial displacement of illegal residence.

In order to get a more complete picture of the operation and effects of immigration detention, future research should try to pay attention to both the detainee and the institutional level (Leerkes and Kox 2016). The decisions that embassies and consulates make on readmission seem to “dilute” the limited specific deterrent effects at the attitudinal level. As was reported, over two-thirds of the respondents were eventually released because of a failed deportation procedure, including 10 of the 17 respondents who preferred going to their country of citizenship. While more research is urgently needed on how receiving states handle requests by governments in the Global North to provide laissez passers to undocumented migrants, a perceived legitimacy deficit of deportation and immigration detention on their part—perhaps in addition to critical assessments regarding the fairness of the immigration policies of the Global North more generally—may well be among the reasons for noncooperation. Another implication of these figures is that most respondents probably stayed in Europe in a worse condition than in which they entered detention, also given the loss of employment that regularly came with it and the health issues that were reported. One would hope that such perverse outcomes create an interest on the part of detaining states to limit the punitive uses of immigration detention, and to find other ways of dealing with the presence of noncitizens who are formally deportable, but are difficult to deport in practice.

This contribution can be read as a case study on the social operation of migration law, with a focus on deportation law. It has broader scientific implications for the study of international migration and legal compliance.

In a world in which international migration is increasingly being regulated by states, it is crucial to obtain better insight in how immigration regimes, including deportation regimes, codetermine migration and mobility. In migration studies, the structuring influence of “immigration law in action is undertheorized and underresearched (see Massey et al. 2005; Massey 2015). So far, most researchers interested in “policy effects” have implicitly or explicitly limited themselves to economic (“rational choice”) compliance models, and have mostly focused on migration decisions taken in countries of origin, such as by asking whether apprehension risks at the U.S. border influence migration decisions (see for example Cornelius and Salehyan 2007; Gathmann 2008; Massey and Riosmena 2010). Ryo (2013) and others (Braithwaite 2010; Van Alphen et al. 2013) have argued that in order to advance our knowledge of contemporary migration, it is essential to incorporate models of legal compliance that also pay attention to (non)compliance on normative grounds. Ryo used a legal compliance framework in order to understand the migration decisions of (potential) migrants in Mexico; we use a similar lens to understand (changes in) migration preferences among deportable migrants. Like Ryo, we find that normative compliance models are important to consider, but we also find that economic models certainly cannot be disregarded. Future research could pay more attention to the complex interrelationship between normative and economic dimensions of compliance, including the possibility that deterrent measures need to be perceived as (somewhat) legitimate in order to be effective, and the tendency of deterrent measures to eventually undercut process legitimacy. While this contribution focuses on migrants, it is also crucial to analyze how professionals working in relevant institutions in the field of immigration law—such as embassy personnel, immigration lawyers, (migration) police, NGOs—structure migration patterns, using a similar lens.

A vast amount of scholarly work has enhanced our knowledge of legal (non)compliance, but little is known on how people give meaning and respond to the laws of foreign governments (see Ryo 2013)—especially when such laws mostly target them, as is the case with immigration law. Existing models in socio-legal studies are useful starting points, but some claims of the compliance literature do not seem to hold, such as the central claim in criminal justice research that perceived process legitimacy is more influential than perceived outcome legitimacy (see Tyler 2003). Compared to other fields of law, there is less agreement that the content of immigration law is fair and serves the “common good.” Using two powerful sociological concepts by James Coleman (1990), one could say that the laws in other fields usually resemble codified “conjoint norms” where those targeted by the norm also benefit from the norm being observed (say: “improved public safety”). Immigration laws, by contrast, are more akin to “disjoint norms,” where those targeted by a norm and those benefitting from it constitute different groups. Immigration law benefits a substantial number of residence permit holders and those obtaining citizenship through naturalization, but also is a repressive, protectionist instrument that is designed to reduce migration, and that has the effect of keeping a significant part of the migrants at the bottom of the social hierarchy in a rather vulnerable position—without extending voting rights to them (also see De Genova 2002; Menjívar and Abrego 2012). As a consequence, the perceived outcome legitimacy of immigration law is much more contested and variant than is found in other fields of law. The central importance of process legitimacy in criminal justice research may well be due to perceived outcome legitimacy being close to a constant there.

More generally, we need to better understand legitimacy issues in immigration law. These need not only be addressed from a formal-legal or normative-philosophical perspective; eventually, we also need to understand how migrants and other relevant actors perceive the outcome and process legitimacy of immigration law in countries of the Global North, how their views are affected by the social structure and cultural factors, and how these may be changing under the influence of globalization and a rise of a global citizenship in countries in the Global South.

Acknowledgment

We are grateful to Daniela DeBono, Joanne van der Leun, Dean Wilson, three anonymous reviewers and the editors for their useful comments on earlier versions of this article. We thank IOM Netherlands for collaborating with us on this project.