Abstract
How should liberal societies select prospective members? A conventional reading of immigration history posits that whereas ascriptive characteristics drove immigration policy in the past, contemporary policy is based on the principle of nondiscrimination. Yet a closer look at the characteristics of those admitted reveals systematic group biases that run counter to liberalism’s core moral commitments. This article first discusses liberal states’ basic moral obligation to treat their citizens with equal respect. It then identifies ways in which the group biases produced by immigration policy violate that principle, when states either deprive their citizens of fundamental rights or stigmatize them through hierarchical constructions of citizenship. Three mechanisms are presented—structural bias, profiling, and positive selection—by which seemingly liberal admissions policies produce illiberal outcomes. The empirical analysis explores the resulting discriminatory group biases in the context of language and income conditionalities on family migration, excessive demand restrictions against economic migrants, and visa waivers for international travelers. We conclude that immigration reforms that mitigate, if not erase, these morally problematic patterns are within the reach of liberal states.
Keywords
Days after his inauguration, President Donald Trump issued Executive Order 13679, which banned citizens from several predominantly Muslim countries from entering the United States. The “Muslim Ban,” as the media dubbed it, triggered vocal opposition from civil society and public officials, including several Republican legislators. Behind the reactions was the conviction that Executive Order 13769 violated the liberal principle of equal respect by explicitly excluding immigrants based on nationality and, implicitly, ethnicity and religion. In this article we show that, although unusual in its explicitness, the so-called Muslim Ban is not the big exception to the rule of otherwise liberal immigration policy. Rather, it coexists with long-standing illiberal practices of immigrant admissions across the Global North. Those practices have tended to be subtler and sometimes (but not always) unintentional. Here we will focus not on this newly surfaced tip of the iceberg but on the much more sizable chunk of submerged ice of illiberal practices in immigrant selection.
The dismantling of the racially stratified immigration systems of Canada, Australia, and the United States in the 1960s and 1970s has been hailed as the beginning of a new era of “nondiscriminatory” and “universal” immigration policy. 1 Although liberal states retain the right to determine who is admitted into their territories, immigrant selection is supposed to be based on clearly stated meritocratic criteria, on the potential contributions that the individual will make to her host community, on her relationship to citizens or permanent residents, and, under certain conditions, on the individual’s need for protection.
This liberal understanding of immigrant selection has led some migration scholars to argue that the era of discriminatory immigration policy is over. Christian Joppke has described the prevalence of skill-based selection as the epitome of nondiscrimination: “The state may consider the individual only for what she does, not for what she is. . . . The individual is selected according to ‘achievement,’ not ‘ascription,’ that is, according to her agency rather than according to what she is immutably born with.” 2 Joppke’s argument reflects a conventional reading of Western immigration history in which the reforms that began in the Civil Rights Era abolished long-standing discriminatory practices. Without denying their import, we must critically examine whether these normative and legal trends have translated into immigration policies that are truly nondiscriminatory.
To be sure, the history of citizenship in liberal societies is marred by examples of exclusionary policies targeting citizens and noncitizens alike. We are certainly not the first to identify ways in which liberal principles and policies have ignored persistent forms of exclusion and discrimination (explicit and implicit, intentional and structural). Feminist, critical race, and critical legal scholars have pointed out numerous contexts in which immigrant admissions reflect systematic group biases along the lines of ethnicity, gender, ability, nationality, religion, and class, even in the most liberal of policy regimes. 3 These arguments, however, often grounded in contested claims about the moral obligations of liberal states toward noncitizens, either lead to embracing an open-borders view or impose costly demands on liberal states (e.g., addressing global structural inequalities). Our project contributes to that body of scholarship by bringing to light some hidden ways many of those policies not only discriminate against noncitizens but also violate the principle of equal respect toward citizens. In this regard, our critique is less radical but more immediately applicable.
We concur, contra Joppke, that immigration policies can be discriminatory even if they are not based on the intentional and explicit exclusion of certain social groups. However, in contrast to critical immigration scholars, we suggest that such policies are problematic even if we recognize the right of states legitimately to restrict immigration and do not assume that they hold moral obligations toward noncitizens. The policies discussed below are illiberal not simply because of policymakers’ discriminatory motivations or because of their effects on prospective immigrants (although a moral argument could certainly be made on these bases), but because they fail to live up to liberal states’ moral obligation to treat their citizens with equal respect. Facially neutral immigration policies discriminate either by protecting the rights of some citizens but not others or by reproducing stigmas associated with ascriptive features such as ethnicity, gender, nationality, religion, class, or ability status that are shared by a subset of citizens. We identify three mechanisms—structural bias, positive selection, and profiling—whereby immigration policies systematically engage in wrongful discrimination. Our ambition is normative, not explanatory. We do not claim that these are the only factors that explain admission biases against certain groups, although their effects are by no means negligible. Our central point is that the biases these mechanisms produce are morally problematic for the same reasons that liberal states have condemned explicitly discriminatory immigration policies since the 1960s.
First, we discuss the principle that the minimum normative parameter for a liberal immigration policy ought to be the state’s moral obligation to treat its citizens with equal respect. Second, we present the three mechanisms whereby facially neutral admissions policies can produce patterns of discrimination that violate this normative standard. Third, we explore the group biases of different policies of immigrant admissions pertaining to family unification, economic recruitment, and international travel. By analyzing these different policy realms, we are able to highlight how states’ interaction with prospective immigrants and international travelers can lead to indirect forms of discrimination against their own citizens. Our analysis shows (1) that these mechanisms account for at least part of the admission bias against certain groups and (2) that these biases violate core moral obligations of liberal states. We conclude that all these policies can be avoided and that replacing them with nondiscriminatory policies is within the reach of liberal states.
Discrimination, Immigrant Selection, and the Rights of Citizens
A conventional reading of the history of immigration policy suggests that the progressive repeal of race-based immigrant admissions since the 1960s has marked the advent of an age of nondiscriminatory immigration policy. 4 The underlying assumption of these claims is that discrimination in immigration policy must be explicit and intentional. Joppke argues that “a truly impartial immigration policy is an impossibility because in including some, all others have to be excluded, and an ethnic pattern in this can always be found.” 5 Thus, for Joppke, immigration policy cannot be discriminatory if it lacks racist intent, even if it has disparate effects across groups. In this section, we argue that policies biased against certain social groups are not immune from the charge of discrimination simply because selection necessitates differentiation. Even if we adopt a minimalist view of the scope of the moral commitments of liberal states and assume that they only hold moral responsibilities toward their own citizens, immigration policies can lead to wrongful discrimination if they protect the rights of some citizens but not others or if they reproduce stigmas associated with certain ascriptive features that are shared by some citizens.
This argument rests on two premises. 6 The first is that all states, including liberal states, have the authority and legitimacy to regulate access to their territories. 7 Second, liberal states do not have absolute discretion when selecting among prospective immigrants since they cannot legitimately violate their constitutive moral principles in the process of, or as a result of, immigrant selection. 8 On these two premises, what are the normative commitments that constrain liberal states when selecting among prospective immigrants?
An “expansive” or “inclusive cosmopolitan” view claims that liberal states have moral responsibilities toward noncitizens. 9 These can include the obligation to receive foreigners that face persecution in their countries of origin, 10 the responsibility to admit inhabitants of former colonies in recognition of historical injustices, 11 the design of immigration policies that do not disrespect immigrants or violate their right to participate in decisions that affect them, 12 or even to open their doors to foreigners to redress global distributive injustices. 13 These claims are vigorously contested within liberalism, since many believe that liberal states are morally obliged to prioritize the interests of their citizens over those of noncitizens. 14
A more “restrictive” or “communitarian” view would remain agnostic about the obligations of liberal states toward noncitizens and focus instead on their normative commitments toward their own citizens. 15 At the very minimum, liberal states are bound by the moral obligation to recognize the equal worth of all their citizens. Of course, equality is a contested concept in political theory. 16 However, the idea of moral equality, which in its most general expression establishes that all citizens have the right to be treated as equals in terms of dignity and respect, 17 is seen as a “shared moral standard” of all modern Western liberal moral and political culture. 18 Indeed, this principle is established in the United Nations Universal Declaration of Human Rights and represents the constitutional bedrock of liberal democracies.
Therefore, following theorists of immigration such as Christopher Heath Wellman 19 and Michael Blake, 20 we argue that policies of immigrant selection must, at the very least, live up to the normative principle that states must recognize their citizens as individuals of equal worth and thus extend equal respect to all citizens. We see this as a “minimal” normative standard that commands consensus about the moral obligations of liberal states. Naturally, if the policies that we describe below are considered discriminatory and illiberal from this “minimal” point of reference, they will also be regarded as discriminatory and illiberal by those who adopt more demanding views about the moral obligations of liberal states toward citizens as well as noncitizens. 21
In other words, whereas Joppke and others 22 see discrimination as a breach of the principle of personal autonomy only through the intentional and explicit exclusion of entire categories of people defined by immutable characteristics, we view discrimination as a violation of the principle of equal respect through policies that demean the inherent dignity and worth of certain citizens. 23 Selection on the basis of ascriptive characteristics is problematic not merely because those features are immutable, but because they have been used in the past to deny the equal moral worth of their bearers, and they continue to be associated with negative stereotypes that systematically disadvantage citizens who share them, such as ethnicity, nationality, gender, sexual orientation, ability status, class, and religion.
Some ascriptive markers refer to immutable properties that individuals cannot change at will. Religion, however, is different from race, gender, or ability status insofar as it is not immutable. But it is also different from nonascriptive parts of one’s sense of self—such as language and profession—because it demands exclusivity: one cannot be a practicing Catholic and Muslim at the same time. Therefore, when it comes to discretionary admissions—in contrast to obligatory admissions based on humanitarian grounds—liberal states can legitimately demand that would-be immigrants speak a language or hold certain professional skills, whereas asking them to embrace a particular religion would demand that they give up a central part of their self- and collective identities. Deselecting prospective immigrants on the basis of religion thus constitutes morally problematic discrimination because it stigmatizes citizens who share the same religion.
Class is also a special kind of ascriptive characteristic. From the perspective of liberal values, class inequalities do not violate the principle of equal respect if they are the result of individual merit. Of course, class inequalities can never be truly reduced to differences in individual achievement, given that they are structurally constituted by historical patterns of wealth acquisition. Inequalities across families in access to economic, cultural, and social capital are generally reproduced across generations; individual class positions (and related attributes such as income, wealth, education, or occupation) are rarely the result of individual merit or lack of merit. The intergenerational reproduction of class inequalities should thus be sufficient to make selection on the basis of class attributes a morally problematic practice. Nonetheless, even if one made the case that class inequalities purely reflect individual achievement, selection on the basis of class could still violate the principle of equal respect for the following reasons. First, class inequalities are often shaped by forms of dispossession and exclusion of certain social groups on the basis of other ascriptive characteristics, such as race or gender, and class attributes are sometimes used as a proxy to exclude individuals with those traits. In those cases, selection on the basis of class entails a denial of the equal worth of the bearers of those ascriptive features even without explicit reference to their race or gender. Second, selection on the basis of class also breaches the principle of equal respect if it leads to the stratification of fundamental individual rights. For example, income restrictions to suffrage or legal immunity for upper-class citizens would still be considered illiberal even if economic disparities were purely the result of individual achievement. In keeping with our goal of carrying out a critique of immigration policies based on liberal values, we deem class a problematic criterion of immigrant selection if it violates the principle of equal respect in either of these two ways. Having said that, we recognize that selection on the basis of class attributes can be considered unjust and undesirable for many other reasons besides this very specific understanding of discrimination. 24
Policies that exclude prospective immigrants on these bases reproduce the stigma associated with those traits and make a “public announcement” of bias against citizens that share them. 25 Such a public announcement of bias violates the principle of equal respect in at least two ways. First, the selective exclusion of immigrants can entail direct violations of the rights of citizens, particularly in relation to citizens’ right to family life, which can be systematically obstructed by policies that restrict the entry of immediate family members on the basis of ascriptive characteristics. 26
Second, states that ban members of certain social groups from admission engage in hierarchical constructions of citizenship. By enacting such policies, the state reproduces negative stereotypes associated with those traits and sends the demeaning message that citizens who share those features are less preferred participants in the political union. One could argue that because, from this perspective, any characteristic unequally distributed in the world will be shared by at least some citizens, any form of immigrant selection will inevitably convey disrespect to some of them. However, the issue here is that ascriptive characteristics are defined by and result from a history of social stigmatization that permeates most Western liberal democracies; and they continue to be a vector of disadvantage today. Consequently, when policies of immigrant selection differentiate among prospective immigrants on the basis of these traits, they deny the equal worth of those citizens who share them. 27
The claim that ascriptive immigrant selection conveys disrespect to citizens who share these characteristics is supported by the history of popular mobilization against discriminatory immigration policies, where pressures to repeal those policies have come from citizens who find them demeaning, denigrating, disempowering, or in outright violation of their rights. 28 One example is the mobilization of societal elites that hastened the dismantling of the racially stratified immigration systems of Canada, Australia, and the United States in the 1960s and 1970s. More recent examples are the legal and political activism of civil society organizations against policies of explicit exclusion based on sexual orientation, religion, and ability status.
Mechanisms of Discrimination in Admissions Policies in Liberal States
There is a broad consensus among liberal thinkers that overtly discriminative policies that exclude individuals based on ascriptive characteristics violate the moral principles that define liberal states. 29 At the same time, neither discriminatory intent nor explicit reference to ascriptive characteristics need be present for admissions policies to violate the principle of equal respect in at least one of the two ways outlined at the end of the previous section. Policymakers often ban certain social groups from admission through what we call pretextual exclusion (see Table 1). Pretextual exclusion describes the reliance on nonascriptive features highly correlated with ascriptive characteristics as pretexts for exclusion. Because these policies systematically and intentionally convey disrespect to certain social groups just as if they had been explicitly formulated in terms of ascriptive selection, they cannot escape the charge of breaching the principle of equal respect.
Types of Illiberal Exclusion.
Although pretextual exclusion reflects intent to exclude members of particular social groups without explicitly referencing their ascriptive identities, discrimination may also be present in facially neutral policies, designed without any illiberal intention, through what we call indirect exclusion (see Table 1). Indirect exclusion occurs when policies unintentionally produce disparate effects that either stratify access to rights or reproduce stigmas associated with ascriptive characteristics. In sum, neither explicit mention of ascriptive features nor discriminatory intent is necessary for a policy to violate the principle of equal respect. Without denying the deplorability of explicit and intentional discrimination, our analysis focuses on cases of pretextual and indirect exclusion rather than on the better-known instances of explicit exclusion.
We identify three mechanisms of group bias that, under certain circumstances, make facially neutral policies violate the principle of equal respect: structural bias, profiling, and positive selection. These mechanisms do not require—although they may entail—discriminatory intent on the part of policymakers. All that matters is that, in practice, they systematically violate the rights of certain groups of citizens or reproduce stigmas associated with certain ascriptive features that withhold respect from those citizens who share them.
First, structural bias occurs where immigrant selection based on facially neutral criteria stratifies citizens’ access to fundamental rights. This stratification is particularly problematic when it is based on ascriptive traits, as it reinforces hierarchical constructions of citizenship that conflict with the principle of equal respect. The mechanism of structural bias often underlies patterns of pretextual exclusion, whereby policymakers intentionally exclude specific social groups by relying on “neutral” requirements that are strongly correlated with “undesirable” ascriptive features. 30 For example, laws that require literacy tests or poll taxes as a requirement for voter registration have been used in many countries to effectively disenfranchise groups that are more likely to be illiterate or to lack resources: women, ethnic and racial minorities, or lower-income citizens. Such policies therefore commit two distinct wrongs: they obstruct access to rights and reinforce hierarchical constructions of citizenship.
However, structural bias can also lead to patterns of indirect exclusion. Where neutral policies unintentionally produce disparate effects by stratifying access to rights on the basis of ascriptive features, they contribute to the stigmatization of the bearers of those traits. Because those policies reproduce hierarchical constructions of citizenship, they are disrespectful of citizens who share those features even if they are unaware of those effects. 31 Take the example of very strict voter ID laws. Even if we assume that they are concerned exclusively with preventing electoral fraud and do not intentionally target any particular group, they interact with structural biases present in society in ways that lead to persistent inequalities in access to a fundamental political right. In doing so, such policies publicly announce that the equal protection of the rights of certain social groups is of secondary importance.
As a second mechanism, profiling (or statistical discrimination) is the result of strategies that address the problems of insufficient information in evaluating the potential costs to society of admitting a particular individual. Risk-averse decision makers may, under conditions of imperfect information, rely on assumptions about the average costs that citizens with similar ascriptive characteristics impose on society when evaluating an individual’s application for admission. In principle, statistical discrimination in immigrant selection can be morally acceptable in situations in which the stakes are limited for the profiled group, the costs of individualized assessments are unaffordable, and reliable statistical parameters are available (e.g., in the use of demographic data to calculate insurance premiums). Profiling is wrongful, by contrast, where it results in the unequal treatment of citizens on matters that may prove central to their life plans, where it can be avoided without imposing excessive costs on the state, or where differential treatment is based on prejudices and stereotypes rather than on reliable data. 32 Even if there is no discriminatory intent, cost and risk assessments based on negative stereotypes reproduce stigmas associated with ascriptive traits. Thus profiling entrenches hierarchical constructions of citizenship that reflect social prejudices about the costs and contributions of the bearers of those traits, thereby violating the principle of equal respect. Moreover, where profiling is used to adjudicate claims that are consequential for citizens’ lives, it infringes on their rights.
As a third mechanism, positive selection occurs where decision makers give priority of admission to prospective immigrants expected to make highly valued contributions to society. 33 This form of selection can be theoretically consistent with liberal principles, as long as societal contributions are based on individual merit. However, meritocratic selection criteria can reinforce hierarchical constructions of citizenship if expectations about social contributions are based on stereotypes rather than on individual assessments or market demand. Certain professional activities, for instance, carry gender and class connotations that influence perceptions about their economic and social value. Activities conventionally associated with female labor (such as child or elder care) or lower-class labor (such as agricultural work) tend to be perceived as contributing less than activities associated with middle-class males—such as white-collar professions—even if labor market demand is equally high for both. 34 Policies of positive selection can thus reproduce stereotypes that entrench hierarchical constructions of citizenship. They not only convey disrespect to citizens who share those ascriptive traits but may also lead to legal differences in the protection of family life, as we demonstrate below.
Table 2 summarizes these mechanisms of discrimination, their effects on citizens, the types of exclusion they enable, and the group biases they produce. The discussion of admissions policies pertaining to family immigration, economic immigration, and international travel that we present in the rest of the article is intended to be neither empirically exhaustive nor representative of a larger population of cases. Rather, we use these cases to illustrate the ways in which contemporary admissions policies produce morally problematic patterns of group bias. We begin each section with a brief policy overview and then analyze its group impacts and the mechanisms driving these biases.
Mechanisms of Illiberal Immigrant Selection.
Family Immigration
The right to family unification 35 follows from the universal right to family life for citizen members of transnational families. The right to family life is widely recognized in international human rights documents (e.g., Article 16 of the United Nations Universal Declaration of Human Rights) and national legislations. States are free to adopt a more expansive definition, but the minimal legal definition of “family” is the nuclear family—two spouses 36 and their dependent children. Therefore, in the admission of the foreign spouses and dependent children of citizens, 37 the right to family life precludes the states’ imposition of conditionalities that do not pertain to the nature of family ties. Family reunification may be restricted only under exceptional conditions, such as where the foreign relative constitutes a threat to public safety. 38 Yet in practice this right is often rendered contingent by states’ reliance on additional restrictions.
Since the early 2000s, family unification has come under sustained attack in parts of Europe as governments have sought to shift their stance on family unification from one of passive acceptance to active selection. Policies aimed at reining in family immigration, preventing forced marriage, and fostering immigrant integration have introduced a new degree of conditionality through measures such as preentry language and civics tests for spouses (and, in some places, even children), 39 higher income thresholds for family sponsors, and increases in the minimum age of marriage. 40 These measures have not only curtailed access to family unification for many nuclear families but also affected applicants disparately depending on their nationality, ethnicity, religion, gender, class, and ability status. In other words, these requirements have imposed unequal entry burdens on applicants with otherwise equally strong family ties. These requirements are morally problematic in two distinct ways. First, they are wrongful insofar as they violate a fundamental human right—the right to family life—regardless of the nature of stratification. Second, if this stratification selectively targets certain groups on the basis of their ascriptive characteristics, it reproduces stigmas associated with those features, sending the message that the rights of certain citizens are more important than those of other citizens.
The Group Impacts of Preentry Language Tests
Reflecting a growing perception of family migrants as unwanted and socioculturally problematic, family unification policy in several EU member states—Germany and the Netherlands in particular—has become conditional on preentry requirements. Since 2007, non-EU applicants for a spousal entry visa to Germany have had to demonstrate their ability to communicate in basic German verbally and in writing before being issued a visa. The impact of the preentry language requirement is far from trivial. Immediately after its introduction, the number of family visas issued by German consular officials fell by 40 percent. 41 Three years later, visa figures were still 25 percent below prereform levels. 42 In addition to impeding family immigration generally, preentry tests have had highly uneven impacts, disproportionately affecting nationals from many non-Western countries. This situation reflects the intersectionality of nationality and class, as the drop in family visas has been most pronounced for the Dominican Republic (−63 percent), Nigeria (−53 percent), and Kazakhstan (−50 percent). Likewise, pass rates have varied greatly across countries. In 2009, the overall pass rate for family visas to Germany—including those who had taken the test multiple times—was 64 percent. 43 However, test takers are at a disproportionate risk of failure if they originate from Syria 44 (pass rate 25 percent), Macedonia (32 percent), or Ethiopia or Sri Lanka (both 37 percent).
Test failure has important implications for the transnational families’ right to family life. Although language-test-related visa denials by German officials account for fewer than 1 percent of applications (165 visas were denied in 2014), 45 each rejection amounts to a categorical denial of the right to family life. Even if applicants ultimately pass the test, repeated test failure leads to long delays in the unification of nuclear families. At best—where applicants can pay tuition, take classes at a Goethe Institute, 46 and gain language proficiency—test preparation adds an average of six months to the period of forced separation. 47 In the Netherlands, the preentry test in combination with higher income requirements has increased the period of forced separation by an average of fifteen months. 48 Thus, it is not surprising that the preentry language test requirement has resulted in high levels of partner and family stress. In the words of an immigrant service provider, “Many people are at [a] breaking point over it, which means that I give these couples advice about family reunification and then transfer them to my colleague in the department of separation and divorce.” 49
As qualitative data from Morocco suggest, 50 high levels of familial stress force families to decide against applying for a family visa and to continue to live apart where preentry test requirements are too difficult to fulfill. Dutch data on the educational composition of test cohorts lend further support to a pattern of self-selection. After the test requirement was introduced in 2006, the proportion of family migrants with high levels of education increased from 20 percent to 33 percent, whereas the proportion of those with low levels of education dropped from 34 percent to 28 percent. 51 It appears likely that those with little formal education not only struggle to pass the test but may also decide not to take or retake it. 52
Class—as an important determinant of education and financial resources—plays a critical role in mediating access to education and the acquisition of a foreign language. After the proficiency level was raised in the Netherlands, fail rates jumped from 29 to 62 percent for poorly educated test takers, compared to a jump from 5 to 16 percent for highly educated applicants, with additional variation based on gender. 53 German data also point to the importance of high-quality language training as a predictor of test success. In 2009, the average pass rate of those participating in a Goethe Institute language course was 81 percent, compared to 61 percent for nonparticipants. 54 However, the vast majority of test takers—80 percent in 2009—do not take classes at a Goethe Institute or a licensed partner institution. Even if the sending country has a German language institute, costs may be prohibitive. Focus group interviews suggest that in many countries tuition alone amounts to 50 percent of the monthly salary. 55 Because reaching the required level of German typically takes three months of full-time study, many applicants must quit their jobs and go without income during this time 56 or must secure alternative childcare arrangements. The mediating effects of class on language acquisition are problematic in this context because they create inequalities between citizens in access to rights. Even though class inequalities can be morally acceptable in liberal societies, the stratification of rights based on class is not. Moreover, class-based stigmas tend to intersect with and reinforce the disadvantages resulting from ethnicity and ability status, among other characteristics.
Finally, applicants who are not literate are particularly disadvantaged. To quote from a communication by the Cairo Goethe Institute: Unfortunately there are no classes for illiterate persons. They should first learn Arabic and then basic English and after that they can try to take a German class at the Goethe Institute. Participants should have knowledge of English or French in order to better learn German.
57
In the absence of a hardship clause 58 that would exempt disadvantaged groups such as older or illiterate adults, applicants from countries such as Morocco, a major sending state where more than 30 percent of the population is illiterate, can face seemingly insurmountable obstacles in applying for a family entry visa. 59 Although it is plausible to suggest that, initially, the absence of a hardship clause may have been an oversight, the German government’s refusal to revise its policy even when confronted with its consequences suggests otherwise. Instead, the government has insisted that the ability to read and write German is essential to integration and has demanded the demonstration of “a willingness of migrating spouses to integrate . . . even when personal circumstances such as illiteracy require a relatively high effort for language acquisition.” 60 There are no policy exemptions, so the test effectively unselects illiterate spouses from family unification. Because illiteracy can result from a lack of educational opportunities or a learning disability, the absence of policy exemptions for illiterate applicants has disparate group impacts based on class, gender, and ability status.
Class membership and education also serve as criteria for positive selection by means of policy exemptions for high-skilled immigrants. In Germany, this policy categorically exempts highly qualified workers from preentry tests, as well as holders of the EU Blue Card 61 and their spouses. With this exemption, the policy bases preferential access to nuclear family unification on a criterion that is unrelated to family ties. As a result, the state violates the principle of equal respect by using class membership as a gatekeeper to family unification.
Finally, preentry test policies in Germany and the Netherlands exempt applicants of certain nationalities. Nationality exemptions are anchored in both supranational and national law. At the supranational level, not only have EU freedom-of-movement provisions precluded imposing any conditionalities on EU nationals, but the European Court of Justice has forced the hand of German policymakers to exempt third-country spouses of EU and Turkish nationals (although Germany so far has refused to implement the Turkish exemption fully). At the same time, German policymakers have voluntarily added thirteen states to the list of national exemptions—Australia, Andorra, Brazil, Canada, El Salvador, Honduras, Israel, Japan, South Korea, Monaco, New Zealand, San Marino, and the United States—further expanding unequal access to family unification based on the nationality of family members living abroad.
Mechanism of discrimination: Structural bias
A closer examination of policy debates shows that European governments have used preentry tests as a pretext for excluding, with often thinly veiled intent, socioculturally “undesirable” family immigrants. For those in the Global South, structural bias operates largely through educational background. Preentry tests disproportionately disadvantage applicants from nonwhite and Muslim-majority countries and, within those countries, women, adults with learning disabilities, and other economically marginalized groups. In Germany, policymakers frequently justified the language requirement as facilitating immigrants’ socioeconomic integration and preventing the occurrence of forced marriages. Lawmakers, however, clearly had one group in mind: young and poorly educated Muslim women from Turkey. Test proponents explicitly justified the targeting of less educated Muslim immigrants with reference to their supposed “cultural backwardness.” 62 Similarly, legislators in the Netherlands acknowledged that preentry tests were designed to limit family immigration by targeting those with “entry problems.” Family migrants from Turkey and Morocco in particular were seen to bear “characteristics that are unfavorable for good integration into Dutch society.” Preentry tests thus facilitate the “postponement or even cancellation of the intended settlement in the Netherlands.” 63
Structural bias not only jeopardizes the right of some citizens to a family life but also reproduces hierarchical constructions of citizenship. A renowned expert on immigration law counseled the Bundestag’s Home Affairs Committee that “in this case, the constitutional protection of family life should not be prioritized because of the weighty public interest in the prevention of the kind of immigration which will impose a lasting burden on the welfare state, as well as in the interest of fighting immigration abuse.” 64 Structural bias based on the state’s interest in preventing the immigration of spouses from undesirable groups not only renders the right to family unification meaningless but also signals to Germans from certain ethnic backgrounds that they are less valued citizens.
Given the prevalence of transnational partner choice in Germany’s Turkish community in particular, 65 there is a high correspondence between the ethnicity of “undesirable” foreign spouses and that of their citizen sponsors. Whereas the hierarchical constructions of citizenship typically amount to a symbolic signaling of lesser social esteem, in this case they communicate the physical undesirability of certain citizens. If citizens of undesirable groups hold dual nationality and therefore can pursue their family life abroad, the denial of a family visa ejects, quite literally, citizens from their national community.
Mechanism of discrimination: Positive selection
Policymakers have used education and nationality as bases for positive selection by exempting certain prospective immigrants from the preentry language requirement. Here positive selection presents the reverse image of structural bias, signaling clear preferences about the social desirability of particular groups and linking admission to criteria unrelated to family ties. The fact that many foreign spouses are exempt from the preentry test solely on the basis of higher education or nationality sends a powerful message to those whose spouses struggle to meet the preentry language requirement. In the words of a German sponsor whose Cuban fiancée had failed the language test, I find it unbearable that there are . . . language test exemptions for the foreign spouses of high-skilled immigrants or of EU citizens living in Germany. . . . I am considering leaving Germany should there be no other possibility for my fiancée and me to live together.
66
A striking feature of both German and Dutch preentry policy is the list of nationality exemptions. Although policymakers had no choice but to exempt EU nationals and their spouses from the requirement (a case of indirect discrimination), the additional exemption of six countries for the Netherlands and thirteen for Germany was a deliberate policy decision to select family immigrants on the basis of nationality. The German legal expert cited above told the Bundestag’s Home Affairs Committee that the preentry test’s nationality exemptions rested on “factual reasons” because of the “differing integration difficulties of nationals of different countries of origin.” 67 This case of positive selection has been criticized as discriminatory on the grounds of nationality by the United Nations’ Committee on the Elimination of Racial Discrimination, the Council of Europe’s Human Rights Commissioner, and the European Commission against Racism and Intolerance. 68 The German government insists that it has special diplomatic and economic interests in the immigration of nationals of these countries, 69 but statements such as the one cited above suggest that preferences for applicants from “culturally similar” sending countries likely played some role in policy development. It seems clear, when we compare these exemptions to the fate of nonexempt spouses for whom the test is an insurmountable hurdle, that for many citizens the difference between a family life and one of forced family separation is determined by the ascriptive matter of their spouse’s nationality.
The Group Impacts of High Income Requirements
Another common reform to family immigration policy has been the tightening of preentry income requirements for family sponsors. Britain and Norway, in particular, are situated at the restrictive end of income requirements. In 2012, Britain raised its income threshold from the social assistance level to 138 percent of the minimum wage. Sponsors are now required to earn ₤18,600 ($26,864) annually to bring in a spouse, ₤22,400 ($32,352) to sponsor a spouse and a non-British child, and ₤2,400 ($3,466) for each additional non-British child. The foreign spouse’s earning capacity is not counted toward the income threshold; neither is any financial support received by third parties, such as parents. Moreover, application costs are extraordinarily high, exceeding ₤6,000 ($8,670) for a spouse without children. 70 It is estimated that 61 percent of employed British women and 32 percent of employed British men are below that income threshold. 71
Class intersects not only with gender but also with ethnicity. Fifty-one percent of white versus 59 percent of nonwhite workers earn below the threshold for sponsoring a spouse and child. 72 The intersection of gender and ethnicity produces even more pronounced differences in the ability to meet sponsorship requirements: whereas the median annual income of a white British man is well above the threshold (₤24,000), the median income of a woman of Pakistani origin is half that required for sponsoring a spouse (₤9,700). 73 Given the severity of British policy, it is not surprising that survey and interview data have revealed widespread detrimental impacts on spouses and their children. 74 The new income rules are estimated to have caused about 15,000 British children to be either separated from a parent or forced to live outside Britain. 75
Among European countries, Norway’s income requirement is the highest in absolute terms. 76 Set at NOK246,136 ($30,060), the required income is twice Norway’s median income. Moreover, the sponsor must meet the threshold not only in the year of application but also during the preceding year. It is telling that since 2010, when it was raised to its current levels, the income threshold has become the most common reason for the rejection of family visas. 77 At the same time, under Norwegian family immigration policy, skilled workers occupy a privileged position. Not only are skilled workers categorically exempt from any income requirement, but their applications for family unification are processed within two months, compared to up to seven months for all other family migrants. 78 Not surprisingly, family reunification rates are highest among those who are exempt from the income rule: skilled workers have a sponsorship success rate of 99 percent. By contrast, Norwegian citizens, permanent residents, and refugees have an acceptance rate of 68 percent. 79 Finally, reflecting deeply entrenched global class hierarchies, success rates are highest for family migrants from North and South America (90 percent) and lowest for East Africans (55 percent). 80
Mechanism of discrimination: Structural bias
With the adoption of higher income thresholds, liberal states have further institutionalized group inequalities in access to family unification. Family migrants and their citizen sponsors, in contrast to highly skilled immigrants, have become associated with unwanted cultural and economic characteristics—“the unemployed of the future.”
81
The British government justified raising the income threshold as part of a broader commitment to reducing net migration by “tens of thousands” and reducing the tax burden imposed by family immigration.
82
It is critical to recognize that the income thresholds discussed in this article are not simply a utilitarian device to prevent immigrant families from accessing social assistance. If that were the case, not only would income thresholds still be set at the social assistance level, but other sources of income, such as the earnings of the spouse to be sponsored, would be factored in. Instead, like preentry language tests, income thresholds exclude, through structural bias and the use of class criteria, intentionally or inadvertently, immigrants with undesirable group memberships that intersect with their labor market status. Policymakers commonly (and, often, correctly) assume that citizen sponsors share the ethnic and class background of the spouses they seek to sponsor and have used income requirements strategically to prevent the immigration of members of “undesirable” groups and also to signal to citizen members of these groups that they are of lesser value. In the words of a German Conservative parliamentarian, There are many naturalized Turks and Arabs . . . .who have basically prepared themselves to live their entire life [on social welfare]. And here the question came up: how could we motivate them to actually get their act together concerning a job? That is the decisive issue, to say: “if you want to have a family reunification, you have to get your act together and work!”
83
Thus, income thresholds intersect with structural biases, not only to undermine the right to family life for citizens from certain class and ethnic backgrounds, but also to indicate that the state holds them in lower esteem than citizens with higher class and nonminority group memberships.
Mechanism of discrimination: Positive selection
Income thresholds both exclude through structural bias and produce discriminatory patterns by means of positive selection. This is the case in Norway, where highly skilled workers are categorically exempt from any income requirements. Here the imperative of international competitiveness trumps all other considerations, as is evident in the following policy statement: “In an arena where the battle over bright minds is great, most will not let themselves be recruited to work in a country which requires that their family stays in the home country for a long time awaiting the processing of an application for a residence permit.” 84
Even if income requirements are an acceptable form of immigrant selection in other contexts (such as economic migration), they are morally problematic when applied to family unification because they stratify access to a fundamental right on the basis of class while at the same time conveying a message of disrespect to less privileged citizens. Whereas positive selection has given rise to a privileged class of families facing few hurdles to unification, structural bias has created second-class citizens with only precarious access to family unification. The unification policies examined here, rather than base admissions squarely on the nature of family ties, reflect the growing racialized and neoliberal logic of citizenship that disregards the moral weight of family unification. In other words, these states are no longer willing to accept family immigration as an expression of the state’s noninterference with, and respect for, the right to family life held by all of their citizens. Instead, we witness a move toward the strategic selection and deselection of family migrants based on ascriptive characteristics that has far-reaching implications for the rights and moral standing of their citizen sponsors.
Economic Immigration: The Group Impacts of Medical Inadmissibility Restrictions
Family and humanitarian migration reflect the “self-limited” sovereignty of liberal states that arises from their commitment to the protection of fundamental human rights—such as family life and physical safety. Economic migration and international travel, on the other hand, allow liberal states to further their national interests by attracting foreigners who can make valuable contributions to society through their skills, assets, or consumption. By offering preferential conditions of admission, liberal states fill labor gaps and attract highly skilled workers and investors. Similarly, international travel is essential for a globalized economy. States have a vested interest in permitting the temporary entry of tourists and business travelers. Economic migration and international travel therefore require states to find nondiscriminatory ways of attracting “desirable” foreigners while keeping away “undesirable” immigrants.
To illustrate how facially neutral policies can lead to wrongful discrimination, we examine the case of medical inadmissibility criteria for economic immigrants in Canada. Human rights organizations such as the Council of Canadians with Disabilities have denounced the ways excessive demand restrictions violate the principle of equal respect. As a result, the Canadian state has progressively removed discriminatory provisions from its immigration policy. This shows that addressing the patterns of discrimination discussed in this article is not beyond the reach of contemporary liberal states.
Most liberal states reserve the right to restrict access to immigration when prospective immigrants or their dependents are likely to generate significant financial burdens on public resources. 85 In Canada, for example, foreign nationals are inadmissible if their condition is likely to put public health and safety at risk or place an “excessive demand” on public services. 86 “Excessive demand” is defined as requiring services that represent higher than average per capita costs for five consecutive years or adding to the existing wait lists for those services. 87 Excessive demand restrictions are estimated to save CAN$135 million in medical costs over a five-year period, or 0.1 percent of provincial and territorial health spending. 88 Refugees and the immediate family members of Canadian citizens and permanent residents have been exempt from these restrictions since 2002, but economic migrants and their family members remain subject to excessive demand considerations. If an applicant or accompanying family member is deemed medically inadmissible, the entire family is barred from admission.
It is common practice among visa officers to base their assessments on average medical profiles, treatment, and service needs. 89 Differences in life conditions and medical histories, however, mean that even if two individuals have the same illness or disability, they may require different services and obtain them in various ways without incurring public expenditure. For example, applicants or their sponsors may commit through affidavits to use their private incomes to finance their medical treatment, or charitable organizations may offer to pay their costs. Moreover, the tax and social security contributions of these prospective immigrants may offset their expected demand for public services.
It is particularly striking that these considerations are often overlooked for economic migrants already being offered preferential admission due to their economic assets. In 2005 the Supreme Court of Canada challenged this implementation of the medical inadmissibility clause in two landmark cases—Hilewitz v. Canada and de Jong v. Canada. The appellants in both cases applied for permanent residency for themselves and their families under the economic-class category (as an “investor” and an “entrepreneur,” respectively). Each had a dependent child with a disability. The Supreme Court ruled in favor of the appellants with the argument that immigration officers were evaluating excessive demand based on the characteristics of the impairment rather than the characteristics of the individuals applying for admission, which represented a form of categorical exclusion. The rulings suggested that a nondiscriminatory implementation of medical inadmissibility must evaluate excessive demand case by case, taking into account the potential contributions that the applicant and her family will make to society, as well as other nonmedical circumstances that will mitigate the costs of their demand for public services.
Following those rulings, immigration officers are required to issue a Procedural Fairness Letter, which explains the reasons for refusal and gives applicants the opportunity to respond to the medical assessment with a second medical opinion or with evidence that their individual circumstances will not lead to excessive demand. Nevertheless, those attenuating factors are frequently ignored, and discriminatory applications of the excessive demand clause are still frequent. 90 Since 2005, the Federal Court and the Federal Court of Appeal have dealt with several cases in which individualized assessments have not been properly implemented. 91 Similar cases have repeatedly appeared in the media, generating public outcry. 92 Although there are no publicly available data on the number of rejected applications by reason, according to news sources, 0.2 percent of applications for permanent residence were denied in 2010 for excessive demand considerations. 93
Liberal states no longer frame their immigration policies in a language of explicit exclusion of people with disabilities. Indeed, the reforms institutionalized by the 2002 Immigration and Refugee Protection Act in Canada made parents and dependent children of citizens and permanent residents exempt from excessive demand restrictions, thus removing obstacles that stratified the right to family life on the basis of illness and disability. The Supreme Court’s 2005 rulings further addressed illiberal aspects of the implementation of excessive demand considerations to economic migrants by demanding individualized assessments, requiring Procedural Fairness Letters from immigration officers, and granting applicants the opportunity to respond.
However, the implementation of medical inadmissibility and excessive demand restrictions has continued to produce illiberal patterns of pretextual exclusion. Profiling occurs when risk-averse visa officers lack direct information about the costs that an individual may impose on society. Thus they rely on the average costs of citizens with similar characteristics to estimate the burden. Assessments based on statistical parameters of a population rather than on individual-level observations can represent a form of categorical exclusion. Once this calculation is carried out, certain disabilities and medical conditions are automatically associated with higher than average demand, and individuals sharing those conditions will inevitably be excluded from admission. In this way, profiling prospective immigrants because of disability and illness brings to the surface ways in which the state perpetuates hierarchical constructions of citizenship.
Restrictive immigration policies against individuals with disabilities originate from illiberal ideologies of eugenics, which saw people with mental and physical disabilities as supposedly inferior human beings. 94 Today, those prejudices continue to inform a “medical model of disability” that frames disability in terms of “inability, incapacity, deprivation, deviation from the normal.” 95 This view obscures the societal contributions of people with disabilities under the generalized perception that they are net receivers of public resources rather than full participants of the political union. 96 When immigration officers rely on profiling and fail to take into account the individual circumstances of the applicant, they not only reproduce but openly legitimize such a medical model of disability. In doing so, the state stigmatizes citizens with disabilities and reproduces hierarchical constructions of citizenship. As Joseph Carens has argued, this type of assessments reflect “popular prejudice and uninformed fear rather than a reasonable calculation of the risks or burdens” that specific individuals are likely to impose on public resources. 97
The fact that these policies are disrespectful has not gone unnoticed by Canadian citizens. In a letter to parliament, the chairman of the Council of Canadians with Disabilities stated, The immigration system’s approach is based on the faulty assumption that people with disabilities are a burden on society and cannot make a contribution. Canada’s immigration practices perpetuate negative stereotypes and devalue Canadians with disabilities.
98
In November 2017, the Liberal government announced plans to revise these policies, even hinting at the possibility of repealing them altogether. 99 It is telling that concerns about the demeaning message excessive demand restrictions send to Canada’s citizens with disabilities are at the center of public debate. 100
International Travel: The Group Impacts of Visa Waver Programs
In this section, we examine the operation of visa waiver programs as a case of migration policy that is fundamentally premised on distinctions based on nationality—and, by extension, ethnicity, religion, and class. Visa waiver policies regulate who can and who cannot enter a territory without a visa on the basis of their country of origin. As nationality is now the prime determinant of international mobility, “the differential ability to move in space—and even more so to have access to opportunities for movement—has become a major stratifying force in the global social hierarchy.” 101 The following discussion highlights how policies that discriminate against noncitizens, even among those who do not seek to immigrate, can violate the principle of equal respect toward citizens by reproducing hierarchical constructions of citizenship.
States impose visa restrictions in an attempt to manage the trade-off between the economic benefits of cross-border movement and the entry of unwanted foreign nationals. Given that visa overstay is the most common pathway to illegal immigration, visa policy is designed to prevent the entry of would-be immigrants (as opposed to genuine visitors). Visa policy further seeks to prevent the entry of individuals who might constitute a risk to national security and is often employed to stop humanitarian flows. Yet cross-border mobility is necessary for tourism, trade, and foreign investment. 102 In order to promote desired forms of mobility and reduce visa-processing costs, states have resorted to visa waiver policies that allow nationals of particular countries to travel visa free for the purpose of short-term stay. As visa waiver programs have proliferated, the distribution of travel possibilities has become increasingly stratified. In 2010, nationals of OECD countries could travel visa free to seventy-four countries, compared to twenty-two countries for non-OECD nationals. 103
The international visa regime is differentiated not only by region, but also by national wealth. The citizens of a very poor country have a predicted probability of facing visa requirements of 89 percent, compared to 34 percent for citizens of a very rich country. 104 In other words, the visa regime “grants nearly unlimited freedom of short-term international movement to nationals of affluent states,” 105 and affluent states only. Intersecting with national wealth, access to international mobility is further stratified by ethnicity and religion. Most countries with black or Muslim majorities face large-scale restrictions on international travel.
Mechanism of Discrimination: Positive Selection
Because “freedom of travel is freedom to trade,” 106 countries of the Global North use visa waivers to maximize the benefits of economic interdependence while minimizing the risk of illegal overstay (on the assumption that citizens of affluent democracies have weak incentives to overstay). Visa waiver policies thus are a form of positive selection by which the individual’s treatment as a traveler is determined by the ascriptive characteristic of nationality.
Visa waiver programs impinge not only on foreign nationals’ cross-border mobility, but also on citizens’ ability to be visited by family members or business associates from abroad. Imagine, for instance, a British-born American and an Iranian-born American who each seek to invite their extended families living abroad to their weddings. While the British-born American will be able to celebrate this occasion surrounded by a large circle of joyous relatives, the Iranian-born American will likely have to ponder the many empty seats that speak of visa denial. Given how ubiquitous and reciprocal visa waiver programs are among—and only among—the countries of the Global North, our fictional Iranian-born American and her American-born family may well experience the denial of visa privileges to their Iranian kin as an expression of disrespect.
Mechanism of Discrimination: Profiling
Visa policies regularly subject individuals to discrimination because their social group is associated with a greater than average likelihood of engaging in undesirable behavior. Visa waiver policies are typically based on low “disqualification rates,” that is, they are conditioned on a low proportion of national passport holders having violated the conditions of their admission (in the United States, 2 percent on average over time). In other words, profiling leads to discrimination in which nationals of countries with relatively high disqualification rates are prevented from visa-free entry because of their group membership, rather than their individual qualifications.
Although one can make a pragmatic case for the use of statistical profiling for the sake of striking a balance between freedom of movement and border control, statistical profiling is employed in deeply flawed ways in visa waiver programs. First, reliable exit controls continue to be remarkably underdeveloped across the Global North. Even for the pioneering US-VISIT entry-exit system, established to identify visa overstays, “full implementation remains a distant goal.” 107 As a result, statistical profiling is based on behavioral assumptions derived from unreliable data. Second, even setting aside the problem of flawed data, given the difficulties of securing a travel visa by applicants from the Global South, the categorical exclusion of entire nationalities from visa-free travel because of extremely low overstay rates is a disproportionate measure: it bases individual treatment on the behavior of 2 percent of a national group.
The assumption that citizens of affluent democracies have weak incentives to overstay is not borne out in practice. Canadians, for example, account for the majority of visa overstays in the United States and, moreover, have the highest overstay rate of all nationalities. 108 However, despite their high likelihood of overstay, Canadian citizens continue to enjoy not only visa-free travel but also remain exempt from US-VISIT controls. In this case, profiling is clearly based on prejudices about certain nationalities rather than on reliable statistical data. Such policies convey disrespect to those US citizens—such as Mexican-Americans—who share the same ethnic background as foreign nationals who are excluded from visa waiver programs.
For the citizen sponsors of visitors from “untrusted” countries, a visa rejection represents a denial of access to temporary family reunion and a loss of business opportunities that contrasts starkly with the privileges enjoyed by sponsors of visitors from “trusted” (although not necessarily trustworthy) countries. Once again, for citizens with the same social group membership as rejected applicants, visa denial can have much broader symbolic connotations, as it signals a withholding of social esteem.
Conclusion
In this normative appraisal of immigration policy, we have adopted a minimal liberal standard that demands that policies not violate the state’s moral duty to treat its own citizens with equal respect. Even from such a comparatively undemanding moral vantage point, we have identified problematic patterns in facially liberal admissions policies that operate even in the absence of discriminatory intent and explicit exclusion. We have identified three mechanisms—structural bias, positive selection, and profiling—whereby seemingly liberal policies violate the state’s duty to treat its citizens with equal respect because they either stigmatize certain citizens or deprive them of basic rights. We have shown how these mechanisms operate in the context of language and income conditionalities on family migration, excessive demand restrictions for economic migrants, and visa waivers for international travelers.
This normative-empirical analysis does not entail that liberal states have the moral duty to address structural inequalities abroad or to open their borders to all foreigners seeking admission, but rather that liberal states should not enact policies that treat their citizens with unequal respect. As the reforms to the medical inadmissibility criteria in the Canadian case show, liberal states are aware of—and, in some cases, concerned with—these patterns of discrimination. The policies discussed in this article are not inevitable. In fact, there are policy alternatives that could either eliminate discrimination (the cases of preentry language tests, income requirements, and excessive demand restrictions) or at the very least mitigate it (the case of visa waivers) without imposing excessive demands on the state. The moral commitment to recognize and protect the equal worth of citizens through immigration policies that do not wrongfully discriminate against certain groups is within the reach of liberal states. Not only that, it emerges from the very normative commitments that make these states liberal.
Unlike more explicit forms of exclusion such as the Muslim Ban, the policies we have discussed rarely generate large public outcry, even at times when antipopulist norms have been widely accepted. Many of these policies are kept in place by bureaucratic decisions that occur at arm’s length of media scrutiny or democratic oversight. As nativist populism once again gains traction in many liberal democracies, we should expect these indirect forms of discrimination to coexist with more overt ones and to be implemented even by mainstream political actors in response to anti-immigrant pressures. In changing political circumstances, the mobilization of citizens affected by these policies will represent the strongest protection against those submerged forms of illiberal backsliding.
Footnotes
Acknowledgements
The authors thank Camille Desmarès, Madeleine Page, and Klaudia Wegschaider for excellent research assistance, as well as copy-editing assistance by Alberto Alcaraz. Sara Kalm, Kimberly Morgan, participants at the 2015 APSA Annual Meeting, and the editorial board of Politics & Society provided very helpful comments on previous versions of the article. We are particularly indebted to Joseph Carens for providing detailed and immensely constructive comments on a previous version.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Social Sciences and Humanities Research Council, Insight Grant #435-2013-1065.