UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Governing refugees Lippert, Randy 1997

You are currently on our download blacklist and unable to view media. You will be unbanned within an hour.
To un-ban yourself please visit the following link and solve the reCAPTCHA, we will then redirect you back here.

Item Metadata


831-ubc_1998-271870.pdf [ 17.23MB ]
JSON: 831-1.0089369.json
JSON-LD: 831-1.0089369-ld.json
RDF/XML (Pretty): 831-1.0089369-rdf.xml
RDF/JSON: 831-1.0089369-rdf.json
Turtle: 831-1.0089369-turtle.txt
N-Triples: 831-1.0089369-rdf-ntriples.txt
Original Record: 831-1.0089369-source.json
Full Text

Full Text

GOVERNING REFUGEES By RANDY LIPPERT B.A., The University of Lethbridge, 1988 M.A., The University of Ottawa, 1991 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY THE FACULTY OF GRADUATE STUDIES (Department of Anthropology and Sociology) We accept this thesis as conforming to the • recruired standard THE UNrVERJ?ITY OF BRITISH (foLUMBlk' NOVEMBER, 1997 (c) Randy Lippert, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of J O C i 0 I 3 q \ / The University of British Columbia Vancouver, Canada Date PcCewi k-</{ ttrfl°l 7 DE-6 (2/88) 11 ABSTRACT Practices directed at refugees emerged in the inter-war period. Thousands of Europeans were maintained in camps, selected, and then resettled in the decades that followed. By the 1960s, what had been ad hoc, small scale, temporary international responses in Europe had become routine, distinctive practices directed at crises and millions of people around the world. In the Canadian context, practices of government directed at refugees have only recently arisen. The present thesis focuses on these Canadian practices as a governmental regime. By adopting Foucault's methods of discourse analysis and genealogy, the overarching research question, 'How is the Canadian refugee regime constituted and governed?', is pursued. Research procedures included forty-eight interviews with authorities active in the regime, attendance at a refugee conference, collection of documents, and examination of indices of the humanities and social sciences. The thesis begins by discussing conditions of possibility of the Canadian regime's emergence. It explores the development of the international refugee regime as one of these conditions, and in so doing, suggests the potential relevance of concepts and themes drawn from governmentality studies to understanding international regimes. The rise of an advanced liberal rationality in Canadian selection, determination, and resettlement practices since the 1970s is then discussed. In all three areas, refugees' conduct and fate can be seen gradually becoming governed less by state agents and more by agents at a distance from political authorities. In Ill resettlement during this period, however, an advanced liberal rationality can be seen deferring to a pastoral rationality, thereby suggesting the Canadian regime has been constituted by more than one rationality. Developments consistent with the ascendancy of advanced liberalism or otherwise making the governance of refugees possible are then explored. These include: the 'partnership' as a form of association; refugee studies; psychological knowledge revealing the resettling refugee's soul; economic knowledge identifying the economically risky refugee; knowledge developing early warning systems: and technologies such as the documentation centre. Finally, the thesis argues that while resistance, understood as obstruction to (liberal) governance, is evident during this period, the presence of a pastoral rationality suggests a more complex view is required. TABLE OF CONTENTS IV ABSTRACT ii TABLE OF CONTENTS iv LIST OF FIGURES vi ACKNOWLEDGEMENT vii INTRODUCTION 1 Purpose 5 Research Questions 6 Statement of Thesis 6 Chapter Organization 10 Endnotes: Introduction. . 12 CHAPTER ONE PREVIOUS RESEARCH • 14 The International Literature 15 The Canadian Literature 20 Historical Studies 20 Legal Studies 21 State-centered Studies 23 SUMMARY AND CONCLUSION 30 Endnotes for Chapter One 32 CHAPTER TWO APPROACHING A REFUGEE REGIME 35 Neo-Marxist Approaches 35 State-centred Approaches 39 A Foucauldian Alternative 44 SUMMARY AND CONCLUSION 59 Endnotes for Chapter Two 60 CHAPTER THREE METHOD AND RESEARCH PROCEDURES 65 Discourse Analysis 65 Genealogy 67 Research Procedures 71 SUMMARY AND CONCLUSION 77 Endnotes for Chapter Three 79 CHAPTER FOUR THE INTERNATIONAL REGIME AND OTHER CONDITIONS OF POSSIBILITY 81 The International Regime 84 The International Refugee 102 Governmentality and the International Regime 105 A Canadian Regime 118 Other Conditions of Possibility 122 SUMMARY AND CONCLUSION 129 Endnotes for Chapter Four 131 V CHAPTER FIVE ASSUMPTIONS, PROGRAMMES, AND RATIONALITIES 137 Assumptions about the Canadian Regime and the Refugee.. 137 Liberal Government 145 Refugee Selection 147 Refugee Determination 165 Refugee Resettlement 188 Pastoral Government 207 SUMMARY AND CONCLUSION 215 Endnotes for Chapter Five 219 CHAPTER SIX ASSOCIATIONS, KNOWLEDGES, TECHNOLOGIES, AND RESISTANCE 231 Associations 231 Knowledges 238 The Birth of Refugee Studies 238 Psychological Knowledge 250 Economic Knowledge 255 Knowledge About Early Warning Systems 261 Technologies 266 Resistance 298 SUMMARY AND CONCLUSION 309 Endnotes for Chapter Six 311 CHAPTER SEVEN CONCLUSION 315 Endnotes for Chapter Seven 327 BIBLIOGRAPHY 328 APPENDIX I 370 APPENDIX II 371 VI LIST OF FIGURES FIGURE 1.1 REFUGEE ICONS 272 Vll ACKNOWLEDGEMENT Writing (and completing) this thesis was a very difficult task. Several people assisted me. My sincerest thanks to Gillian Creese who, despite differences in theoretical perspective, was solidly supportive, from my first letter from Ottawa that mused about the thesis subject matter, to its completion six years later. During this period her constructive comments and friendly encouragement seemed to always come at the right time. I am indebted to Richard V. Ericson for his willingness to become part of the thesis endeavour and his valuable suggestions when its direction had (finally) become clearer to me. Thanks also to David Schweitzer for all his thoughtful comments and encouragement and to Brian Elliott for helping to get the thesis going when it seemed headed in a different direction. I am also grateful to the many interviewees across Canada who gave up their time to speak with me (including the person who telephoned every hotel in Thunder Bay to locate me to change the time of the interview, only to keep her appointment anyway (at the time I was lounging at one of many telephoneless, late 1950s decor, 'fleabag' motels that I came to know during my cross-Canada excursion to conduct interviews)). Thanks to my friend, Jesse Seary, who during the past six years taught me, through sometimes intoxicating dialogue, about social theory, and by example, about the meaning of the 'Glass Bead Game'. Finally, a very special thanks to Francine Milne Lippert, without whose patience, love, and support this thesis would not have been possible. V 1 INTRODUCTION: GOVERNING REFUGEES Human beings have undoubtedly migrated from place to place to seek refuge from harm since time immemorial. Practices of government directed at refugees have only recently emerged. It is only in the inter-war period that what had been ad hoc, temporary responses of relief, selection, and resettlement of various European groups began to occur. Hundreds of thousands of Europeans were maintained in camps, selected, and then resettled in the decades following the First and Second World Wars (Zolberg, et al., 1989). By the 1960s, ad hoc, small scale, temporary international responses to events and human beings in Europe were becoming routine, distinctive practices directed at events and human beings around the world. In the Canadian context practices of government directed at refugees have also only arisen recently. Between the inter-war period and the early 1970s there were several Canadian responses, in conjunction with international reactions, to refugee crises and groups, including the Displaced Persons in 1947, Hungarians in 1956, Czechoslovakians in 1968, and Ugandans in 1972. These Canadian responses were made possible only by sudden, temporary suspension of standard operating immigration procedures and relaxing of screening and selection criteria through Orders-in-Council (Dirks, 1995: 61). Though eventually resulting in admission and resettlement of large blocs of human beings, these Canadian intervention efforts were regarded by authorities of various kinds as ad hoc, temporary, not necessarily to be repeated 2 in kind. These efforts were not predicated on arrangements specifically targeting refugees beforehand. It was not until the mid-1960s that Canadian authorities even began to make a distinction between the admittance and resettlement of refugees and that of immigrants. Dirks (1985: 124) writes: Until very recently, many of the refugees who have been admitted to Canada have not been regarded as distinguishable from routine economic migrants. The government historically did not classify newcomers according to either their reasons for leaving their states of origin or for choosing Canada as a new homeland (emphasis added). The 1966 White Paper on Immigration, the programmatic statement of the period, referred to Canadian responses to refugees abroad (Dirks, 1977: 229). Following this there were attempts to define a refugee, to carve out and classify an object/identity to be governed from a Canadian perspective. In 1969 Canada became a signatory to the 1951 Geneva Convention and the 1967 Protocol on refugees and adopted the corresponding refugee definition. In 1973 'refugee1 was introduced into Canadian legislation for the first time in an amendment to the Immigration Appeal Board Act (Canada, 1995b: 2). Following release of the programmatic Green Paper on Immigration in the mid-1970s, 'refugee' was written into a new Immigration Act as a distinct legal category, separate from 'immigrant' (Lanphier, 1981). In 1978 the regulations accompanying this Act referred, for the first time, to procedures for annual refugee plans and the determination of Convention refugee status, as well as provisions for selection of refugees abroad and their resettlement (Hawkins, 1989). With these developments an 3 assortment of state and non-state Canadian authorities and agencies has emerged. Some are refugee-specific but before the early 1970s none were both refugee-specific and assumed permanent. Since the late 1970s Canadian practices directed at refugees have been undergoing significant change. The present Canadian refugee regime comprises a variety of practices directed at refugees both beyond and within Canada's territorial and conceptual borders. These include Canadian Mennonite women taking Hmong refugee women to shopping malls to show them how to shop; psychiatrists employing specialized techniques to diagnose and treat refugees eliciting symptoms of post-traumatic stress disorder; legal professionals presenting their refugee claimant's cases in status determination hearings; visa officers examining human beings in refugee camps abroad to determine their suitability for selection; Canadian refugee experts conducting longitudinal studies to monitor adaptation and adjustment of refugees in Canada (e.g., Johnson and Beiser, 1994); Canadian immigration officials consulting representatives of refugee advocacy and other organizations to set annual refugee selection intake levels; Refugee coordinators promoting refugee sponsorship among Christian church congregations; and innumerable others. These practices can be grouped into threeJ types: selection, determination, and resettlement. Refugee selection occurs abroad and involves Canadian visa officers periodically visiting refugee camps, detention centres, and transit points. They select human 4 beings that pass security screening and fit Canadian refugee categories. Once selected and screened, refugees are transported to Canada for resettlement. Determination involves human beings who have travelled to Canada to make a legal claim to Convention refugee status. Once in Canada the refugee claimant is represented by legal professionals who develop the claim before it is presented at an oral hearing. Usually after several months, the refugee claimant is either deemed to be a refugee, granted legal immigrant status, and resettled or, following appeals, literally removed from Canada. Finally, refugee resettlement occurs within Canada. Those selected and transported to Canada are either sponsored by the Canadian state or by private refugee sponsorship organizations and groups. Once transported, refugees are met at ports of entry and then helped to obtain housing, employment, education, medical coverage, and financial support and directed to conduct themselves as workers, consumers, and citizens. In his later writings on 'governmentality1 Michel Foucault characterized the practice of government in the West as a 'conduct of conduct' (Gordon, 1991: 2). For Foucault, government involved authorities seeking to shape the direction of citizens' lives. By focusing on programmes, rationalities, technologies and knowledges rather than the state, Foucault and others writing about 'governmentality1 (e.g., Rose and Miller, 1992) offer a way of making sense of the varied Canadian practices directed at refugees, as well as how they have been changing. 5 PURPOSE This thesis seeks to make intelligible these varied Canadian practices directed toward refugees, and how they have been changing, as a governmental regime. Simultaneously it attempts to lend insight into governmentality generally. Using Michel Fducault's methods of discourse analysis and genealogy this thesis attempts to problematize" these practices, to render them strange, to get beyond their self-evidence, to destabilize and unsettle comfortable views about them, including their connections with 7 necessity, humanity, and progress. This thesis takes up Foucault's challenge: to engage in research and analysis that is critical -in that it problematizes practices and corresponding objects/identities taken as given- and effective -in that it refuses to provide refuge to other practices and their related objects/identities (cf. Dean, 1994c: 164). Such an endeavour does not seek chaos, irrationality or despair, but inventiveness, the possible creation of new ways of governing and resisting, and new forms of identity. But if there is another purpose of this thesis, it is this. If the varied practices that make up this regime and the ways in which they have been changing are seen to have pernicious effects, such as inequitable identification and treatment of refugees based on perceived gender, class, race, ethnicity, age, sexual orientation, or world region; the production and reproduction of the refugee, refugee claimant, and other non-citizen objects/identities within Canada; or if there is a broader dissatisfaction with relating to human beings as refugees and 6 0 refugee claimants -whatever negative effects might be perceived-then it makes sense to focus attention on the constitution and workings of this emergent regime before seeking strategies aimed at ameliorative change. RESEARCH QUESTION(S) Given these purposes, the overarching research question of this thesis becomes: How is the Canadian refugee regime constituted and governed?5 This question implies at least six others: (1) What were the conditions of possibility under which the Canadian regime began to emerge? (2) What are the current discursive assumptions about the refugee object/identity among relevant authorities? (3) What are the programmes and rationalities that have constituted and organized the regime since its emergence? (4) What is the form of association among authorities that is consistent with one of these current rationalities? (5) What are the forms of knowledge and the technologies that have made the regime and changes within it possible? (6) How is resistance to be conceived? STATEMENT OF THESIS In previous research, including recent studies in governmentality, there has not been recognition of Canadian refugee selection, determination, and resettlement practices as comprising a unique historical governmental regime. Nor has there been problematization of these practices as such. There has been little effort to make this regime intelligible by elucidating that which constitutes it: rationalities, forms of association, knowledges, and technologies. Instead of being problematized, the Canadian 7 refugee regime has been taken for granted, most often seen as an exceptional humanitarian development devoid of politics, power, and resistance, but also as necessarily relevant to racism, reducible to the inner workings of the Canadian state, fully explained by reference to Canada's international relations with other nation-states, or fitting with the logic of capitalism. International refugee programmes and organizations began in Europe following the First World War as temporary, small-scale, and specific to groups defined by ethnicity or nationality. By the 1960s these programmes had become permanent, large-scale, and directed at a universal category. A brief review of this international regime since the First World War reveals governmental programmes, rationalities, and technologies making possible and shaping relevant practices, thereby suggesting that the themes and concepts of governmentality studies may be relevant to attempting to understand both national and international regimes. It is in this international sphere that 'refugeeness' started to become more useful to Western authorities after 1947 within intensifying Cold War relations with the Soviet Union and the Eastern Bloc nations exemplified in international responses to events in Hungary in 1956 and later within the context of de-colonization in Africa and elsewhere. 'Refugeeness' can be seen as a Western moral-political tactic in international power relations. The fact that a Canadian refugee regime emerged in the period between the mid-1960s and early 1970s was not a matter of course. It was contingent on certain conditions, the crystallization of this international o 8 refugee regime being only one. Others include Canada's attempt to become a more influential citizen in the international community; a renewed interest in Canada's population size and composition; and a recognized scarcity of desired labour for the national economy. The Canadian refugee regime is assumed largely distinct from immigration, as is the refugee from the immigrant, the latter distinction being predicated on the basis of a capacity for choice. Since the 1970s changes in three types of practices of the refugee regime consistent with the rise of an advanced liberal rationality are evident. During this period refugees' conduct and fate can be seen gradually becoming governed less by state agents and more by a variety of agents and authorities at a distance from political authorities. Though consistent with this, the resettlement practices of private sponsors and Host volunteers, mostly from Christian churches, that also emerged during this period, are of a character that suggest the presence of a Christian pastorate. Advanced liberalism has in this instance deferred to a pastoral rationality, a broad discourse that to date has received little attention in governmentality studies. Consistent with the ascendency of an advanced liberal rationality is the rise to dominance of the 'partnership', a flexible, active form of association among authorities that is present throughout the regime. A new science of refugee studies has also arisen during the period in which the Canadian refugee regime has emerged and can be seen making Canadian (and international) practices directed at refugees possible. New emphases on economic, psychological and 9 early warning knowledge can be seen accompanying and making possible significant changes in selection, resettlement and, to some extent, determination (and international) practices respectively. A variety of technologies including statistics, icons, camps, and examinations can also be seen making practices directed at refugees possible. Among those that have come to complement the ascendency of an advanced liberal rationality are special psychological assessment and intervention techniques in resettlement and the documentation centre in determination. Resistance is evident in relation to the Canadian refugee regime and can be understood as obstruction to governance, as leading to failure of governmental programmes and occasionally, as in determination during the 1980s, crisis. Resistance from at least three sources converged at this point to spawn a new determination programme consistent with an advanced liberal rationality. The notion of the presence of more than one rationality in a regime, in this case a pastoral and an advanced liberal rationality in the refugee regime, suggests, however, that resistance is more complex than this. Overall this thesis contends that the Canadian practices directed at refugees and the changes they have been undergoing since the 1970s are best seen as a contingent governmental regime that first came into being as a result of certain conditions that happened to come together in the late 1960s and early 1970s. This regime is a shifting historical assemblage of heterogeneous rationalities, forms of association, technologies, and knowledges. 10 Besides lending support to what others have observed about liberal government in other contexts, the exploration of this regime contributes to governmentality studies by highlighting the relevance of migration, as well as the significance of an international sphere of politics and a pastoral rationality. Those who perceive pernicious effects stemming from the practices that make up the regime, and the ways in which they have been mutating, would do well to recognize the new arenas of contestation that have opened up and those that have been dismantled. CHAPTER ORGANIZATION The chapters of this thesis are organized in the following manner. Chapter one describes previous research on refugee policy and practices with an emphasis on the Canadian context. Chapter two moves from this previous research to begin to elaborate an appropriate conceptual approach exemplified by the work of Nikolas Rose, Peter Miller, Mitchell Dean, Ian Hacking, and others who have extended the insights contained in Michel Foucault's writings, especially those on 'governmentality'. This is accomplished in part by discussing themes and features of the historical sociology of the state and state theory. Chapter three discusses discourse analysis and genealogy as methods consistent with such an approach. Following this, but in the same chapter, the research procedures carried out for this thesis are detailed. The remaining chapters that discuss the results of these procedures are largely organized around the research questions above. Chapter four discusses the formation of the international regime since the First World War and 11 its object, the relevance of themes and concepts drawn from the governmentality literature to it, and then other conditions of possibility of the emergence of the Canadian regime. After briefly discussing the distinctiveness of the Canadian regime, as seen in the appearance of refugee-specific authorities and agents and assumptions about the refugee among authorities, chapter five examines relevant Canadian refugee programmes and rationalities that have constituted and organized the refugee regime, and how they have changed, since the 1970s. Chapter six proceeds to discuss the form of association among authorities -the 'partnership'- consistent with an advanced liberal rationality, a logic that presently constitutes this regime. This chapter also discusses the appearance of refugee studies as a new science and other forms of knowledge, technologies, and, finally, resistance. Chapter seven, the concluding chapter, discusses how the thesis contributes to existing research literatures. It suggests that this research contributes to refugee studies by offering an alternative account of Canadian refugee policy and practices and to governmentality studies by lending support to what previously has been noted about liberal rule and law in other contexts and pointing to significant omissions, especially attention to a pastoral or other rationalities and an international sphere of politics. 12 1. Formal refugee bureaucratic structures of determination, selection, and resettlement are rarely found outside the West. 2. Directing attention to the Canadian refugee regime, as this thesis does, undoubtedly risks exaggerating its importance relative to other domains within Canada. It should be made clear at the outset, therefore, that the volume of human and material resources associated with these varied Canadian practices is undoubtedly far less than longer-standing fields such as 'the national economy' (Miller and Rose, 1990b). Similarly, the branch of the Department of Immigration that explicitly targets refugees is dwarfed by the remainder that targets immigrants. 3. These three types form the 'core' of Canadian refugee policy (cf. Adelman, 1984a: 1-2) and it is for this reason they are focused on. A fourth type not focused on would be the provision of funding, through the Canadian International Development Agency, for international agencies such as the International Red Cross and the United Nations High Commission for Refugees that deliver refugee emergency aid and development. There is no distinction to be made between Canadian and international emergency aid and development practices. 4. Convention refugee status determination also takes place at Canadian embassies and consulates abroad prior to selection. 5. With few notable exceptions, until the mid-1980s Foucault's writings have been used as a supplement for a variety of theoretical (e.g., Giddens, 1984) and political (e.g., Fraser, 1989) projects to adjust or refine a particular theoretical or historical claim within a larger endeavour (see also Hunter, 1996). Accompanying this has been a clamouring in the background suggesting that for these writings to be of any (further) use, they must be 'positioned' (cf. Leonard, 1990: 3; Garland, 1990: 173)- it must be determined beforehand whether they are, for instance, structuralist, anti-structuralist, neo-conservative, Marxist, feminist, anti-racist, anarchist, or whatever. I am not so sure about this suggestion. As Dean (1994d: 36) argues, perhaps this lack of positioning is less an 'Achilles' Heel1 and more a strength. It is for this reason that I seek to avoid grounding this thesis in a system of' values (Dean, 1994d: 36), while simultaneously abstaining from the old liberal view that knowledge and values are somehow separate. 6. Burchell (1993: 276-77) suggests this problematization stems from a specific experience. He writes: There often appears to be a motivating experience for adopting the kind of approach that Foucault called the 'history of the present' which seems to involve the experience of not being a citizen of the community or republic of thought and action in which one is, 13 nevertheless, unavoidably implicated or involved. It is an experience of being in a goldfish bowl in which one is obliged to live but in which it seems impossible to live, that is, to think and act... The experience is not at all just a matter of holding a different opinion than everyone else, but of finding oneself not knowing what or how to think. . . This experience. . . calls. . . for a kind of criticism by which our view from inside of our goldfish bowl is made to appear as not more than the historically contingent effect of a kind of selective determination by a particular outside of practices. Foucault's work provides us with a number of splendid examples of 'ways out' in relation to certain features of our goldfish bowl. His genealogies work in this way by revealing to us the (often quite recent) inventedness of our world. His descriptions enable us to discern the broken lines of the irregular contours of our goldfish bowl, of our present, taking shape in all their necessarily contingent exteriority. 7. Precisely because of these connections, problematizing the Canadian refugee regime risks eliciting the same reaction as sending a slice of a neighbour's apple pie to the local health department for chemical analysis, or volunteering a family pet for medical research experiments (Waldron, 1987: 1)- there is little doubt that some readers will find it distasteful. 8. If there is dissatisfaction with relating to human beings as refugees or refugee claimants requiring regulation, one might note Hunt's (1993: 316) passing comment that "if objects of regulation can be created, then they can similarly be dismantled and abandoned." 9. As seen in chapter three, this question is in part an historical one, as regimes are constituted by historical elements. The distinction it makes is analytical, as regimes do not exist separate from their governance. 14 CHAPTER ONE: PREVIOUS RESEARCH Though there are no governmentality studies that have focused on international or Canadian policy and practices directed at refugees, there is other previous research that has. This chapter critically reviews this research with an eye to identifying similarities with and deficiencies in relation to the approach adopted for this thesis. In seeking to describe and assess this previous research on refugee policy and practices, two basic types are immediately discovered: one that approaches these subjects as an international concern, and the other that sees them as a concern of particular Western nations. There is a large body of research that takes a decidedly international approach, and distinct literatures, themselves internally fragmented, that pertain to the policy and practices of single nations (e.g., research on British refugee policy and practices). The international literature is dealt with first. Following this, attention is paid to research that relates directly to Canadian refugee policy and practices, thereby excluding that which relates to similar policy and practices of other nations, such as Germany, the United States, Australia, France, or Britain. I have chosen in both instances not to limit discussion to previous research that proclaims itself sociological, as such research -particularly regarding refugee selection and determination- can be characterized as meager at best. With these qualifiers in mind, as well as by limiting discussion to the English language literature, in what follows I describe and assess previous research on refugee policy and 15 practices. THE INTERNATIONAL LITERATURE The international literature focuses in some way on the international refugee regime that began to emerge in Western Europe following the First World War and that has since become extended across the world. It includes a burgeoning body of research on international refugee law (cf. Hathaway, 1991, 1992; Goodwyn-Gil, 1983, 1990; Grahl-Madsen, 1990) that explores the legal ramifications of Conventions and agreements about refugees in various world regions. It also consists in historical studies that centre on the development of relevant international organizations and their practices (cf. Woodbridge, 1950; Holborn, 1956; 1975; Marrus, 1985). Other studies that adopt an international approach, some of which are historical (Gordenker, 1987; Sjoberg, 1991; 7 Skran, 1995), draw on international relations theory (Loescher, 1993; 1988; Loescher and Scanlan, 1985; Loescher and Monahan, 1989; Weiner, 1993). These works tend to view refugees as weapons used by nations or alliances of nations, here conceived as rational actors in an international field of politics, to embarrass, destabilize, or otherwise weaken enemy nations or alliances. There is also a body of research making up part of the international literature that discusses refugee policy and practices in the context of international migration (cf. Zolberg et al., 1983; 1989; Kunz, 1969; 1973; 1981). Some studies of this kind pledge their allegiance to social geography, but also to the 'sociology of involuntary migration' (Harrell-Bond, 1988). The 16 latter is part of the broader sociological study of migration (cf. Richmond, 1988; 1993; 1994), a fragmented sub-field compared with other areas of sociological interest (Marshall, 1994: 329). While some studies remain mired in an outmoded 'push-pull' perspective that directs attention to individual motivations and interprets refugee policies as an intervening 'variable' (cf. Lee, 1966) in an international refugee migration process (e.g., Kunz, 1981), others, such as Zolberg (1983; 1989), are historical in character and point to the formation and.reproduction of nations and resulting conflict between them as an important source of refugee migrations. Neo-Marxist research on international migration also makes up part of the international literature. Here migrants are conceived as a reserve army of docile, 'unfree' labour (cf. Cohen, 1987; Miles, 1987). 'Unfree' (the neo-Marxist equivalent to 'involuntary') refers to a subordinate status under which workers cannot sell their labour-power as a commodity in the market of their own free-will, but are instead legally or physically forced to do so. The capitalist mode of production is said to entail "free and unfree labour regimes" not, as Marx had argued, free labour exclusively (Cohen, 1987: 25). Unfree labour takes several forms, including slavery, serfdom, prison and convict labour and, more recently, migrant contract labour. Large blocs of human beings are constituted as unfree forms of labour to be exploited either in the core or the periphery of regional political economies (e.g., North America) (Cohen, 1987).8 Policies of Western nations operating in the core are seen making possible the recruitment of 17 unfree labour; the legitimation of this labour force; and the management of the boundaries of core zones (Cohen, 1987). In such discussions of the role of states in international migration, refugee policies and practices are often mentioned, but they are then either awkwardly dismissed as outside the area of interest (e.g., Satzewich, 1991b: 8; Cohen, 1987: 31; Sassen-Koob, 1981: 68n; see Richmond, 1988: 9) or dubiously claimed to be indistinguishable from other state policies and corresponding forms of migrant labour (e.g., Castles and Kosack, 1985: 11) (cf. Miles and Satzewich, 1990). With the former tendency, refugee policies and refugees are viewed as inherently political rather than economic phenomena and therefore an exception to the rules regarding unfree labour. With the latter, refugee policies are reduced to the logic of broader immigration policies that merely respond to the changing needs of a regional or global political economy by creating and admitting or excluding a docile, unfree labour source as required. A recent detailed historical study by Kay and Miles (1992; 1988), entitled Refugees or Migrant Workers?, situates itself in this structuralist neo-Marxist problematic that uses the concept of unfree labour. It represents a rare exception to both tendencies above in that it neither dismisses refugee policies nor reduces them to the logic of immigration. In rich, historical detail the authors describe the implementation of the British state's temporary programme, the European Volunteer Worker (EVW) scheme that targeted the recruitment of large numbers of persons from 18 European Displaced Persons camps from 1946 to 1951. The authors show how the British state, through the introduction of this programme, brought about the migration of docile, unfree labour to counter an extreme labour shortage Britain was experiencing in the years immediately following World War Two. The authors also describe the resistance of those targeted by the scheme after their arrival in Britain (Kay and Miles, 1992: 95-119). The authors conclude that those selected from Displaced Person camps in this instance were neither obvious migrant contract labourers nor refugees and, therefore, that the EVW migration was of a "hybrid nature" (Kay and Miles, 1992: 6). The authors point to three kinds of political process that can form such a surplus population: conflict over national borders; change in a national government; and "the dynamic process of nationalisation" (Kay and Miles, 1992: 185). Kay and Miles (1988: 186) conclude: if the surplus population was created in the political sphere, its insertion into British capitalism was dictated primarily by an economic imperative. They argue, then, that "surplus populations can also be constituted by predominantly, but not exclusively, political processes" (Kay and Miles, 1992: 185; see also 1988: 231). The process, in this instance, is the division between communist and capitalist nations in Europe (Kay and Miles, 1992: 6). But all of this raises a key question: if the relevant surplus population was created mostly within political processes, why attempt, as these authors do, to reduce the EVW programme or policy to the telos of capitalism? While those occupying refugee 19 categories have undoubtedly served as a source of labour of one kind or other within Canada's capitalist economy after their selection and resettlement, it is very questionable whether reducing the practices directed at them, and how they have been changing, to an unfree labour regime of capitalism could render them intelligible. For one thing, as is seen in chapter five, since the 1970s refugees are clearly imagined within contemporary Canadian refugee selection and resettlement programmes not only as q a source of labour, but as potential liberal citizens. In such schemes they are imagined not merely becoming submissive, docile workers, but acquiring the subtle skills and knowledge required to exercise their citizenship in a liberal manner. While this thesis does not seek to investigate particular refugee policy and practices to construct or refine universal theories of international law, relations, or migration, choosing to deal with them as a contingent national governmental regime instead, the research above is helpful nevertheless. The work in international relations, in particular, points to the existence of an international field of politics that, while not of immediate interest here, is necessarily discussed in chapter four and occasionally referred to in later chapters. Attempting to accomplish the latter by relying exclusively on international relations, legal, or migration theory, however, would be futile, as one kind of practice at issue here -resettlement- occurs within rather than between nations. The perspective of Kay and Miles (1992) is also helpful in terms of its emphasis on politics, power, 20 and resistance, which are also concepts and themes central to governmentality studies. For this reason it will be useful and instructive to compare the latter two approaches in the next chapter, to take a closer look at their similarities and differences before proceeding to investigate Canadian refugee policy and practices. THE CANADIAN LITERATURE Previous Canadian research can be divided for analytical purposes into historical, legal, and state-centred studies. These and some of their findings are discussed below. Historical Studies Some studies have been decidedly historical in their orientation (e.g., Kapriellan-Churchill, 1986; 1990; 1994; Bassler, 1992; Luciuk, 1986; Abella and Troper, 1983). Abella and Troper (1983) -undoubtedly the most widely read study of this type to date- examine Canadian practices as they pertained to Jews attempting to flee Nazi persecution in Europe from 1933 to 1948. The authors describe in detail how Department of Immigration officials systematically prevented Jews from entering Canada. They conclude that exclusionary practices in this instance can be better explained by racist attitudes among Canadian state agents than changing domestic economic conditions, namely the onset of the Great Depression. In a very similar study, Bassler (1992) focuses on Newfoundland's restrictive practices in relation to Jews between 1906 and 1949 when it became a province of Canada. Kapriellian-Churchill (1986; 1990; 1994) similarly documents the reluctant 21 Canadian response to Armenian women refugees from 1919 to 1930 following the breakup of the Ottoman Empire, and Luciuk (1986) the response to Ukrainians immediately following World War Two. These works tend to concentrate on refugee selection in relation to a particular nationality or ethnic group over a circumscribed period. These studies describe past Canadian practices in detail. They highlight state agents' perceptions of refugees' race, political ideology or gender as the most important determinants of policy. The main point is usually the same in each- Canadian policy was not what it ought to have been for a particular group, to the extreme detriment (ie. death or continued persecution) of the excluded refugees. Unfortunately, these studies focus exclusively on the historical context surrounding selection of a particular group and offer little detail about resettlement. While these studies are historical, they are not informed by a coherent conceptual approach. Furthermore, they say little about power, politics or resistance explicitly, all of which are concerns in studies in governmentality. Like several studies below, they examine Canadian refugee policy and practices in ways that take the present institutional framework for granted, rather than call it into doubt. Legal Studies There are a variety of studies that have adopted a broad legal approach to Canadian refugee policy and practices (cf. Matas and Simon, 1989; Mandel, 1994; Barsky, 1994; Macklin, 1995). Matas and Simon (1989), for example, examine Canadian refugee policy and 22 practices from 1978 to 1989. The authors highlight the Department of Immigration officials' apparent racist practices and mistakes, primarily concerning the determination of refugees. They contend that Canadian state agents have denied access to legal refugee status by employing 'techniques of denial1: using a narrow legal definition of a refugee that excludes those fleeing 'generalized violence'; implementing arbitrary legal procedures of refugee determination; deterring those who might reach Canadian borders through visa restrictions abroad; and making diplomatic asylum and selection unavailable. A major conclusion is that while the Canadian state accepts the ideal of refugee protection, it denies it in practice. In a more recent study, Barsky (1994) attempts to make sense of refugee determination hearings in Canada by reference to ' the othering process'. Drawing primarily from the discourse theory of Mikhail Bakhtin, Barsky traces in considerable detail the transformation, over three distinct stages of the hearing process, of a (whole) human being into a constructed 'other' (ie. a Canadian refugee claimant). This is accomplished by analyzing hearing transcripts from two Chilean refugee claimants who arrived in Canada in 1987. In so doing he demonstrates the utility of discourse theory for legal studies and suggests several specific reforms of the refugee determination process. Mandel (1994) studies Canadian refugee determination policy and practices since the enactment of the Charter as one of several examples of the 'legalization of polities'. That the Singh 23 decision in 1985 resulted in rights for refugee claimants, an oral i hearing and other features of the new determination process, is interpreted as "an essential and expensive element of the sophisticated fortifications that Canada and other wealthy nations have thrown up around themselves" (Mandel, 1994: 257). He argues that in this instance the effect of the Charter has been to "legitimate repression by reconciling it with our ideals of humanity" (Mandel, 1994: 257). In other words, law is merely ideological cover for repressive practices directed at those existing outside affluent nations such as Canada. This emphasis on the maintenance of legitimacy is also seen in some state-centred studies below. Not surprisingly, these legal studies focus almost entirely on determination, as this is the area of the regime where law has become intimately involved. Except for Barsky (1994) and Mandel (1994), they do not proceed with an explicit conceptual approach, or make reference to power, politics or resistance. Barsky's attention to discourse, however, does resonate with studies in gove rnmenta1i ty. State-centred Studies There are also studies that have sought to make sense of Canadian refugee policy and practices, usually over a specific period, by referring to the objective interests or workings of the Canadian state. By the late 1970s Dirks (1977) was undoubtedly the most comprehensive study of this or any kind. It traces Canadian refugee policy and practices from 1789 -the period of the Loyalist 24 migration- to the present and attempts to see why Canadians sometimes actively welcomed and admitted refugees and, at other times, were indifferent to their situations. Dirks interprets refugee policy, mostly in the decade preceding 1977, as the result of alliances between government departments and the efforts of individuals within the Canadian state, though he grants some importance to prevailing domestic economic conditions. In a more recent but similar work, Dirks (1995) describes what has occurred during the 1980s in relation to refugee policy and practices as part of a larger immigration policy. In a similar way, Hawkins (1989; see also 1991) researches Canadian refugee policy from 1945 to 1986, with an emphasis on the period from 1972 onward, and as part of a broader immigration policy. She attempts to make sense of refugee policy by reference to the endeavours of individual political authorities and Department of Immigration officials. Though remarkably detailed, both Dirks (1977; 1995) and Hawkins (1989) tend to depict state agents as largely unconstrained actors shaping progressive policies as they see fit. Canadian refugee practices, and how they have changed, are to be understood by reference to the foresight and failings of a succession of Ministers and high-ranking federal state bureaucrats. Perhaps as a result, these studies are devoid of explicit references to power or resistance. Howard (1980) examines Canadian refugee policy and practices during the period 1945 to 1979, though with an emphasis on the two years before 1977. She attempts to assess the relative influence 25 of humanitarian, racial, economic, and ideological factors on both determination and selection practices. She concludes that, before 1945, perceived race and domestic economic conditions were the most important factors. She suggests, however, that the relative weight of state agents' perceptions of refugees' race and political ideology changed in the late 1970s, with perceived race becoming less and perceived political ideology becoming more important influences. Howard also suggests that since 1945 the minor ideological bias in selection and determination has been against socialist refugees fleeing right-wing governments and on behalf of right-wing refugees fleeing socialist governments. Dewitt and Kirton (1983) interpret refugee policy as part of broader immigration and foreign policies. The authors examine refugee policy in the context of six crises abroad: Displaced Persons of Europe, Hungary, Czechoslovakia, Uganda, Chile, and Indochina. The authors argue that since Canada was apparently not responding to the dictates of powerful nations such as the United States, and Canadian political and economic interests were part of the decision to intervene in each crisis, refugee policy can be understood less as an instance of racism, as seen in the selection of White European refugees, and more a reflection of the neo-realist pursuit of national interests. They also conclude that taken together these cases suggest humanitarian ism was an important determining factor. Creese (1991) examines change in Canadian refugee policy and practices during the 1980s with an emphasis on the period 26 immediately before passage of new refugee legislation in 1988. She interprets this new legislation as the culmination of the Canadian state's efforts, begun during the arrival of large numbers of refugees who had not been pre-screened and selected as such, to reduce the levels and types (based on their skill and political ideology) of refugees entering Canada. Like other authors above, Creese suggests the introduction of this refugee legislation was also consistent with ethnic and racial politics internal to Canada, and with its external political economic relations with other nations. As part of this study, Creese conducted a content analysis of newspaper articles about refugees published during a crucial period prior to 1988. She found that most articles advanced an image depicting refugees as criminals and a stress on Canadian society. For Creese the creation of this negative image for the public supported the state's goal of passing the new legislation. She therefore views the Canadian state as actively attempting to shape civil society in a way that allows its interests to be realized. In these ways Creese attends -to use a less than eloquent phrase- to the 'hows of rule', that is, how rule is possible in a particular domain. Though approached in a different manner, this is also a theme in governmentality studies. Like Barsky (1994), Creese's attention in this study to discourse, as seen in the content analysis, also resonates with the approach adopted for this thesis. Unfortunately, Creese's critical conceptual approach remains largely implicit. Whitaker (1987) explores refugee policy as part of a broader 27 immigration policy from 1945 to 1987. Like several authors above, Whitaker suggests political and ideological factors are more important than racist attitudes among state agents in influencing Canadian refugee policy and practices. He suggests there is a systematic bias, a 'double standard', in favour of refugees fleeing socialist regimes and against refugees fleeing right-wing governments. Besides the perceived political ideology and race of refugees, he also points to national security and domestic economic need as factors affecting Canadian refugee policy during this period. By conceiving of the Canadian state as a nation, Whitaker contends that changes in refugee policy, as seen in the new legislation, reflect patterns in political Cold War arrangements among Western nations. Unfortunately, like Dewitt and Kirton (1983), Whitaker tends to reduce refugee policy and practices to the logic of Canada's immigration policies and, in turn, to external political relations with other Western nations. Simmons and Keohane (1992) is the final state-centred study to be discussed. The authors examine the making of Canadian refugee policy as an example of immigration policy between 1987 and 1988 and before passage of the new legislation mentioned above. The authors document, in detail, how state agents systematically shaped and limited debate over this legislation during consultation procedures with various non-state actors. In so doing, and unlike Hawkins (1989) and Dirks (1977), Simmons and Keohane do not presume Department of Immigration officials to be unconstrained actors shaping policy at will. Instead they see them as considerably less 28 powerful agents who have to negotiate with and persuade groups in civil society to achieve the state's objective goals. In part by drawing on the work of neo-Marxists such as Offe and Ronge (1982) and Habermas (1975), the authors adopt an explicit approach that imagines the state as having three broad, objective goals or interests: (1) sustaining economic well-being for itself and society; (2) pursuing a rational-bureaucratic agenda; and (3) attempting to maintain legitimacy, which is dependent on (1) and (2). They assert the "state's hegemony... must always be worked out within a larger cultural context over which it has only limited control" (Simmons and Keohane, 1992: 427). They argue the state's goal, as it pertains to refugee policy, as part of a larger immigration policy, is to "exercise legitimate power and control" (Simmons and Keohane, 1992: 428). More specifically the Canadian state . . . gears immigration policy toward the smooth running of the economy, the mediation of major social conflicts, and the preemption of crises and challenges from oppositional social movements. Policy over recent years may thus be interpreted as the operation of these strategies for state legitimacy in a changing national and international social context (Simmons and Keohane, 1992: 428). This study is more explicit about the conceptual approach it adopts than most work above and is important in the way it highlights how various non-state authorities are brought into alignment with the interests of the state in one context. In this respect, like Creese (1991), it also explores the 'hows of rule1. Unlike Kay and Miles (1992), this study avoids interpreting refugee policy and practices simply as the creation of an unfree labour source to 29 satisfy economic needs, in that it also points to the state's effort to maintain legitimacy as a separate (though dependent) interest. These interests the authors describe, however, are static rather than historical- they are assumed to remain the same over time. Though this study's reference to non-state authorities is important, the 'oppositional social movements' phrase it employs raises a question- about whether these agents and authorities are always 'oppositional1, or whether they are occasionally brought into alignment with the efforts of state and other agencies and authorities to achieve certain goals. What is to be made of all these Canadian studies? Though none adopt a governmentality approach, they nevertheless make an important contribution. Except those which rely heavily on humanitarianism as an explanation, by studying refugee policy and practices as a legal or historical phenomenon, as the will of the state or, like their international counterparts, as fitting the logic of capitalism, these studies together suggest that these phenomena are not demanding of an exceptional conceptional approach. This is quite contrary to popular accounts of refugee policy and practices, in which they are depicted as purely humanitarian and devoid of power, politics, and resistance. The historical studies in particular show that aspects of Canadian refugee policy and practices have a history, that they have changed over time. Other kinds of studies above point to the significance of discourse in understanding how rule is accomplished in this context. Finally, all studies mentioned above are important in 30 that they themselves serve as data to be drawn on to explore the regime and point to other relevant sources -both documentary and human- that can be consulted in carrying out the present research endeavour. Despite their contribution and importance, however, there are significant deficiencies evident in the studies above. First, several conflate Canada's refugee and immigration policies and practices. Though there have been exchanges between them, as seen in chapter five, the distinction between them has become a discursive fact. Second, most studies above ignore the relevance of non-state actors in relation to refugee policy and practices. While all studies above address either determination or selection (though the latter only as it pertains to a particular nationality or ethnic group), another type of practice, resettlement, is rarely mentioned in these works. Third, virtually all studies above depict the refugee as essential, as though the refugee was other than a contingent and historically constructed object/identity. Finally, some studies above are reductionist. They describe Canadian refugee policy and practices as necessarily about racism (Matas and Simon, 1989; Abella and Troper, 1983; see also Richmond, 1994); the inner workings of the Canadian state; external relations with other nations; or capitalism. SUMMARY AND CONCLUSIONS Previous research on refugee policy and practices is both international in its orientation and specific to particular Western nations such as Canada. The relevant international literature 31 includes research on international law, relations, and migration. The Canadian literature includes historical studies, legal studies, and those that place the state at the centre of analysis. Given the deficiencies of previous research, to make sense of Canadian refugee policy and practices, and how they have been changing, a governmentality approach must be adopted, one that places them "within a wider discursive field in which conceptions of the proper ends and means of government are articulated" (Miller and Rose, 1990b: 5). In seeking to elaborate a governmentality approach in the next chapter, studies that focus on the state, because of their attention to power, politics and resistance, will be an excellent place to begin. Some previous research above has used a metaphor of a door (e.g., Adelman, 1991: 176; Knowles, 1992; Matas and Simon, 1989; Canada, 1988a; Malarek, 1987; Zucker and Zucker, 1987; Loescher and Scanlan, 1986) to refer to refugee policy and practices. This thesis is concerned not so much with why this metaphorical door opened or closed and even less with why it should be opened or closed; it is preoccupied with the door itself. It seeks to take it off its metaphorical hinges to find out where it came from, what it is made of, and how it works. To accomplish this a different approach is required, one that avoids placing the state at the centre of analysis. 32 1. It should be mentioned at this point that though I discuss previous research here primarily to determine what is collectively lacking and to begin to forge an appropriate critical approach, each of the studies mentioned is a form of knowledge referred to in chapter six. 2. Sjoberg (1991: 18), cited in chapter four, similarly divides the scholarly literature. 3. Refugee aid and development practices of particular nations are rarely included here. Instead they are subjects taken up in policy research carried out in relation to the international regime, studies that, like their Canadian counterparts, I have chosen not to discuss. 4. A rationale for choosing the Canadian context to explore, rather than that of another Western nation, is that it is convenient to do so- this is where I happen to live and most data about Western contexts available here centre on the Canadian one. As a graduate student I simply lack the resources to access previous research, documentary materials, and authorities and agencies organized in relation to other Western nations. 5. See, for example, Zucker and Zucker (1987) and Loescher and Scanlan (1986) on the United States and Cohen (1988) on Britain. Canada's refugee policy and practices have been systematically compared to Australia's (see Hawkins (1989; 1991) and Adelman et al. (1994) but also see Robinson (1993) on Canada and Britain, Hardy (1994) on Canada, Britain and Denmark, Adelman (1991) on Canada and the United States, and Lanphier (1987) on Canada, France and the United States). Most of these are theoretically impoverished policy studies that I have chosen not to discuss here. Refugee policies associated with 'settler societies' (Anderson, 1991: 19) such as Canada undoubtedly differ from those of Western European nations in that the former tend to include annual refugee selection in their repertoire. On the other hand, like Canada, most Western European nations have refugee determination systems and provide regular funding to international organizations (e.g., the UNHCR) for emergency refugee aid and development. Some neo-Marxists, such as Gaucher (1987), claim treating the Canadian 'social formation' as similar to European social formations conflicts with Marx's call for specificity. This may be so, but it remains difficult to see where this specificity ought to begin or end, or why the conceptual borders of a 'social formation' (ie. a nation-state) ought to be singled out as defining the parameters of a study if, following this logic, as Trotsky pointed out, 'socialism in One Country' is a contradiction (Magnusson and Walker, 1988: 48). Perhaps the nation-state is best seen as "one, albeit special, agency among others" (bean, 1994c: 165; see also Hirst and Thompson, 1995). It was nevertheless with some trepidation that I chose to recognize the nation-state form as a defining parameter of this thesis (an issue separate from which 33 Western nation1s refugee regime to examine). Whether Canada can be assumed to elicit the same general patterns as other Western nations is not resolvable. This thesis leans more to the idiographic pole that emphasizes uniqueness and that proceeds from the assumption that a given nation has particular historical features, than to the nomothetic pole that assumes a given nation's historical trajectory is merely an instance of larger regularities (Marshall, 1994: 233). 6. That which has been carried out is fragmented and proceeds from varied assumptions. The fact there is a profound lack of acknowledgement of previous research within the text of any given study is evidence of this. Harrell-Bond (1988: 1) posits several mundane reasons for its traditional neglect within sociology. Another reason -especially concerning determination and selection practices- may be that two of three writers traditionally viewed as the founders of sociology, Emile Durkheim and Karl Marx, said little about the modern state as a nation-state involved in political relations with other nation-states (Giddens, 1987: 22). The practices at issue here come into view when placed in such an interpretive frame. 7. Conventional and well-known works in international relations and the political economy of international relations (see Keohane and Nye, 1977; Waltz, 1979; Krasner, 1978; Keohane, 1984) have ignored refugee policies and practices. 8. Cohen (1987) departs from world-systems theory (cf. Wallerstein, 1979) here by arguing that 'unfree' forms of labour exist both in zones (often previously colonized non-Western spaces) on the periphery of regional political economies and in core areas of capitalist development- typically metropolitan areas of the modern West. 9. Immigration programmes since 1950 have similarly imagined immigrants as potential citizens (see, for example, Vernant, 1953). 10. These three categories are not mutually exclusive, of course. For instance, there are works that I have called state-centred, such as Dirks (1977), that are historical in their orientation. They are not, however, carried out by historians per se and seem to better fit the other two categories. A work that does not fit any of these categories is Harp (1981), a study of 'Project 4000', an ad hoc Canadian refugee advocacy group that arose in conjunction with the resettlement effort directed at tens of thousands of Indochinese refugees in 1979 and 1980. A more sophisticated social movement study of this ilk (see also Matas and Simon, 1989: 216-30; Stastny and Tyrnauer, 1993), though carried out in an American context, is Coutin (1993), an ethnographic exploration of the sanctuary movement in the United States during the 1980s. Other similar American studies, though directed more to a popular than an academic readership, are Golden and McConnell (1986), Tomsho 34 (1987), and Bau (1985). There are also numerous Canadian policy studies that evaluate refugee policy by uncovering factors that led to its success in achieving its objectives or that reveal the miscalculations that resulted in its failure, but these are more a source of data for later chapters than works worth considering here. 35 CHAPTER TWO: APPROACHING THE CANADIAN REFUGEE REGIME In the previous chapter it was suggested that the conceptual approach adopted for this thesis would take its lead from studies that focus on the state. In this chapter the scope and focus of this conceptual approach and the reasons for adopting it need clarification. This is best accomplished by briefly discussing first, neo-Marxist state theory and the historical sociology of the state, and then, a critical Foucauldian alternative. In so doing, the first part of this chapter also attempts to show that despite efforts to draw rigid distinctions among these kinds of approaches, there is, in their assumptions and themes, overlap among them. NEO-MARXIST APPROACHES In the late 1960s the state as an object of contemporary social theorizing came into view. This overcame pluralist conceptions of the state predominant in sociology since World War Two (Brodie, 1990: 62). There has since been a large volume of research and writing produced about the role of states in the West. Much of this contemporary work has stemmed directly from the writings of neo-Marxists where attention has been paid to, among other issues, the definition of the state and how state power is related to the mode of production (Dean, 1994a: 142). Several competing conceptions and lines of argument have emerged. One of these is instrumentalist, a conception of the state associated with Marx himself that suggests the state is "a committee for managing the affairs of the whole bourgeoisie" (Marx 36 and Engels, 1967: 82). In contemporary work this conception is associated with the writings of Miliband (1969). The state is seen to operate as a passive instrument of capital, at its behest (Ratner, et al., 1987: 85). Such a conception highlights the overlap of the economic and the political, the class nature of the power of the state. This conception is now widely seen as outmoded, reductionist, economic determinist, and lacking specificity. Another basic line of argument is referred to as structuralist and is most often associated with the writings of Poulantzas (1973). The structuralist argument suggests there is an objective relation between capital and the state, whereby the latter operates in the long term interests of the former. This occurs irrespectively of proponents of capital being actively involved in state action. To maintain the political economic status quo, the state acts on behalf of capital, rather than at its behest. State policies are reducible 'in the last instance' to the struggle between labour and capital, and serve two functions: legitimation and accumulation. The state is seen to have 'relative autonomy'. For Poulantzas, the state is relatively autonomous from classes. For other structuralists, like Offe and Ronge (1982), the state is relatively autonomous from capital accumulation processes. The concept of relative autonomy is a much debated one, and there are several structuralist positions represented within this debate. One way to critically approach Canadian refugee policy and practices, then, would be to adopt a structuralist neo-Marxist 37 position (see Poulantzas, 1973; Habermas, 1975; Offe and Ronge, 1982; Panitch, 1977; Albo and Jenson, 1989). As suggested in the previous chapter, in a structuralist approach refugees would be understood as a form of docile, unfree labour and refugee policies and practices as ultimately responding to the needs of capital by creating and controlling such a labour source (cf. Sassen-Koob, 1981; Miles, 1987; Cohen, 1987). Here refugee policy is perceived to be completely consistent with the logic of immigration policy. In recent years, structuralist approaches have come under fire and are also now widely seen as outmoded and lacking specificity (Mahon, 1991). A neo-Marxist line of argument garnering considerable attention of late is associated with the writings of Antonio Gramsci. This approach has been called "class-conflict" (cf. Ratner et al., 1987: 94). Here the state "has a central function in developing a moral (class-based) consensus that organizes social, civil, and intellectual life around the structural tendencies set by the economy" (Ratner et al., 1987: 94). This conception is allied to the concept of 'hegemony', "a historically specific organization of consent" (Carroll, 1992: 9). Hegemony refers to the notion that, in the modern West, ruling class ideas become universal ideas. Rule is accomplished not so much by coercion, as by popular consent to authority (or authorities) (Hall, 1988). As Carroll (1992: 9) writes: In Gramsci's formulation, power is both centralized in the coercive apparatuses of the state and diffused across other institutional sites such as the Church, the family and the school. 38 In this approach the capitalist economy and objective class interests do not decide the direction of state policies and practices (Hall, 1988: 54). The character of the state is not deducible from a general theory. Instead, the state is a location where hegemony is struggled for and can be understood in its specific, historical context (Ratner et al., 1987: 98). The state and civil society are shaped, not through monopoly of the means of repressive violence, but through (non-violent) ideas. In such an approach attention is paid to the role of ideology (ie. discourse), understood here as constitutive, in assembling alliances among varied authorities in civil society (Hall, 1988: 53). In discussing this aspect of Gramscian analysis, one of its leading proponents, Hall (1988: 51), writes: ...we do not in any way refuse the advances made by the development of the analysis of the discursive. No social practice exists outside of the domains of the semiotic-the practice and production of meaning. This necessarily modifies, in a radical way, the traditional material/ideal, base/superstructure dichotomies of classical theories of ideology (emphasis in original). From the above, a set of similarities between a neo-Marxist class-conflict approach and the approach adopted for this thesis can be seen. Governmentality studies also proceed with a focus on the specificity of national contexts (Barry, Osborne, and Rose, 1996: 11); with the assumption that rule is accomplished rather than given and dependent on non-state authorities; and with a conception of discourse as constitutive. 39 STATE-CENTRED APPROACHES In the late 1970s and early 1980s there was a move toward 'bringing the state back into1 analysis (Jessop, 1990: 278). The varied contemporary works in this area (cf. Poggi, 1978: 1990; Katzenstein, 1978; Block, 1980; Mann, 1988; Giddens, 1985; 1990a) 7 have been characterized as anti-Marxist (Jessop, 1990: 278), neo-Weberian, and neo-institutional (Brodie, 1990: 66). Skocpol's (1979) study of the French, Russian, and Chinese Revolutions is an important text associated with this intellectual movement (Mahon, 1991: 123n). Skocpol directs her critique toward neo-Marxist theorists of the state, criticizing their structuralist assumption that state agents automatically possess an awareness of the long term interests of capital (Mahon, 1991: 123). Such studies approach the state as an independent source of power, that is, independent from capital and forces in civil society (Jessop, 1990: 278). The assumption is that "[t]he state is a force in its own right and does not simply reflect the dynamic of the economy and/or civil society" (Jessop, 1990: 279). It is assumed state agents have their own interests and that there is a capacity for autonomous (not relatively autonomous) state action. Skocpol (1985: 20-21), for instance, writes: ...states conceived as organizations claiming control over territories and people may formulate and pursue goals that are not simply reflective of the demands or interests of social groups, classes or society. A theme in Skocpol's (1979) study above was the state as a nation involved in external political relations, including war, with other nations. This external set of factors is seen to affect 40 the internal development of the nation and itself leading to some independence for state agents from domestic capital or other forces. This concern with the state as a nation in the 1980s gradually led to "a more profound questioning of issues of power, domination, and rule" (Dean, 1994a: 145). Inquiry was increasingly directed toward how the state obtains independent or autonomous power, and how this power is brought to bear on civil society. Regarding the first issue, for example, Mann (1988: x) writes that "states are essentially centralized and territorial, and that autonomous state power derives from these twin characteristics." Mann (1986; 1988) attempts to account for state power in terms of the mechanisms of state rule (Dean, 1994a: 146). He distinguishes 'despotic power' -characterized by the king shouting 'off with their heads' and having that order obeyed- from the 1 infrastructural power' of the state that seeks to penetrate the far reaches of civil society and the lives of citizens. Mann (1988) directs attention to the logistical techniques from which this infrastructural power stems. These techniques include literacy, rapid communication and transportation, taxation, information collection, and welfare provisions (Dean, 1994a: 146). Giddens (1985; 1990a) work on the nation-state includes similar themes in addressing this form of power, here called 0 'administrative power'. Giddens points to the centrality of practices of communication, transportation, information-gathering, and discipline. Dominant in his analysis is the concept of 'power containers', which are "circumscribed arenas for the generation of 41 administrative power" (Giddens, 1985: 13). Examples include the prison, factory, and school, where what Giddens calls allocative and authoritative resources are concentrated. According to Giddens, in this sense the nation-state is itself a power container, albeit a special one. The exercise of administrative power by state agents, then, consists of the surveillance of the population within the borders of the nation-state, but also of external political conditions and relations. Surveillance, put simply, is either direct supervision of human beings or indirect information control and is seen to flow from administrative power (Giddens, 1990a: 59). In Giddens's view, communication and transportation technologies are resources for the concentration of administrative power in the nation-state, though such power is apparently also generated in much smaller organizations (or institutions) (cf. Giddens, 1990a: 58). In keeping with a state-centred approach Giddens (1990a: 57) writes: The characteristics of the nation-state... must be explained and analyzed separately from discussion of the nature of either capitalism or industrialism. The administrative system of the capitalist state, and of modern states in general, has to be interpreted in terms of the coordinated control over delimited territorial arenas which it achieves. For Giddens (1985; 1990a), administrative power, the surveillance that flows from it, and the nation-state itself cannot be reduced to the workings and history of capitalism (Giddens, 1990a: 62). This departs from neo-Marxist approaches that see warfare and political relations among nations as elements of class struggle (Dandeker, 1990). 42 So in this and other state-centred research attention has been paid to mechanisms, techniques, and the logistics of rule, particularly as they comprise state power. It is in this attention that a set of similarities can be seen between governmental ity and state-centred approaches. There is considerable overlap, for example, between what Giddens identifies as 'power-containers' (e.g., the school) and what Rose and Miller (1992), advocates of a governmentality approach, understand as 'technologies'. There is a risk of overstating the difference between neo-Marxist and state-centred approaches briefly described above. As Jessop (1990: 91) writes: [A]lthough adherents of the movement to 'bring the state back in' sometimes suggest that their approach is both novel and somehow anti-Marxist in its implications, many so-called statist themes can also be found within the Marxist tradition. This is especially true when one moves away from extreme positions on state autonomy (Jessop, 1990: 93). For example, both state-centred and neo-Marxist theorists suggest the state's need for tax revenue to finance its own projects and activities leads to policies that support capital accumulation which is, ultimately, the source of such revenues (Jessop, 1990: 93). An instance of this can be seen in the previous chapter in the work of Simmons and Keohane (1992). It will be recalled that while claiming to depart from neo-Marxist structuralist approaches, the authors of this particular study asserted the state has three general interests: (1) sustaining economic well-being for itself and society; (2) pursuing a rational-bureaucratic agenda; and (3) maintaining 43 legitimacy. Given that the state's economic well-being can be achieved in part through immigration policy (by creating and legitimating a docile labour source on behalf of capital) and by confiscating tax revenues from the resulting accumulation process, does this study embody an approach that is state-centred or neo-Marxist? This lack of distinction is evident in another Canadian example. Ratner et al. (1987: 100), cited earlier, ask: "What are the state's sources of power to be able to achieve relative autonomy in a capitalist society?" In response the authors point to three basic sources: accumulation, storage, and distribution of information about the population that leads to far-reaching surveillance capacities; monopoly over the legitimate use of violence; and the state's contradictory character (1987: 100-104). But is this response remarkably different from one expected from the (state-centred) works of Mann (1986) or Giddens (1985)? It is in this theme of the logistics, techniques, and mechanisms of rule that some overlap between neo-Marxist and state-centred approaches becomes apparent as well. There are, then, similarities or overlap among neo-Marxist, state-centred, and governmentality approaches. The originality of the latter approach adopted for this thesis, described in the pages that follow, should therefore not be overstated. Nor, however, should the findings that result be assumed irrelevant to these other traditions of thought. It will be prudent to keep similarities above in mind as the 'hows of rule' or, more 44 specifically, answers to the overarching question stated at the outset, 'How is the Canadian refugee regime constituted and governed?1, are explored. Unlike neo-Marxist and state-centred approaches, and despite similarities and overlap, seeking these answers does not require placing the state at the centre of analysis. DE-CENTRING THE STATE: A FOUCAULDIAN ALTERNATIVE Just as the state as an object of inquiry had seemed to have become more important to historical sociology and social theory, through the varied work of Foucault (1977a), Smith (1987), Laclau and Mouffe (1985), Bowles and Gintis (1986), Magnusson and Walker (1988), and several other writers in the 1980s, it underwent a de-centring (Mahon, 1991). This development occurred alongside a wider recognition of sites of politics and resistance outside the state, such as the workplace and the domestic sphere (Mahon, 1991). It should be made clear, however, that de-centring does not deny the state's existence. For example, Laclau and Mouffe (1985: 162) assert at one point that the creation of 'public spaces' is carried out not in the form of subordination. It is here that we have to look for the terrain on which numerous struggles emerge against bureaucratic forms of state power (emphasis added). Similarly, Bowles and Gintis (1986: 184) write about Marxist theory's "overarching focus on class exploitation at the expense of more sustained attention to state despotism" (emphasis added). Nor does Foucault in his later writings on governmentality (cf. Foucault, 1991b), and writers such as Nikolas Rose, Peter Miller, 45 Vikki Bell, Ian Hacking, and Mitchell Dean who have drawn from them (see below), anywhere assert that the state does not exist. Instead, in their writings the state is seen more as the historical effect, resultant, or residue of governmental practice. Rather than denying its existence, the latter writers suggest the state be attributed less importance and attention directed beyond or behind it (cf. Valverde, 1994a: vii; See, for example. Rose and Miller, 1992: 189). A critical alternative to approaches that focus on the state, then, is one inspired by Michel Foucault's later writings on 'governmentality', which he describes as the ...ensemble formed by the institutions, procedures, analyses, and reflections, the calculations and tactics that allow the exercise of this very specific albeit complex form of power... (Foucault, 1991b: 102) This approach1 has developed partly in response to neo-Marxist criticisms of Foucault's apparent preclusion of the state (Miller, 1987: 207; Gordon, 1987: 299). Exemplars of this approach include Donzelot (1979) on the family, Hunter (1988) on literary education, Hacking (1990; 1991) on statistics, Rose (1990; 1984) on psychology, Castel (1988) on psychiatry, Dean (1991) on the British Poor Law, and Hunt (1996) on British Sumptuary Law. Other studies include Bell (1993) on child abuse, Ashenden (1996) on child sexual abuse, Procacci (1991) on poverty, Osborne (1993) on the medical profession, Dillon (1989) on international deterrence, Cruikshank (1994; 1996) on empowerment and self-esteem, Hopwood and Miller (1994) on accounting, Kinsman (1996) on persons with AIDS, Stenson (1993a; 1993b) on community policing and social work, Minson (1993) 46 on sexual harassment, O'Malley (1992) and O'Malley and Palmer (1996) on crime prevention, Simon (1994) on campus life, Murdoch and Ward (1997) on agriculture, and Walters (1994) and Dean (1995) on unemployment. As suggested in the previous chapter, neither the Canadian nor any other (international or national) refugee domain has been approached in the ways exemplified in these studies. Taken together these studies do not constitute a grand theory to be applied or confirmed (Valverde, 1996: 367). Instead, they have been described simply as a Lakatosian 'research programme' (Miller and Rose, 1995a: 592). Some of these studies have focused on specific rationalities. Liberalism as a rationality of rule in its classical, welfare, and advanced modes has been explored to a considerable degree (cf. Minson, 1985; Dean, 1991; Rose and Miller, 1992; Burchell, 1993). Other studies have paid close attention to specific governmental technologies that complement liberal rationalities, such as accounting (Hopwood and Miller, 1994) and the public inquiry (Ashenden, 1996). Still others have centred on ethical self-formation (Foucault, 1993) or what Dean (1994c: 162) calls the 'self-self game1 (e.g., Dean, 1995; Baistow, 1995; Cruikshank, 1994; 1996; Valverde, 1996). This 'third axis' (Dean, 1995: 560) refers to "practices, techniques, and discourse of the government of the self by the self by means of which individuals seek to know ...and act on themselves" (Dean, 1994c: 156). While making ethical self-formation intelligible is an important endeavour, this thesis limits its attention to governmental rationalities, technologies, 47 and knowledges in one sector or domain. It is necessary to point out the significant differences between this Foucauldian approach, and the state-centred and neo-Marxist approaches above. As suggested, one of these concerns the state. Toward this end, Gordon (1991: 4) writes: State theory attempted to deduce the modern activities of government from essential properties and propensities of the state, in particular its supposed propensity to grow and to swallow up or colonize everything outside itself. Foucault holds that the state has no such inherent propensities; more generally, the state has no essence (emphasis added). In his later writings, then, Foucault is not developing a general theory of the state (or of government), but is rather advocating analysis of how governmental power works, and through which techniques and schemes (Dean, 1994b: 179). In keeping with this, Rose and Miller (1992: 177) write: [T]he state can be seen as a specific way in which the problem of government is discursively codified, a way of dividing a 'political sphere1, with its particular characteristics of rule, from other 'non-political spheres' to which it must be related, and a way in which certain technologies of government are given a temporary institutional durability and brought into particular kinds of relations with one another. Such an approach, then, avoids treating the state as an actor, as an anthropomorphic, essential entity. There is a move here from the state as an actor to ...a broader field identified by policies, programs, strategies, projects, and tactics; the 'practices' thus generated extend well beyond the traditional boundaries of the state apparatuses (Hunt, 1993: 311). To employ the broadest of definitions, government is ...the historically constituted matrix within which are articulated all those dreams, schemes, strategies, and 48 manoeuvres of authorities that seek to shape the beliefs and conduct of others in desired directions by acting upon their will, their circumstances, or their environment (Rose and Miller, 1992: 175). It follows that analyses of governmental practice cannot be carried out in terms of the historical development of states and the accompanying concepts of law, legitimacy and sovereignty (Dean, 1994b: 179). The concept of 'social control', often associated with state theory and conventional political sociology (cf. Cohen, 1985; Cohen and Scull, 1983) is similarly inadequate. At present 'social control' is obscure and ill-used. Cohen (1985: 2) suggests that "[i]n everyday language, it has no resonant or clear meaning at all." Rothman (1983: 106) similarly writes: "For as much as the term is currently used, so often is it abused, pasted on with little or no effort at clarification." But the problem is not merely one of clarification or, as Mayer (1983) suggests, of a need to 'broaden' the concept to refer to the 'broader field' that Hunt (1993) mentions. Whatever analytical value the concept may still have, it is simply too blunt to connote the varied historical elements of particular governmental domains (cf. Hindess, 1996: 144-5). Indeed, proponents of this Foucauldian alternative explicitly seek to separate their endeavours from such a concept (cf. Valverde, 1994b: 25; Miller and Rose, 1988: 172; Rose, 1987: 72). Furthermore, if what has been termed 'advanced liberalism', a rationality of relevance to contemporary domains, attempts to "govern without governing society" (Rose, 1996a: 61), and tries instead to structure outcomes through the choices of independent agents, one wonders what the 'social' in 'social control' would 49 then refer to, if this latter concept were used. 'Regulation' is Hunt's (1993) term of choice. It too is more limited in its connotations than 'govern1 in that it suggests control through the law exclusively. Seen this way, 'social control' and 'regulation', are perhaps better understood as particular ways of governing the conduct of human beings, rather than the only way. This Foucauldian approach focuses on how practices of government (and the accompanying categories of human beings they specify) are made possible by rationalities, technologies, programmes, and knowledges rather than, for example, their formal legal codification; how they represent political authorities' trans-historical quest for legitimacy; or how they exemplify 'social control' or 'regulation'. This Foucauldian approach differs from conventional state-centred and neo-Marxist approaches in other important ways (Rose and Miller, 1992: 177-8). First, it differs in terms of sociological realism.15 The conventional approaches discussed in the first part of this chapter, other than those which draw from Gramsci, promote the discovery and identification of changes in, usually during specific periods, the actual configurations of and relations between individuals, groups, organizations, and events. Foucault's project (and the conceptual approach it has inspired), on the other hand, is as much "about the changing shape of the thinkable, as it is about the 'actually existing1" (Gordon, 1991: 8). A second difference, related to the first, concerns a Foucauldian attention to knowledge (Rose and Miller, 1992: 177). 50 The production of knowledge is necessary for a domain to emerge and governmental practices are assumed dependent upon knowing their objects (Miller and Rose, 1988: 174; Hunt, 1993: 317-18). An analysis of government involves asking questions about what knowledges have emerged along with a regime or subsequent changes within it. In an analysis of government, the focus is not only on ideas, as in nineteenth century idealism, but also on more practical knowledges, what Rose and Miller (1992) call 'technologies'. A third important difference concerns discourse. In most of the conventional approaches described above the necessary task would be to determine, for instance, how the language of political authorities or state agents mystifies, legitimizes, and covers over real objective state interests (e.g., capital accumulation or repression). In other words, there is a focus on discourse as the legitimation of power relations, as a kind of camouflage for the exercise of power. An analysis of government involves an interest in discourse, but this is not so much an attention to ideology as an attempt to discover and make intelligible "systems of thought and systems of action" (Rose and Miller, 1992: 177). In the Canadian refugee regime there are undoubtedly striking discrepancies at certain sites and periods between, on the one hand, images of practices directed at refugees as seen, for example, in the accounts of authorities and, on the other, the practices themselves. This disjunction, however, is simply not the focus here: an analytic of governmental practice is not about 51 exploring the discrepancy between the rhetoric and the reality of the Canadian refugee regime. Besides these differences, it should also be noted that this Foucauldian conceptual approach highlights the idea that government is dependent, not only on political authorities and state agents, but a variety of other kinds of authorities in the form of specific organizations, groups, and individual identities. Though this is widely acknowledged among state theorists (recall Carroll's (1992) description of Gramsci's formulation of power above), if previous research in the context of the refugee policy and practices is any example, this idea has been ignored. Dean (1994c: 152) writes: This restriction to the state... ignores the multiplicity of agencies and authorities involved in the governance of the life-conduct of individuals, families, groups, and populations. This is clearly illustrated by the multiple and overlapping jurisdictions involving local, regional, national, international, and global authorities within which actors are located. To be consistent with this, and as described in the next chapter, it is non-state authorities and agencies that some research procedures are directed at. Miller (1987: 207) writes that governmentality "refers to the practices which characterize the form of supervision a state exercises over its citizens." As seen in chapters that follow, the targets of 'the form of supervision a state exercises' are not limited to 'citizens', if by 'citizens' Miller means self-regulating entities assigned legal citizenship status within a liberal order. It also extends to 'non-citizen' categories within and between the conceptual and territorial borders of nations. 52 Those human beings who occupy, seize, fall, or are forced into these categories may or may not be eventually transformed -primarily through resettlement programmes- into citizens. What else should the approach described above attend to? First, governmental practices can be analyzed in terms of governmental rationalities. These rationalities are defined as: the changing discursive fields within which the exercise of power is conceptualized, the moral justifications for particular ways of exercising power by diverse authorities, notions of the appropriate forms, objects, and limits of politics, and conceptions of the proper distribution of such tasks among various sectors (Rose and Miller, 1992: 175). They include thoughts about the proper targets and limits of politics and the suitable allotment of tasks to various authorities (Rose and Miller, 1992). More specifically, a rationality is ...a way or system of thinking about the nature of government (who can govern; what governing is; what or who is governed) capable of making some form of that activity thinkable and practicable both to its practitioners and to those upon whom it was practised (Gordon, 1991: 3). Rationalities are not merely theories or political philosophies. They neither automatically arise when political authorities consult their objective interests, nor do they come from some natural or essential capacity for making rational decisions. Instead, rationalities are historically developed -they come from the past-though not in unmodified forms (Dean, 1994b: 182). A rationality can be thought of as a necessary but not sufficient condition of governmental practice. The other side of governmentality, alluded to earlier and 17 conseguently a focus of this approach, are technologies. These 53 are instruments, means, and mechanisms that make different forms of rule possible (Dean, 1994b: 188). They include: ...techniques of notation, computation and calculation; procedures of examination and assessment; the invention of devices such as surveys and presentational forms such as tables; the standardisation of systems for training and the inculcation of habits; the inauguration of professional specialisms and vocabularies; building designs and architectural forms... (Miller and Rose, 1990b: 8). Technologies of government are assembled within particular programmes by rationalities. Technologies and rationalities are therefore complementary. A particular technology can be used for different purposes, depending on the rationalities they articulate with (Valverde, 1996: 358). Technologies are either logically or historically associated with specific rationalities (Valverde, 1996: 358). Concerning these technologies, a rough distinction can be made between "representation of and intervention into specific governmental domains" (Dean, 1994b: 188). In this way, governmental and scientific practice are analogous, as Hacking (1983: 31) writes: Theories try to say how the world is. Experiment and subsequent technology change the world. We represent and we intervene. We represent in order to intervene, and we intervene in the light of representations... As seen in chapter six, mundane technologies that represent and intervene have made the different sites of the Canadian refugee regime governable. Besides knowledge, then, governmental practices are made possible by rationalities and technologies. Several other terms need to be defined before proceeding. One of these is 'programme'. 'Programmes' are idealized schema for the 54 ordering of social life (Miller and Rose, 1990b: 14). They articulate knowledge of a particular domain (Gordon, 1980: 243). Programmes "constitute a space within which the objectives of government are elaborated, and where plans to implement them are dreamed up" (Miller and Rose, 1990b: 14). Programmes are more than wishes or intentions of individual programmers (Gordon, 1980: 248). Programmes assume a sector of reality is "subject to certain determinants, rules, norms and processes that can be acted upon and improved by authorities" (Rose and Miller, 1992: 183). Governmental programmes and rationalities are translated into one another to form specific regimes (Miller and Rose, 1992: 181). 'Regime' is another important term, one which is consistent with Foucault's (1991a: 75) well-known phrase, 'regime of practices'. It is meant here to refer to an 'administrable domain' ...which is formed when it is possible to delineate a specific historical complex of various practices including theoretical discourse, more 'practical' knowledges, state and non-state interventions, and administrative techniques - which constitute a relatively coherent and distinct field, with definite objects of knowledge and targets of intervention, specific agents, modes of organization, and aims (Dean, 1991: 221n) . Canada's national economy is an example of a regime and, as the chapters that follow show, Canadian practices directed at refugees are another. Another term used in the chapters that follow is 'discourse'-a central feature of a Foucauldian alternative. The term 'discourse' has several different meanings in the social sciences. This definitional issue is part and parcel of theoretical debates among proponents of competing theoretical 55 perspectives and about ontology or theories of reality. When considering the status of discourse, a central issue that has arisen is whether it should be considered as representational or constitutive (Wetherell and Potter, 1992: 9; Eagleton, 1991: 206). In a representational understanding, discourse is assumed to represent, name, show, indicate, point to, stand for, or passively reflect, some underlying pre-existing reality (Eagleton, 1991: 206). In this conception, discourse is seen as something of a medium for reality to express itself. This means a recognized change in discourse (ie. the signifier) indicates or reflects a change in, most often, a political economic reality (ie. the signified). In this conception, discourse is strictly determined by power relations. This conception associated with classical Marxism is one I wish to explicitly reject. The conception of discourse as constitutive, though relevant to the neo-Marxist class-conflict approach mentioned earlier, is most often associated with perspectives variously referred to as post-structuralist and post-Marxist. It is also associated with Foucault's writings and others who have extended his insights. With the emergence of perspectives under these labels, there has been a concomitant 'withering away1 or 'displacement' of classical Marxist conceptions of ideology (Zhao, 1993: 82; Eagleton, 1991: 8; Hall, 1988: 51).20 As stated earlier, interest in discourse, in the approach adopted for this thesis, is not tantamount to an interest in ideology understood in the sense of pure legitimation of power 56 relations. Instead, discourse is understood as performative and constitutive. Miller and Rose (1990a: 180) write that ...an analysis of political rationalities requires an attention to the constitutive role of language. The discourse of politics is more than mystification, legitimation or ideology, more than the thin surface of contact between social interests and the exercise of domination. Valverde (1991b: 10) writes that discourses are "organized sets of signifying-practices that often cross the nineteenth-century boundary between 'reality' and 'language'." Conceiving of discourse as constitutive is therefore not tantamount to adopting V) an idealist and rejecting a materialist position (cf. Miller and Rose, 1995a: 592; Hall, 1988: 51; Minson, 1985: 122). As Valverde (1991b: 10) notes, in this conception of discourse ...language is not a transparent window giving access to the world but is rather itself a part of the world, a kind of object among objects, [an insight which] eliminates the old dichotomy between idealism and materialism. Minson (1985: 124) similarly writes: The radical bottom line of Foucault's reworking of the term 'discourse' is that knowledges a.re no longer to be regarded as part of the order of ^representation or signification. The idea is to treat discursive materials less as representing things on which they bear, than as means of attempting to organize them. The knowledges that target refugees discussed in chapter six, for example, can be understood as discourses that seek to organize raw materials -bodies, conditions, processes, memories and so on- to form the object of a distinct assemblage of governmental practice. But discourses conceived in this way should not be seen as causal or determinative agents in their own right. Seeing them this way 57 comes close to seeing events as other than the outcome of human subjects. Though in later chapters analysis is pitched at the broad use of language, it does not necessarily follow that there is some automatic effect. Discourses are constitutive in that they at once organize and help produce a world of social, political and material practices. 'Practices' can be defined as "common groupings of discourses and technologies" (Wickham, 1983: 480). 'Discourses' can be separated analytically from 'practices' while simultaneously recognizing they overlap and constitute one another (Ericson and Haggarty, 1997: 84). Understood in this sense, governmental rationalities are broad discourses, a 'sphere of language' that sets out in a programmed way, the proper means by which government of a particular population is to be carried out (Miller and Rose, 1990a: 166). Contemporary liberal rationalities, for example, presuppose that government is to be carried out in terms of 'policies', such as refugee policy, and directed at 'claimants' and 'clients', such as the refugee claimant. Two more terms that will be used throughout the thesis need to be defined: 'refugee' and 'refugee claimant'. In this thesis I take 'refugee' not to signify a given or natural entity, but as the name of a historical construct. The status of 'refugee' is therefore directly analogous to 'madness', 'sexuality', and 'delinquency' in Foucault's well-known histories (ie., 1965, 1978, 1977a), in that it is used without presupposing the essential characteristics presently attached to it. At present 'refugee' is 58 simultaneously an object of the exercise of governmental power and a contingent, historically constructed identity (hence the use later on of the admittedly awkward term 'object/identity') . As noted previously, this is definitely not how previous research has understood 'refugee1. In this regard and in the chapters that follow, therefore, it will be useful to keep in mind the United Nations Convention definition adopted by Canada in 1969, which is as follows: Any person who, by reason of a well-founded fear of persecution for reasons of race, religion, nationality, membership in a particular social group or political opinion, (a) is outside the country of his nationality and is unable or, by reason of such fear, is unwilling to avail himself of the protection of that country, or, (b) not having a country of nationality is outside the country of his former habitual residence and is unable or, by reason of such fear, is unwilling to return to that country (UN Convention Relating to the Status of Refugees, 1951 as cited in Dirks, 1984: 280). 'Refugee claimant1, a category that has emerged more recently, is also at once an object of governmental practice and a contingent, historically constructed identity. It refers to human beings who have claimed Convention refugee status under Canadian law and are awaiting determination. These persons may or may not eventually be determined to be refugees following their processing through the Canadian refugee determination system. Finally, one might ask how it is this Foucauldian alternative to neo-Marxist and state-centred approaches is critical, especially given its rejection of the concept of ideology. Toward this end, Foucault (1988b: 154) writes: A critique is not a matter of saying that things are not right as they are. It is a matter of pointing out on 59 what kinds of assumptions, what kinds of familiar, unchallenged, unconsidered modes of thought the practices we accept rest. Though this issue is taken up in the next chapter, it should be noted here that the Foucauldian approach adopted in the chapters that follow is critical in that it is genealogical, and that genealogy is a form of 'exemplary' critique (Owen, 1995). SUMMARY AND CONCLUSION There is overlap between a Foucauldian approach and at least some neo-Marxist and state-centred approaches (the latter two also overlap) in terms of an attention to discourse as constitutive, the mechanisms of rule, the role of non-state authorities, and historical specificity. In this way the Foucauldian alternative outlined in this chapter, that looks beyond or behind the state toward knowledges, rationalities, technologies, and programmes, can be understood to have the potential to address some themes found within research that has placed the state at the centre of analysis. The next chapter discusses methods and research procedures appropriate to such an approach. 60 1. Significantly, a recent issue of Economy and Society was devoted to studies that involved "articulating elements of the governmentality literature with aspects of other traditions (most notably feminism and Marxism)" (Pearce and Valverde, 1996: 307). Pearce and Valverde (1996: 309) describe the distinction made between a Foucauldian alternative and these more conventional approaches as sometimes "hostile." 2. In keeping with the effort to point out links between neo-Marxist and state-centred approaches and a Foucauldian alternative in this chapter, it should be noted that in his later writings Poulantzas drew directly from Foucault's writings (Jessop, 1990). In a chapter within his larger work on state theory, Jessop goes as far as to suggest that Poulantzas "actually was far less 'Marxist' and far more 'Foucauldian' than most of his critics suggest" (Jessop, 1990: 223). 3. Mahon (1991: 121) suggests this common portrayal of the debate as instrumentalist-structuralist oversimplifies the position of Miliband, and that the Marxist debate in the 1970s was instead about how to conceive of 'relative autonomy'. 4. For excellent summaries of the various positions regarding the relative autonomy of the state, see Ratner et al. (1987: 85-125), Jessop (1990: 79-104), or Albo and Jenson (1989). 5. Renewed interest in Gramsci is undoubtedly as much a response to the failings and deficiencies of classical Marxism as the writings of Foucault and his followers. It seems that for advocates such as Hall (1988), however, pointing to faults in classical Marxism sometimes is not enough to justify a turn to Gramsci, and it is then that Foucault is taken out to the shed and placed in position of 'whipping boy1 (Foucault has been placed in such a position by advocates of different approaches but for similar reasons- see Ashley (1990) on Jurgen Habermas or Nancy Fraser (1989)). Those taking Foucauldian insights seriously are alleged to have been placed under the spell of their "permanently radical gloss", "gospel", or "magical terms" (Hall, 1988: 52), the latter of which, when compared to Gramsci's concepts, are alleged to be "thin, undernourished abstractions" (Hall, 1988: 56). In this chapter, instead of resorting to this kind of sniping, coupled with grand pronouncements of the "superiority" (Hall, 1988: 53) of one thinker's concepts (including Foucault's) over all others, I attempt to point out some links between conventional approaches and a Foucauldian approach- the reader is left to decide questions of superiority. It should be said, however, that while Hall (1988) grants some importance to Foucault's Archeology of Knowledge and Discipline and Punish in lending insight into Thatcherism as a discursive formation, he might have discussed Foucault's later writings on governmentality and the similar direction that Gramsci's writings and the Foucauldian alternative discussed in this chapter can be taken. For an example of a study that uses 61 Gramscian and Foucauldian insights, see Pearce and Tombs (1996). Hall's treatment of Foucault's writings here is expressive of how the latter has been received and criticized in Britain generally (Gordon, 1996). Parenthetically, it was not the polished, Judeo-Christian, or mystic qualities of Foucault's writings that seduced and led me to adopt a governmentality approach, it had more to do with certain experiences. 6. The need for such a shift was an illusion because the state had already been a subject within historical sociology where it was explored in its concrete forms across time and in a loose comparative way (cf. Eisenstadt, 1963; Moore, 1966; Anderson, 1974; Tilly, 1975). 7. While Jessop (1990: 278) uses the term 'anti-Marxist' to refer to a wide variety of writings that centre on the state, it should be emphasized that some of these approaches are sympathetic to Marxism. The irony here is that Jessop's (1990) own strategic-relational approach has been criticized precisely because it is difficult to see in what way it is Marxist (Block, 1992: 872). 8. It could also be referred to as bureaucratic power, which is, of course, associated with Weber's trenchant analyses of the bureaucratic state. What seems to underlie attention paid to this form of power, at least in the work of Mann (1988) and Giddens (1985), is a suggestion that the aim of critical social theory cannot be limited to aiding the neo-Marxist goal of a transformation from capitalism to socialism (or hypothetical radical transformations associated with gender, ethnicity and race), that it must also address an associated potential for totalitarianism (Giddens, 1985: 294-341; cf. Arendt, 1973). This suggestion is misplaced, at least at this historical juncture-'advanced liberal' government (see chapter five) relies upon domains of freedom that must not be hindered by political authorities. 9. This seems to be the main point of a recent neo-Marxist critique (Curtis, 1995) of Rose and Miller (1992). 10. Foucault's (1977a) popular work on disciplinary power and the prison, as well as several earlier works, of course, are not entirely consistent with his later writings on governmentality. Despite this, Giddens (1984; 1987) uses the former popular work to argue that disciplinary power is merely a type of administrative power. Giddens depicts Foucault (1977a) in a manner that would not fit what is described here as a Foucauldian alternative. It is similarly interesting that Mahon (1991) -cited earlier in this chapter- attempts to resurrect state theory through Jessop (1990) and along the way conveniently refers only to Foucault (1977a), rather than his later writings. 62 11. See also Valverde, 1991a; 1994a; 1994b; 1996; Miller, 1987; Rose, 1987; 1993; 1995; 1996a; 1996b; Miller and Rose, 1988; 1990a; 1990b; 1995b; Gordon, 1991; Rose and Miller, 1992; Hunt, 1993; Hunt and Wickham, 1994; Dean, 1991; 1994a; 1994b; 1995; Hunter, 1990; 1994; 1996; Hacking, 1982; 1986; 1995; O'Malley, 1996a; 1996b; Burchell, 1993; Burchell et al., 1991; Hindess, 1993; 1996; Barry, Osborne, and Rose, 1993; 1996. 12. This thesis is not focused on ethical self-formation specifically or those undoubtedly most directly affected by the Canadian refugee regime generally. These would include asylum-seekers prevented from seeking refuge in Canada or at its posts abroad, those who seek refuge but are systematically prevented from making claims or deemed not to be refugees and, lastly, those deemed to be refugees by determination or selection practices abroad and who are then resettled in Canada. Exploring the practices of human beings who occupy, seize, or fall into these categories carried out upon their selves would be worthwhile, not merely in Canada during 'adaptation1, as in resettlement research that examines their mental health (e.g., Johnson and Beiser, 1994), but rather before and after using a Foucauldian approach. Doing so would be fraught with significant logistical problems (see Yu and Liu (1986) for a discussion of such problems in an American context), however, including locating those willing to relate their perceptions of their practices. It is because of such problems that interviews for this thesis were limited to focusing on other authorities of various types (see chapter three) who were more accessible. 13. The intellectual roots of the term 'social control' lie in nineteenth century classical sociology (Cohen, 1985: 5). Cohen and Scull (1983: 5) relate that "the concept of social control was at the centre of the enterprise- both in relating sociology to political philosophy and in solving the emergent debates of macro-sociology." During this period it referred "to the capacity of a social group to regulate itself." The use of the term in this sense continues but along the way it has been re-defined to refer to "either socialization or social repression" (Janowitz, 1975: 82). In the early twentieth century thought of G.H. Mead and E.A. Ross, the term 'social control' was used to refer to "how people associated together to perform common endeavours, how voluntaristic cooperation marked social activity" (Rothman, 1983: 107). Rothman (1983: 108) points out: "At its birth, something very conservative clung to the notion of social control, both in its content and its orientation" and that "those ready to explore the concept wished to understand the roots of order, not to examine the roots of change, and certainly not to foment change" (Rothman, 1983: 108). The members of the Chicago school of the 1920s and 1930s fostered the understanding of social control as socialization, particularly as it applied to the conduct of new immigrants in America. It is through such American sociology that the "connection between social control and a contemplation of the state became weaker and weaker" 63 (Cohen, 1985: 5). After a hiatus in the early 1940s, the concept of social control came back into use (cf. Parsons, 1951). By the 1960s, mostly through sociologists associated with the societal reaction (or 'labelling') perspective (e.g., Becker, 1963), the connotations of coercion, conflict, and repression were attached to social control (Rothman, 1983: 109). The use of the term in this latter sense, of course, has continued in the tradition of the 'new' or neo-Marxist criminology of the 1970s and 1980s (cf. Ratner and McMullan, 1983). At present the meaning of social control remains unclear. I doubt that if 'deviant behaviour', on which concepts of 'social control' tend to rely, were defined (cf. Cohen, 1985; Horwitz, 1990) that this would clearly demarcate differences in meaning. The concept of 'deviant behaviour' was left behind in the early 1970s (Sumner, 1990), 'social control' ought to be thrown into the conceptual dustbin with it. 14. There is some variance in the terms used in the governmental ity literature. For example, Hunt (1993) uses 'regulate' and Hunter (1988), Rose (1990), and Dean (1995) use 'govern'. Regarding a recent 'Rethinking Regulation' workshop, Valverde (1994b: v) writes that "[o]ne of the recurring debates... concerned the implications and connotations of choosing to use a term such as 'governance' as opposed to terms such as 'regulation' and 'social control'" (See Valverde (1994b: 25-26) for an excerpt from one such debate). 15. One might, notice an absence of the terms 'description', 'explanation', and 'cause' associated with realist social science in the text of this thesis. To be consistent with a Foucauldian alternative I have used 'make intelligible', 'make sense o f , 'elucidate', and 'conditions of possibility' throughout. The word 'undoubtedly' is a similar device that is agreeable with the latter approach whenever reality needs to be referred to. 16. This approach is reacting to a classical Marxist view of ideology as static and expressive, as coming directly out of objective social interests. Were it a reaction to much recent concrete neo-Marxist research, including that which draws on Gramsci, it would be inappropriate. In this sense the foil chosen is admittedly a facile one. 17. While Dean (1994b: 187-8) briefly distinguishes among technologies, techniques, and intellectual technologies of government, for simplicity 'technologies' is used in chapter six. What the introduction of 'intellectual technologies' suggests, however, is that the difference between governmental rationalities and technologies is not a straightforward mental-manual one because this distinction can apply to technologies themselves (Dean, 1994b: 188) . 18. Mitchell Dean takes 'administrable domain' from Castel (1988) who coined it in his study of psychiatry in France. 'Regime' is also a theoretical concept widely used in the field of 64 international relations (See, for example, Skran, 1995: 7; Walker, 1993: 163) which, given the discussion of international programmes in chapter four, is fitting. See Keeley's (1990) excellent attempt to develop the concept of 'regime' in the context of international relations. 19. Discussion of this term arose in a 'Rethinking Regulation' workshop (mentioned in endnote 14; see Valverde (1994b)) attended by several major proponents of a governmentality approach. This suggests it is necessary to discuss this term a bit further here. 20. In studies of ideology of the neo-Marxist variety there has been a move toward seeing discourse as constitutive (Wetherell and Potter, 1992). This is seen much earlier in the work of Althusser (1971), but also more recently in the writings of many neo-Marxists as varied as Hall (1988) and Jessop (1990). In this understanding, ideology is seen as a material process, whereby it becomes forceful through other practices (MacDonell, 1986). This conception of ideology is consistent with elaborate, sustained responses to the 'end of ideology1 thesis (See Eagleton, 1991; Zhao, 1993; Simons and Billig, 1994). 21. I began this thesis intending to use a concept of ideology in the context of a non-class realm, the nation-state system generally and the Canadian refugee regime specifically. Zhao's (1993: 79) statement that "[i]deology not only works along class lines, but also along other axes, for example, sex, race, and nation-state" (my emphasis) supported this intention. Despite numerous attempts, however, I failed to arrive at an adequate understanding of the essential functions of the state as they relate to refugee policy and practices and I therefore abandoned the idea of using a concept of ideology. 22. To the extent that it is concerned with how objects/identities of government are made possible, this approach is consistent with the philosophical doctrine of dynamic nominalism (cf. Hacking, 1986). 23. Giddens (1990b: 214) goes as far as to say that "Foucault's history tends to have no active subjects at all. It is history with the agency removed." This is a misplaced criticism as Foucault's later writings are inconsistent with it. As Gordon (1991: 5) writes: "In his 1982 essay 'The subject and power', Foucault affirms, on the contrary, that power is only power... when addressed to individuals who are free to act in one way or another. Power... presupposes rather than annuls their capacity as agents." Power associated with liberalism, for example, works through these capacities. 24. The term 'political rationality' is also used in governmentality studies. 65 CHAPTER THREE: METHOD AND RESEARCH PROCEDURES So far in this thesis I have discussed previous research about Canadian refugee policy and practices and an appropriate critical conceptual approach to study these topics. In this chapter I elaborate a method and research procedures for this thesis. In the introduction I mentioned Foucauit's methods of discourse analysis and genealogy. Now I should like to say more about them. Following this I discuss the research procedures carried out to respond to the research question(s) posed earlier. DISCOURSE ANALYSIS I mentioned 'discourse analysis' in the introduction and discussed 'discourse' in chapter two. Like 'discourse', 'discourse analysis' has a variety of meanings. Seidel (1987), for instance, describes no less than seven forms of political discourse analysis. It will be recalled that Creese (1991) and Barsky (1994), mentioned in chapter one, both employed discourse analysis, one a content analysis of Canadian news articles about refugees and the other a Bakhtinian analysis of Canadian refugee determination hearing transcripts. These types of discourse analysis, however, are not what I have in mind here. Because of this, it is obviously important to elaborate what I mean by 'discourse analysis'. Toward this end, it is worth repeating Potter and Wetherell's (1994) instructive typology of four basic kinds of 'discourse analysis'. The first type of discourse analysis Potter and Wetherell describe stems from what is known as speech act theory. Such analysis examines the way conversational exchanges are organized. 66 The second type is firmly established in the discipline of psychology and examines the structure and effects of discourse on, for example, an individual's ability to recall and understand specific items on a questionnaire. The third stems from the sociology of science and explores how scientific authorities attempt to make their accounts persuasive. The fourth type of discourse analysis is exemplified by the early work of Foucault (1971), who refers to his distinctive 'archaeology' as 'discourse analysis' (Potter and Wetherell, 1994). This type is also commonly referred to as 'semiology' or 'post-structuralism' (cf. Barthes, 1973, 1977; Derrida, 1976). The kind of discourse analysis employed in this thesis is most akin to this last type. So what does this type of discourse analysis entail? To identify a discourse an analyst should look for recurrently used systems of terms used for characterizing and evaluating actions, events, and other phenomena... [often] organized around specific metaphors and figures of speech (Potter and Wetherell, 1987: 149). It involves exploring, broadly, what is common in a set of materials and then finding out how these discourses work to constitute particular domains. Since a discourse is less than everything that is said and written, not all talk and text falls into one or another discourse (Minson, 1985: 124). An analyst, therefore, should not expect to find entire discourses in a set of materials, but rather 'pieces of discourse' (Parker, 1990a: 193). Discourse analysis is not tantamount to the notion of refusing to accept the image of the world that it presents of itself, associated with, for example, the sociologist Mills (1959). The 67 concern here is not with images (or ideologies) of the world. The effort is not to 'scrape off the surface' or 'peel back the layers' to reveal an essence lying underneath images or discourses. If discourses are understood as constitutive, then there is a sense in which there is nothing underneath. A discourse analyst is interested in constitution, not in image. To deploy our previous metaphor, it is not the shadow that the door casts that is interesting to the discourse analyst, it is the door itself: what is it made of and how does it work? GENEALOGY Genealogy was also mentioned in the introduction. What does it entail (cf. Foucault, 1977b; Drefus and Rabinow, 1983; Minson, 1985; Poster, 1989; Leonard, 1990; Wickham, 1990; Prado, 1995; Viskar, 1995)? First and foremost, genealogy should be understood as a form of critique (Sarup, 1993). The neo-Marxist studies in chapter one involve what is termed 'legislative' critique (Owen, 1995). They are critical by their reference to Marxist humanist values or scientific laws. Genealogy cannot be critical in this way because it does not recognize, for example, an explicit system of laws (Owen, 1995: 492). It cannot draw on laws for purposes of critique because it has nothing to fashion them with in the first place (Owen, 1995: 492). Genealogy is instead an instance of 'exemplary' critique (Owen, 1995). In Foucault's studies, genealogy exemplifies the value of autonomy by what it says and by what it shows (Owen, 1995: 491). More specifically, genealogy "attempts to reveal the 68 multiplicity of factors behind an event and the fragility of historical forms" (Sarup, 1993: 59). The principle historic forms at issue here are the Canadian refugee regime and the refugee object/identity. In Foucault's (1988b: 155) words, the aim of genealogy is to show "that things are not as self-evident as one believed, to see that what is accepted as self-evident is no longer accepted as such." Connolly (1981: 155) writes that "genealogy strives to distance us from the rationale implicit in past and present practices." It attempts to do so without invoking the nostalgic, conservative suggestion that the past is a viable alternative to the present (Poster, 1989: 92). As Garland (1990: 136n) writes: "The point... is to cast light on a contemporary issue or institution by investigating those historical conditions that brought it about" in order to "problematize and destabilize the present." A genealogist should focus on the 'emergence' and the 'descent' of an issue or, in this case, regime of practices (Prado, 1995: 36). How does genealogy compare with orthodox history (cf. Foucault, 1977b: 152)? Although the alleged founders of sociology -Emile Durkheim, Karl Marx, and Max Weber- engaged in historical inquiry (Skocppl, 1984), until the 1980s the relationship between sociology and the discipline of history could be described as a "dialogue of the deaf" (Burke, 1980: 13). At this point, however, historical sociology underwent a significant resurgence (cf. Abrams, 1982; Skocpol, 1984; Tilly, 1981). The historical sociology of the state mentioned in chapter two is part and parcel 69 of this development. Skocpol (1984: 362) identifies three major strategies used by historical sociologists: applying a given theoretical model to historical facts; discovering causal regularities that account for defined outcomes; and using concepts to construct a meaningful historical interpretation. Of the three, genealogy is most similar to this last type, in that it too rejects grand theoretical models (or Theory) and causal hypotheses present in the former two methods. 'Cause' is replaced in genealogical analyses by 'conditions of possibility', a phrase used throughout chapter four and, to some extent, five, to highlight Foucault's (1977b: 148-52) understanding of 'emergence'. In this way, genealogy can be seen to oppose history, if history is a quest for origins, or is essence-seeking (Foucault, 1977b: 144). It is the beginnings of the Canadian regime, not its 'origins', that are mostly discussed in chapter four. Like those of punishment (Foucault, 1977a), the purposes of what I call ' refugeeness' have not stayed the same over time. Genealogy, in other words, rejects "the idea that behind events is a guiding hand or set of regulatory principles that are the grand determinant of the present" (Prado, 1995: 33-4). Genealogy refuses, in Foucault words (1977a: 31), "the unity of the past in terms of the present."' The genealogical enterprise is directed by questions of intelligibility rather than exhaustiveness, and seeks progressively to understand particular dimensions of the past, which are always open to revision and extension, rather than a complete reconstruction (Dean, 1991: 220; cf. Dillon, 1989). 70 Genealogy is, therefore, a modest project compared to much orthodox history in that it does not seek exhaustive reconstruction. Genealogy requires history, however, if the latter is understood as the collection and analysis of documents of the past (Prado, 1995). Whereas historical inquiries tend to adopt a style of writing characterized as a "narrative description of a sequence of events" (Giddens, 1990b: 213) or a "tracing of chronological developments" (Hewitt, 1983: 226), writing associated with genealogy can be organized around certain elements such as a particular programme or series of programmes, with only passing attention paid to sequence and chronology. The remaining chapters of the thesis are organized with this in mind. How, then, does discourse analysis compare with genealogy? Discourse analysis or 'archaeology' is associated with Foucault's earlier and genealogy with his later writings (Macdonell, 1986; Bevis, Cohen, and Kendall, 1993). Genealogy requires 'archeology' and adds a serial component (Dean, 1994d: 34): 'Archeology' is contained within genealogy. The two are, therefore, not contradictory methods (Connolly, 1981: 155). Because governmental regimes are constituted by historical elements, such as specific rationalities, knowledges, and technologies, the overarching research question, 'How is the refugee regime constituted and governed?', is partly an historical one. The chapters that follow include suggestions -although very brief in places- about the emergence and descent of the constitutive elements of the present Canadian refugee regime. 71 Given this, it is genealogy* rather than discourse analysis, that best characterizes the general method deployed in this thesis. Overall this thesis is genealogical in that it problematizes, it questions, it calls into doubt, that which is assumed to have always been the case -the present Canadian refugee regime and refugee object/identity- by drawing attention to that which has made it possible (cf. Poster, 1989). To return to our metaphor one last time, in addition to the questions of the discourse analyst, the genealogist asks: where did the components -the hinges, knob, wood, lock, and screws- from this dismantled door come from? RESEARCH PROCEDURES With the above discussion in mind, coupled with a recognition that a Foucauldian approach does not insist on standardized research procedures, I carried out a variety of qualitative manoeuvres: I attended a conference, conducted interviews, sought out a variety of relevant documentary materials, including secondary historical sources, and searched several indices of the humanities and social sciences. The broad purpose of these procedures was to get to know the Canadian refugee regime, how it has changed, and to discover that which has made the regime and this change possible. Refugee Conference I attended, as an observer, a four-day bi-annual conference held in late May, 1995 by a national refugee advocacy organization. It was attended by over a hundred Canadian representatives of organizations, groups, and provincial and federal levels of state. 72 Attendance at this conference allowed for the identification of authorities and agencies active in the Canadian refugee regime; the creation of a common point of reference for interviews later on; and the collection of relevant documents of various groups and organizations available at a conference display table and not easily accessible otherwise. Interviews As suggested in chapters one and two, governmental practices directed at refugees are not restricted to the efforts and activities of state agents and political authorities. The interviews discussed below were, therefore, directed primarily at non-state agencies and authorities presently active in the Canadian refugee regime. More specifically the interviews (see Appendix II for the schedule) were intended: (1) to obtain a broad sense of the scope and features of the refugee regime; (2) to reveal rationalities, knowledges, forms of association, and technologies that constitute it; and, more pragmatically, (3) to facilitate access to documents of a given authority not available through other means that would also help to realize (1) and (2). The interviews consisted of three basic parts. The first gathered background information about when a given authority or agency began to emerge and their practices in relation to refugees. The second part involved asking broad questions about their practices in relation to Canadian refugee policy, as well as, where relevant, their associations with other authorities in this regard. The third involved asking broad questions to encourage 73 interviewees to talk about various contemporary facets of Canadian refugee policy and their understanding of the refugee category. With regard to (2), it was assumed that relevant discourses would be present, although in 'pieces', in both the talk of these interviewees and in relevant documentary materials or texts. This is in keeping with Foucault's main emphasis: "a broad middle range (between spontaneous speech and abstract scientific theory)" (Dean, 1994d: 32), in that this talk was analyzed not in relation to its specific, concrete context (an interview situation with a university-affiliated researcher), but rather in a general way and 'pooled' with documentary materials or texts. The interviews were, therefore, thought of as one, rather than the source of data for this thesis. To conduct these interviews as described I first had to identify prospective individuals, groups, and organizations with significant involvement in the Canadian refugee regime. Besides attending the conference, to accomplish this I consulted previous research described earlier (e.g., Simmons and Keohane, 1992; Hawkins, 1991), as well as Canadian newspaper articles, a Canadian associations directory (Associations Canada, 1995), and a directory of services (The Red Book, 1994) pertaining to the lower mainland area of British Columbia. Once several individuals, groups and organizations were identified, I then consulted these two directories, as well as relevant metropolitan telephone directories to find addresses and telephone numbers. When this procedure failed to produce a reasonable number of relevant groups, 74 organizations and individuals, I then contacted a spokesperson of one of the major refugee advocacy organizations (who later agreed to be interviewed) to request their extensive membership list that contained relatively up-to-date addresses, telephone numbers, and names of representatives. Early interviews sometimes also revealed names, addresses and telephone numbers (or only one of these) of additional prospective interviewees. Once they were identified I contacted prospective interviewees by letter using University of British Columbia letterhead in which I identified myself; indicated the topic and purpose of the research; assured the confidentiality and anonymity of prospective interviewees; and indicated that an abstract of the final research report would be made available upon its completion (see Appendix I). Approximately one week after mailing the letter I telephoned these individuals to inquire about a response and to answer any questions raised by the letter. If the prospective interviewee's response was positive, I then arranged a time and place for the interview. Some groups and organizations had offices in the Vancouver area and, hence, representatives who could be interviewed. Others, however, had representatives or offices only in Ottawa-Hull, Toronto, Montreal or Winnipeg metropolitan areas (or several of these). To simplify access to these representatives, travel to these areas was required. Following a brief introduction and the signing of the consent form (see Appendix II) by the interviewee, I asked their permission to audio-tape the interview to facilitate analysis. In the three 75 instances where interviewees were unwilling to be audio-taped, I took notes and elaborated them immediately following the interview. At the end of the interview, and only if it seemed appropriate at the time, I asked interviewees about relevant documents that might provide additional background or that described practices involving them and/or their organization or group. By employing these research procedures described above, between May, 1995 and August, 1996, I conducted 48 interviews with individuals or representatives of groups and organizations with significant involvement -at the time of the interview- in the Canadian refugee regime. Interviewees resided in five provinces: British Columbia, Alberta, Manitoba, Ontario, and Quebec, although there was a definite emphasis (31 of 48) on individuals and representatives of groups and organizations in the lower mainland area of British Columbia. The aim was not to generate a random sample, but to cover a wide range of important authorities and agencies active in the regime in part to obtain a sense of the regime's scope and features, how it had been changing, and those elements that have constituted it. Interviewees included prominent lawyers and academics/experts, as well as representatives of refugee-specific organizations, women's groups, Judeo-Christian organizations, immigrant settlement organizations, an anti-racist group, ethno-cultural organizations, international development and humanitarian organizations, and federal government bureaucracies and agencies. Most interviews were carried out at interviewee's places of work or residences, although several were conducted in 76 more public locations: three in coffee shops, and one each at a restaurant, high school, and university campus. The duration of interviews ranged from forty-five minutes to three hours, with the majority taking about an hour to complete. Documentary Materials I collected a variety of documents suggested by previous research on Canadian refugee policy and refugees, and by interviewees. These included: the White Paper, the Green Paper, government research reports, federal government commissioned public opinion polls, press releases, minutes of meetings, research proposals, speeches, pamphlets, as well as recent mission statements, correspondence, newsletters, legal briefs, and research reports associated with individuals, groups and organizations with significant involvement in the Canadian refugee regime. Of particular note here was the periodical, Refuge, which was consulted extensively. I also examined existing histories of international and Canadian refugee policy, several of which were cited in chapter one. This latter procedure is consistent with other studies in governmentality (e.g., Dean, 1991; Rose, 1984) and genealogical analyses generally. Technologies and Knowledges Within both interview and documentary data there were references made to particular technologies and knowledges. To explore technologies of government and knowledges I did not employ one particular research procedure. Instead, I pursued each in an ad hoc manner using whatever means were available. 77 For technologies, once identified, I made further inquiries about them. For example, an authority indicated during an interview that the creation of the documentation centres in Canada had been 'a major advancement1 in refugee policy (interview 28/07/95). To learn more about this technology I arranged a two-hour guided tour of a centre with one of its practitioners during g which I took fieldnotes. I then arranged an interview0 with an individual authority who had been intimately involved in the centres' development. In addition, I collected and drew from available accounts contained in previous research such as legal reports (e.g., LRCC, 1991; Houle, 1994). For other technologies, such as the refugee camp, inquiries were limited to consulting secondary sources. To aid in the identification and examination of forms of knowledge produced about refugees I consulted several indices of the humanities and social sciences in electronic database (Sociofile, PAIS, and Psychlnfo) and, since most do not cover the period that precedes 1980, their equivalent3 paper form (e.g., the paper equivalent of 'Sociofile' is 'Sociological Abstracts'). Together these indices cover an array of disciplines. For each I searched for the subject 'refugee' over the past three decades. From there I established which knowledges were represented in the generated records and when they emerged. SUMMARY AND CONCLUSION Genealogy, which requires discourse analysis and has a serial element, best describes the method used for this thesis. 78 Qualitative and varied best describe the research procedures for this thesis, as they include attending a conference, conducting interviews, accumulating relevant documents, and searching indices of the humanities and social sciences. In the chapters of the thesis that remain, the results of these procedures are presented. 79 1. How are technologies, the other side of governmentality, to be analyzed? Miller and Rose (1990b: 8) provide only limited insight here: "The analysis of such technologies of government requires a 'microphysics of power1, an attention to the complex of relays and interdependencies which enable programmes of government to act upon and intervene upon those places, persons and populations which are the i r concern." 2. For another example of a study that employs this form of discourse analysis as it relates to media coverage of refugees -although in an American context- see Hufker and Cavender (1991). 3. Two terms often used in discussions of discourse analysis are 'discourse1 and 'text'. It is important to distinguish the two. A discourse is actually realized in or finds its expression in text (Parker, 1990a: 194). One text, therefore, may be the expression or realization of several contradictory or competing discourses (Kress, 1985: 27). Discourse analysis is not meant to refer here to a method drawn from hermeneutics or, more specifically, the work of Gadamer (1975) (cf. Sharpies, 1994). In drawing attention to discourse this thesis is not concerned with "the problem of meaning" (Miller and Rose, 1990b: 7). 4. It is important to note that, like 'the State', 'policy' is an historical effect of certain governmental rationalities. I used 'policy' to have a common point of reference for the interviews. 5. An alternative to this part of the interview would have been to observe authorities doing relevant refugee work (ie. determination, selection and resettlement). This would have presented great difficulties, not the least of which would have been the prohibitive cost and convincing university bureaucrats that such an activity would be in line with ethical guidelines. 6. I experienced several minor difficulties while conducting these interviews: Some interviewees were simply unwilling to be interviewed; prospective interviewees, after agreeing to an interview, failed to show up at the time and place agreed to; and several interviewees chose not to respond to the occasional question. 7. A positivist social scientist might refer to the resulting sample of interviewees as a nonprobability purposive sample (cf. Weisberg et al., 1989: 32-3). 8. When I directed inquiries to this authority and another whom I had not interviewed previously, I first identified myself, my university affiliation, and briefly described my research endeavour before asking questions about a given technology. I also had them read and complete a consent form. 80 9. It is, of course, bordering on the contradictory to use these databases "as a tool, a technological fix, that perfectly reproduces printed information" as though it will "provide a transparency of knowledge or data reproduction without any loss of meaning or corruption of text" (Poster, 1990: 96) and then in chapter six problematize determination practices in terms of the documentation centres that are highly dependent on similar databases. 81 CHAPTER FOUR: THE INTERNATIONAL REFUGEE REGIME AND OTHER CONDITIONS OF POSSIBILITY Particular governmental domains emerge under certain conditions of possibility. Circumstances in apparently unrelated realms can converge to give rise to a new regime. The Canadian refugee regime is an example. At present it co-exists with a massive international refugee regime and regimes of other nations. In the area of refugee aid and development there is little distinction to be made between the two. In contrast, refugee selection, determination, and resettlement programmes are presently understood as national (ie. Canadian) and distinct from the continuing international governance of refugee crises and situations. While the Canadian regime really only began to emerge with the release of the White Paper on Immigration in 1966, an international regime had already begun to take shape around the time of the First World War and had fully formed by the mid-1960s. While the Canadian regime is the main focus of this thesis, the formation of this international refugee regime and its object warrants some attention in order to show how it was to become a condition of possibility under which the Canadian regime could begin to emerge. At the same time, discussion of the formation of the international refugee regime presents an opportunity to examine the relevance of concepts and themes drawn from the governmentality literature in analyzing an international (as opposed to a national) regime. The latter has been ignored in studies in governmentality, remaining the sole preserve of 'international relations theory1. 82 This chapter begins with a discussion of the emergence of the international regime and its object, the refugee,3 as well as Canada's (mostly lack of) participation within this regime as a nation from the First World War to the 1960s. Some concepts and themes drawn from the governmentality literature are then elaborated in the context of this international regime. Finally, other conditions of possibility for the emergence of a Canadian refugee regime -during a period between the mid-1960s and the early 1970s- are elaborated: a renewed concern about Canada's population size and composition; a recognized scarcity of desired labour in the West for the national capitalist economy; and a renewed interest in becoming an influential citizen in the community of nations. Throughout most of the nineteenth century Europeans did not regard large masses of human beings forced to migrate to seek refuge from persecution as experiencing victimization of a kind distinguishable from other forms of oppression (Marrus, 1985: 8). As the historian Marrus (1985: 8) writes: People were used to being ruled by foreigners of one sort or another in the days when the next valley or a nearby town was 'foreign' in most respects. Assuming refugees survived to find refuge, their trials merged with the history of Jews, Moriscos, Protestants and Catholics, other resident aliens, or whomever; they would not have understood our modern preoccupation with a condition they shared- that of the refugee. Zolberg et al. (1989: 3), in discussing this matter in regard to the late nineteenth century, similarly note: "The need for precise categories did not arise under conditions of unlimited immigration, and so the distinction was not made explicit." But while these 83 Europeans may not have 'understood our modern preoccupation' with the refugee condition, and there may not have been a 'need for precise categories' during or before this period, it is my contention that there is a sense in which there was no refugee condition and no such distinction to be made. Put simply, there were no refugees until the late nineteenth or early twentieth centuries. At first this may appear startling, for it seems to suggest oppression, migration, war, mental illness, impoverishment, physical torture and death -all those things presently associated with the refugee condition- did not exist before this period. This is not my intention. Nor is it to argue that refugees are 'social constructions' -as that phrase has come to be understood -, that they had always existed, but * increasingly came to be recognized by Western authorities only after this period. Instead, this statement is meant to convey the notion that before the late nineteenth or early twentieth centuries, one simply could not be a refugee as one can be one today, that it was not an 'opportunity' that presented itself. It is to say that after the late nineteenth or early twentieth centuries refugees came into being at the same time as the category of refugee was being invented, that the two went hand in hand (cf. Hacking, 1986: 228). During the nineteenth century there were no significant administrative barriers to migration between nations in Europe (Marrus, 1985: 92). Regarding Britain, for example, the historian Porter (1979: 4) writes: For the best part of the nineteenth century. . . the British government deliberately denied itself any control 84 over immigration, and appeared indeed for the most part to take no interest in it. There were a few individuals, known as 'exiles', during this period. However, upon becoming such, these individuals tended to be well-received by political authorities and the citizenry of liberal European nations (Sjoberg, 1991: 24; Marrus, 1985: 14-15). A somewhat glaring exception to this was briefly seen in Britain in the mid-Victorian period when legislation to bar their entry appeared. But even at this point, as Porter (1979: 3) notes: ...these Acts were considered to be justified only by very extraordinary circumstances, and in 'normal' times it was not thought proper that governments should have any powers at all to exclude or expel aliens, except under extradition treaties for crimes committed abroad. Consequently from 1826 until 1848, and again from 1850 to 1905, there was nothing on the statute book to enable the executive to prevent aliens from coming and staying in Britain as they like. This short-lived 1848 British legislation was not an element of a programme that imagined doing something about masses of innocents forced to seek refuge, entities who now would be understood as refugees. Around the time of the First World War, however, something quite remarkable started to happen. An international refugee regime and the refugee, a domain and a 'kind' of person never seen before, began to emerge. THE INTERNATIONAL REGIME Immediately following the First World War international relief organizations such as the Red Cross, Near East Relief, the American Relief Administration, and Save the Children Fund delivered relief to the thousands in Europe who had left Russia after the Russian Revolution, civil war, and famine (Simpson, 1939: 198-9). But 85 relief resources were nearing exhaustion (Sjoberg, 1991: 24-5; Marrus, 1985: 58) and these masses could not simply be repatriated to Russia because of a 1921 All Russian Central Executive Committee decree that revoked Soviet citizenship for those who had resided outside the Soviet Union for greater than five years or had left Russia after November 7, 1917 without permission (Kapriellan-Churchill, 1994: 283; Marrus, 1985: 88). Thousands of those receiving relief were without valid passports. These documents had become widely required again for travel from one nation to another at the start of the War (Gordenker, 1987: 20; Cohen, 1987: 145; Marrus, 1985: 92). Passports were required before this, of course, including within the territories of Tsarist and Ottoman Empires, but when travelling from one European nation to another they generally were not essential (Marrus, 1985: 92). Because of this sudden development, these masses experienced severe difficulties living and working in a particular European nation or migrating to another (KaprielIan-Churchill, 1994: 283). Among the varied authorities already dealing with these masses, this was deemed a technical problem requiring a practical solution. Though there was no fanfare surrounding it, a new type of organization and a new 'kind' of person were about to be brought into being in Europe. In response to this problem, in February, 1921, a meeting was held by the International Committee of the Red Cross and League of Red Cross Societies (Simpson, 1939: 199). It was decided at this meeting that an office of High Commissioner for Refugees should be 86 created under the League of Nations. The Office of the "High Commissioner on behalf of the League in connection with the problems of Russian refugees in Europe" (Sjoberg, 1991: 25) was subsequently created in September, 1921 (Simpson, 1939: 198). Within the Office programme, besides coordinating the delivery of relief, the Office was imagined easing the resettlement of Russians who had no identity documents and were being refused entry by other nations (Sjoberg, 1991: 26). This was to be accomplished through the invention of an identity document for refugees, the 'Nansen Passport' or 'Nansen Certificate' (Kaprielian-Churchill, 1994; Marrus, 1985: 94). The League Office was to become involved in the examination and screening of refugees, and the issuing of this document. From there, refugees were foreseen being admitted to live and work in nations such as France and Belgium (Kaprielian-Churchill, 1994: 303). The use of this special passport gradually expanded. In 1924, thirty-eight nations were accepting these documents for Armenians who were without national passports following their flight from Armenia and by 1928 fifty-one nations had agreed to issue and recognize this document for Russians (Marrus, 1985: 94-5). A condition of possibility for the emergence of an international refugee regime was the rise of the unique features of the European nation-state form (Zolberg, 1983), one of which was citizenship. A notion of citizenship began to be developed earlier within what has been termed the 'absolutist state' form (Anderson, 1974), but it was in the nation-state that it was fully realized 87 (Giddens, 1985: 83-121). Before this, persons did not necessarily understand themselves as citizens, as Giddens (1985: 210) writes: In many cases the mass of the population of traditional states did not know themselves to be 'citizens' of those states, nor did it matter particularly to the continuity of power within them. With the rise of the nation-state form also came intensively surveilled borders where there were previously only frontiers (Giddens, 1985: 4). One way that both citizenship and the policing of borders associated with the nation-state form were made possible was through the invention of the passport. Though there were undoubtedly other conditions of possibility of the rise of an international refugee regime, none of which will be discussed here, it is the case that without the passport and the prevalence of notions of citizenship and borders there would not have been a technical problem in this instance demanding a practical solution. The principles for the Office that were laid down by the League in 1921 included "protection of certain classes of refugees" (Simpson, 1939: 192). From the outset, the Office's programme imagined not a universal category of person, but specific groups defined by ethnicity or nationality. In fact, throughout the 1920s the Office's programme was limited to several specific groups, at first Russians, and beginning in May, 1924, Armenians (Kaprielian-Churchill, 1994: 284). Another principle established in 1921 was that League intervention would be "'on a temporary basis.'" (as cited in Simpson, 1939: 192). The programme associated with this Office did not address a permanent problem requiring continuous intervention (Sjoberg, 1991: 37). The High Commission Office had 88 only a ten-year mandate. It was inconceivable at this juncture that the specific refugee problems being targeted would continue beyond this period (Sjoberg, 1991: 37). This first programme was exceedingly modest. In 1922, for example, the High Commissioner was to be granted a trivial operating budget of four thousand francs and no salary (Simpson, 1939: 200). It was another explicit principle of this League Office's programme that private international voluntary organizations such as the Red Cross and others mentioned above would continue to have responsibility for funding the relief effort (Sjoberg, 1991: 25; Marrus, 1985: 89; Holborn, 1975: 7). 'Continue', because immediately before the creation of the League Office, for example, during the repatriation of Russian prisoners of war, "it was the International Red Cross that organized assembly camps, provided food, set up disinfecting stations and supplied transport for repatriates" (Marrus, 1985: 88). As the historian Sjoberg (1991: 227) writes: "At the time of the outbreak of World War One all help for refugees was still based on private philanthropic efforts" (cf. Bramwell, 1988: 2). Relief to maintain refugees' lives was not funded through public donations of Western nations, later in the 1920s, either; it remained a philanthropic endeavour financed through private donations, with the Office merely coordinating delivery among organizations of various types. This public/private opposition, this discursive fact, will be seen again and again within programmes targeting the international refugee. Its significance is discussed later in this chapter. 89 In the late 1920s another agency, the Nansen International Office for Refugees, replaced the first League Office (Sjoberg, 1991: 37). The new Office was "concerned with Russian, Armenian, and some smaller groups of Refugees" (Simpson, 1939: 191). The League Assembly of 1931 set the last day of 1938 as the date of its termination (Simpson, 1939: 207). Two other organizations created by the League Assembly followed: the 'Nansen International Office for Refugees' and the 'High Commissioner for Refugees coming from Germany'. Each of these League organizations was also created as temporary (Sjoberg, 1991: 28). In May, 1939 a final League organization, the 'High Commissioner for Refugees under the protection of the League of Nations', was established on a similar temporary basis: It was planned to last only five years (Sjoberg, 1991: 37). There was never an attempt associated with these League organizations to create a general definition of a refugee (Sjoberg, 1991: 37-38). At this point refugee problems were atypical, ethnic or nationality-specific, and limited to Europe. There is a sense in which there were no permanent refugee problems outside Europe until the 1950s. By sending a one-time twenty-five thousand dollars to the High Commissioner for Refugees to provide housing and food in the field in September, 1922 -the first year of the operation of the Office-Canada had provided some temporary, limited public funding (Kaprielian-Churchill, 1994: 293-4). Canadian authorities, however, refused to accept the Nansen Passport created under this Office (Kaprielian-Churchill, 1994: 285). Their rationale for this 90 refusal centred on 'returnability' or the capacity to deport holders of the Nansen Passport after their admittance to Canada. The suggestion was that if such individuals stepped off the path to becoming liberal citizens once in Canada, they could not be easily disposed of (See Kaprielian-Churchill, 1994: 286-92). An official of the Canadian Dominion later signed an agreement at a European conference on the matter of Nansen Passports in 1926, but unlike other Western nations whose representatives had signed, Canadian Parliament did not immediately ratify the agreement (Kaprielian-Churchill, 1994: 286). Instead, authorities of the Dominion made clear they would accept holders of these documents only if they could be admitted under the same conditions and through the same mechanisms as immigrants (Kaprielian-Churchill, 1994: 286-7). As the historian Kaprielian-Churchill (1994: 288) writes: Refugees, the Immigration Department held, must comply with all regulations governing voluntary immigrants, including the valid passport ruling. They were not to be 'considered apart from other ordinary immigrants'. From the creation of the first League Office in the 1920s through to 1959, Canadian state officials did not keep refugee-specific statistics or records (Simpson, 1939: 489; Vernant, 1953: 557; Fairclough, 1963: 2; Kaprielian-Churchill, 1994: 298; Canada, 1995b: 1; cf. Dirks, 1977: 228-30) and it was not until the early 1970s that there was to be a refugee-specific category in Canadian 7 8 law and a refugee-specific state0 agency. As Dirks (1995: 61) writes: For most of Canada's immigration history, neither politicians nor officials made any distinction between immigrants and refugees. The reasons for people's 91 departures from their homelands seldom interested officials responsible for processing those who wanted to settle in Canada. During the 1920s Canadian officials selected and admitted Mennonites from Russia (Dirks, 1977: 38-9) and over one thousand Armenians (Kapriellian-Churchill, 1994). But while both admitted groups have been referred to as refugees by Canadian Department of Immigration officials and Canadian historians (e.g., Dirks, 1977; Kapriellan-Churchill, 1994) since the emergence in the late 1960s and early 1970s of a Canadian regime, it is important to recognize that they were selected and admitted to fill immigrant categories-they were deemed either to have fulfilled immigrant requirements themselves and/or to have had an immigrant sponsor within Canada at the time (Kaprielian-Churchill, 1994: 303). This suggests that for most of the twentieth century there is a sense in which there were no Canadian refugees. By 1938 the Nansen Refugee Passport had not been recognized by Canadian political authorities, nor had the 1926 international agreement been ratified in Parliament (Kaprielian-Churchill, 1994: 292). Following an international conference in Evian, France in March, 1938 the Inter-governmental Committee on Refugees (IGCR) was established (Sjoberg, 1991: 13). The IGCR was not part of a larger programme to deal with refugees as a universal category. Like the previous League organizations it targeted specific groups, namely the thousands of German and Austrian Jews fleeing Nazi-dominated territories (Rystad, 1989: 8). The IGCR programme foresaw resettling these groups, in part by seeking to mediate between 92 German political authorities and representatives of European nations of potential resettlement (Sjoberg, 1991: 14). The IGCR was the first international refugee agency created on a permanent foundation (Sjoberg, 1991: 13). In 1943 this agency was re-organized and its mandate broadened to deal with all refugees from the continuing Second World War (Rystad, 1989: 8). The IGCR became at this point the first refugee programme in the modern West to deploy a 'universal' definition of a refugee (Rystad, 1989: 8). It was not geographically universal, however, as the definition continued to be limited to 'events in Europe1 (Sjoberg, 1991: 16). The IGCR was not expected to become "a powerhouse of assistance and protection" (Gordenker, 1987: 22), but this programmatic re-organization does mark the first time regular public or state funding was explicitly marked for actual refugee relief (Sjoberg, 1991: 16). The IGCR, despite having been established as permanent, was dissolved in July, 1947 (Holborn, 1975: 30). What is important, however, is that it was expected to be permanent when it was invented, indicating assumptions about international refugee problems and refugees were changing. Canadian delegates attended the conference in Evian, France in 1938 along with representatives of thirty-two other nations. According to the historian Skran (1995: 213): The Canadian delegation adopted the position that it was prepared to participate in a discussion of the refugee question, but under no circumstances would it agree to a special admissions quota for refugees. For Canadian political authorities, then, the 'refugee question' at this point continued to be regarded as a European problem, not one 93 that demanded assistance through resettlement from non-European nations such as Canada (Kelley, 1987: 5). Canada was a member of the IGCR until its expiration (Dirks, 1977: 102). The UN Relief and Rehabilitation Administration (UNRRA) was created in November, 1943 by representatives of forty-four nations in the midst of the Second World War (Gordenker, 1987: 22). Though in recent years this organization has been depicted as a prototype of the United Nations High Commission for Refugees (UNHCR), an organization discussed below, the UNRRA was not explicitly associated with a refugee programme. Easing the resettlement of refugees outside their nation of origin, for example, was not part of this organization's mandate (Gallagher, 1989: 579). As Gordenker (1987: 22-3) writes: A major part of its anticipated work was to be directed at displaced persons, not refugees with political fears... The UNRRA only incidentally provided assistance for refugees escaping from untenable political situations. The UNRRA programme was to entail "relief, rehabilitation and repatriation" (Gallagher, 1989: 579). To the extent that the UNRRA incidentally targeted those deemed to be refugees within international refugee schemes, it should be noted that it imagined sharing responsibility with private international voluntary organizations in the management of camps at this juncture (Gordenker, 1987: 23). Thus, in this instance as well, private organizations were foreseen working with an organization receiving public funding from Western nations. The UNRRA had a temporary mission and, like the IGCR, ceased operations in July, 1947 94 (Gordenker, 1987: 24-5). In 1947 the International Refugee Organization (IRO) was established "as a non-permanent Specialized Agency of the UN" (Holborn, 1975: 30). Therefore, like organizations that had come before -other than the IGCR- the IRO was created as temporary, explicitly avoiding permanent obligations for member nations within its programme. Initially it was granted three years of life. This was later extended by eighteen months (Gordenker, 1987: 26). Despite its name, its programme imagined maintaining the lives and aiding the resettlement of 'displaced persons' (Gallagher, 1989: 579; Gordenker, 1987: 27; Holborn, 1975: 32), not necessarily 'refugees' (Gallagher, 1989: 579). Like the UNRRA and other organizations discussed above, the IRO programme foresaw the involvement of both privately and publicly funded agencies. According to the historian Holborn (1975: 31-2): This complex operation could be accomplished only through joint effort by the member nations of the IRO, the governments of asylum and resettlement, the international and national voluntary agencies, and several UN organizations. The IRO led to the resettlement of more than one million human beings and was dissolved in 1951 (Gallagher, 1989: 579). Canada was one of the eighteen original member nations of the IRO (Sjoberg, 1991: 222n). In November, 1946 Canadian political authorities issued emergency orders to bring 'displaced persons' to Canada (Wyman, 1988: 190; Hawkins, 1972: 90) and two Canadian selection programmes subsequently appeared: a bulk labour system and a close-relatives plan (Wyman, 1988: 190). In January, 1947 a 95 Canadian immigration official arrived in Germany and by March two teams were actively selecting people from camps (Wyman, 1988: 191). Initially, selection and admission of displaced persons were carried out by reference to existing immigration legislation covering dependents of Canadians (Green, 1976: 28). In June, 1947, however, an Order-in-Council, immediately followed by several others, proclaimed the specified admission of 'displaced persons', which all subsequent arrivals were referred to (Vernant, 1953: 559). Regarding this selection, according to the historian Wyman (See also Luciuk, 1986: 476-7), there was a "lack of public debate on the issue; the federal cabinet made all major decisions on which nationalities and ethnic groups could be admitted" (Wyman, 1988: 190). The Canadian federal cabinet had decided among themselves, as a critic at the time asserted, "'to let these touchy questions be settled in private'" (Wyman, 1988: 191). As seen in the next chapter, this is not unlike how refugees have been selected abroad for resettlement in Canada since the mid-1970s. Later in July, 1947 Canada became part of the IRO General Council and Executive Committee (Dirks, 1977: 121). Though more than one hundred and fifty thousand displaced persons were eventually selected and resettled (Marrus, 1985: 344-5), Canada's response, like the IRO itself, was assumed to be special and temporary. During its life, Canada contributed seventy staff and some eighteen million dollars to IRO operations (Canada, 1974: 101). As the IRO was dissolving, new international refugee schemes foresaw the invention of new international agencies.10 In the fall of 1951 political 96 authorities of the United States sponsored a conference in Brussels to discuss the creation of one of these organizations (Dirks, 1977: 186). In 1952, the Inter-governmental Committee for European Migration (ICEM) was established by Western nations to assist in the transportation and resettlement of European refugees and other migrants (Gallagher, 1989: 580). ICEM took over much of the staff and technologies that the defunct IRO had left behind (Dirks, 1977: 186; Marrus, 1985: 365). ICEM was foreseen screening refugees and providing air and ocean transport facilities. It was also imagined mediating with political authorities of nations, such as West Germany and Italy, within which refugee camps were situated ( „ ™ 1 9SS: 365). These tas.s were ™ * to ^ o u ^ e t he UNHCR programme that had appeared a year earlier (see below) (Dirks, 1977: 185-6). / Canada was a founding member of ICEM (Dirks, 1977: 187). Canada's representatives in this organization were drawn from the Canadian Department of Immigration. In 1953, and again in 1955, these officials advocated Canada's withdrawal. The rationale for this was the inconsistency in the interests of the two bureaucracies evinced by ICEM representatives' efforts to take over responsibility for selecting refugees and immigrants destined for resettlement in Canada (Dirks, 1977: 187-8). Canadian Department of Immigration officials regarded such selection to be a national concern, not an international responsibility (Dirks, 1977: 1987). Canadian Department of External Affairs officials, however, occasionally resisted withdrawal, suggesting in 1953 that doing so 97 would "no doubt engender much ill will, particularly on the part of some of our overcrowded NATO allies" (as cited in Dirks, 1977: 187). Canada, they argued, would be viewed by other Western nations as rejecting its responsibilities. Immigration officials argued in 1955 that Canada's public funding for ICEM should not subsidize transportation of selected refugees to other nations such as Australia and that this was interfering with efforts to select those thought desirable for Canada. A decision was made at this time to withdraw from ICEM to become effective in 1957, but an interdepartmental committee comprising representatives from both Departments -Immigration and External Affairs- decided at a meeting in 1957 to suggest to Cabinet that Canadian participation continue. In 1962 Canada finally withdrew, the rationale this time being that the refugee problem in Europe had largely diminished and that ICEM was at this point mostly assisting in the transportation of migrants within Europe (Dirks, 1977: 226). In 1970, rejecting the idea of rejoining ICEM1 as a full member, Canada's Minister of Immigration wrote in a background paper that "this might lead to the introduction of foreign matters which would be contrary to the principle that immigration is essentially a matter of domestic concern" (as cited in Hawkins, 1989: 160). This concern over Canada's membership in ICEM shows what in later chapters will become important- that Canadian selection of refugees from abroad lies between two broad spheres of power relations, an international and a national one. In 1951 the United Nations approved a proposal for the 98 creation of the UNHCR (Holborn, 1975: 87). At this point the UNHCR, like the majority of refugee organizations that preceded it, was created as temporary, to last only three years. It had tentative beginnings (Gallagher, 1989: 580; Canada, 1986e). As the historian Gordenker (1987: 28) writes, to form the UNHCR ...they rescued some bits and pieces of IRO practice while allowing others to disappear. From the rubble [it] was constructed, slowly and at irregular intervals... The UNHCR was created as 'non-operational' which meant that, unlike the IRO, it was not expected to deliver emergency aid to displaced persons or refugees directly (Adelman, 1982: 32). It was initially granted a "miniscule" (Gordenker, 1987: 29) operating budget in part to facilitate "the coordination of the efforts of private organizations concerned with the welfare of refugees" (Holborn, 1975: 101). Here again the public/private opposition is seen. The UNHCR was not initially granted authority to set up refugee camps or arrange for the transportation of refugees (Gallagher, 1989: 581). The UNHCR programme distinguished voluntary repatriation, local integration, and third nation resettlement as 'solutions' to international refugee problems (Gallagher, 1989: 581). In the years between 1951 and 1956 there was little support for it from its member nations (Marrus, 1985: 351; Gallagher, 1989: 582). In 1956, the UNHCR's 'new' temporary mandate was again set to expire, but it happened that this was the same year hundreds of thousands of Hungarians fled from Hungary to Austria and Yugoslavia (Gallagher, 1989: 582). After a request from Austria, the UN General Assembly passed a resolution in November, 1956 allowing the 99 UNHCR to continue to receive public funding and to coordinate emergency relief for Hungarians in these areas (Gallagher, 1989: 582). Canadian representatives eventually selected and admitted more than thirty thousand Hungarian refugees in conjunction with this event (Dirks, 1984: 125). As with other responses that occurred prior to release of the Green Paper in 1975, such as those associated with World Refugee Year in 1959, the 1968 crisis in Czechoslovakia, and the 1973 crisis in Chile, this effort was made possible through a special, temporary, Order-in-Council (Green, 1976: 32) that, as such, was not expected to be repeated in kind. In keeping with this, at this point Canada had not acceded to the 1951 UN Convention, although Canadian representatives had been involved in writing it. Indeed Canada did not sign it or the 1967 Protocol until September, 196912 (Dirks, 1977: 188; Immigration Manual. IS. 3.05; Hathaway, 1992: 73). Included in this 1951 international agreement was a provision known as non-refoulement: the prohibition of the return of refugees to "the frontiers or territories where life or freedom would be threatened on account of race, religion, nationality, membership of a particular social group, or political opinion" (as cited in Hathaway, 1992). As had happened concerning the League Office and the ratification of an international agreement in relation to the Nansen Passport decades earlier, the technical issue of ' returnability' -or capacity to deport individuals- was the Department of Immigration officials' rationale for resisting efforts to accede during this period 100 (Hathaway, 1992: 73; Kelley, 1987: 5). If Canadian representatives were to sign the Convention, Inunigration officials argued, refugees who went astray after being admitted to Canada could not be deported because of the non-refoulement provision. This resistance to a formal, explicit refugee policy among Department of Immigration officials has continued in the decades that have followed. Canadian political authorities denied an invitation to become part of the Executive Committee of the UNHCR in 1955 (Dirks, 1977: 184), doing so only after the events in Hungary in 1957. Since then Canada has served as a member of the UNHCR's Executive Committee. In 1952 Canadian political authorities rejected a proposal that would have seen a representative of the UNHCR operate in Canada (Dirks, 1977: 185). A UNHCR 'correspondent' was briefly allowed to do so during Canada's temporary World Refugee Year programme (Dirks, 1977: 185) and a part-time 'correspondent' was permitted to enter and function in Canada beginning in 1962 (Adelman, 1982: 41). It was only following release of the Green Paper in 1975, however, -the point at which 'refugee' became recognized as a distinct category in Canadian law and refugee-specific state bodies, namely Refugee Affairs and the Refugee Status Advisory Committee, were established- that a full-time UNHCR representative was permitted to regularly operate in Ottawa (interview 30/08/96; Adelman, 1982: 41). Immediately following intervention associated with this Hungarian movement, in 1957 the UNHCR began emergency refugee 101 efforts to aid Algerians in Morocco and Tunisia (Rystad, 1989: 9). This was to be the first time an international refugee organization targeted refugees in a socio-political space outside Europe (Rystad, 1989: 9). The UN Assembly followed this by allowing the UNHCR to aid Chinese who had migrated to Hong Kong in 1959 (Gallagher, 1989: 582). Besides Africa and Asia, the UNHCR subsequently began emergency refugee operations in other non-Western regions of the world such as Latin America (Gallagher, 1989: 583; Melander, 1988: 10). In 1964 it was decided that both the time and geographical limitations on the UNHCR programme would be removed. This took the form of the 1967 UN Protocol (Gallagher, 1989: 583). In the 1960s as the refugee problem disappeared in Europe, refugee crises, movements, and refugees continued to appear in various non-Western spaces. The 1951 UN Refugee Convention, through which the UNHCR had operated, referred to human beings who had been affected by "events occurring in Europe before 1 January 1951," but as the historian Hawkins (1989: 158) writes: Within a few years of the framing of this definition, it was becoming increasingly clear that all refugee movements did not originate in World War Two or its aftermath, and that refugees might well be a continuing feature of the international scene. From a "non-operational agency with limited resources" (Rystad, 1989: 7) in 1951, the UNHCR has since grown in scope and size. The present UNHCR's mandate remains 'non-political1, but its resources have grown considerably: In 1996 the annual budget was 1.43 billion dollars U.S. (UNHCR, 1996: 12). The present organization is headed 102 by a High Commissioner appointed by the United Nations Security Council. The High Commissioner reports to the Executive Committee (Excom), currently comprising forty-four nations, that approve the annual UNHCR budget (Cuncliffe, 1995: 283). While the administrative budget comes directly from the United Nations, the operational budget comes from voluntary donations from public sources of EXCOM nations, most of which -other than Japan- are Western (Gordenker, 1987: 35).14 The scope of UNHCR refugee programmes has also increased significantly since 1951 (Adelman, 1982: 32; Cuncliffe, 1995). By the 1970s the UNHCR, as well as regional, national, and international governmental and non-governmental organizations, had become linked into a massive refugee aid and development network (Gallagher, 1989: 584). Whereas in 1962 ninety-two organizations were involved in UNHCR programmes in sixteen nations (Tabori, 1972: 298), in 1996 the UNHCR has four hundred and ninety-five "operating partners" (NGOs), as well as two hundred and fifty-five offices in one hundred and twenty-three nations, targeting the lives of millions of refugees on several continents (UNHCR, 1996: 12). THE INTERNATIONAL REFUGEE Contemporaneous with and inseparable from the emergence of the international refugee regime described above was the emergence of the refugee as an object of the aspirations of Western authorities and a new 'kind' of person or identity. Objects of government are not self-evident, ready-made, or natural (Hunt, 1993: 316) and each has a unique historical trajectory (Hacking, 1986). Just as there 103 is no original international refugee organization that determines the present form of the UNHCR, there is no 'proto-refugee' that determines the present character of the international refugee (Malkki, 1995: 497). The term 'refugee' was originally coined in the West to specify French Protestants who fled from the forced conversion policy of the French state in the late seventeenth century (Wong, 1989: 279). In the eighteenth century the term 'emigres' was used by individuals who had left France following the Revolution to refer to themselves (Marrus, 1985: 9). It was not until the end of the eighteenth century, following the American Revolution, that 'refugee' began to be used to refer to human beings that 'leave their country in times of distress' . Despite this usage, through the first half of the nineteenth century French and English dictionaries continued to refer to 'refugees' as those people who were victims of the revocation of the Edict of Nantes (Marrus, 1985: 9). Since this was an event obviously limited in time and space, this suggests 'refugee' did not yet have a universal meaning. Throughout the nineteenth century German dictionaries included the French word 'refugies' and this earlier reference to the French Protestants (Marrus, 1985: 9). It was only after the First World War that 'Fluchtling' -the modern German term for refugee- came to dominate other German terms, including 'Heimatlos' or 'staatenlos' (Marrus, 1985: 9). To this point, then, 'refugee' did not refer to a legitimate object of government and to a 'kind' of person. As suggested earlier, the nineteenth century was to 104 involve 'exiles', understood as "individuals who left their native country for political reasons" (Marrus, 1985: 9). Regarding the difference between the two Said (1984: 160) writes: Exile originated in the age-old practice of banishment. Once banished, the exile lives an anomalous and miserable life, with the stigma of being an outsider. Refugees, on the other hand, are the creation of the twentieth century state. The word 'refugee' has become a political one, suggesting large herds of innocent and bewildered people requiring urgent international assistance, whereas 'exile', carries with it, I think, a touch of solitude and spirituality. The millions of Jews who left Tsarist Russia and Eastern Europe in the years immediately preceding the First World War were occasionally called 'refugees', but they were also referred to as 'emigres' (Marrus, 1985: 9; Skran, 1995: 13). The latter term was later widely used prior to and during the Second World War to refer to Jews who had fled to the United States from Nazi-dominated territories in Europe, but it is only infrequently used today. The discourse of 'exile' presently has currency, but mostly in referring to a condition in which individuals find themselves (e.g., Bauman, 1992: 225) or to a stage within what is now presumed to be a typical 'refugee experience' (e.g., Stein, 1986; Immigration Selection Manual), rather than to large masses of innocent human beings forced by something or someone to seek refuge and to related programmes that target their lives (cf. Wong, 1989: 279; cf. Llambias-Wolff, 1990; Kunz, 1981; Tabori, 1972; See also Refuge. 1986 6(2): 11). Immediately following the First World War, as refugee organizations were invented within international programmes, 105 'refugee1 began to subsume -albeit not completely- 'emigres', 'exile' and similar terms (cf. Zolberg et al., 1989: 6). At this point, however, the refugee was still not (yet) a self-evident object/identity. As the historian Kapriellan-Churchill (1994: 298) writes about the mid-1920s in Europe, [w]hile the term 'refugee' was widely used in immigration circles... [it] was used inconsistently by international agencies, government officials, and non-government organizations alike. Immediately following the Second World War, 'displaced person' was introduced to refer to groups of Europeans outside their nations of origin in need of relief and resettlement (Gordenker, 1987: 23). After 1947 'displaced person' also began to be subsumed by 'refugee'. By the 1960s an international refugee regime had solidified, a new administrable domain had fully formed. So had a new 'kind' of person. The international refugee had become 'always-already' there, its existence as an object of government and an identity widely and fully acknowledged. GOVERNMENTALITY AND THE INTERNATIONAL REFUGEE REGIME The discussion above also shows that international refugee programmes did not emerge separate from certain governmental rationalities, as though this realm was one of anarchy, as international relations theorists might suggest. What is consistently seen in these programmes above is an assumed opposition between either 'public' and 'private1, 'governmental' and 'non-governmental', or 'political' and 'non-political' authorities and organizations. From the inception of international refugee programmes, private international organizations, later to 106 be referred to as 'NGOs', were imagined carrying out tasks distinct from those thought appropriate to governmental or primarily public bodies. The presence of this opposition has continued. The following excerpt from a 1981 article describing NGO/UNHCR consultations held in Geneva in Refuge, 'Canada's national newsletter on refugees', for example, suggests how these NGOs have been conceived within these programmes. It notes ...the advantages of the diversity and independence of NGOs... the flexibility to assist people... to act quickly and pragmatically, since they often already have connections within the relevant country, and to deal with refugees in an informal, person-to-person manner (Refuge. 1981 1(3): 6). In a study of American refugee programmes in the post-war era, Nichols (1988: 87) gives a similar sense of NGOs' imagined role: The NGOs could move more swiftly, arrive at a disinterested appraisal of a situation free of political influence, experiment with an open style, and most importantly, make contact with the people at a level well below official governmental contacts. The division or opposition between, for example, 'governmental' and 'non-governmental' in the field between nations is not dissimilar to the public-private division assumed within Western nations such as Canada where liberalism prevails as a rationality of rule (see chapter five). Within the latter, only certain realms and authorities are assumed 'political' while others, such as the family and the church, are assumed 'non-political' and private (Valverde, 1996; Rose, 1987). Hirst and Thompson (1995: 432) have recently suggested that "the category of non-governmental organization [NGO] is a misnomer" because they also 'govern' as seen, for example, in the practices of Greenpeace and Oxfam in 107 relation to the ocean environment and to famine relief respectively. But 'NGO' is not a 'misnomer', an error or mistake in naming; it is the logical extension of liberal rationalities that are present in the international sphere. In this way international refugee programmes between nations during the twentieth century can be seen as broadly comparable with poor relief programmes within Canada (cf. Valverde, 1991a; Strange, 1995), and undoubtedly other Western nations. Within Canada, from the nineteenth century through to the 1920s, it was primarily private philanthropic organizations that sought to govern the conduct of 'the poor1. This was followed in the 1940s, as seen in the rise of the Welfare State that reached its zenith in the late 1960s (Valverde, 1991a), by a significant movement of responsibility for their conduct to the state. In relief of 'the poor1 since the 1970s, however, a shift in rationalities is evident that has involved a gradual, though uneven, shift in emphasis (back) to the private sector (e.g., introduction of provisions for private food banks and tax breaks allowing increases in donations to private charities). As seen above, through the 1920s and 1930s, as the international refugee regime was beginning to form, refugee practices were a decidedly private philanthropic endeavour, with public funds and agents being used mostly to coordinate efforts of voluntary private international organizations (or 'NGOs') or facilitate resettlement. With the creation of the IGCR in 1938, and within the IRO, UNHCR and other organizations that followed, 108 international refugee practice began to depend more on public funding and agents drawn from Western nations, although NGOs 10 continued to be imagined playing an important role. Despite its initial limited mandate, for example, the UNHCR gradually began, from at least 1957 on, to carry out refugee aid and development in the field on several continents (Tabori, 1972: 298). To the delivery of emergency aid or relief -shelter, food, water, medical intervention- were added extensive education and social services for refugees. In the mid-1980s, however, a shift in rationalities can begin to be seen. This is particularly transparent in relation to programmes that targeted African and Indo-chinese refugees (cf. Harrell-Bond, 1989: 55-6). In Africa this shift is seen in the appearance, in 1983, of a Voluntary Repatriation programme regarding Ethiopian refugees in refugee camps in Djibouti, one that was stepped-up beginning in 1986 (Harrell-Bond, 1987: 8-10; See also Kibreab, 1993: 328-9 on Somalia during this period). This was accompanied by claims that the regular dispensing of aid within refugee camps was ruining the potential of refugees, leading them to adopt a 'dependence mentality1 or experience 'dependency syndrome' (See, for example, Tucker, 1982; Harrell-Bond, 1986; Luciuk, 1986).19 It is also visible in the appearance in 1990 (Refuge. 1990 10(2): 18)20 of ?1 related programmes to 'empower' inhabitants of refugee camps in Africa, whereby refugees are imagined becoming more responsible for their own relief and development and less dependent on regular aid. The resemblance here to the contemporary critique of welfare and 109 its effects on 'the poor' within nations of the modern West (cf. Rose and Miller, 1992: 198; Dean, 1995: 578-9) is remarkable. Indo-chinese refugees had previously garnered considerable media attention through resettlement programmes launched by several Western nations including Canada (see chapter five) in the late 1970s and early 1980s. In the mid-1980s, however, voluntary 11 repatriation began to be touted more and resettlement less a 'solution' for the situation of those refugees still living in Hong Kong camps (local integration was apparently never deemed a 'solution' in this context) (Gilad, 1990b: 138; Canada, 1991: 7). A UNHCR Voluntary Repatriation programme appeared here in early 1989 (Hitchcox, 1990: 117; Cuncliffe, 1995: 286). NGOs operating in the region, in conjunction with the UNHCR programme, then sought to implement a new approach "under which the Khmer themselves took a far greater share of responsibility for the programmes at a management and decision-making level" (Fordham, 1990: 5). This was accompanied in 1992 by a large (25 percent) reduction in UNHCR funding for the educational and social programmes that had targeted Hong Kong refugee camps (Cuncliffe, 1995: 288). Though discussed in the next chapter in relation to the Canadian regime where similar mutations are occurring, it has to be noted here that what can be called 'advanced liberalism' is a rationality that demands a more active or enterprising subject (Rose, 1992: 145). The appearance of discourses of 'empowerment', 'voluntary' repatriation, and 'self-management' in Africa and Asia in the mid-1980s are undoubtedly manifestations of its rise in the 110 international refugee regime. Other general and more recent changes consistent with a shift from a 'liberal welfare1 to an 'advanced liberal' rationality (see chapter five) in this international realm are evident. A reduction of the operating budget of the UNHCR was first seen in 1985 (Kibreab, 1993: 329). In 1989 for the first time donor (mostly Western) nations refused to approve the UNHCR's annual operating budget and began calling for a further scaling back of its relief operations (Refuge. 1989 9(2): 2). During the early 1990s there have been shortfalls of one hundred and fifty million dollars between the predicted cost of UNHCR programmes and the available resources (Cuncliffe, 1995: 279). Following creation of a task force comprising representatives from donor nations and a subsequent audit of UNHCR records, it was decided to significantly reduce public contributions (by 20 percent) to the UNHCR (Cuncliffe, 1995: 285). 4 These mutations have been closely accompanied by a shift in knowledge, a new emphasis on early warning discussed in chapter six. Governmental programmes described above did not emerge separate from the invention of certain technologies, such as the refugee camp (see chapter six) and the refugee passport. These technologies were borrowed from other domains and modified within programmes of the League organizations, IRO, ICEM, UNHCR, and private voluntary international organizations such as the Red Cross. In myriad ways these technologies allowed these programmes to intervene upon diverse populations and places. Ill Take the refugee passport for example. The Nansen refugee passports created in July, 1922 specifically for refugees from Russia were a modification of the national passport. With this document the holder could move from one nation to another to work or join up with dispersed family members. The Nansen passport did not replace a national citizen's passport, however, in that it did not give its holder the right to return to the nation that issued the document without a special provision to that effect approved by the issuing nation (Kaprielian-Churchill, 1994: 283). The predecessor of the national passport was the passeport, a document that allowed a sailing vessel to leave or enter a seaport. The predecessor of the passeport was the 'safe-conduct', a written pass giving permission to individuals to travel through a dangerous area without risk of arrest by authorities. While in North America national passports were issued before the American Revolution (Rule et al., 1983: 225), prior to 1914 they were not generally required by authorities to travel from one nation to another in the modern West (Cohen, 1987: 145; Rule et al., 1983: 225; Gordenker, 1987: 20). The national passport was a technology that promised to produce certainty about individuals (Rule et al., 1983: 223), about who was a citizen, and therefore entitled to certain rights and benefits, and who was not. At the time of its invention the passport had several important features. It allowed for the storage of information specific to the Western citizen. The passport translated the characteristics and features of the citizen into material form. Besides a date and place of birth, given 112 names, and citizenship, an image of the bearer in the form of a photograph was permanently affixed to it. The passport served as a multi-page journal through which travel patterns could be recorded: which countries were entered, on what dates, and in what sequence (it was not, therefore, as dependent on files stored elsewhere as other kinds of documentary identification, such as the birth certificate). This was accomplished through the use of specially-designed and restricted ink stamps representing a given nation. As a slim, compact booklet this technology could be easily carried in the pocket or in the limited belongings of the citizen travelling abroad. Those travellers who lost their passport knew their identity could be placed in official jeopardy: "The consul banged on the table and said; 'If you've got no passport you're officially dead'" (Malkki, 1995: 495). Following the First World War "refugees were non-persons. They usually had no valid legal identification" (Gordenker, 1987: 20). But through the loss of a national passport an individual's identity was not only officially in doubt- the passport was officially and literally an identity document. Precisely like statistics (Hacking, 1986) that are discussed in chapter six, the national passport became a device for 'making up' the modern citizen and, with the slight modification mentioned above, the refugee. After the First World War national passports became part of a technical problem and, with a minor adjustment by Nansen, promised to become part of the solution. The appearance of a programme following the First World War that imagined solving a technical problem through the creation of 113 a special organization and passport was followed by identification of its failure. In fact, 'failure' was explicit in the rationale used to justify the establishment of each subsequent international refugee organization, including the UNHCR. This suggests 'failure' was a kind of catalyst or impetus for the re-thinking and re-engineering of programmes (O'Malley, 1996b: 310). Indeed, Miller and Rose (1990: 4) write that the 'failure' of one policy or set of policies is always linked to attempts to devise or propose programmes that would work better... the identification of failure is thus a central element in governmentality. In each instance above 'failure' led, not to programmes that foresaw giving up on the international refugee organization, passport and camp as ways of dealing with international refugee crises and situations, but to the invention of new programmes. In this way the international refugee organization, passport, and camp are similar to the modern prison. As Foucault (1977a: 277) writes: So successful has the prison been that, after a century and a half of 'failures', the prison still exists, producing the same results and there is the greatest reluctance to dispense with it. 'Failure', as seen in later chapters, is also consistently evident in the Canadian refugee regime. The consistent presence of the identification of 'failure' here, however, shows the relevance of a theme from the governmentality literature in making an international regime intelligible. What was it about ' refugeeness' that permitted it to gain such currency and to gradually subsume the other discourses mentioned above despite a succession of 'failures'? Whereas 'displaced 114 person' suggested a temporary aberration requiring 'replacement' and related terms such as 'exile' and 'emigres' similarly avoided the connotation of a cause of the condition, 'refugee' better suggested masses of innocents, of victims, forced to flee from something or someone. It was above all this feature that would allow it to be so versatile and to gain such currency. Though effectively deployed at the birth of the Soviet Union following the First World War, with the start of the Cold War, when the proliferation of nuclear weapons had rendered war between the international alliances a non-strategy (cf. Dillon, 1989; Hirst and Thompson, 1995), 'refugeeness' became more useful. This is especially so after the appearance of the UN Convention and the invention of the UNHCR and ICEM in the early 1950s. The identification of and responses to groups fleeing (or trying to flee) the Soviet Union and Eastern Bloc nations by the Western alliance of nations as refugees highlighted differences between the two alliances and their accompanying systems of government and citizens' ways of conducting themselves. It fit with a notion that the Soviet Union and nations of the Eastern Bloc, as well as their rulers and citizens -the 'something or someone' in this instance-were illiberal and uncivilized. If the type of governing and way of living that was predominant in these nations were adequate, why would their citizens attempt to flee to the West with nothing but a few belongings? Identification of refugees and responding to them as such simultaneously fostered the cohesion of the Western alliance of nations and rulers by assuring them that their approach 115 to government was the civilized and decidedly liberal one. 1Refugeeness' would later be useful during the rise of nationalism in Africa in the 1950s and 1960s, as well as in other non-Western spaces where previously there had been colonial arrangements. Since de-colonization Western authorities have assumed these new nations (with their rulers and citizens) to be deficient. Typical is Gasarasi's (1996: 4) comment in a recent article in Refuge: The word 'state' is written in quotation marks to indicate that there is doubt that most refugee-generating countries, particularly in Africa, qualify to be called as states, at least in the modern liberal sense (emphasis added). Post-colonial African nations have been assumed to require aid and development of one kind or other so as to be brought into line with requirements of liberal nationhood and associated ways of ruling and conduct among their citizens. The deployment of 'refugeeness' allowed Western 'non-political' intervention through the UNHCR and NGOs in these regions during this period where 'political' interference would have been impossible, first through emergency refugee aid and later through additional programmes of refugee development (cf. Malkki, 1995: 506). 'Refugeeness' was a way in which such intervention was made possible and justified. Since something or someone forced refugees in these socio-political spaces to flee it could be assumed that 'something or someone' was deficient and needed correction through intervention. Consider, for example, the following excerpt from a report on Africa from the mid-1960s by the High Commissioner for Refugees: 116 . . .the process of de-colonization and the internal strife that all too often accompanies it- these problems call for solutions that are humane, practical, immediate and adapted to the realities of life in Africa. In response to the drama of hundreds of thousands of men, women, and children fleeing their villages and their homelands, the international community has entrusted my Office with the task of helping these uprooted people, first to survive and then to create a new existence if it is clear there is no hope of their being able to return voluntarily to the homes they have left behind. For the countries that so spontaneously and so generously welcome them are in general beset with their own development problems and are not in a position to meet even the most vital needs of all the refugees that cross their borders unless they receive outside assistance (as cited in Tabori, 1972: 268). This is a prime example of governmental discourse in the international sphere. These technical 'problems' need 'practical' and 'immediate' 'solutions' provided by the UNHCR on behalf of an international community dominated by the West. But the UNHCR is not expected to merely solve these problems; it is foreseen doing no less than aiding the 'creation of a new existence'. The assumption above is that since de-colonization, African nations, and the ways of ruling and the conduct of citizens that prevail within them, continue to be inferior. In fact, in this instance both the new nations fled from and those fled to are deficient, for while the former are obviously so (otherwise why would citizens flee their villages with only a few belongings to cross into another nation?), the latter are also inadequate, unable to 'meet even the most vital needs', carrying the burden of 'their own development problems' as they are. Through 'outside assistance' from the UNHCR and NGOs, these newly de-colonized spaces and their inhabitants could be, in effect, re-colonized by the West, but this 117 time in a decidedly 'non-political' way. 'Outside assistance' for refugees in these spaces, precisely like 'non-political' philanthropy and later welfare for 'the poor' within Western liberal nations, promised to help create this new existence, to bring these nations, and the kinds of ruling and conduct taking place within them, into line and up to grade. Therefore, both in the intensifying international Cold War with the 'Second World', and in the nationalizing 'Third World' (which, of course, were not mutually exclusive historical events) 'refugeeness' became a Western moral-political tactic within a much broader objective, a way of relating to and intervening into that which was deemed illiberal and uncivilized and it was the accompanying assumption of being forced to flee from something or someone that allowed it to be so. Consistent with this assertion it will be noted that the relevant programmes and organizations described above were -without exception- products of Western rather than a community of equal nations (cf. Tabori, 1972: 293). When the first League Office of the High Commissioner was created in 1921, for example, the Soviet Republic, the nation of origin of the Russian refugees was neither a member of the League, nor recognized by political authorities of Western nations (Simpson, 1939: 191). Later the United States and Britain were to fund the operational expenditures of the IGCR after its reorganization exclusively (Sjoberg, 1991: 18n) and the U.S. was to be the largest contributor of funds to both the short-lived IRO and UNRRA (Gordenker, 1987: 24-5). The Soviet Union and its Eastern European allies did not 118 participate in the UNRRA, IGCR, IRO, or ICEM. Only Yugoslavia was involved in the UNHCR's operations, and by 1972 no nation from Asia had acceded to the UN Convention (Holborn, 1975: 177).30 Non-Western nations were uninvolved, were mere members -with the exception of Japan- rather than benefactors of the League and the latter organizations, or simply did not yet exist as nations. 1Refugeeness' has been deployed through these Western 'non-political' international organizations and directed at nations and groups in non-Western socio-political spaces or -in more recent decades- at individuals who have fled these spaces to the West. This is not simply to suggest ' ref ugeeness' has been a fraud, merely ideological cover for the exercise of power and, more specifically, camouflage for Western nations' domination of the non-Western world. Instead, it is to say that in the international sphere of power relations 'refugeeness' has been at once a target of governmental power and a tactic. Though this point is returned to briefly in the final chapter, the foregoing discussion should begin to suggest the relevance of concepts and themes drawn from the governmentality literature in making sense of international regimes. A CANADIAN REFUGEE REGIME Having made an extended excursion into the international refugee regime and its object, it is time to return to the Canadian refugee regime and the main focus of this thesis. By the mid-1960s international refugee crises and movements, those events deemed to be uncivilized, illiberal, and unpredictable, and refugees, their 119 human products, had not been eradicated. By this point Western political authorities generally had begun to recognize that the world was in such a state that it could be relied on to regularly produce refugee crises, refugee movements, and refugees. This is not to suggest a specific refugee crisis (e.g., sudden social and political upheaval in Uganda) was viewed by authorities as permanent, as lacking a capacity for resolution, but that refugee crises and movements were now generally expected to routinely and regularly arise. A single crisis had become regarded as typical rather than unique. From the mid-1960s Canadian programmes of selection and resettlement became predicated on this assumption, on the notion that there would be refugee crises in the world followed by mass movements of refugees into camps outside Europe who could then be selected and then resettled in Canada. It was at this juncture that permanent Canadian refugee programmes imagining the continuous selection of refugees became possible, for prior to this the product of these crises, the refugee, was assumed to stem only from a fleeting, atypical problem. In other words, before the 1960s the international regime and its object, the refugee, were still forming rather than self-evident. It is in this sense that the international regime was a condition of possibility for the emergence of a Canadian refugee regime. In the same way, it was not much before the 1960s that the Canadian Minister of Immigration could declare that "[g]reater attention will be given to the acceptance of refugees for resettlement in other parts of the world" (as cited in Dirks, 1977: 230). This is because before the 120 mid-1950s, -while there was undoubtedly oppression, migration, death, physical torture and so on-, there is a sense in which, just as there were no international refugee organizations, refugee camps and refugee passports, there were no refugees 'in other parts of the world' .32 Before the 1970s Canadian refugee inland determination did not exist. Before the mid-1960s refugee selection and resettlement -as seen above- were episodic and ad hoc. Until the early 1970s Canadian political authorities occasionally responded through Orders-in-Council, temporary selection guidelines (Dirks, 1977: 182; 225), and arrangements designed for the resettlement of immigrants (Interview 30/08/96). Before this there simply were no permanent, coherent Canadian refugee programmes of selection and resettlement. Beginning in the mid-1960s, however, there was a sign that the temporary, ad hoc character of responses to refugees was shifting and that a new Canadian governmental regime was developing. A programmatic statement, the 1966 White Paper on Immigration, explicitly refers to refugees: There will be continuing obligation to accept individuals or families who have fled their own country for one reason or another. However, neither the extent of the obligation nor our capacity to fulfil it can be predicted with any accuracy (as cited in Dirks, 1977: 229; emphasis added). Nevertheless the White Paper proceeds to suggest that [t]he past policies on refugees have been reactions to emergencies such as the Hungarian revolution or to appeals for aid from the United Nations High Commissioner for Refugees. If Canada is to accept its fair share of international responsibility for refugees... more formal arrangements than now exist are reguired, including the annual appropriation of funds on a continuing basis (as 121 cited in Dirks, 1977: 229; emphasis added). Canadian responses to refugees were no longer regarded as temporary and ad hoc, they were to be 'continuing1 and 'formally arranged'. Following the White Paper, however, no specific provisions for selecting, resettling and determining refugees were introduced in the revised immigration regulations introduced in October, 1967 (Canada, 1995b: 2; 1974: 104).33 A major document prepared by the federal Department of Immigration a few years later, the 1975 four-volume Green Paper, was a precursor to the 1976 Immigration Act. A chapter simply entitled 'Refugees' provides a historical account of Canada's sporadic, ad hoc responses to refugees. Another volume of the Paper, 'Immigration Policy Prospectives' , includes a recommendation that legal provisions for admitting refugees be created. Similar to the thrust of the White Paper, the programmatic Green Paper states: "Whatever the responses made by Canada or the international community in the past to refugee crises, there is no reason for complacency about the future" (Canada, 1974: 117; emphasis added). The Green Paper proceeds to suggest that "[t]he only certainty for which Canadian policy in this area must be prepared is the unhappy prospect that grave refugee crises will continue to erupt" (Canada, 1974: 117; emphasis added). Canadian refugee programmes suggested by the Green Paper foresaw the regular and continuous selection, resettlement, and determination of refugees to match the assumed permanence of international refugee crises and refugees that continued to appear despite the sustained efforts of international 122 refugee organizations. OTHER CONDITIONS OF POSSIBILITY Besides the full formation of an international refugee regime and its object, there were changes in several long-standing Canadian governmental fields that served as conditions of possibility for the rise of a Canadian regime during a period between the mid-1960s and the early 1970s. Population, the National Economy, and Immigration Population, the collective mass of citizens, displaced the family as an object of government in the West in the eighteenth century (Miller and Rose, 1990a: 167). Through the science of demography, population was rendered intelligible and made amenable to intervention. The size, rate of growth, distribution and composition of the population from then on could be thought and dreamed about. Immigration programmes imagined intervening in this domain. They sought to adjust the size of a national population and its rate of growth by adding or subtracting numbers of bodies and to adjust its age and sex composition to proportions demographic knowledge had revealed desirable. A field that had also developed long before the mid-1960s as an object of government and knowledge was the national economy (Miller and Rose, 1990b). Since it was consistently deemed unavailable within Canada, immigration programmes envisaged recruitment of labour of one kind or other from abroad for the national capitalist economy from the nineteenth through the twentieth centuries (Green, 1976; Satzewich, 1991a; 1991b). More 123 recent immigration programmes have also foreseen assisting the economy by explicitly recruiting capital(ists) through investor and entrepreneur categories (Simmons and Keohane, 1992: 423). Immigration programmes have targeted, then, both the composition and growth of the Canadian population and the well-being of the national economy. While enhancing the population was certainly recognized to be related to the well-being of the national economy and vice versa, one concern was not necessarily reduced to the other at every juncture. In the early to mid-1970s demography revealed the Canadian nation to have a declining birth rate and a renewed interest in population issues appeared (Hawkins, 1976: 46). Consistent with this latter development Canadian representatives attended the United Nations World Population Conference held at Bucharest during this period (Hawkins, 1988: 382). Although in 1947 there had been renewed interest in population growth through immigration (Vernant, 1953: 570; Kelley, 1987: 4), in the decades that followed there was less concern about it, as Dirks (1995: 113) writes: Interest in determining an optimal population size for Canada diminished substantially during the 1950s and 1960s, pushed aside partly by unprecedented economic growth and partly by a birth rate that, at the height of the 'baby boom1, approximated three children for every woman of child-bearing age. It is significant, therefore, that the Green Paper released in 1975 after two years of consultations with a variety of state and non-state organizations and agencies was entitled the 'Canadian Immigration and Population Study' (Canada, 1974). This programmatic Paper promised to stimulate a national debate on 124 immigration practices that included the development of a population policy (cf. Hawkins, 1988: 381). In 1975 a Special Joint Committee of Parliament -established to transform the Green Paper's recommendations into law (Canada, 1986d: 3)- also concluded that immigration and population policy were closely related (Hawkins, 1991a: 399). Reflecting this renewed interest in population, in a statement accompanying release of the Green Paper the Minister of Immigration suggested: "'We need a set of flexible guidelines to which policies which affect our population future may be related'" (as cited in Hawkins, 1988: 381). The historian Hawkins (1976: 46) remarks that "[n]othing so explicit had been said about the role of immigration in population growth throughout the postwar period." Securing the demographic needs of the nation would become one major objective of the Immigration Act a year later (Immigration Act, 1976)." Concern about population and an interest in comprehensive demographic analyses has continued since the 1970s (Dirks, 1995: 114) .36 During the time of the Green Paper incidents involving the sudden admittance to Canada of large numbers of refugees had been deemed to be increasing. The sudden selection and admittance of the Czechoslovakians, Tibetans, Ugandans, and Chileans had all occurred in a seven-year period preceding its release. The unpredictable character of these arrivals resulting from decisions made in the international sphere of politics and morality was hardly consistent with a renewed interest at this juncture in rationally managing the growth and composition of the population of 125 the nation. A coherent, visible refugee programme, however, was. During the period between the mid-1960s and early 1970s it had also been recognized that desired labour was scarce in the West, and it was concluded that it would have to be regularly recruited by Canada and other Western nations from outside (Hawkins, 1974: 146; Castles and Kosack, 1985). At this point the recognition that the world was producing crises, movements and, eventually, refugees in camps outside the West, fit well with this conclusion in that recruitment efforts would not necessarily have to be as far-ranging or costly as they might be otherwise (Interview 30/08/96)- refugees would already be assembled, sorted, and rendered docile in these disciplinary enclosures (see chapter six). These programmes anticipated taking advantage of those already in camps in various spaces outside the West through the 'successful establishment1 requirement that was based, in part, on a conception of desired labour. Foreign Affairs Foreign affairs, as a governmental domain, has been deemed to affect the health and security of the population. Responses to the 1956 Hungarian, 1968 Czechoslovakian, and other refugee movements in the decades following the Second World War had been recognized among some authorities to have enhanced Canada's role in relation to other Western nations. A regular, visible Canadian refugee selection programme could be even more advantageous in relations with these nations. A regular Canadian refugee programme along with accession to the 1951 Convention could do what an immigration 126 programme simply could not. It could regularly assist in alleviating the international refugee 'burden' (Girard, 1990: 113) that had by this time been recognized as permanent. This would allow Canada to assert itself as a nation where other -mostly Western- nations could not (Dirks, 1990: 95), not as a super-power such as the United States, but as a capable liberal citizen of an imagined international community. Such a programme that included accession to the Convention was promoted by officials of the Department of External Affairs before this. As seen earlier and as Hathaway (1992: 74) writes: Since at least 1954 External Affairs appealed to the Department of Citizenship and Immigration to agree to the Refugee Convention as a means of securing gains within 'the broad context of international relations'. Department of Immigration officials, however, had resisted such a commitment, arguing that it -like participation in ICEM that was deemed to lead to a loss of control over selection- would later lead to the inevitable production of a Canadian 'burden' since during resettlement refugees might not follow a path to becoming liberal citizens and could not then be easily disposed of because of non-refoulement (ie. their 'returnability' would be in question under the Convention). In the late 1960s, however, there was a sudden renewed interest among authorities in pursuing Canada's well-being and security within the imagined international community, in making the nation a credible, capable citizen on the international stage (Satzewich, 1991a: 306; Hathaway, 1992: 75; interview 30/08/96; cf. Page, 1977; Bothwell, 1977). This fit well with the prospect of acceding to the United Nations Convention and 127 introduction of a regular, visible refugee selection programme. The Deputy Minister of Immigration suggested to External Affairs that Canada ought to adhere to the Convention in 1964 (Hathaway, 1992: 74). As a senior Department of Immigration official wrote at this point, Canada was already carrying all the burdens of responsibility in refugee matters, but [its] highly legalistic refusal to accede to the 1951 Convention has prevented [Canada] from getting much of the credit [it] should have (as cited in Hathaway, 1992: 75). The circumstances in these and undoubtedly other domains happened to come together during this period to give rise to a Canadian refugee regime. In 1976 'refugee' appeared in the new Immigration Act as a distinct legal category for the first time (Lanphier, 1981). Regulations accompanying this Act were enacted in 1978. They spelled out procedures for selection; an Annual V] Refugee (and Immigration) Plan; resettlement; and Convention refugee status determination within Canada (Creese, 1991; Hawkins, 1989; Dirks, 1977). Regarding the Annual Refugee Plan a senior Department of Immigration official active throughout the 1970s relates: ...until we began very deliberately to table an immigration plan with numbers, we tended always to respond to situations in an ad hoc [way] as far as refugees were concerned... the government wanted more predictability. . . it meant as well that in a year when we wouldn't necessarily have a crisis per se, there still would be a refugee response going on (interview 09/01/96; emphasis added). The new Canadian refugee programmes that followed release of the Green Paper foresaw refugee levels being set annually and a two-step process of refugee selection, the second of which would be 128 determining whether the person was capable of "successful establishment" (Immigration Manual. IS 3.02). For this the requirements of the immigrant points system (See Hawkins, 1989: 295-7), first put in place in the 1967 and continued in the 1978 regulations, would be "relaxed" (Immigration Manual. IS 3.02). Through the setting of annual refugee levels these programmes sought to render predictable the otherwise unpredictable responses to crises abroad that tended to result in sudden influxes of large but varied numbers of bodies that could be out of line with precise needs of the Canadian population rendered knowable by demographic analyses. From this point onward, political authorities were no longer to react by passing Orders-in-Council; respond without knowing the numbers of refugees that should be selected; rely on existing welfare arrangements for 'the poor' or those designed solely for maintaining the well-being of immigrants; or dispose of unwanted non-citizens without a way of considering possible harm befalling them from events, deemed to be illiberal and uncivilized, taking place elsewhere. The chapters that follow show, among other things, that these new Canadian refugee programmes did not anticipate eradicating international refugee problems as the first League Office had; they were merely about averaging and spreading out their effects on the Canadian population, 'making-up' new Canadian citizens by taming their individual human products, and later surveilling and disciplining non-Western nations and populations where the source of these problems was assumed to lie and whence their products came. 129 SUMMARY AND CONCLUSION International refugee programmes and organizations began in Europe following the First World War as temporary, small-scale, and specific to groups defined by ethnicity or nationality. By the 1960s the international refugee programmes and organizations had become permanent, large-scale, and directed at a universal category. Contemporaneous with this the international refugee arose as an object of authorities' interest and as a new 'kind' of human being. Besides suggesting the newness of refugee policies and practices in the Canadian context and how the international regime served as a condition of possibility for their emergence, the brief discussion of international refugee programmes and organizations above, also reveals the relevance of concepts and themes drawn from the governmentality literature in attempting to understand an international regime. Certain rationalities and technologies can be seen shaping and making possible this international regime. The 1 failure1-new programme-'failure'-new programme sequence seen in this context suggests the present international refugee regime should not be simply understood as the end result of inevitable progress in responding to a growing, intractable international refugee problem (a Foucauldian 'facile gesture' if there ever was one). Instead, the fact that international refugee programmes and organizations were not abandoned or subsumed in each instance of 'failure' ought to be seen as due to their usefulness in international power relations. 'Refugeeness' started to become 130 more useful to Western authorities after 1947 within intensifying Cold War relations with the Soviet Union and the Eastern Bloc nations exemplified in international responses to events in Hungary in 1956 and later in relation to de-colonization in Africa (and other socio-political spaces), first seen in a response to events in Algeria and Morocco in 1957. In these instances -between the Western and Eastern alliances (or 'First' and 'Second Worlds') and between North and South (or 'First' and 'Third Worlds')-'refugeeness' can be seen as a Western moral-political tactic, a kind of 'transfer point' (Foucault, 1978: 103) in international power relations. As seen in the next chapter, in the late 1960s and early 1970s 'refugeeness' was to become useful to the Canadian nation. In the 1980s in relation to determination, however, it was to become problematic and unruly. The fact that a Canadian refugee regime emerged in the period between the mid-1960s and early 1970s was not a matter of course. It was contingent on certain conditions besides the crystallization of an international refugee regime and its object. Among them are a renewed interest in seeing Canada become a more influential citizen in the international community; a renewed interest in Canada's population size and composition; and a recognized scarcity of desired labour in the West for the national economy. 131 1. Canadian state agencies such as the Canadian International Development Agency have been directly involved in refugee aid and development through bilateral arrangements since the emergence of the Canadian refugee regime, but more often they have provided funding for agencies constituted as ' international' such as the UNHCR (Adelman, 1982: 42). Similarly, Canadian private agencies such as Canadian Christian organizations and the Canadian Red Cross operate through their international affiliates, the International Red Cross and the World Council of Churches (Adelman, 1982: 37). These and a complex array of non-state organizations are the "operational arms" (Adelman, 1982: 37) or, to use the advanced liberal phrase, the "implementing partners" (Rogers, 1992: 1116) of the UNHCR and Western nations operating in this sphere. 2. There is some overlap between them at certain sites, however. Nor is this to suggest that Western nations have not entered into bilateral or multilateral agreements to harmonize determination. Western European nations have entered into such agreements among themselves and Canada and the U.S. have recently signed a similar agreement (Canada, 1995d). Canada is also part of the multilateral "Comprehensive Plan of Action" (CPA), a temporary international programme involving refugee determination in countries of first asylum; forced repatriation of those deemed not to be refugees; and third-country resettlement in nations such as Canada (Canada, 1995b: 7). The CPA is scheduled to be completed in mid-1996 (Canada, 1995b: 7). 3. Attention is paid to international programmes and organizations rather than practices here because previous historical research pertaining to the former is readily available. The history of international refugee practices, particularly within refugee camps, is, with few exceptions (e.g., Wyman, 1988), one that is yet to be written. 4. There were no refugee movements, practices, or refugees in the same sense that there was no sexuality until the eighteenth (Foucault, 1978) or human beings with multiple personality disorder until the late nineteenth centuries (Hacking, 1995: 4). 5. See Hacking (1986: 226) and Pires and Acosta (1994: 20-22) for insight into the differences between the two perspectives. The discussion that follows is not an attempt to respond to the question, 'Are refugees real1?, which is nonsensical as it stands because it requires both a noun and an alternative (Hacking, 1995: 11). To supply a noun would be to employ the following question structure: 'Is a refugee a real N?' (or 'Are refugees real N's?'). To make it intelligible, one is also required to suggest how a refugee fails to be an N. It is to ask, for example, 'Is a refugee a real N as opposed to an X'? 132 6. There were large movements of human beings outside Europe in the two decades that followed the 1920s within colonies and nations. For example, there was a mass migration from 1937 to 1939 of some 30 million Chinese during a conflict with Japan (Skran, 1995: 3). 7. This is not necessarily the case regarding Newfoundland prior to Confederation with Canada in 1949. According to the historian Bassler (1992: 223), unlike Canada, Newfoundland had a 'refugee clause' in its 1906 Aliens Act. One might note, however, that clause 1(d) of this Act was not referred to as a 'refugee clause' at the time. It referred instead to 'an immigrant who proves that he is seeking admission to this Colony solely to avoid persecution or punishment on religious or political grounds' (as cited in Bassler, 1992: 40). No one was admitted under this provision between 1919 and 1930 (Bassler, 1992: 225). 8. There was, however, a non-State refugee-specific organization-the relatively short-lived Canadian National Committee on Refugees (CNCR) created on December 6, 1938 under the auspices of the League of Nations Society in Canada (Dirks, 1977: 62; Knowles, 1992: 112; Holborn, 1975: 597). It established branches in several Canadian metropolitan areas to respond to the temporary European international refugee situation. Members of the CNCR included church groups, the Business and Professional Women's Association, and the YMCA (Dirks, 1977: 64). After 1947 it was disbanded (Holborn, 1975: 597). A similar temporary organization, the Canadian Christian Council for the Resettlement of Refugees (CCCRR), was created in 1947 after Canadian selection of Displaced Persons abroad began (Vernant, 1953: 564; Adelman, 1980: 107). It included six Canadian Christian and ethno-cultural organizations: Catholic Immigrant Aid Society, the German Baptist Colonization and Immigration Society, the Canadian Lutheran Relief, the Latvian Relief Fund of Canada, the Canadian Mennonite Board of Colonization, and the Sudetan Committee (Hawkins, 1972: 304-5). The CCCRR's mandate was to "'organize the assembly abroad, selection, presentation to Canadian offices, and onward movement to Canada of refugees and displaced persons who did not come within the mandate of the IRO'" (as cited in Hawkins, 1972: 304). In other words, the CCCRR was imagined making referrals to Canadian visa officers (or their equivalent at the time) for selection abroad. It later became one of four groups of the new Church Approved Program established in 1953 for a brief period allowing selection of unsponsored immigrants and approve and process sponsored ones (Hawkins, 1972: 305). 9. The efforts of the IRO were combined with those of some sixty private voluntary agencies (Gallagher, 1989: 579). 10. The United Nations Relief and Works Agency (UNRWA) was established in December, 1949 and commenced its operations in 1950 (Tabori, 1972: 288). The United Nations Korean Reconstruction Agency (UNKRA) was established a year later in December, 1950. The 133 first programme imagined directing aid to Palestinians who had fled from the events of 1948 in what is now the state of Israel. The second sought to "assist the Korean people to relieve the sufferings and to repair the devastation caused by aggression" (As cited in Tabori, 1972: 288). Like the UNRRA neither was necessarily directed at refugees and although both were assumed to be temporary at the outset (Tabori, 1972: 289) the UNRWA continues today. 11. As refugee problems appeared outside Europe, the ICEM dropped 'European1 from its name. More recently it became the International Organization for Migration (IOM) (Rogers, 1992: 1115). 12. By 1954 the six nations that had signed the Convention were: Denmark, Norway, Belgium, Luxembourg, West Germany, and Australia (Holborn, 1975: 177). 13. Following the events in Hungary, Canada was contributing from between a hundred and twenty-five and two hundred thousand dollars to the UNHCR annually for emergency refugee aid (Dirks, 1977: 183). This reached three hundred and fifty thousand dollars in 1965 (Dirks, 1977: 183), seven hundred and fifty thousand in 1977, and two million in 1979 (Adelman, 1982: 40). By 1991 Canada was contributing over seven million dollars of public funding to the UNHCR through CIDA annually (Adelman and Cox, 1994: 262), a rather remarkable difference from the temporary, limited public funding sent to the first League Office seventy years earlier. 14. Only about one percent of the UNHCR's budget, therefore, is guaranteed (interview 10/04/96). 15. As noted in the third edition of the Encyclopedia Britannica in 1796 (Marrus, 1985: 8). 16. Zolberg et al. (1989: 8) mention in passing the development of 'a community of discourse1 about the refugee as having "...followed a 'Kantian' progression, beginning with particularistic judgements. . . contingent on the circumstances prevailing in a given country at a given time; to which, over time, other categories were added; and gradually becoming more universalistic until, after World War II, the general criteria for identifying refugees were institutionalized in a body of international law administered by a bureaucratic organization staffed by appropriate experts." This is consistent with the foregoing discussion. 17. Of course some NGOs, such as Oxfam that began in 1963, have received up to half their funding from public sources, in the case of Canada through matching funds from the Canadian International Development Agency (Oxfam, 1995). 134 18. See, for example, Simpson's (1938: 549-50) discussion of the "steps which cannot be taken by private organizations" from this era. 19. This was a consistent theme during a workshop held in December, 1985 that was jointly sponsored by the International Council of Volunteer Agencies and the UNHCR (Kibreab, 1993: 323). 20. In October 1990 there was a seminar organized by the Centre for Refugee Studies at York University regarding literacy and education in refugee camps (Refuge. 1990 10(2): 18-23). A participant notes: "It is important to find out what the refugees themselves perceive to be their basic needs and to let them define their own needs on a priority basis." Another notes: "We need to empower refugees to take control of their own situation and not become passive recipients of international aid" Refuge. 1990 10(2): 19). At the third international symposium in Britain "Refugees from at least eleven different countries currently residing in the UK or Canada, also participated in the overall programme" (Refuge. 1988 8(3): 10). 21. For discussions of 'empowerment' in the context of governmentality, see Cruikshank, 1994; Baistow, 1995; Rose, 1996b: 348. 22. In 1992 the High Commissioner for Refugees announced that the 1990s was to be no less than the 'decade of repatriation1 (Rogers, 1992: 1129). At this point Canada's Department of Immigration was also touting this renewed emphasis on repatriation on the part of the UNHCR in conjunction with statements about refugees levels (Dialogue. 1992 1(1): 3). 23. Recently the UNHCR has begun referring to NGOs as 'partners' (cf. UNHCR, 1996: 12; Rogers, 1992; 1116) and 'empowerment' has become a theme in international refugee research (e.g., Cooper, 1994: 7) both of which, as seen in chapters six and five respectively, is consistent with a broad advanced liberal discourse. 24. Canada's Department of External Affairs and CIDA had undertaken a less thorough audit of UNHCR activities and expenditures earlier in the 1980s (Adelman, 1982: 42). The greater dependence on the audit is itself a feature of advanced liberal governance (Valverde, 1994b). 25. See Sjoberg (1991: 39) on the last League organization; Sjoberg (1991: 236) on the IGCR; Holborn (1975: 57) and Gordenker (1987: 27) on the IRO; Takkenberg (1991) on the UNRWA that followed the UNRPR (United Nations Relief for Palestine Refugees); and Dirks (1977: 185) and Hawkins (1972: 18) on the ICEM. 135 26. In this way the term 'displaced person1 can be understood to have been invented "for the expressed purpose of liquidating statelessness once and for all by ignoring its existence" (Arendt, 1973: 279). 27. Giddens (1985: 272), for example, suggests regarding post-colonial nations in Africa that compared to Western nations "it is obvious enough that their origins and character are mostly quite discrepant." 28. Refugee development, as opposed to emergency refugee aid, was first seen in experimental work projects in Palestinian camps after 1948 (Gorman, 1993: 8), but it was not until the early 1960s in Africa that refugee development began to supplement refugee aid programmes on a massive scale (Gorman, 1993: 8). 29. None of this is to suggest that this tactic has not been resisted from time to time by the inhabitants of these spaces, as seen in relation to New Guinea during the Macias dictatorship, Uganda in 1982 (see Gordenker, 1987: 113), and more recently in African nations when entrance by representatives of the UNHCR and NGOs has been denied by political authorities. 30. Yugoslavia was, however, by 1985 making contributions to the UNHCR (Marrus, 1985: 370). 31. Significantly, by 1969 African nations had formed their own refugee definition that is presently not recognized by Western nations: "The term refugee shall also apply to every person who, owing to external aggression, occupation, foreign domination or events seriously disturbing public order in either part or the whole of his country of origin or nationality, is compelled to leave his places of habitual residence to seek refuge in another place outside his country of origin or nationality" (Holborn, 1975: 189) . 32. This was not lost on Department of Immigration officials and the Minister of Immigration (cf. Girard, 1988; Weiner, 1988). 33. Though a small-scale, temporary Handicapped Refugees Program was introduced at this point (Canada, 1974: 106). 34. See Daly (1976: 61-5) on what he refers to as a "population policy movement" that appeared beginning around 1971. 35. According to Dirks (1995: 114): "In the late 1970s and the early 1980s immigration officials attempted to weigh demographic factors to some extent when processing applicants at overseas posts" He adds that "their efforts were not especially successful." 136 36. For example, in 1985 the Parliamentary Standing Committee on Labour, Employment and Immigration suggested that immigration be used to smooth out the age imbalance of the Canadian population (Kelley, 1987: 8); Canadian representatives attended the United Nations International Conference on Population in Mexico City in August 1984 (Canada, 1988c: 24); and the comprehensive National Demographic Review was subsequently carried out by Health and Welfare Canada between 1987 and 1992 (Canada, 1989a). See also Canada (1990d). 37. According to a senior Department of Immigration bureaucrat, this arrangement ...provided comfort to the community because they had some coherent touchstones and it made policy visible... as ongoing and not just ad hoc. But it also made it easier to budget for the settlement services that refugees needed because you would say: 'well we're going to admit 13,000 refugees this year under the government plan so we need this much money set aside to finance their selection, their language training, their transportation to Canada and their basic settlement services on arrival (interview 09/01/96; emphasis added). 137 CHAPTER FIVE: ASSUMPTIONS, PROGRAMMES, AND RATIONALITIES Canadian refugee programmes that appeared in the 1970s imagined protecting refugees from immediate physical damage and death through selection and determination and from moral and mental dangers through resettlement. In this rather lengthy chapter the Canadian refugee regime is broken down for purposes of analysis into three basic types of practices- selection, determination and resettlement. Taking the regime apart in this way will better allow rationalities and programmes, and how they have been changing, to be discerned. The discussion that follows, therefore, is not an exhaustive, detailed account of these three types, but a broad inspection of the ways of thinking about government that have made them possible, and the schema in which they have been imagined since the release of the Green Paper in the 1970s. Before proceeding it is first necessary, however, to show that the refugee regime is assumed distinct among relevant authorities, that it is thought irreducible to other regimes and their objects, especially to immigration and immigrants. It is also necessary to look at authorities1 assumptions about the refugee as an object/identity. This is accomplished primarily by drawing from interviews with authorities. Assumptions about the Canadian Regime and the Refugee There are presently both Canadian state and non-state refugee-specific agencies or bodies in existence. The Refugee Status Advisory Committee (RSAC), consisting of representatives from the Departments of External Affairs and Immigration, was created in 138 1975 to deal with determination of refugee claims. In 1989 it was replaced by the Convention Refugee Determination Division (CRDD) of the Immigration and Refugee Board (IRB) that continues at present. At the federal level of the state there is also presently a permanent Refugee Branch of the federal Department of Immigration51 that deals with refugee policy. From its creation in 1975 (interview 30/08/96), until 1988, 'Refugee Affairs' was a division within a branch directed by a low-ranking bureaucrat (Dirks, 1995: 132) . The Department of Immigration underwent significant re-organization in 1983 when five branches were established at its national headquarters. Refugee Affairs was then re-located as a division within the 'Policy and Program Development' branch (Dirks, 1995: 130). Another re-organization occurred in 1988 leaving national headquarters with four major branches. One of these became 'Refugee Affairs and Settlement' (Dirks, 1995: 131). In 1990, when policy and operations within the Department were split and six branches added (Dirks, 1995: 134), Refugee Affairs then became a branch unto itself. In 1993 Employment and Immigration became two separate departments -Human Resources Development and the Department of Citizenship and Immigration- and two branches were created, 'Refugee Affairs' and 'International Refugee and Migration Policy'. The former was to cover operations; the latter policy. In 1995, yet another re-organization was announced that imagined the eventual partition of the Department of Citizenship and Immigration into three sections, each consisting of several 139 branches. The core of these sections is to be 'Service Lines' and the 'Refugee Branch', one of this section's four branches (Canada, 1996a: 3). Throughout the period since release of the Green Paper, then, there have been distinctive state or state-funded bodies specifically concerned with refugees. There are also national, non-state, refugee-specific organizations operating at present: the Canadian Council for Refugees (CCR) based in Montreal and the Inter-Church Committee for Refugees (ICCR) located in Toronto. The CCR is a small, permanent umbrella advocacy organization established in 1978 during meetings about the selection of Indo-Chinese refugees. It was volunteer-based until 1989 at which time two and one-half permanent salaried staff were added (Hardy, 1994: 285). It presently represents one hundred and thirty-eight Christian, citizenship, immigrant settlement, ethno-cultural, regional refugee advocacy, and local refugee sponsorship organizations, as well as refugee research centres, refugee lawyer associations, and groups of (rather than 'for') refugees (CCR, 1995a).' The ICCR is a slightly larger umbrella organization created in 1980 that presently has ten salaried staff (ICCR, 1995a). Its membership currently consists of ten major Christian churches. Prior to the establishment of refugee sponsorship provisions in the 1978 immigration regulations (see below), there was a National Inter-faith Immigration Committee8 operating, but no similar refugee-specific q organization. In Canadian urban areas there are also presently permanent 140 regional organizations that have adopted similar refugee advocacy roles. These include the Toronto Refugee Affairs Council created in 1983 (interview 22/08/95) and the Vancouver Refugee Council and Table de concertation de Montreal pour les Refugies also established in the 1980s (CCR, 1995b; Refuge. 1984 3(4): 17). In addition, over the past two decades Amnesty International has had a small Refugee Affairs unit within its English Section that has advocated in Canada on behalf of individual refugee claimants (Refuge. 1987 7(2): 9; 1989 9(2): 4). From their inception, representatives of the CCR, ICCR, and these various advocacy organizations have made oral and written submissions to annual levels consultations (see below) and various legislative committees, and have participated in numerous public fora regarding refugee policy (cf. Simmons and Keohane, 1992: 432). Since release of the Green Paper,10 then, there have been distinctive non-state agencies and bodies specifically concerned with refugees in Canada. Associated with these state and non-state bodies and agencies are a variety of individual identities that similarly show the distinctiveness of the regime. At present these include: refugee board member, refugee hearing officer, refugee coordinator, refugee sponsor, refugee lawyer, refugee consultant, refugee worker, and refugee expert. Across the talk and texts of various authorities the distinctiveness of the regime is also evident. A recent Department of Immigration strategy document, for example, notes: The refugee program will be managed separately to distinguish its protection and burden-sharing objectives 141 from the different objectives of the immigration program (Canada, 1994b: 2; emphasis added). A senior Department of Immigration bureaucrat similarly observes in passing: ...more today then ever before the refugee program is a separate program from the... overall levels, (interview 06/11/95). A refugee coordinator operating on behalf of a major Christian church notes: I'm not dealing with the context of immigrants coming into the country... I'm merely interested only in the refugee part (interview 28/08/95). Though 'immigration' and 'immigrant' are not present in interview guestions (see Appendix II), most authorities interviewed make perpetual reference to immigration and immigrants in talking about refugee policy and refugees, but they do so without reducing the latter to the former. Consider these two excerpts from interviews with representatives of ethno-cultural organizations: ...so we deal a lot with post-traumatic stress disorder, people that came from countries, you know, torn by war etc. civil wars, and they need, they're not a normal immigrant... (interview 03/10/95; emphasis added). ...a refugee is different because I think refugees, because they come under a difficult situation, they don't have the financial resources to be self-sufficient right away, immigration is different, immigrants are, if you are talking of immigration as a whole. . . The refugees are bottom of the ladder and usually they have to be dependent on, you know, like government welfare at the beginning (interview 16/10/95; emphasis added). Absent are suggestions that refugee issues are fully reducible to or 'really about1 immigration, or for that matter, international development, the economy, foreign affairs, or some other longer-142 standing domain. At present the refugee is instantiated in the talk and texts of various authorities of the Canadian refugee regime in a distinctive way. In a major report a few years ago, Canadian refugee experts, for example, note that ...people usually do not choose to become refugees: they are either forced out of their homelands by hostile governments... or forced to leave because of fear of political reprisals... (Beiser et al., 1989: 190; emphasis added). In the talk and texts of these authorities the refugee is assumed to lack voluntarism. A representative of an ethno-cultural organization, for example, observes: These people are different and what goes unrecognized is the fact that they didn't choose, they are forced out, and that fact is very important (interview 17/10/95; emphasis added). The refugee is distinguished by a lack of choice and this is assumed to be important for, among other things, their integration into Canadian society. A Christian church-based refugee worker and representative of a national refugee advocacy organization relates: An immigrant has choices and power and a refugee has none of those things... they are in, put in a position in their home country prior to flight where they are deprived of choice and are deprived of power and so that I think is a fundamental difference. I believe that refugees are persons who do not choose to leave their country and come to Canada, that they are persons who are put into a position where they have no choice but to leave. They suffer tremendous losses, in every aspect of their life... by leaving. And this affects the way in which they resettle because there was no choice to leave, therefore, there is no choice to come, and no choice to be here (interview 23/06/95a; emphasis added). A representative of an ethno-cultural organization similarly relates: 143 . . .they're people that come here because they didn't have any other choices, sometimes leaving all their families behind and their support. In addition to having to overcome all the sort of so-called normal integration problems as happens to everybody, they have their own personal problems to deal with... So its... a very specific type of population I should say (interview 03/10/95; emphasis added). From these examples it is evident that the refugee is assumed to be an entity lacking choice that is contrasted with the choice assumed to be exercised by the immigrant. In North America the immigrant has become an established object/identity in its own right and has its own historical trajectory. As Wong (1989: 278) writes regarding the United States: The attribution of status and identity by virtue of the act of migration emerged only in the wake of the establishment of the United States of America as an independent nation-state... Not until 1868, however, did the term 'immigrant' acquire officials' recognition in that the lists of steerage passengers on incoming ships were referred to as alien immigrants rather than alien passengers. The term 'immigrant' thus emerged as a 'discursive fact', as denoting a particular type, not of population movement, but of population of a particular breed of humanity... The refugee has emerged only recently as a discursive fact in Canada, the United States, and other nations of the modern West, and, is in a sense, carved out of the immigrant object/identity on the basis of a capacity to exercise choice. A representative of an immigrant settlement organization notes: ...refugees and immigrants, which often get lumped together under the banner of newcomer or immigrant, have very distinctive needs. One chooses, one group of newcomer chooses to come to the country, the other one [sic] are forced or do not have any choice in that matter and we have continually strived to... promote the unique distinction between those groups and the corresponding challenges and needs of those two groups because they are very different (interview 14/09/95; emphasis added). 144 A widely disseminated Canadian 'Resource Kit' concerning immigrant women with a separate section on refugee policy and refugees reads: There is a lack of specific funding for programs geared to meet the needs of refugees. It is assumed that their needs do not differ from the needs of immigrants. Programs for refugees should be integrated to address their fears for loved ones back home, their emotional stability and deteriorated physical health. The programs must also address their experiences of persecution, incarceration, and torture. Refugees do not choose to emigrate (Moreno and Persad, 1990: 46; emphasis added). It is in this way, then, that the refugee is defined against the immigrant: the latter chooses to migrate and then is assumed to undergo normal integration while the former does not. Consider two more excerpts from interviews that illustrate this. A representative of an ethno-cultural organization relates: ...they're different in that they didn't choose to come here, they were forced out of their countries, big difference between the average immigrant that sells everything and has planned to come to Canada. . . (interview 17/10/95; emphasis added). A psychiatrist specializing in the treatment of immigrants and refugees during their resettlement notes that ...the basic difference between an immigrant and a refugee is an immigrant has time to prepare for their trip... Refugee is [sic] completely unprepared. When the right time comes for them, they just leave from their own country either by air, by boat, walk, or swim or God knows [and]... end up in a refugee camp. For sometime they have no idea where they would go to. They have no choice about what their future is like, about what the settlement country is like... of course this group of people cannot contribute much when they first come here, you know, given what they went through, it take them [sic]... a long time even to function... (interview 29/01/96; emphasis added). These examples are enough to show that in the texts and talk of a variety of Canadian authorities, the refugee is assumed to be a 145 distinctive object/identity that lacks a capacity for choice. This is important because it is this assumption that underlies all efforts to govern refugees as refugees. It is consistent with the notion of the refugee within international programmes as an entity forced to flee from something or someone. In selection it is what has made the refugee useful to the Canadian nation in the international sphere of politics; in determination it is the moral-political causes of this assumed lack of choice that programmes foresee identifying; and in resettlement it is what justifies special -as opposed to normal- intervention in refugees' lives after they arrive in Canada. LIBERAL GOVERNMENT It will be seen below that when the three types of practices of the regime are investigated, liberal rationalities can be discerned. It is necessary to draw attention to several distinguishing features of liberalism as a rationality of rule before proceeding to show the varied ways in which it constitutes and shapes the regime. Governing in a liberal manner -since at least the nineteenth century- has assumed an opposition between public and private realms (Rose, 1987) or between the state and civil society (Burchell, 1991). Liberal government places limits on political intervention and presumes a domain of freedom and action outside the legitimate reach of politics (Rose and Miller, 1992: 179). It confronts subjects in civil society who are presupposed to possess rights that must not be violated by immediately political authorities (Rose, 1993: 289). Civil society 146 is nevertheless to be shaped in preferred ways by 'non-political' authorities (Rose, 1993: 290). Another feature of liberalism is its insistence that subjects be rulers of their own conduct and fate, that they be self-governing entities (Valverde, 1996: 357). Liberal government is not arbitrary government. Instead it entails a perpetual critique, an on-going suspicion of political authorities who might impose their will across a territory (Rose and Miller, 1992: 180; Rose, 1996a: 44).^ It spawns the perpetual questions: Is government proceeding as planned, is it intervening enough, is it interfering too much? Liberal government is reflexive and seeks to ensure the continued existence of institutional spaces such as universities, public fora, mass media, and advocacy organizations, through which reflections on the actions of political authorities and on public policy become possible (Rose, 1993). With the aim of good, proper government, academics, citizens, journalists, capitalists, trade unionists, and advocates of one kind or other are encouraged, and in a sense, required to criticize these actions and policies in civilized ways. It is in this way that liberalism relies on expertise of various kinds, as well as the social sciences (Rose, 1993: 291). In a wide range of domains from health care to higher education to policing in Canada and other Western nations, as well as in international domains (as suggested in the previous chapter), a shift in governmental rationalities continues to take place. While each of these domains has its own specific trajectory and conditions of possibility, the way they are transforming is 147 consistent with a shift from what can be called a 'liberal welfare rationality1 (cf. Ashenden, 1996: 85n), characterized by what has been traditionally understood as the Welfare State, to a rationality variously referred to as 'neo-liberal', 'post-Keynesian1 and 'advanced neo-liberal1 (Rose and Miller, 1992: 198-201; O'Malley and Palmer, 1996; and Valverde, 1996 respectively), but perhaps best termed "advanced liberal'12 (Rose, 1993: 294-6). It was mentioned in the previous chapter that one feature of 'advanced liberalism' as a rationality of rule is that it presupposes a more active or enterprising subject (Rose, 1992: 145). Governing in an advanced liberal way specifies actors who are "subjects of responsibility, autonomy, and choice" (Rose, 1996a: 53-4). Another characteristic is that it entails less reliance on the social sciences and more on knowledges such as accounting and econometrics and related techniques like the audit and 'marketization' (O'Malley and Palmer, 1996: 150; Rose, 1996a: 54-5). It also involves 'localization', the movement of responsibility to more local levels (O'Malley and Palmer, 1996: 141-2). In general, advanced liberal rule involves the introduction of more distance between decisions of political authorities and a variety of social actors (Rose, 1996a: 53). While there is no clear-cut point at which this shift commenced in the context of the Canadian refugee regime, as is seen below, signs can begin to be seen in the mid-1970s. Refugee Selection From the release of the Green Paper to the present, the 148 Canadian selection programme has envisaged visa officers in Canadian embassies and consulates entering refugee camps and crisis situations abroad (Hawkins, 1989: 293) to select refugees to fulfil annually set levels. Within the selection programme two basic types of refugees have been imagined: the Convention refugee and the Designated Class refugee (Canada, 1995b: 7; Nash, 1989: 40). The former refers to the UN Convention definition; the latter to groups designated by the Minister of Immigration to be in 'refugee-like' situations who might not otherwise fit the Convention definition (Canada, 1995b: 7; Nash, 1989: 41). These two types are to be selected abroad using special criteria, one of which is the potential for 'successful establishment' in Canada, as mentioned previously (Hawkins, 1989: 295-7; Nash, 1989: 40-1; Simmons and Keohane, 1992: 431; Lanphier, 1990: 82).16 The first designated classes to be established in 1978 were the 'Latin American Designated Class' (referring to persons in Chile, Argentina and Uruguay); the 'Self-Exiled Designated Class* (referring to non-Yugoslav Eastern Europeans); and the 'Indo-Chinese Designated Class' (referring to Cambodians, Laotians, and Vietnamese) (Canada, 1995b: 7-8). The first of these was renamed -when Poland was added- the Political Prisoners and Oppressed Persons Designated Class in 1982 (Canada, 1990b: 10; 1995b: 7-8). Guatemala and El Salvador were added to this Class after 1982 and remain the only nations to which it now refers (Canada, 1995b: 8 ) . ^ In 1990 the second of these was altogether cancelled and the third replaced by an 'Indo-chinese Designated Class, 149 10 Transitional'. Another significant development during this period has been the addition, in 1988, within these two basic types, of a Women at Risk programme that foresaw the selection of individual women refugees deemed to be in immediate danger of harassment, physical/sexual abuse, or refoulement within refugee camps (Canada, 1995b: 8; Macklin, 1995: 220).19 Prior to this, Canadian refugee programmes had constituted the refugee as an uncivilized, immoral man, sometimes accompanied by his wife and children. Women were mentioned within these refugee programmes to the extent they were a source of labour outside the domestic sphere in Canada, but even then they were referred to as 'wives' of refugees, rather then as 'refugee women1 (e.g., Canada, 1973; 1976a). This point is returned to in chapter six when resistance is discussed. When the Immigration Act was proclaimed in 1978, annual consultations with provincial representatives regarding levels of refugees and immigrants to be selected from abroad, preceding the presentation of an Annual Plan to Canadian Parliament, became required (Nash, 1989; 72). Consultations for an Annual Refugee Plan were imagined being carried out at the same time and in the same way as an Annual Immigration Plan (Canada, 1986d: iii). Besides provincial representatives, the Act also allowed non-governmental organizations to be included in these consultations to be decided on a discretionary basis by the Minister of 71 Immigration. Non-state refugee-specific organizations, such as the CCR, 150 were briefly consulted through special meetings on selection levels of Indo-Chinese refugees in 1978 and have been regularly consulted since 1980 (Canada, 1986d: 1,4). Department of Immigration officials carry out these annual consultations first by contacting representatives of this and other organizations and groups by letter. Several broad questions about levels are put forth and responses through written submissions and/or attendance at meetings hosted by immigration officials in various regions of Canada are encouraged (Canada, 1986c: 15). Since 1981 Department of Immigration regional representatives (e.g., Ontario, Quebec, or Alberta/N.W.T. regions) -as opposed to those from the Department's central planning headquarters in Ottawa exclusively- have become involved in both these aspects of consultations (Canada, 1988c: 15). In 1982 through to 1984 the number of groups and organizations consulted in the manner above significantly increased (Canada, 1986d: 4). Refugee experts and other academics were invited to participate in consultations by Department of Immigration officials beginning in 1984. J Refugee levels were identified by world area for the first time in 1980 (Canada, 1988c: 12). The world regions specified within these consultations were consistent with those implied by the Designated Classes mentioned above. Initially they included 1Indo-china', 'Eastern Europe', and 'Latin America/Caribbean', as well as an 'Other' and a 'Reserve'.24 In 1981 'Africa' and in 1982 'Middle East/West Asia' regions were added (Basok and Simmons, 1993: 140). These specified regions have, for the most part. 151 remained in place through the 1990s. The presence of these annual consultations reveals selection's liberal character. These yearly routines of receiving letters containing several questions about levels and responding through attendance at meetings and/or composing and submitting written responses are instances of political subjectification that seek to 'make' liberal citizens out of the varied participants. These procedures seek to summon participants' assumed potential to exercise their choice. These routines endeavour to encourage and conjure up participants' opinions, objections, arguments, calculations, and grievances and in so doing simultaneously seek to ensure that the setting of levels is scrutinized. Liberal forms of government seek to prevent the collapse of types of rule into mere domination by invoking the capacities and powers of the self-governing individual while at the same time undertaking to foster, shape and use those capacities and powers (Dean, 1994c: 163). These exercises make this possible in the area of selection. Liberal government can demand that even the routines themselves be reflected upon in an attempt to ensure that levels are not arbitrarily set. In 1986, for example, following concern expressed by some participants about whether the Department determined levels in advance of consultations and whether their input was being seriously considered (Canada, 1986d: 8), a comprehensive review to "streamline and deepen" the consultation process was almost immediately initiated and studies (e.g., Canada, 1986d; 1988c) of varied aspects of the consultative process undertaken (Canada, 152 1986d: 7). In conjunction with this effort, as an addition to the 1987 levels consultations and standard questions about proper levels, those receiving letters were encouraged to respond to eight questions about the levels planning process itself (Canada, 1986d: 7). But these annual levels consultations are just that- they target levels, the numbers of refugees to be selected in the coming year(s). Criteria for deciding the distribution of these numbers abroad are not necessarily made transparent or discussed in this or other contexts. Within the selection programme there is virtually no regular public debate about distribution of these spaces or, in other words, discussion over possible answers to the inherently difficult question of 'who is a refugee?1 (ie. which nations are refugee-producing or which group has been or was likely to be persecuted by this or that group within a particular nation) . At most, debate regarding levels for the private-sponsored component of the overall Refugee Plan within a particular world region has occurred within the context of the consultations (interview 21/06/95a; Refuge. 1981 1(3): 6; 1984 4(2): 4). As Adelman (1984: 1) writes: Selection of refugees abroad is a matter of government policy. The private sector may advocate a specific number in its distribution, and may be consulted by government on that number and distribution. The private sector may augment those numbers and affect the numbers brought from a particular source country through private sponsorship... [b]ut the primary basis of selection policy is rooted in government decisions. In the global approach to planning and allocating a limited number of spaces, the government must distribute those spaces among many source countries (emphasis added). 153 The target of the selection programme, then, is an annual refugee level, the numbers of refugees to be selected in the coming year(s). Since annual refugee planning began, the extent to which the numbers of individual refugees actually selected abroad has been in line with these set annual levels for a specific world area in which visa officers operate has been closely monitored (interview 30/08/96; Dialogue. 1(3): 8 1992). From within Canada, however, it appears distribution has been left to the whims of Canadian political authorities and visa officers. Canadian refugee selection, outside of a concern over levels, seems to have a decidedly despotic character. Despotism is defined by the limits of liberal liberty (Valverde, 1996: 359) and is more evident from what is absent in the realm of selection than what is present. This is a point that requires further discussion. How refugee spaces are allocated, in sharp contrast with determination and resettlement discussed later (as well as the setting of levels above), is remarkably unreflexive. Consider that since 1979 no less than five major state or state-commissioned, widely disseminated, comprehensive studies of determination have been carried out: Robinson (1981), Ratushny (1984), Plaut (1985), LRCC (1991), and Hathaway and MacMillan (1993) (cf. Refuge. 1989 9(2): 2). There have also been large-scale state-commissioned public opinion research conducted focusing on perceptions of determination (e.g., Canada, 1985c; 1989b; 1990c). During the same period there have been numerous large-scale, state (mostly state-commissioned), widely distributed studies of refugee resettlement 154 (e.g., Neuwirth, 1984a; 1984b; Canada, 1981a; 1981b; 1982a; 1982b; 1985; 1986b; 1987a; 1989c; 1989d; Nann et al., 1984; CRESH, 1984; Adelman, 1984b; Johnson, 1984; Samuel, 1987; Dorais, 1991; Johnson and Beiser, 1994). During this same period the Department of Immigration has hosted: a 'National Symposium on Refugee Determination1 in 1982 (Refuge. 1981 1(7): 5); an "Ontario Consultation Conference on Refugee Resettlement' in 1984 (Refuge. 1984 4(2): 16); public consultations on the Plaut Report (1985) in 1985; a comprehensive 'Review of Private Sponsorship of Refugees Program' begun in 1991 (Refuge. 1992 12(3): 4-5); Canada, 1990b: 11.; and numerous other public fora pertaining to these two types of practices. But neither the Department of Immigration, External Affairs, nor any other state agency, has sponsored or conducted public opinion or comprehensive research, or hosted a public symposium or conference, focusing on how Canadian refugee spaces are distributed abroad (interviews 30/08/96; 28/07/95).28 The prevailing assumption is that how refugeeness is deployed abroad by Canadian authorities does not require continual scrutiny through the production of formal knowledge (the kind of knowledge that is abstract and systematic), forms of expertise, or public debate in the free spaces of Canadian society. To be sure, there have been several scattered and very limited studies of selection over the past two decades by non-state researchers, but these can be understood as brief instances of resistance (see chapter six) rather than an integral component imagined within the regular selection programme. Only one of these 155 studies, carried out by the CCR, has been comprehensive in scope (CCR, 1992a). The response of political authorities and state agents to this study, however, is itself suggestive. After meeting with External Affairs and Department of Immigration officials in 1989 to discuss the possible creation of a task force that would eventually hold public hearings and prepare a report about all aspects of refugee selection practices (including how refugee levels are distributed), CCR representatives received a short letter containing the following response: "'I regret that I can agree with neither your estimate of the need for such a task force, nor with the fundamental premises of your recommendation1" (Refuge. 1989 9(2): 12). After the study commenced -without cooperation of relevant political authorities and Department officials- the CCR again requested a meeting with Canadian visa officers temporarily recalled to Ottawa for briefings "to talk with them about the processing of refugee claimants abroad" (Refuge. 1989 9(2): 12). This request was flatly denied, officials simply suggesting that "visa officers were too busy" (Refuge. 1989 9(2): 12). When the seventy-seven page CCR study was completed in September, 1992 and sent to Department of Immigration officials for comment, included was a major recommendation that Canadian visa officers select all those persons referred (see below) to them by UNHCR officials abroad. This received the following formal response from the Executive Director of Immigration Policy: "Decisions as to whether referrals meet the Convention refugee definition are left to sovereign states, like Canada, which have signed the United Nations 156 Convention relating to the Status of Refugees" (Dialogue. 1(3): 7 1992). The formal response to a suggestion in the CCR report that "the system of overseas protection of refugees is not working" reads: "We agree that this program is not beyond criticism. The fact that the announced target was not met in 1991 is a source of serious concern that we are addressing" (emphasis added; Dialogue. 1(3): 7 1992). Note that it is the fulfilment of the "announced target" or refugee level that is a "serious concern," the only 0(1 aspect of selection mentioned as such in this instance. This is significant because it supports the notion above that it is set refugee levels that are to be monitored within the Canadian selection programme not how, other than within world regions, these spaces are allocated in practice. This comprehensive study of Canadian refugee selection, when completed, nevertheless illustrated the apparently despotic character of how refugee spaces are distributed by visa officers. The study concludes that their decisions are "arbitrary and capricious" and that there are "inexplicable variations between visa offices and even between visa officers at the same visa post" (CCR, 1992a: 4). It also points out that there is no review or appeal process associated with or legal counsel allowed during individual refugee selection interviews (Refuge. 1993 12(3): 4; CCR, 1994: 2). In addition it notes that Canadian visa officers are not subjected to a competency test before being allowed to carry out refugee determinations (Refuge. 1992 11(1): 9)32 and that refugee expertise is generally not attributed to these agents (Refuge. 1992 11(1): 9). 157 In practice visa officers have used Designated Class categories in place of the Convention Refugee category to fulfil a level for a world region within the Annual Refugee Plan (Nash, 1989: 65). Because they refer to 'refugee-like' situations, Designated Class categories can be used without the visa officer having to carry out an individual refugee determination (Gilad, 1990b: 125). It will be recalled that nationalities and groups are 'Designated' by Canadian political authorities at another time and place. Individual refugee selection interviews often take place after Canadian political authorities -who are freed to deploy refugeeness in the international sphere- decide which nations are refugee-producing and which groups are refugees, and then visa officers are directed to appropriate camps and crisis situations to carry out individual selections. The distribution and sequence of tasks in the realm of refugee selection is not necessarily kept secret or otherwise concealed from the Canadian citizenry. During interviews, for example, many Canadian authorities active in the refugee regime demonstrate an awareness of these arrangements. They do not, however, regard them as an important area of concern. In more than forty interviews with authorities, only the refugee advocate quoted above raised serious concerns about them. Consistent with this, Dirks (1984: 280) writes that35 ...individuals outside Canada seeking to gain entry as refugees through application and processing at immigration offices around the world... normally cause no major problems for the Canadian government (emphasis added). Two recent developments also illustrate the apparently 158 despotic way in which refugee spaces have been allocated abroad. In 1993 members of the CRDD of the IRB became required to provide written reasons for not complying with the IRB Guidelines on "Women Refugee Claimants Fearing Gender-Related Persecution" that were established that same year (Canada, 1993b). But establishment of a similar requirement that would have allowed monitoring of visa officers' application of this criterion was not instituted (CCR, 1994: 1; ICCR, 1995b: 2-3).36 Similarly, in 1996 a formal agreement was made between the IRB and the Department of Immigration regarding knowledge-sharing such that "the IRB can now make specific inquiries to CIC [the Department of Immigration] about refugee claimants and the circumstances surrounding their claim" (Canada, 1996a: 7). But in programmatic statements this 'sharing' agreement is foreseen enhancing determination in Canada, not selection abroad. Whereas those involved in the determination process are presumed to require additional knowledge produced by visa officers to make proper deliberations, visa officers are not thought to require additional knowledge produced or relayed about these 'circumstances' by those in the determination process or the IRB's documentation centres (see determination below and chapter six) in order to improve the quality of refugee selections. Instead, the guidelines for refugee determination abroad instruct visa officers to rely primarily on informal knowledge of these conditions (Immigration Manual IS 3.24; Gilad, 1990b: 130). These two absences -of a requirement to monitor application of new guidelines abroad and of a need for 159 additional knowledge for determination abroad- are consistent with the notion that visa officers are imagined deciding less which human beings are refugees, and more which refugees in refugee camps or specific crisis situations in a particular world region should be selected to come to Canada to live permanently. Visa officers are pursuing not so much the question of 'who is a refugee?', as 'which refugees should be selected?' or, in other words, which refugees will work submissively, produce and raise moral children, make their consumer choices wisely, save for a rainy day, vote from time to time, and otherwise exercise their citizenship in a civilized manner if selected to come to Canada (e.g., Gilad, 1990b: 129; Hathaway, 1992: 81). The 'successful establishment' criterion, as it is the second of a two-step selection process, is entirely consistent with this notion (cf. Hathaway, 1992: 78, 81; Mandel, 1994: 253). As a refugee advocate notes concerning the emphasis and knowledge of state agents involved in refugee selection abroad: Visa officers are mostly doing immigration cases so that they think in immigration terms and that's basically their knowledge and experience... they're diplomats abroad and like all diplomats abroad basically they know what people sitting at hotels and bars in foreign countries know, which is very little... the visa offices don't go through that many processes. They don't know. They don't have access to information... (interview 28/07/95; emphasis added). Similarly, Dirks (1985: 124) has noted regarding Canadian refugee selection practices: Immigration officials, when screening and processing applicants, primarily focused attention upon the health, age, and skills of the individuals. Whether the applicant has been politically persecuted in his state of 160 former residence was of no consequence (emphasis added). The distribution of refugee spaces abroad is unreflexive and not necessarily reliant on expertise or formal knowledge in the form of the social sciences. In making their individual selection decisions visa officers do rely, however, on informal knowledge in the form of referrals from agencies such as the UNHCR, the International Red Cross, and international Christian organizations operating abroad (interviews 18/08/95; 09/01/96; 06/11/95; Refuge. 1982 2(2): 8; 1989 9(2): 13-14; 12(3): 5 1992; CRESH, 1984: 33; Canada, 1982c: 11; 1995e; CCR, 1995c: 3; Hitchcox, 1990: 123). As the Minister of Immigration related in 1984: It must also be remembered that there are special logistical problems in operating a refugee program which is truly global in scope, and in which we often cooperate with third parties -such as the UNHCR and church groups -in locating and interviewing those persons who are in need of resettlement (emphasis added).38 In sharp contrast with the setting of levels, as well as determination and resettlement discussed below, formal, regular relations are not established between state agents (ie. visa officers) and these non-state agents who make referrals. They are not imagined, for example, entering legal agreements or annually consulting one another in public fora (interviews 09/01/96; 30/08/96). Referring to these non-state agents a career Department of Immigration official relates: They will, as well, if they're asked, provide referral services with certain profiles, but generally speaking there's no uniformity in that respect we, we have always said, as have most other recipient countries, we will make our own selections. You can do some pre-screening for us. If you say a person's a refugee, we won't necessarily agree with you, we're not even bound to 161 accept that they're a refugee, and we won't necessarily admit them either, but we often solicited their help in the referral mechanism (interview 09/01/96). In practice, regarding fulfilment of the private sponsorship (see resettlement below) component of the Annual Refugee Plan, camp inhabitants and those in crisis situations who have a willing Canadian sponsor are referred to visa officers (Hathaway, 1992: 78-9; Refuge. 1990 9(3): 2-3; Hitchcox, 1990: 125; Canada, 1982c: 11). This sponsorship obviously brings them more in line with the 'successful establishment' criterion (cf. Hathaway, 1992: 78-9): their referral assists visa officers in deciding not so much who is a refugee, as which refugees should be selected to live to Canada. It should be recognized that how refugee spaces are distributed abroad by Canadian authorities is not formally limited by the dictates of international refugee law and the UNHCR (cf. Nash, 1989: 68; Immigration Manual. IS 3.06). The 1951 UN Convention and 1967 Protocol that Canada acceded to in 1969 suggests it is "the responsibility of governments, not the UNHCR, to determine who qualified under the subjective definition of refugee owing to a 'well founded fear of persecution'" and that these decisions are presumed to be "a sovereign area of discretion" (Cuncliffe, 1995: 282). Nor does the UNHCR dictate to Canadian political authorities which nations are refugee-producing or which groups are refugees (interview 22/08/95), although as mentioned above, its agents will refer individuals to Canadian visa officers for possible selection. Instead, since its creation, the UNHCR has tended to commence its aid (and referral) activities in a 162 particular socio-political space after political authorities of Western nations -freed to identify and address a refugee situation as such- request the UNHCR assist (interview 22/08/95; Adelman, 1982: 38; Dirks, 1977: 191; Tabori, 1972: 298n). The UNHCR's reports of the numbers of refugees in need of resettlement in a given world region for a given year have been consistently used as a rationale for the refugee levels that are finally decided upon in the Annual Refugee Plan. They are used to justify to what extent and sometimes how -only in terms of distribution among world regions- refugeeness will be deployed in the coming year(s). But it should be noted that prior to the appearance of Annual Refugee Plans, the UNHCR did not report on those refugees in need of resettlement on an annual basis and by world region, a format that now precisely fits the demands of the Canadian selection programme (interview 30/08/96; UNHCR, 1995). Canadian Department of Immigration officials at this juncture encouraged the UNHCR in Geneva to begin publicly reporting these numbers in this manner (interview 30/08/96). This clearly shows the UNHCR has not had the task of formally limiting Canadian selection since then, but instead has been one more non-political organization whose autonomy, in this instance liberal government, seeks to at once preserve and use. Within the Canadian selection programme, therefore, liberal government shapes selection in such a way that there are annual levels consultations, but this is the point at which it reaches its limit. From there it defers to what from within a Canadian or 163 national perspective appears as despotism- political authorities distributing refugee spaces abroad as they so desire. From an international perspective, however, of nations as actors operating in another realm of politics and morality, the Canadian distribution of spaces can be seen not to occur in a despotic way, but one consistent with liberal rationalities present in that sphere. For it is exclusively non-Western nations (or groups within them) which have been 'Designated' and from which Convention refugees have been selected- nations assumed to be something less than the ideal of liberal nationhood (as suggested in chapter four). Consistent with the presence of liberal rationalities in the international sphere and, as seen above, outside the Canadian selection programme, the deployment of refugeeness abroad has allowed Canada to relate to these non-Western nations and groups deemed to be illiberal, uncivilized -and until 1989- enemies, in ways discussed in the previous chapter. Simultaneously, this deployment has allowed the fostering of alliances with other Western nations (e.g., either single nations such as Britain and the U.S. or groups such as the Commonwealth and NATO), as they have sometimes lacked the capacity for such responses on their own. In this way, refugeeness can be seen as a moral-political tactic associated with Canadian authorities, and sometimes other Western nations with which Canada is assumed to have much in common (cf. Anderson, 1991: 26), operating in an international field. It is important to recognize that it is the existence of refugees once identified and selected as such, their collective 164 (either national or group) persecution that has been useful in the international domain, not the fact that any given individual was persecuted. Nor within the Canadian selection programme have persons in refugee camps or crisis situations abroad been presumed to be liberal subjects, rational agents whose freedom must not be interdicted by Canadian authorities. Instead, here refugees have been assumed to be little more than a reliable (assembled and disciplined in refugee camps as they are) non-Western immigrant resource to be mined by visa officers to fulfil levels. As is seen below, it is only after these human beings leave or are transported from these camps and crisis situations and -in one way or other-come into the range of Canadian determination or resettlement programmes that they are assumed to be juridical subjects or entities with a potential to become liberal citizens and confronted accordingly. The move to include, in regular levels consultations, representatives of the provinces and of various Regional headquarters of the Department of Immigration, rather than merely the Department's central planning headquarters, as forms of 'localization', show a shift to an advanced liberal rationality taking place. This is also evident in a gradual though uneven reduction in the planned or programmed levels of state-sponsored refugees in the Annual Refugee Plan since 1979. This reduction began in 1982 and continued in 1983 (Canada, 1986e; 1988c: 16). The 1983 levels were continued in 1984 (Canada, 1988c: 19), but were subsequently lowered slightly in 1985. For both 1986 and 1987 165 numbers were returned to 1984 levels (Canada, 1988c: 27; 1985b: 13) and for 1988 raised slightly (Canada, 1986c: 11). The levels of state-sponsored refugees planned for 1988 were continued until 1993 when they were significantly reduced (by 23 percent) to the lowest mark since 1979 when planning began (Canada, 1991: 6). For 1994 the planned levels were again significantly reduced by the same amount as the previous year (Canada, 1994c: 3), thereby bringing refugee levels in the 1994 Annual Refugee Plan to about half 1984 and a small fraction of 1981 levels. From 1994 to the present these significantly reduced planned levels have been maintained (Canada, 1995a: 13; 1996b: 5). This shift in liberal rationalities is also evident in the regular inclusion of refugee experts and other academics in levels consultations commencing in 1984-expertise is seen moving out of the state. Inseparable from these specific changes in the area of selection and the more general shift in liberal rationalities has been the appearance during this period of a new emphasis on econometric analyses that endeavour to quantify the contributions and costs to the economy of each type of entrant or migrant (e.g., refugee, independent) after their arrival in Canada (e.g., DeVoretz, 1995). This point is elaborated in chapter six when knowledge is discussed. Refugee Determination Unlike refugee selection, determination underwent considerable change during the period from its inception to the present. Though no specific provisions regarding determination appeared in the new 166 immigration regulations in 1967, the Immigration Appeal Board was created that year to deal with deportation case appeals launched by non-citizens (Dirks, 1977: 230; Nash, 1989: 38). As mentioned in the previous chapter, Canada acceded to the UN Convention in 1969 and this fit well with Canada's renewed attempt to assert itself as a nation in the international sphere. Accession announced to all that Canada was adopting a regular, visible refugee selection programme. This would allow (but not require) Canada as a nation to wield refugeeness abroad on its own and begin receiving the accolades from other Western nations, especially those without such a refugee policy. But now what was required as a result of accession was a determination programme targeting those who had fled to Canada to escape persecution elsewhere, one that would be in line with the UN Convention and Protocol and the non-refoulement provision. At this moment, such a determination programme was merely an afterthought, little more than a minor consequence of accession, largely hidden from the Canadian citizenry within the state and potentially relevant to but a handful of individuals per year. By the 1980s, however, determination would become the site of a rather dramatic reversal in broad relations of power. New guidelines were subsequently established in 1970 and in 1973 an amendment was made to the Immigration Appeal Act that explicitly recognized refugees. The latter amendment allowed -but did not require- Immigration Appeal Board members to quash deportation orders if they deemed the non-citizen in question was a Convention refugee (Canada, 1974: 106, 116; Hathaway, 1992: 75). 167 As this was not a requirement, at this juncture refugee status was deemed a privilege rather than a right in Canada (Mandel, 1994: 241). Also established in 1973 was an ad hoc interdepartmental committee to review refugee claims and make recommendations regarding status to the Minister of Immigration (Nash, 1989: 39; Canada, 1988b). At this point, the programme imagined a two-stage process, the responsibility for which was located solely within the corridors of the state. As is seen below, the rise of an advanced liberal rationality has since encouraged a gradual re-shaping of determination, such that there has been considerable movement of responsibility for determining refugee status in Canada out of the state. A new determination programme, made possible through the 1976 Immigration Act, foresaw the creation of a four-stage process and RSAC (Canada, 1974: 116). At this point refugee claimants were assumed to have a right to counsel and to receive a copy of the written transcript of the refugee claim made during an interview with a Department of Immigration official (Nash, 1989: 42). Drawing on knowledge of refugee-producing conditions in the world generated through experience with previous deportation cases or specific ad hoc inquiries through its small research unit (Goodwyn-Gil, 1987: 27), it was foreseen that the RSAC would examine these written transcripts and present advisory opinions to the Minister of Immigration regarding status (Canada, 1974: 116; Nash, 1989: 42; Hathaway, 1992: 81). To accomplish this, every RSAC member would also receive training sessions from an expert in international 168 refugee law (interview 10/04/96). The quasi-diplomatic UNHCR representative, mentioned in the previous chapter and who began operating in Canada in 1975, was to be a resource person who would review claims and then provide an opinion -but who would not vote-during RSAC sessions (Adelman, 1982: 41; Nash, 1989: 43). The rationale for these arrangements was to reduce the discretion of the Immigration Appeal Board in this area (Nash, 1989: 41). Significantly, this new determination programme also foresaw the refugee determination status process beginning before a deportation order was made. This was deemed to be an improvement that would reduce "the tribulations of a person who may indeed be a refugee, and eliminate unnecessary work for the immigration service and Immigration Appeal Board" (Canada, 1976b: 29). This suggests that, virtually from the outset, determination was seen to be creating problems for on-going immigration schemes. A few years later this improvement would be observed to be causing difficulties considerably more serious than 'unnecessary work' for the Department of Immigration. Immediately following release of the Green Paper and enactment of the Immigration Act, then, Canadian determination programmes envisaged RSAC officials deciding whether non-citizens who had made claims were indeed refugees and whether, therefore, they had no choice but to remain in Canada. Beginning in 1980, the numbers entering Canada's refugee determination process began to increase dramatically from a few hundred, to thousands annually (and by the late 1980s to more than ten thousand annually). This immense 169 increase in refugee claims constituted a crisis of resources (cf. Ashenden, 1996: 66): It was imagined that future increases could be avoided by deploying a proportionate volume of staffing and funds and that, with a few minor reforms, the existing system could be made adequate (interview 21/02/96; Simmons and Keohane, 1992: 432; cf. Jackman, 1988; Creese, 1991: 11-12). Besides the invention of a determination programme, this crisis arose under certain conditions of possibility during the early 1980s. Among them was a decrease in cost and an increase in availability of international air travel from outside the West to North America via Europe, thus making populations in these regions more mobile and Canada more accessible (cf. Dirks, 1985; Gilad, 1990b). Another condition was considerable Canadian and international mass media coverage of Canada's effort to select and resettle more than fifty thousand Indo-chinese refugees in 1979 and 1980 -subsequently touted to be the most generous effort among the Western nations involved (cf. Canada, 1982c)- that made Canada appear as more of a haven for refugees than previously (cf. Adelman, 1980). This was followed, beginning in 1983 and continuing through 1985, by a reduction, not only in planned levels of state-sponsored refugees, as seen above, but also of immigrants. The latter were reduced to the lowest levels since World War Two. This potentially extended the waiting time to immigrate by years (Nash, 1989: 70, 86; Beaujot, 1991: 145; Simmons and Keohane, 1992: 423; Dirks, 1995: 83). If some refugee claimants were later 'queue-jumpers' (unfairly moving ahead of those waiting patiently 170 in the immigrant line ('queue') by entering ('jumping into') the faster-moving refugee line on the way to receiving immigrant status), this was undoubtedly because the queue itself was imagined being squeezed at this point, effectively pushing those within it out at its seams, and otherwise encouraging those seeking immigrant status to find alternative ways to come to Canada. After 1985 another condition was a decrease in access to determination in Europe through changes exemplified by the implementation of the Schengen agreement among European nations (Hathaway, 1992: 91n) that encouraged those fleeing persecution to continue on to Canada or to the U.S. and then Canada. Systematic mass arrest and deportation of hundreds and later thousands of would-be refugee claimants as they arrived at Canadian ports of entry (e.g., airports, land border-crossings, or harbours) to enter the determination process over this period would have been inconsistent with liberal government. There is an .Enforcement Branch of the Department of Immigration in each of the Regions that regularly carries out similar practices on a small scale, but undertakings such as this -the naked force of political authorites and the state- have an inherent potential to upset the Canadian citizenry. They can be unstable in their effects and if constantly used on a mass scale can call into question the assumptions upon which liberal government rests (cf. Stenson, 1993a: 375; Dean, 1991; Valverde, 1991a: 105). While arrest followed by deportation certainly has been carried out to dispose of immigrants deemed to have gone seriously astray (Roberts, 1989; Avery, 1979), and was 171 later used to remove some arrivals eventually determined not to be refugees, such efforts can be understood more as a last resort than a liberal response to crisis. In the period following 1980, liberal government showed itself to be much more inventive and experimental than this. A task force on determination was created in September, 1980 that culminated in the Robinson (1981) Report, mentioned above in relation to selection. Among its recommendations were for RSAC to be made independent of the Department of Immigration and External Affairs and for every refugee claimant to be granted an oral hearing as a central part of the determination process. It was further suggested, that while oral hearings for all claimants would be more expensive than an administrative review, minor changes in the process would make up for this increase by improving efficiency (Nash, 1989: 45). In February, 1982 the Minister of Immigration convened a National Symposium on Refugee Determination in Toronto announcing structural changes that would effectively remove RSAC from the Department of Immigration (Refuge. 1982 1(7): 4). The creation of RSAC had already placed determination further away from the influence of the Departments of Immigration and External Affairs than it had been at the outset (ie. when the ad hoc committee was in operation), but this arrangement would place it even further away. The head of RSAC, however, was foreseen continuing to report to the Minister of Immigration (Dirks, 1995: 84). In line with the task force's recommendations, the Minister of Immigration also announced in May, 1983 an Oral Hearings Pilot 172 Project for Montreal and Toronto (Canada, 1988b; Refuge. 1983 3(1): 12; Ratushny, 1984: 9; Plaut, 1985: 45-46). Following introduction of these programmes and termination of the practice of delivering AAP benefits (see resettlement below) to refugee claimants in the Quebec Region in October, 1982 (Globe and Mail, 1982: 9), refugee determination continued to be deemed to be creating problems for continuing immigration programmes. In 1984 another major report observed that these problems were serious, that determination ...has been the subject of abuse by some claimants with little chance of success, who claimed refugee status in order to be permitted to remain in Canada while their claims were being processed. The laborious procedures imposed by the Act, together with a dramatic increase in claims, have created serious delays in the disposition of cases, with repercussions throughout our system of immigration enforcement (Ratushny, 1984: 10-11; emphasis added). Nevertheless, in the same report the pilot projects were judged to lead to fewer adjournments, oral interviews of shorter duration, and overall improved efficiency (Ratushny, 1984: 10). The Plaut (1985) Report, also mentioned earlier, was commissioned in 1984. Among its many recommendations when it was released was to allow refugee hearings to take place before a 'refugee board1 that would be part of a larger Immigration and Refugee Board independent from political influence; to train members of the Board through the inclusion of an Education Division; to allow two oral hearings and an appeal; and to make hearings before this board public unless the claimant requested otherwise (Plaut, 1985: 60-143). Liberal government demanded that 173 as determination became both visible and problematic that it be intensively and continuously scrutinized through: increased production of knowledge by an Education Division; liberal law in the form of hearings, appeals, and legal representation on both sides; public display; and a board freed5 from arbitrary intervention by political authorities through, in particular, the Department of Immigration. In April, 1985 the Singh decision was rendered in the Supreme Court. It was decided that the "everyone" in section seven of the Canadian Charter of Rights and Freedoms, that happened to be entrenched in 1982, referred to everyone physically present in Canada as well as "anyone seeking entry at a port of entry" (Mandel, 1994: 241). It required refugee determination in Canada to be brought in line with a notion of 'fundamental justice' (Canada, 1995b: 5; Jackman, 1988: 321). In practical terms this meant that at least one oral hearing had to be granted to refugee claimants in place of a minor administrative review of a written transcript representing the claim (Nash, 1989: 48). Hence by 1985, refugee claimants, upon reaching Canadian territory and unlike the human beings in camps and crisis situations abroad targeted by the Canadian selection programme, had become juridical units with rights that had to be respected by political authorities. Broad power relations had undergone something of a reversal. In 1985 another task force was created to begin drafting legislation for a new refugee status determination system (Dirks, 1995: 88). At this moment, sovereign efforts were undertaken to limit access 174 to determination from abroad (Creese, 1991: 13). Explicitly intended to restrict the numbers accessing determination, in June, 1985, transit visa requirements were put in place for fourteen nations (Creese, 1991: 20). These allowed visa officers abroad to pre-screen prospective visitors to Canada, including those who might later claim refugee status. Despite this, in practice in the months that followed the backlog of refugee claimants entering the system continued to grow. There was considerable mass media coverage of the crisis between 1981 and 1985, but beginning in 1986, until passage of new legislation in the summer of 1988, it increased significantly (Creese, 1991). The image of the crisis became one of Canada's territorial borders being flooded by non-citizens making refugee claims ultimately to collect welfare or otherwise gain status to live permanently in Canada (Creese, 1991: 13). The four thousand Portuguese and two thousand Turks who arrived in 1985 and 1986, the eight hundred Brazilians in early 1987 (cf. Creese, 1991: 21-2, 24; Hawkes, 1988: 255; Jackman, 1988: 323), as well as the more dramatic arrival of one hundred and fifty-five Tamils off Newfoundland's coast in lifeboats in August, 1986 and a similar number of Sikhs on the shores of Nova Scotia in July, 1987, became the focus of the crisis during this period (cf. Mandel, 1994: 246;Dirks, 1995: 90; Goodwyn-Gil, 1987: 27). Following their arrival these and other groups were thought to have made fraudulent refugee claims or to have travelled to Canada using fraudulent documents, or both. These events produced a general image of these 175 refugee claimants as criminals and queue-jumpers (cf. Hawkes, 1988: 254; Creese, 1991: 13; Mandel, 1994: 248). Here again refugee determination was seen as a problem for on-going immigration schemes. But besides immigration, at this juncture the increase in the numbers of refugee claimants also began to be perceived as pressuring Canadian provincial (and municipal) welfare programmes by, for example, dramatically increasing the numbers of welfare claims (See Creese, 1991: 23; Dirks, 1995: 87-8).47 It is in 1986, then, that the crisis began to expand from a mere concern over the arrival of increasing numbers of refugee claimants, to the determination process itself. The crisis of resources began shifting to one of governability (cf. Ashenden, 1996). A crisis of governability is not tantamount to a 'crisis of the State' (Simon, 1994: 17), as that phrase has come to be understood. Nor, in this instance, was it simply a crisis 'whipped up' by Department of Immigration officials through mass media, as Creese (1991: 4) argues, to justify gaining greater control of the types and numbers of refugees entering Canada. Consonant with the problems determination was assumed to be creating for both immigration and welfare programmes, the crisis that had been centred on a need for allocation of more staff and funds to RSAC to deal with increased volume, now expanded to the capacity of the determination system to identify refugees. As more instances of claims deemed fraudulent became known and were then dispersed through mass media, the question became: can the system distinguish the motivations of those travelling to Canada and entering the 176 process as refugee claimants? Did these claimants have no other choice but to enter the process because they feared persecution or had been persecuted? Had they merely done so because they had chosen to collect welfare ('welfare burdens') or to gain Canadian permanent resident status ('queue-jumpers')? Had they merely done so as a result of choosing to flee prosecution from political authorities elsewhere ('criminals'), or was it some combination of the latter three scenarios? The system was increasingly deemed to lack the capacity to make such distinctions. In March, 1986 a new effort foresaw dealing with the backlog that had formed from the increase in appeals of RSAC determination decisions by expanding the Immigration Appeal Board from eighteen to fifty members. This was imagined being accomplished by the fall of 1986 through passage of new legislation (Dirks, 1995: 82; Nash, 1989: 51). Legislation was announced in May, 1986 (Jackman, 1988: 322) that also foresaw a new determination system being put in place. At the same time, an Administrative Review was introduced that envisaged the transfer of all refugee claims from the RSAC backlog to a temporary ad hoc body (Canada, 1995b: 11; Creese, 1991: 12; Nash, 1989: 48). Such a programme imagined Department of Immigration officials deciding whether claimants demonstrated 'successful establishment1 in Canada (Canada, 1986e: 50; Nash, 1989: 49). The following criteria were to be used: stability in employment, the length of time employed, frequency and reasons for changes in employment, present income and future prospects, and family obligations (Canada, 1986e: 50; Nash, 1989: 49). If this 177 establishment could not be shown, then the potential for establishment was to be considered using criteria similar to those used for selecting immigrants abroad. This Review was not, therefore, a new refugee determination programme: Those receiving positive decisions were to be selected and counted as immigrants. Those receiving negative decisions from Department officials were to be transferred back to the regular refugee determination process to have their claims decided by RSAC (Canada, 1986e: 51; 1995b: 11). This was not a whole-sale amnesty programme either. Instead, it sought to circumvent the difficult question of 'who is a refugee?1 while awaiting effects of sovereign measures used abroad that promised to reduce the numbers of claimants accessing the determination process. It sought to deal with these claimants in a way comparable to the manner in which they would have been approached by state visa officers working abroad -as potential citizens of a liberal order- while simultaneously avoiding a full amnesty -'amnesty' being completely absent in programmatic statements- (e.g., Canada, 1986e: 50-1) that was assumed to have a potential of attracting still more refugee claimants to Canada Aft (Canada, 1986e). It sought to "avoid creating incentives for a new influx of claimants in the transition period" (Canada, 1986e: 50). More than twenty thousand claimants were dealt with in this manner in practice. Those persons not covered by this temporary Administrative Review (claimants arriving after May, 1986), if from nations on a list of nations assembled by political authorities (that became 178 known as the 'B-l' list), were imagined entering a process involving the automatic granting of Minister Permits and work permits. This would allow them to remain in Canada while preventing them from entering the existing backlogged determination process (Canada, 1986e: 51). The 'B-l' list was claimed to include nations that Canada did not deport persons to as a result of their having been deemed to be refugee-producing by RSAC in the past (Nash, 1989: 51). In effect, 'who is a refugee?1 would now be answered on the basis of RSAC's past practices. The Designated Classes decided by the Minister and mentioned earlier in this chapter comprised many of the nations on the B-l list. Those persons not covered by the Review and whose nationality was not on this list were envisaged entering a 'fast-track' determination process. Additional resources were to be allocated to the latter (Canada, 1986e: 51). In other words, resources thought to be required for several years here and in relation to the Immigration Appeal Board were finally to be increased, but by this point it was too late- the crisis had already changed in kind. It was no longer simply about a lack of resources, it was about the way in which determination was carried out, the rationality that guided it. Despite the introduction of sovereign measures, mentioned earlier, and these new programmes, by the summer of 1986 another backlog was quickly forming. The number of claims, particularly those associated with the 'B-l' list, continued to grow (Nash, 1989: 51). In February, 1987 the expansion of sovereign measures was announced. Those nations requiring visitor visas -now extended 179 to ninety-eight nations- were now required to have transit visas (visas allowing persons to enter Canadian territory by aircraft to await for travel elsewhere) (Canada, 1987c: 1). The rationale for this was to "reduce the number of non-bona-f ide visitors who abuse the transit privilege to claim refugee status" (Canada, 1987c: 1). Another measure was then announced that foresaw private international airline personnel being assisted in detecting false documents of travellers in order to stop them abroad before their travel to Canada (Canada, 1987c: 1-3). There was also cancellation at this time of the distinction between claimants represented in the 'B-1' list and those entering the 'fast-track' process, thereby requiring all claimants to be dealt with in the same way. The rationale for the elimination of this distinction was the 'failure' of the previous blanket approach to distinguish the motivations of those entering the process: It made "no distinction between economic migrants and refugees" (Canada, 1987b: 2). This new programme was also introduced because of: Fairness - protection was granted to the oppressor as well as the oppressed. Further, people from unaffected regions of a country benefited from the policy. Case-by-case review will allow the government to remove people where there are no personal consequences from this action. . . it recognizes that we need to view each individual's need within the specific context of the country to which they might be returned. . . [c]ase-by-case procedures will allow the government to remove those who are not genuine refugees i.e. economic migrants, or those who are undesirables because they are criminal offenders or security risks (emphasis in original; Canada, 1987b: 2-3). The 'failure' of the previous approach can be seen above to be centred on the inability of the system to distinguish motivations 180 of claimants so as to answer the question of 'who is a refugee?1 and, by implication, who could be simply disposed of. Sufficient answers to this question were now deemed to require considerably more detailed knowledge of specific contexts from which the individual claimant had left. Allowing political authorities to dictate answers to 'who is a refugee?' through use of a 'B-l' list after only eight months was deemed to have 'failed'. This reinforced the perception of a need for a new system. Unlike those human beings in camp and crisis situations within the gaze of visa officers, the refugee claimant within the determination process in Canada had, in 1985, become a juridical subject equipped with rights that were not to be interfered with by political authorities. It should be recognized that in determination after the refugee claimant came to be deemed as such and a crisis of governability occurred, what became important was individual victimization, rather than a mix of collective (either nation or group) and individual victimization, as seen for the eight months preceding February 1987, or collective (either nation or group) victimization exclusively, as in selection abroad. As refugee claimants were now assumed to be juridical units in Canada, liberal government demanded that the question of 'who is a refugee?' be continually scrutinized. The Refugee Reform Act (Bill C-55) was passed in the summer of 1988 during an emergency sitting of Parliament that followed extensive mass media coverage of the dramatic arrival of the Sikhs mentioned above (Creese, 1991). A few months prior to its passage, 181 however, one last effort that would allow political authorities to dictate answers to the question of 'who is a refugee?1 is seen. This would occur in May, 1988 when the debate over the Bill in the Senate came to be centred on a proposed 'safe third country' provision (Dirks, 1995: 94) that would allow refugee claimants to be returned to a 'safe country' without having their claims determined. Which nations were 'safe' was foreseen in the provisions of the Bill to be dictated by Canadian political authorities. A 'safe third country' list, like the earlier Administrative Review, promised to allow Canadian political authorities to side-step the difficult question of 'who is a refugee?', answering instead 'which nations are not refugee-producing? ' . The latter question was recognized, however, as having the same potential effect as the former because the number of nations at any moment is finite. As Nash (1989: 56) reports: Witnesses before the Committee pointed out that this might 'bring refugee issues into the political arena1 because the absence of countries from the list would involve a public judgement on those countries (emphasis added). As a compromise the Senate Committee recommended a 'safe third country' list be compiled, not by political authorities, but by the Chairman of the proposed independent IRB in consultation with the CRDD, the Department of Immigration, External Affairs, the UNHCR, and refugee experts. In doing so these latter authorities were foreseen taking into account the human rights records of nations (Nash, 1989: 56-7). On July 15, 1988 the Minister of Immigration, 182 rejecting this compromise, amended the 'safe third country1 provisions in the Act just prior to its passage four days later, such that "claimants could be sent back only to countries that they had 'lawful authority to be in'" (Nash, 1989: 57). As political authorities of other nations normally decide who could 'lawfully' be within their national territories, in effect this amendment imagined them, rather than Canadian authorities, deciding which nations were safe third countries for the purpose of the Canadian determination system (Nash, 1989: 57; Dirks, 1995: 94). Once again, the inherently difficult question of 'who is a refugee?' had been side-stepped. After passage of the Bill, another temporary administrative review (Dirks, 1995: 95-6) was created to deal with the backlog of over eighty thousand claims that had appeared by the end of 1988 (it was completed three years later) (Canada, 1995b: 11). It too managed to avoid the question of 'who is a refugee?'. The period between 1986 and 1988 demonstrates, perhaps more than anything else, the remarkable inventiveness of liberal government in dealing with opposition and crisis. The continuing crisis of governability over this period had encouraged a new programme that anticipated a three-stage process 4Q to replace RSAC beginning in 1989. A large, quasi-judicial decision-making body, the Immigration and Refugee Board (IRB) was invented. A division (CRDD) of this independent 'non-political' body would oversee refugee hearings and held out the promise of effectively and continually responding to the question of 'who is a refugee?' in Canada. This programme imagined legal professionals 183 representing refugee claimants through the provision of legal aid and developing their claims prior to presentation at an oral hearing. These hearings would be attended by a refugee hearing officer and overseen by two members of the Convention Refugee Determination Division (CRDD) of the IRB, of which only one needed to decide in favour of a claimant for status to be granted (Hathaway, 1992: 82).50 The members would be appointed by political authorities for multi-year terms. Whereas within RSAC there was little or no emphasis on the need for formal knowledge51 about refugee-producing nations and specific contexts from which refugee claimants had migrated, the new process was to be highly reliant on the production of such knowledge outside the state. A new technology, the documentation centre, was deployed within this programme to allow IRB members, refugee hearing officers, refugee claimants' counsel and, more significantly, the Canadian citizenry to this knowledge (see chapter six). Considerably more resources would be directed toward the CRDD for staff and these centres than had been foreseen required for RSAC to counter the crisis of resources before 1986 (Dirks, 1995: 90). Through the documentation centres the facts would be provided and then subjected to legal judgement in the hearings. The new process, therefore, sought to govern the difficult question of 'who is a refugee?1 by transforming it into a technical (in the documentation centre) and then a legal question (during oral hearings). Formal knowledge and law would seek 'the Truth' of 'who is a refugee?' and, by implication, who was merely a criminal 184 fleeing prosecution, welfare burden, queue-jumper, or some combination of these objects/identities assumed within a liberal order to be worthy of quick disposal. Once claims were determined, the refugee claimant would be either granted immigrant status and resettled or, following appeals, encouraged to leave and later literally removed from Canadian territory. Such a programme held out the hope of rendering determination governable again. As seen in the next chapter, through the introduction of the documentation centre that promised to allow this new system to operate, something else was accomplished besides. Before the 1980s refugeeness was deployed in the international sphere by Canadian political authorities to great advantage (fostering alliances with other Western nations, managing and replenishing the Canadian population, fulfilling the economy's labour needs, and so on). But during the 1980s there occurred something of a reversal in this broad power relation: Refugeeness, previously implanted in populations outside the West by Western authorities to the latter's advantage, could now be used by these same populations who sought to enter Canada to escape persecution or to otherwise improve their well-being by prolonging or securing their stay in Canada to demand certain rights. With the introduction of the documentation centre, however, a panoptic device that promises to allow non-Western regions, nations and populations to be intensively surveilled (see chapter six), as well as other changes that have followed its appearance, this particularly broad relation has begun to be reversed again. To see 185 this, as well as several other instances of the inventiveness of liberal government in the context of determination, changes since 1989 require discussion. After three years of operation, the determination system was to be significantly altered again. Through passage of Bill C-86 in 1992 the initial 'credibility' stage of the new process was removed. Deferment of landing until satisfactory identification of refugees and the entering into agreements with other nations over responsibility for the examination of refugee claimants were also to be made possible through this legislation (Canada, 1992a). The latter measure sought to deter those who would otherwise travel to Canada without passports or other identity documents and then enter the determination process. The Bill also allowed for the introduction of guidelines for determination to be used by the IRB. This amendment to the Immigration Act came into force in February, 1993 (Canada, 1992a). The rationale for these changes was to improve the efficiency and effectiveness of the refugee status determination system without compromising the protection which Canada has provided to refugees. The Minister hopes to avoid future backlogs which weaken the determination process and delay the acceptance of genuine refugees (Dialogue. 1(3): 1 1992). These changes were not aimed at restoring governability to determination- this had been already accomplished. Instead, they sought to respond to specific opposition, similar in kind to that levelled against the RSAC process early in the 1980s. Since the coming into force of this amendment, the guidelines on "Women Refugee Claimants Fearing Gender-Related Persecution" men