Deporting social capital: Implications for immigrant communities in the United States

Abstract

The United States currently removes approximately 400,000 individual migrants each year, which represents close to an eightfold increase since the mid-1990s. While scholars have studied the consequences of such policies for children and families, this article posits broader effects on communities through the reduction of immigrant social and human capital. Using findings from three studies of immigrant communities and Salvadoran deportees, we show that current deportation practices remove individuals with a wide range of socio-economic resources and ties to local communities. When they are removed from economic, family, social, and civic networks, the individuals and communities left behind are impoverished in important ways. This is particularly consequential for low-resource immigrant communities, which under the best of circumstances encounter obstacles to economic advancement, social integration, and political engagement. In addition, we consider the potential harm to the institutions in which immigrants participate, such as businesses and churches, which has implications for the economy and society more generally.

1. Introduction

As strangers in a foreign land, new immigrants have long relied on social capital to navigate an unfamiliar society. Historical and contemporary accounts of immigrant America abound with narratives of immigrant groups relying on the knowledge, experience, and economic resources of their established co-ethnics to locate housing, access job referrals, and provide start-up capital for ethnic businesses (Thomas and Znaniecki 1918; Massey 1987; Light and Bonacich 1988; Portes and Sensenbrenner 1993). This social capital is critical in guiding newcomers to vital institutions that will enable them to integrate in the host country.

In this article, we argue that social capital is declining in some immigrant communities in the United States due to the unprecedented mass deportations of Mexicans and Central Americans, many of whom have cultivated economic, educational, and civic ties within local US institutions. This removal of valuable social capital from established immigrant communities is thus hampering the integration of those left behind. Drawing on findings from three studies designed to assess the effects of these rising removals on immigrant communities, we suggest that deportation not only reduces bonded social capital built on strong co-ethnic ties but also—and perhaps more importantly—weakens the bridging social capital based on looser ties with non-ethnics and host society institutions that are vital to integration into US society. While previous studies have examined the consequences of mass deportation for immigrants and their families, the implications for social capital are largely unexplored.

Approximately 400,000 foreign-born persons in the United States now face formal deportation proceedings each year, which represents close to an eightfold increase since the mid-1990s. The rapid increase in deportations in the United States since the mid-1990s has resulted in scholarship focusing on the theoretical and legal foundations of the deportation regime (Flores 2003; Gibney and Hansen 2005; Buff 2008; Kanstroom 2007; Ellerman 2009; De Genova and Peutz 2010); the implementation and policing of deportation operations at national, regional, and local levels (Phillips, Hagan and Rodriguez 2006; Lewis and Ramakrishnan 2007; Varsanyi 2008; Southern Poverty Law Center 2009; Weissman, Headen and Lewis Parker 2009; Armenta 2012; Capps et al. 2011); and the human and psychological costs of this enforcement practice for deportees and their families (Capps et al. 2007; Hagan, Eschabch and Rodriguez 2008; Mendoza and Olivas 2009; Dreby 2012).

In this article, we add to the existing literature to address the social capital consequences of mass deportations for Mexican and Central American immigrant communities. Our article is organized into four sections. We first provide a brief historical overview of mass deportation operations in the United States and the extant research that has been conducted to investigate the origins and effects of the phenomenon. We then survey the literature on the role of social capital in immigration and community development. We next describe the research sites, methodologies, and findings of the three studies on which our conceptual analysis is based. In the subsequent section, we draw on the collective findings of these studies to demonstrate how migration policies of control, such as deportation operations, can potentially weaken bonding and bridging social capital in immigrant communities.

2. Mass deportation trends in the United States

The United States has a long history of mass deportation of foreign-born persons. And most of these operations of migration control have targeted poor working-class migrant groups, especially Mexicans and Central Americans. While we can trace the toughening of the US deportation regime back to the early 1900s, three mass deportations have taken place since World War I. The first transpired during the Great Depression when Mexicans became the targets of America’s economic troubles, resulting in the repatriation of over 400,000 Mexican migrants (Hoffman 1974). The second mass deportation took place in 1954 when the US government removed over one million Mexican migrants through ‘Operation Wetback’ (Lytle Hernandez 2006; US Department of Homeland Security 2010: Table 36). Since the mid-1990s, changes in immigration policy and deportation law have ushered in a third wave of mass deportation. What distinguishes this third wave from earlier ones is the presence of a vast state bureaucracy and technology that enables the melding of two different methods of deportation: extended border control and interior social control (Kanstroom 2007).¹

The present-day period of mass deportation is largely a product of a new series of laws in combination with this growing bureaucracy and technology (Kanstroom 2007). One law in particular, the Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA) of 1996, expanded the criteria for detaining and deporting migrants, restricted the ability of migrants to appeal deportation orders, and limited the power of immigration judges to grant relief to migrants convicted of relatively minor crimes (Aleinkoff, Martin and Motomura 2001). The Antiterrorism and Effective Death Penalty Act (AEDPA) of 1996 further increased the enforcement authority of the federal government by almost eliminating judicial review for most categories of immigrants subject to deportation. By weakening judicial review, AEDPA removed the legal constraints that protected many non-citizens from deportation by eliminating relief for non-citizens with family ties in the United States (Kanstroom 2007). In 2001, Congress passed the USA PATRIOT Act, which bolstered administrative authority to apprehend, detain, and deport immigrants who are perceived to be threats to national security (Salter 2004).

The creation of the bureau of Immigration and Customs Enforcement (ICE) in the Department of Homeland Security in 2003 has also contributed to contemporary mass deportations. Working closely with state and local agencies, ICE initiated several interior enforcement programs devoted to apprehending, detaining, and deporting ‘criminal and fugitive’ non-citizens long after their arrival (Kanstroom 2007). The Criminal Alien Program (CAP) enables ICE officials to enter state prisons to conduct immigrant screening and detain and remove criminals. The Secure Communities Program (SCP) requires police to enter prints of arrestees into a joint FBI and ICE database. The 287(g) program (an IIRIRA provision that has been de-emphasized as SCP became more expansive and operative) is used by ICE to train state and local police to identify, process and detain suspect immigrants whom they encounter during their regular law enforcement activities. Both programs have been highly controversial. Studies have documented that they facilitate racial profiling of Latinos, as well as weaken security in immigrant-heavy neighborhoods (Weissman et al. 2009; Capps et al. 2011). When immigrants suspect that police are partnering with ICE, they are less likely to seek assistance from police or to report crimes.

The recoding of civil violations into criminal acts and the expansion of enforcement operations from the border to a massive state technology of federal and local enforcement agencies in the interior United States has resulted in the current period of mass deportations. Just how large is the recent increase in deportations? Until the mid-1990s, annual deportations averaged fewer than 50,000. After the enactment of IIRIRA, deportations increased sharply. From 1996 to 2005, yearly removals averaged around 180,000; from 2006 to 2010, they increased further, reaching 414,481 in 2014 (US Department of Homeland Security 2011: Table 39; US Immigration and Customs Enforcement 2014).

According to the Migration Policy Institute,² Department of Homeland Security (DHS) data show year-to-year variations in the number of deportations, but the overall trend is clear. For instance, we see a very small upward trend from 1986 to 1996, which likely reflects growth in the total number of unauthorized immigrants in the nation, not specific changes in public policy. The numbers sharply increase in 1997 and 1998, however, which are followed by a brief plateau for several years. The strong upward trajectory resumes by 2003 and continues for the remainder of the decade.

While several factors likely influence the year-to-year variation in deportation data, the growing numbers since the passage and implementation of IIRIRA indicate that this law is the primary explanation. In addition, some of the yearly changes are not very meaningful. For instance, fluctuations can reflect administrative updates of data and not changes in actual deportation numbers. Fluctuations in the annual deportation data can also occur because a specialized deportation program (e.g. the Criminal Alien Program) is transferred to an administrative unit that does not normally report deportation counts (US Department of Homeland Security 2008). Changes in deportation hearing schedules, the shifting of detainees across centers, and differences in how judges conduct trials involving unauthorized immigrants also cause fluctuations in annual deportation counts.³

Like earlier periods of deportation, Latin Americans are the focus. Seventy-three percent of the immigrants formally removed from the United States in FY 2010 were from Mexico and another 20 percent from Central America (US Department of Homeland Security 2011, Table 41). Moreover, less than half (44 percent) of those deported in 2009 were removed for criminal violations (US Department of Homeland Security 2011: Table 41), despite the ‘criminal alien’ narrative that is the primary political justification for the program.⁴

The dramatic increase in deportation levels in the United States has resulted in a growing body of scholarship focusing on the origins and politics of massive deportation and the human costs of this practice for deportees and their families. Debates on the use of deportation as a method to control migration have grown. Mathew Gibney and Randall Hansen (2005), for example, argue that deportation embodies what they term the ‘liberal democratic paradox’. On the one hand, the capacity to control borders is fundamental to the sovereignty of the state. On the other hand, immigration must also embody the rights of its residents. The tensions between support for sovereignty and individual rights bears on the course of the deportation regime. Antje Ellermann (2009), Buff (2008), and Flores (2003) identify political actors and conditions that have historically produced different patterns of migration control. Flores identifies the ways in which anti-immigrant sentiment and US deportation operations have followed periods of racialized portrayals of immigrants. Buff (2008) contends that the growing and vast state technology, which includes multiple entities charged with immigration control, has produced an operation of deportation terror. Daniel Kanstroom (2007) introduces the concept of ‘post-entry social control’ and explains increased deportations in the interior United States as a result of failed control of the border.

Other scholars have concentrated their attention on the human cost of the deportation regime for deportees and their families. Legal and social science scholars have argued that the conditions under which immigrants are arrested and detained may fall short of US and international standards. Detainees spend long periods in detention facilities and are often subject to racial slurs and excessive force (Phillips, Hagan and Rodriguez 2006; Southern Poverty Law Center 2009; Weissman, Headen and Lewis Parker 2009; Capps et al. 2011). Some governments to which migrants are repatriated regularly stigmatize and criminalize deportees, designing programs to regulate their mobility upon arrival and long afterwards (Hagan et al. 2008).

Although the targets of deportation are non-citizens, the practice results in ‘collateral damage’ to other individuals as well as to immigrant communities more generally (Buff 2008; Human Rights Watch 2009; Nguyen and Gill 2010; Capps et al. 2011). Using census data and figures released by the Pew Hispanic Center, a recent Human Rights Watch report (2009) estimates that more than one million family members have been separated from one another through deportation. Another source places the number of US-born citizens with deported parents at 200,000 (Colorlines 2012). Deportation produces emotional and psychological hardship for deportees and traumatic effects among children and spouses left behind (Capps et al. 2007; Mendoza and Olivas 2009; Dreby 2012). Not surprisingly, social ties to spouses and children remaining in the United States increase the likelihood that deportees will plan to migrate again (Hagan, Eschbach and Rodriguez 2008).

Although this scholarship informs our research, the primary purpose of our article is not to extend the theoretical and legal discussions of deportation, nor to examine its effects on immigrant families. Rather, we extend conceptual and empirical research on deportation effects to address the consequences of mass deportations for Mexican and Central American immigrant communities as their available social capital declines. We draw on the findings from three previous immigration studies to argue that the US policy of mass deportations—which deports many immigrants who would not have been deported under policies prior to the enactment of IIRIRA in 1996—is removing large numbers of immigrants whose accumulated resources are important to the political, civic, and economic development of migrant communities. While deportations remove dangerous criminals from immigrant communities, such policies may also affect the ability of immigrant communities to adapt to core institutions in the host society. In this way, many individuals are harmed by the deportation of non-citizens, either directly as relatives of deported migrants or indirectly through the weakening of bonding and bridging social capital in immigrant communities. In addition, we consider the implications for institutions that frequently include immigrant employees, members, and participants, such as businesses and churches.

3. Social capital and immigrant communities

The concept of social capital in contemporary scholarship can be traced to Pierre Bourdieu (1980, 1985) and James Coleman (1988), who argued that social relationships and membership in communities provide individuals with valuable resources and potential benefits. For Bourdieu, social capital refers to relations developed instrumentally with the intention of capturing future benefits. Moreover, the profits that accrued to individuals formed the basis of group solidarity, with the implication that networks are constructed through a process of institutionalizing group relations that subsequently become a usable, reliable source of benefits. Coleman sees social capital as also encompassing obligations and expectations, information channels, and social norms. This broader conception of social capital was adopted and adapted by Putnam (1995) and Fukuyama (1999).

Putnam defined social capital as networks, norms, and social trust that facilitate coordination and cooperation for mutual benefit (1995: 67). Networks of civil engagement create norms of generalized reciprocity, which leads to social trust, thereby allowing for the resolution of collective action dilemmas. Moreover, family is the most fundamental form of social capital (1995: 73), although some measures of social capital include neighborliness and membership in associations. For Fukuyama, social capital is a norm that promotes cooperation between individuals, with the only caveat that these must be actualized norms. Trust, networks, and civil society are the results of social capital. Social capital leads to cooperation in groups, as well as norms of honesty, reciprocity, and reliability, and it serves an economic function by reducing transaction costs.

In his discussion of social capital, Hardin (1999) found that sociologists conceive of social capital as structures within society—social relationships, specifically—that allow individuals to cooperate and achieve. This is an instrumental view of social capital, one that may be important to individuals but is not necessarily connected to politics or civic outcomes. Hardin argues that political scientists advance an alternative perspective that begins with the individual. This treats trust at the individual level as an independent variable that allows social institutions to work; social capital is therefore a critical factor in collective outcomes.

Douglas Massey (1987) was among the first to apply the concept to the study of immigration in the United States. He found that social capital, in the form of relations to individuals with migrant experience and employer contacts, was a resource newcomer migrants could convert into better jobs. Michael Aguilera and Douglas Massey (2003) concluded that different outcomes among documented and undocumented migrants could be explained by the latter’s lack of social capital.

Some scholars study the relationship between social capital and immigrant incorporation. Many networks are built on extended family and informal hometown relationships of trust and reciprocity that generate social capital and help newcomers find jobs and housing, and ultimately forge linkages to the larger society. Jimmy Sanders and Victor Nee (1996) focus on the family as a form of social capital that immigrants can use in pursuit of economic gains. Interactions with family members represent a social exchange characterized by mutual dependence and expectations in the context of a moral order marked by the accumulation of obligations among members (1996: 233). Gil Epstein (2009) focused on the difference between a bonding network, which decreases incorporation, and a bridging network, which increases incorporation and social capital. Concha et al. (2013) examined how measures of social capital—support from parents, spouses, children, relatives, and friends—affected the acculturation-related stress of Latino immigrants in Miami-Dade. They found that the results were not constant over time—in the first, second, and third year after migration, different relationships played different roles in increasing or reducing reported stress. Relatedly, Hoyt (2009) observed that the 2006 immigrant rights demonstrations grew out of deep networks of social solidarity within immigrant communities.

From these social capital perspectives, immigrants who have accumulated knowledge, experience, and economic resources in a new society can play an important role through social networks in assisting other newcomers (especially family members) to navigate through, and participate in, a host of activities that some see as amounting to ‘effective citizenship’ or social membership. As Bauböck (2001) explains, although transnational communities are important survival and adaptive mechanisms, they are no substitute for the useful and eventually necessary modes of entry into core arenas of the host society where newcomers can access security and rights. The social capital accumulated by established immigrants can help bridge newcomers to economic, educational, civic, and other institutional arenas of the host society.

The one study to examine immigration and social capital in our contemporary enforcement regime is Valdez, Padilla and Valentine (2013). They used focus groups in an Arizona city to explore how the state’s widely publicized enforcement policies affected the social capital ties of Mexican mothers. They noted that the ‘[c]onsequences of restrictive immigration policy on unauthorized immigrants’ social capital has yet to be examined’. Their interviews indicated mixed effects—while enforcement could increase social capital (such as a greater reliance on family networks), it also led to decreased capital (such as less parental involvement in schools). These findings reflect the results of our research in North Carolina and Texas sites (described below), whereby stronger enforcement leads unauthorized immigrants to withdraw from public life as much as feasible.

4. Three studies of national and local enforcement effects

Since the passage of IIRIRA in 1996, two of the authors have independently conducted a number of studies to investigate the effects of escalated enforcement operations on US immigrant communities and those deported from them (Rodriguez and Hagan 2004; Hagan, Eschbach and Rodriguez 2008; Hagan, Rodriguez and Mullis 2011). In this article, we recast the collective findings in these studies to develop a new perspective on how contemporary deportations affect immigrant communities—particularly through the removal of individuals who contribute to social capital. This has implications for immigrant integration, as scholars have noted the importance of social capital for connecting new arrivals to US institutions and sectors.

The first study was launched in 1998 and conducted in five established immigrant communities in Texas (El Paso, Hidalgo, Laredo, Houston, and Fort Worth). In this project, researchers sought the views of social service providers and community-based organizations about the impact of enforcement measures on their migrant client populations. They also interviewed about one hundred households with immigrant members in each of the five research sites to determine the effect of enforcement efforts (Rodriguez and Hagan 2004; 510 total interviews).

The second community study was conducted in 2009 in the relatively new immigrant destination of Johnston County, North Carolina (Hagan, Rodriguez and Mullis 2011). In 1990, Latinos made up 1.5 percent of the population. By 2010, they represented 11 percent of the county’s nearly 160,000 people. The Latinos in the county include newcomer and established Mexicans and Central/South Americans. The involvement of local law enforcement in immigration issues has been pronounced in new immigrant states like North Carolina and in rural areas, like Johnston County, where established residents are less tolerant than their urban counterparts (Bada et al. 2010). The county sheriff was an advocate of local immigration enforcement operations, including the 287(g) program. The North Carolina study was conducted to understand the effects of local enforcement initiatives on immigrants and the community in which they lived and worked. In fall 2008 and spring 2009, fifty Latino immigrants were interviewed, mostly business owners and established residents (Hagan, Rodriguez and Mullis 2011).

A third study focused on the deportees themselves and was undertaken in 2002 in El Salvador. The study was implemented through the cooperation of a consortium of community-based organizations coordinated by Catholic Relief Services (CRS)—El Salvador that operated a reception program (Bienvenido a Casa, or Welcome Home) for deportees arriving from the United States at the San Salvador airport. When Bienvenido a Casa announced in 2002 that it planned to meet with deportees across El Salvador to assess their employment situations, the researchers organized a survey to be conducted with a random sample of deportees attending the meetings. Questions in the instrument concerned the migration experience to the United States; work conditions in the United States; family characteristics in the United States and El Salvador; conditions during apprehension, detention, and deportation; and post-deportation economic experiences and future plans. Three hundred interviews were conducted through systematic random sampling of deportees attending the meetings between June and November of 2002 (Hagan, Eschbach and Rodriguez 2008).

5. Findings

5.1. Effects and perceptions of increased immigration enforcement on US immigrant communities

The household surveys in the five Texas communities found evidence that IIRIRA was affecting individuals and families, but there were variations across the research sites (Rodriguez and Hagan 2004). Overall, 16.5 percent of the 510 respondents reported that Border Patrol and other federal immigration agents had stopped and questioned them regarding their citizenship status. Thirty-nine percent of those questioned were arrested for an immigration violation. About two-thirds of the respondents who stated that INS agents had questioned them or a household member lived in one of the three border research sites of Hidalgo, Laredo, and El Paso. This finding corresponded with the expansion of the Border Patrol force in the border region.

Across the Texas research sites, fear of authorities and the related responses of limiting or circumventing public activity emerged as a common theme among many of the survey respondents. The research found that some of the interviewed households significantly reduced their time away from home except for time spent at work. For example, one young couple in Houston reported that they never traveled outside their apartment together in order to reduce the risk of being arrested together and deported—and thus being separated completely from their two small children. Other survey respondents commented that they avoided public settings where Hispanic immigrants concentrated because they feared these settings were more likely to be targeted for raids by US immigration agents. In the border cities of El Paso and Hidalgo, where the number of Border Patrol agents increased sharply, many survey respondents also indicated fear and apprehension of being out in public; yet, as a few respondents commented, this was part of normal life in border areas where US border agents are stationed.

Across all the Texas research sites, social service providers expressed concern that increased enforcement would frighten away immigrants who were eligible to receive benefits. Public health providers were worried that increased enforcement would make unauthorized immigrants reluctant to immunize their children at public health clinics or receive maternal and child health services. In addition, one health service provider in a Houston community outreach program observed that after immigration enforcement and deportations increased, the Latino participants in the program became more likely to give their identity as ‘white’ rather than ‘Hispanic’.

Educators also expressed concerns. In the Fort Worth and El Paso research sites, school administrators reported enrollment declines in some district schools that they believed were due to new immigration enforcement measures (the Fort Worth district also experienced increased enrollment in other schools). In El Paso, student enrollment decline had started with the implementation of Operation Hold the Line in 1992, which had greatly intensified border enforcement in the area. In the Houston site, an administrator of a school in an immigrant neighborhood blamed a drop in parent attendance at school meetings on fears of heightened enforcement among unauthorized immigrant families.

In contrast to times prior to IIRIRA implementation, the immigrant communities in the Texas research settings clearly demonstrated a heightened fear of increased immigration enforcement. Community institutions such as Spanish-language radio and television stations additionally circulated the social anxieties through their call-in and talk-show programing that often focused on the heightened enforcement. In Houston, a popular Spanish-language radio program reported the street locations of sightings of immigration officers in order to alert its immigrant audience.

In Johnston County, North Carolina, immigrants also reported concern about heightened enforcement efforts, but they were especially anxious about local police participation in the 287(g) program. The local sheriff was a strong advocate of local immigration enforcement. In his position as chairman of the state’s sherriff association, he helped make North Carolina the lead state to partner with ICE deportation operations. Traffic stops essentially became the new enforcement strategies in the county, and many immigrants drove in fear because they could not obtain a North Carolina driver’s license without a valid Social Security number. Consequently, traffic stops became routine. Respondents reported that they were profiled and pulled over by police for no clear reason. More than three-fourths of the fifty persons interviewed knew of someone who had been arrested by a local police officer for violations such as speeding or driving without a license and subsequently deported.

How have new and established immigrants and long-term residents and community leaders in the county responded to expanded local enforcement? What strategies, if any, have immigrants and established residents developed to circumvent enforcement operations? How effective are these operations, and what are their implications for levels of accumulated social capital in the community? We find that although established Latinos and immigrants have forged networks of civil engagement over concerns about local enforcement operations, these networks are not strong enough to mitigate the fear and anxiety that prevails among many immigrants. In fact, many have withdrawn from public life and some have left Johnston County.

As in the case of other communities in new immigrant destination areas with little history of immigration, Latino immigrants in Johnston County do not have an established political base to turn to for advocacy or assistance. Yet, in the aftermath of the immigrant marches of 2006, we found that alliances were forming between newcomer Latinos and established elements of civil society. Foreign and native-born Latinos founded a chapter of Yes We Can, an organization dedicated to immigration reform. The group, comprised of business owners, immigrants, and native-born Latinos, met regularly at a local church to discuss the community costs of local enforcement dragnets and racial profiling. Informal linkages have also developed over the issue of driving without a license. The fear of apprehension and deportation because of a traffic stop has become so great that people with licenses started driving unlicensed immigrants to work. Fliers posted on bulletin boards in many Latino businesses advertise transportation assistance (Hagan, Rodriguez and Mullis 2011).

While some migrants are able to cope and survive the atmosphere of massive deportations through formal and informal intergroup alliances, others live in fear, remain excluded from public life and core institutions, and are sometimes apprehended and deported. Owners of business establishments that cater to Latinos, for example, reported that their sales declined between 30 and 90 percent in the two-year period from 2008 to 2009, and they attributed the bulk of the decline to the loss of Latino customers. Restaurant owners and managers complained that their sales declined substantially because Latinos fear apprehension if they drive to Latino establishments.

Some immigrants have responded by withdrawing from the community. Settled immigrants reported avoiding public spaces, like parks, and no longer attending community events. Immigrants without valid driver’s licenses spent most of their non-working hours at home. This means they no longer went out to eat at local restaurants nor regularly frequented churches or stores. In this way, we can see how deportation practices serve to remove individuals from the networks that constitute social capital.

Some immigrants have stopped investing in a future in North Carolina, and some of the more successful immigrants sent their earnings to family members in Latin America. Others reported that they stopped shopping for clothes, furniture, cars, and other expenditures that link them to life in the United States. They reasoned that by doing this they will have savings waiting for them back home if deported. Fearing deportation and family separation, some immigrants have left the United States. By some community leader accounts, an estimated 30 to 40 percent of the Latino immigrant population had left North Carolina; some were forced home through deportation orders; others returned home to Mexico or Central America fearing the risk of deportation; and still others moved to other states.

State data suggest similar trends. From 2007 to 2009, the North Carolina undocumented population fell from 350,000 to 275,000. This decline reflects a variety of dynamics—deportation, relocation for employment reasons, moving to states with less stringent enforcement, and a return to home nations. The latter is unlikely to be a substantial share of total population loss, as scholars have found little evidence for the strategy of ‘deportation by attrition’, also known as ‘self-deportation’. For instance, Filindra (2012) criticized the behavioral logic behind this theory, and others have observed few voluntary returns (Pedroza 2011; Orlinoff 2012; Parrado 2012) and argue that returns better reflect the recession than enforcement (Massey and Riosmaria 2010). Nevertheless, Latino community leaders believed that pressured returns constitute a part of the North Carolina immigration story.

Collectively, the findings from the Texas and North Carolina community studies indicate that the new immigration enforcement measures implemented since IIRIRA had significant effects, particularly on Latino immigrants and the communities in which they work and reside. Migrants hesitated to participate in public life and feared arrest and deportation. The increased level of immigration enforcement was a daily reality, as immigration officers, sometimes accompanied by local police, stopped and inspected Latinos in interior communities as well as the border region. An ultimate effect of the new enforcement activity is that it reduces the propensity of many immigrants to participate in a wide range of community activities, which has implications not only for individuals and families but also for the social capital of communities.

5.2. Survey findings of Salvadoran deportees

The survey of 300 deportees in El Salvador provided rare evidence about the social and demographic characteristics of deportees removed from the United States after the implementation of IIRIRA. While the US government reports annual data on deportees by nationality and type of deportable violations (criminal or non-criminal), the reports do not cover the social and demographic characteristics of the deportees. Such information is important in determining their levels of integration into US society, and thus to learning about the resources that deported migrants possess and could share with others in immigrant communities.

To better show how the deportation of long-term (non-criminal) immigrants removes valuable sources of human and social capital from immigrant communities, the presentation of the survey findings is differentiated by the length of time spent in the United States prior to deportation. Five categories of time are used in Tables 1–2: lived in the United States for more than 0 years, for more than 1 year, for more than 3 years, for more than 5 years, and for more than 10 years.

Table 1.

Individual characteristics of Salvadoran deportees (percentages)

	Years spent in USA before deportation
	more than 0 (n = 300)	more than 1 (n = 188)	more than 3 (n = 164)	more than 5 (n = 150)	more than 10 (n = 106)
Gender
Female	5.3	2.7	3.1	3.4	1.9
Male	93.7	97.3	96.9	96.7	98.1
Missing	−3.0	−1.0	−1.0	−1.0	0.0
Mean age	29.0	30.0	30.0	31.0	32.0
Age distribution
18–29	51.0	46.0	42.7	40.9	36.8
30–39	32.1	35.3	38.0	38.9	40.6
40–62	16.9	18.7	19.0	20.1	22.6
Missing	−4.0	−1.0	−1.0	−1.0	0.0
Migrant status in USA
Not authorized	67.0	52.2	47.7	44.7	33.3
Authorized	33.0	47.8	52.3	55.3	66.7
Missing	−15.0	−10.0	−9.0	−9.0	−7.0
Education
0 to 6	33.1	28.8	27.8	26.7	22.3
7 to 12	65.7	69.3	70.1	71.0	74.5
13 to 16	1.2	1.6	2.1	2.3	3.2
Missing	−49.0	−25.0	−20.0	−19.0	−12.0
Speak English
None	27.2	7.7	5.6	4.1	3.8
A little	19.4	21.1	17.3	15.5	10.5
More or less well	22.8	28.1	29.0	28.4	21.9
Fluently	30.6	43.2	48.1	52.0	63.8
Missing	−6.0	−3.0	−2.0	−2.0	−1.0
Employed in USA
No	22.2	9.9	9.4	10.2	5.8
Yes	77.8	90.1	90.6	89.8	94.2
Missing	−48.0	−7.0	−5.0	−3.0	−2.0
Years worked in USA
1 year or less	15.6	5.6	2.8	2.3	2.1
1 to 5	31.8	35.0	29.8	24.0	16.8
6 to 10	24.5	26.3	29.8	32.6	25.3
More than 10	28.1	33.1	37.6	41.1	55.8
Missing	−108.0	−28.0	−23.0	−21.0	−11.0
Income per week
0 to $299	27.3	25.8	24.8	24.8	26.3
$300 to $599	51.9	51.6	50.4	49.5	44.7
$600 to $899	12.3	13.3	14.2	14.9	15.8
$900+	8.4	9.4	10.6	10.9	13.2
Missing	−146.0	−60.0	−51.0	−49.0	−30.0

Source: Hagan, Eschbach and Rodriguez (2008)

Table 2.

Family characteristics of Salvadoran deportees (percentages)

	Years lived in USA prior to deportation
	more than 0 (n = 300)	more than 1 (n = 188)	more than 3 (n = 164)	more than 5 (n = 150)	more than 10 (n = 106)
Married
No	48.7	42.3	42.5	42.5	43.3
Yes	51.3	57.7	57.5	57.5	56.7
Missing	−23.0	−6.0	−4.0	−4.0	−2.0
If married, spouse in USA?
No	40.1	24.8	19.1	18.5	12.1
Yes	59.9	75.2	80.9	81.5	87.9
Missing	−163.0	−87.0	−75.0	−69.0	−48.0
Status of spouse in USA
US citizen	40.0	44.1	46.4	49.0	46.2
Green card	36.9	33.9	33.9	31.4	35.9
Other authorized status	10.8	10.2	10.7	11.8	12.8
Unauthorized status	12.3	11.9	8.9	7.8	5.1
Missing	−235.0	−129.0	−108.0	−99.0	−67.0
Do you have children?
No	37.6	28.6	29.9	30.0	26.0
Yes	62.4	71.4	70.1	70.0	74.0
Missing	−45.0	−13.0	−10.0	−10.0	−6.0
Children younger than 18 in USA?
None	13.7	8.2	6.3	6.7	3.3
One	43.2	44.7	43.8	44.0	39.3
More than one	43.2	47.1	50.0	49.3	57.4
Missing	−205.0	−103.0	−84.0	−75.0	−45.0
Children (<18) is US citizen
None	8.9	7.0	5.2	4.5	4.3
One	50.9	49.0	49.0	47.7	41.4
More than one	40.2	44.0	45.8	47.7	54.3
Missing	−188.0	−88.0	−68.0	−62.0	−36.0
If sent money to family/other in El Salvador, how much?
$25–$200 per week	59.8	59.6	64.4	64.5	68.2
$201–$500 per week	35.7	34.8	32.9	33.9	31.8
>$500 per week	4.5	5.6	2.7	1.6	0.0
Missing	−188.0	−99.0	−91.0	−88.0	−62.0

Source: Hagan, Eschbach and Rodriguez (2008)

As Table 1 indicates, deportees tend to be overwhelmingly male, although this is changing due to interior enforcement efforts and the increasing governmental capacity to detain women and families. While the sample deportee median age was 29 years, the median ages became older as the length oftime spent in the United States increased, since more recently arrived migrants tended to be younger. Moreover, as Table 1 indicates, the proportion of authorized migrants increased as the time spent living in the United States lengthened.

Deportee educational levels also increased with time spent in the United States. Some of the deportees with longer residence in the United States attended US schools, which provide more years of compulsory education than do schools in many Central American areas. Not surprisingly, English-speaking skills increased with time spent in the United States.

The findings reported in Table 1 also indicate that time in the United States is associated with economic incorporation. Locating a job, participating in the labor force, and undergoing income mobility transpire across time. Nonetheless, there appears to be some degree of compression in the migrant labor force as the increased proportions in the two highest income categories are not dramatic (although there are substantial missing values across all time categories).

Table 2 contains data on family relations, which reflect the effects of time. Given that recent migrants tend to be younger, and that marriage likelihood increases over time, migrants with more years in the United States tended to be married. It is not surprising that migrants with more years in the United States were more likely to have spouses in the United States, as presumably many of these migrants have located their family formation in the country where they were settled. The findings also indicate that migrants with more years in the United States were more likely to have spouses with legal status (citizen, green card, etc.). Migrants with more years in the United States also tended to have children at a higher proportion than the most recent (and younger) migrants. Moreover, migrants who had lived longer in the United States were more likely to have children living in the United States.

Among those deportees who reported having to leave behind in the United States children younger than 18 or a spouse, 38 percent indicated that they planned to re-migrate to the United States. The comments made by some of the interviewed deportees indicated their enduring bonds with family members as well as the practical importance of social capital. For instance, many of the married deportees who planned to re-migrate spoke of practical family necessities—such as income and social support for spouses—that made their return to the United States urgent. Although they understood the risk of a prison sentence if apprehended for illegal re-entry, this did not appear to be a major deterrent.

Findings from the survey of 300 Salvadoran deportees therefore indicate that the mass removal of Latino migrants is affecting deep layers of the Latin American origin population in the United States. This forced repatriation is removing categories of migrants who have become established residents of the United States with more than five or ten years of living in the country. ICE argues that it is removing criminal aliens, but according to the data presented by the government in annual immigration statistical yearbooks, the majority of deportees are removed for non-criminal violations (see e.g. US Department of Homeland Security 2011: Table 41).

6. Consequences of deportation for immigrant social capital

The findings listed in Table 1 indicate that deportations are removing key human capital resources that are important to the integration of migrant communities. Many of those deported have authorized status, relatively high levels of education, fluency in English, durability in the labor force, and relatively high income levels. Migrant communities incur considerable disadvantages when they lose these resources. When examining the 106 individuals in the sample of Salvadoran deportees who lived in the United States for more than ten years, we see that they have relatively high levels of economic and linguistic skills and are connected to employers and families. They collectively constitute a significant source of human and social capital.

Table 2 indicates how deportations remove resources critical for family household stability. The majority of deportees are being removed from family units in which spouses are likely to have legal status and in which children younger than 18 are present (and who are likely US citizens). As scholars have noted, removing a migrant from families can have devastating effects in terms of loss of income and parental authority. The transmission of social knowledge and life skills also has an intergenerational component that, when absent or disrupted, can have maladaptive consequences (Mannheim 1993). In addition to the fragmentation of family units, the physical removal of family members by force, especially parents, can have long-lasting traumatic effects on children (Capps et al. 2007).

The findings with implications for social capital are those in the Texas and North Carolina studies. To summarize, we found that the fear of deportation caused many immigrants to withdraw from public life. As deportation often begins with a simple traffic stop, many immigrants now avoid driving to churches, restaurants, and other community centers. Such reduced community participation breaks social capital networks or prevents their formation, and thereby has important implications for individuals and communities.

More generally, we know from the literature that immigrants normally participate in a range of community organizations. For instance, although non-citizens are not eligible to vote, there are no prohibitions on a wide variety of other political and civic activities. Researchers have found that Latino non-citizens are engaged in non-electoral political activities (Leal 2002), participate in labor union organizing (Delgado 1996), are networked into community political activities, often by women (Hardy-Fanta 1993), and are actively involved in religious organizations (Levitt 2002; Leal 2010). As Levitt (2002: 158) noted, the Catholic Church is ‘a vast, interconnected network of activities throughout the world’, and religious networks are both national and transnational. Such activities contribute to the creation of social capital, but in the new enforcement climate every decision to participate entails the risk of deportation.

We can therefore draw several conclusions based on our data and the larger scholarly understanding of immigrant communities. First, deportation and involuntary return migration removes an individual from family and business networks. Some of these links may be within immigrant communities, a form of bonding capital, but other links may connect individuals with people and organizations outside the community, a form of bridging capital. In both cases, the community loses an individual who may have been able to uniquely provide material assistance to friends, family, and colleagues. For immigrants, such networks are particularly important, as newcomers may have few other ways to access potentially helpful individuals and institutions—especially in terms of finding jobs. In addition, we saw that a fear of driving causes individuals to withdraw from community life, which further inhibits the growth of social capital networks.

Second, mass deportation may lower levels of community trust, particularly in government. For those who know, work with, or are related to a deportee, the government may become an object of fear. In addition, individuals may be reluctant to contact police or social service providers if they believe this might put them or an unauthorized friend or relative at risk of deportation. Unauthorized individuals who know of a deportation may make special efforts to avoid contact with others in order to minimize awareness of their status. As with the fear of driving, a lowering of trust can have negative consequences for immigrants and their communities (which include citizens).

Third, unauthorized individuals may decide to focus on creating bonding capital instead of developing bridging capital. Interactions with other immigrants, or with institutions familiar with immigrants, may appear less risky than interacting with those less knowledgeable of, and possibly less sympathetic toward, immigrants. While this may lower opportunities for immigrants and their communities, it may be seen as a way to minimize the odds of deportation.

Fourth, as is consistent with past research, deportation affects families. As our data show, many deportees leave behind spouses and children. Aside from the emotional dimension, this reduces the resources available to the family and therefore increases the likelihood that the family will live in poverty. This, in turn, diminishes the life opportunities of children (who are often citizens) and increases stress on the remaining parent. In this way, the deportation of social capital can negatively affect generations of American citizens.

Lastly, we might consider how deportation affects institutions that are relevant to the wider community. When immigrants with social capital are deported, the harm is felt not only by the individuals noted above but also by the organizations in which they participated. Businesses lose workers, churches lose members, and schools lose parents. Such institutions thereby become part of the collateral damage of mass deportation, which in turn has implications for society more generally. If businesses lose experienced workers, this could have negative consequences for the local economy. If churches lose members, this may reduce their ability to engage in religious outreach or community service activities. When parents are deported, fewer volunteers may be available to help with school activities. While this article has focused on the likely harm of deportation to immigrant communities, we should not overlook these additional, wider implications.

If deportations lower social capital, one consequence may be reduced capacity to address community needs. Scholars have hypothesized that one of the functions of social capital is to overcome collective action problems. In communities that already face significant obstacles, a vigorous deportation policy may impede grassroots efforts to combat crime, improve education, enhance the quality of public spaces, and otherwise act on problems that affect everyone yet prove challenging for any individual to fix. If social capital is a web of relationships and concomitant trust that allows communities and individuals to overcome obstacles, solve problems, and make socio-economic gains, then deportation has the potential to tear these fragile threads. For those communities with few other resources, this loss may be considerable.

We also note that deportations vary across the nation. Some locales are involved with the 287(g) program, for instance, while others are not. Immigration laws apply nationally, but some communities do not wish to directly involve local law enforcement in such efforts. This may differentially affect certain types of immigrant communities. If enforcement and deportation are more common in the ‘new destination’ locations with little history of immigration, like North Carolina, then the immigrant communities that most need social capital might be those where it is most often removed.

On the other hand, renewed deportation efforts by ICE have undoubtedly removed many dangerous criminals from the United States, and therefore from immigrant communities. If immigrants and their families cannot live safely in their homes, cannot venture outside without fear, are the subjects of fraud, and are harmed physically, emotionally, and financially by crime, then social capital and community well-being will be negatively affected.

One key issue is to what degree contemporary policies lead to the deportation of ‘criminal aliens’ versus those who have committed relatively minor offenses. While a targeting of violent criminals would not impede—and could even enhance—social capital and immigrant integration, a focus on some other categories of offenses might have the opposite effect. While all crime needs to be taken seriously, the deportation of those with resources—especially legal permanent residents—requires some additional reflection about family costs, community consequences, and proportionality.

In addition, not all crimes that lead to deportation are equal. According to DHS data, most deportees did not commit violent crimes or similar offenses. In fact, the largest categories involve immigration, traffic-related, and drug offenses. The crimes that come to mind when the term ‘criminal alien’ is used—such as assault, robbery, and sexual assault—constitute only single-digit percentages of total deportations. For example, in Fiscal Year 2012, the largest categories were ‘Immigration’ (23.8 percent), ‘Criminal Traffic Offenses’ (23.1 percent), and ‘Dangerous Drugs’ (21.4 percent). By contrast, the percentage for ‘Assault’ was 6.5 percent, for ‘Robbery’ was 1.8 percent, and for ‘Sexual Assault’ was 1.7 percent. This means that the United States deported 12,962 individuals convicted of assault and 3,349 convicted of sexual assault, which undoubtedly benefits the nation generally and immigrant communities in particular. However, the totality of deportations does not reflect the national security and crime narratives used to justify such programs and policies.

Another consideration is the effect of deportation on migration. It might appear that deportation reduces future migration, but as Massey, Durand and Malone (2002: 19) observed, social capital serves to enhance migration because ‘migrant networks are an important source of social capital for people contemplating a move abroad’. For example, they noted that the Bracero program enhanced migration because returning migrants created social capital among their friends and family in Mexico, which ‘reduced the costs and risks of their own international movements and increased their access to US employment’ (2002: 42).

Today, the deportation of hundreds of thousands of individuals may serve to increase and deepen such international social capital networks, which could lead to more migration in the long run. In another context, scholars have observed that deporting gang members can strengthen international criminal networks, with El Salvador a prime example (Vigil 2002; Strickland 2009). When contacts are maintained between the deported and those remaining in the United States, this form of social capital can facilitate illicit future activities.

7. Conclusion

The research discussed in this article indicates that the US policy of massive deportation is carried out at significant cost not only to migrants and their families but also to immigrant communities. Many of those who are deported possess relatively high levels of human capital and are embedded in the family and workplace networks that constitute social capital. Among the deportees to El Salvador, some live legally in the United States; have lived in the United States for some time; have been employed for much of this time; earn incomes that can support a family; and have high school and even college educations. Such connections and relationships are helpful not only to these individuals and their families but also to those they know in their communities. These deportees may also be among the most likely to have developed bridging social capital.

In addition, enforcement practices that target drivers have negative consequences for social capital. Immigrants in North Carolina reported an avoidance of churches, restaurants, and other community venues because driving can start a chain of events that leads to deportation. They therefore withdrew from their communities as much as possible, sometimes paying licensed drivers for rides to work but otherwise staying at home. Such communities are thus fractured by fear, which prevents the development and maintenance of the networks at the heart of social capital. Because immigrant communities are in need of social and human capital, massive deportation imposes substantial costs on individuals as well as communities.

Similarly, current policies deter civic and community mobilization. For instance, while non-citizens cannot vote, the literature finds that immigrants do participate in various political, union, religious, and civic activities. Hardy-Fanta (1993: 120–3) described the mobilization of undocumented immigrants to address neighborhood problems in Boston; they even held rallies, with police officers, in front of drug houses. Such efforts are less likely in the current environment, thereby harming communities directly by reducing local activism and indirectly by preventing the resulting social capital linkages.

Furthermore, we should keep in mind the broader harm to society that can result from a loss of immigrant social and human capital. When businesses lose workers, churches lose parishioners, and schools lose parents, these organizations may not be able to operate as efficiently or effectively. This may have negative implications for the local economy, for religious life, and for local education. Mass deportation may thereby not only affect immigrant communities but also localities more generally. Regardless of race, ethnicity, or legal status, individuals may find local institutions less robust and thereby see their quality of life diminished.

On the other hand, current policies serve to deport some people who have committed serious crimes, and whose removal benefits immigrant communities and American society more broadly. However, it is difficult to judge how this balances the collateral effects of mass deportation described in this article.