《国际政治经济学》文献资料（Political Economy of Migration）No Vacancy The Political Geography of Immigration Control in Advanced Industrial Countries

International Organization http://journals.cambridge.orq/INO Additional services for International Organization: Email alerts:Click here Subscriptions:Click here Commercial reprints:Click here Terms of use:Click here No Vacancy:The Political Geography of Immigration Control in Advanced Industrial Countries Jeannette Money International Organization Volume 51 Issue 04 September 1997,pp 685-720 D0:10.1162/002081897550492,Published online:18 August2003 Link to this article:http:/liournals cambridge orgl abstract S0020818397440225 How to cite this article: Jeannette Money (1997).No Vacancy:The Political Geography of Immigration Control in Advanced Industrial Countries.International rganization,51,pp685-720doi:10.1162/002081897550492 Request Permissions:Click here CANRIKE JOURNALS Downloaded from http://journals.cambridge.org/INO,IP address:211.80.95.69 on 21 Oct 2013

International Organization http://journals.cambridge.org/INO Additional services for International Organization: Email alerts: Click here Subscriptions: Click here Commercial reprints: Click here Terms of use : Click here No Vacancy: The Political Geography of Immigration Control in Advanced Industrial Countries Jeannette Money International Organization / Volume 51 / Issue 04 / September 1997, pp 685 - 720 DOI: 10.1162/002081897550492, Published online: 18 August 2003 Link to this article: http://journals.cambridge.org/ abstract_S0020818397440225 How to cite this article: Jeannette Money (1997). No Vacancy: The Political Geography of Immigration Control in Advanced Industrial Countries. International Organization, 51, pp 685-720 doi:10.1162/002081897550492 Request Permissions : Click here Downloaded from http://journals.cambridge.org/INO, IP address: 211.80.95.69 on 21 Oct 2013

No Vacancy:The Political Geography of Immigration Control in Advanced Industrial Countries Jeannette Money This article examines the domestic political sources of immigration control in ad- vanced market economy countries after World War II.'Immigration control can be distinguished from the broader concept of immigration policy by its emphasis on state policies that define the permissible level of resident alien admissions.2 The analysis is based on the well-established fact that immigrant communities are geo- graphically concentrated.I argue that this geographic concentration creates an un- even distribution of costs and benefits,providing a spatial context for immigration politics.In this context,net public demand for tighter immigration control increases in localities where immigrants concentrate when those areas experience higher unem- ployment,rapid increases in immigration,higher immigrant proportions,and more generous immigrant access to social services.Each of these conditions aggravates competition between immigrants and natives,and hence native hostility,in these communities while employer support for immigration usually diminishes.Yet na- tional politicians may ignore changes in the demand for immigration control unless these constituencies are also able to swing a national election from one party to another.The larger and less "safe"the local constituencies,the greater their influ- ence in this sense.Evidence from the United Kingdom between 1955 and 1981 is consistent with these propositions. I would like to acknowledge the financial support of the Center for International Relations at the Univer- sity of California,Los Angeles;the faculty senate of the University of California,Davis;the Center for German and European Studies(CGES)at the University of California,Berkeley;and the Institute on Global Conflict and Cooperation at the University of California,San Diego.I would also like to thank the various individuals who supported this research and commented on different aspects of the project.in- cluding Gary Freeman.Robert Jackman,David Lake,Philip Martin.Ted Perlmutter,Richard Rosecrance, and members of the working group on Comparative Immigration and Integration Policy associated with CGES,as well as John Odell and two anonymous reviewers.Finally,I would like to thank Jan Breemer, Stacy Bumett.and Tressa Tabares for their research assistance. 1.Although immigration control includes temporary visitors who enter the country for a short time. such as tourists and business visitors.I focus only on those who enter on a nontemporary basis.I also exclude from the analysis policies controlling refugees and asylum seekers. 2.Hammar 1985. International Organization 51,4,Autumn 1997,pp.685-720 1997 by The IO Foundation and the Massachusetts Institute of Technology

No Vacancy: The Political Geography of Immigration Control in Advanced Industrial Countries Jeannette Money This article examines the domestic political sources of immigration control in ad￾vanced market economy countries after World War II.1 Immigration control can be distinguished from the broader concept of immigration policy by its emphasis on state policies that deŽ ne the permissible level of resident alien admissions.2 The analysis is based on the well-established fact that immigrant communities are geo￾graphically concentrated. I argue that this geographic concentration creates an un￾even distribution of costs and beneŽ ts, providing a spatial context for immigration politics. In this context, net public demand for tighter immigration control increases in localitieswhere immigrants concentratewhen those areas experience higher unem￾ployment, rapid increases in immigration, higher immigrant proportions, and more generous immigrant access to social services. Each of these conditions aggravates competition between immigrants and natives, and hence native hostility, in these communities while employer support for immigration usually diminishes. Yet na￾tional politicians may ignore changes in the demand for immigration control unless these constituencies are also able to swing a national election from one party to another. The larger and less ‘‘safe’’ the local constituencies, the greater their in u￾ence in this sense. Evidence from the United Kingdom between 1955 and 1981 is consistent with these propositions. I would like to acknowledge the Ž nancial support of the Center for InternationalRelations at the Univer￾sity of California, Los Angeles; the faculty senate of the University of California, Davis; the Center for German and European Studies (CGES) at the University of California, Berkeley; and the Institute on Global Con ict and Cooperation at the University of California, San Diego. I would also like to thank the various individuals who supported this research and commented on different aspects of the project, in￾cluding Gary Freeman, Robert Jackman, David Lake, Philip Martin, Ted Perlmutter, Richard Rosecrance, and members of the working group on Comparative Immigration and Integration Policy associated with CGES, as well as John Odell and two anonymousreviewers. Finally, I would like to thank Jan Breemer, Stacy Burnett, and Tressa Tabares for their research assistance. 1. Although immigration control includes temporary visitors who enter the country for a short time, such as tourists and business visitors, I focus only on those who enter on a nontemporary basis. I also exclude from the analysis policies controlling refugees and asylum seekers. 2. Hammar 1985. International Organization 51, 4, Autumn 1997, pp. 685–720 r 1997 by The IO Foundation and the Massachusetts Institute of Technology

686 International Organization In advanced market economy countries in the late 1990s,the importance of under- standing the politics of immigration control may seem self-evident.Immigration is a Janus-faced policy with both domestic and international consequences.Domesti- cally,many politicians are confronting a politically powerful backlash against for- eign residents of all types:undocumented and documented,foreign born and foreign by ethnicity,workers and dependents,voluntary migrants and involuntary refugees and asylum seekers.At the same time,employers continue to petition for greater access to migrant labor,both skilled and unskilled.Internationally,conflict arises over disparate national interests,such as the desire by developing countries for greater emigration versus the preference for limited immigration by advanced industrial coun- tries.Alternatively,conflict arises when one country implements policies that transfer its control problems toother nations.The 1993 German policy to reject asylum seekers transit- ing"safe"countriesen route to Germany reduced Germany's intake of asylum seek- ers but only at the expense of an increase of immigrants in neighboring countries. Empirically,national patterns of immigration control vary widely,despite similar positions of advanced industrial countries in the global economy.From the mid- 1950s to the mid-1970s,for example,many OECD (Organization for Economic Co- operation and Development)countries experienced low unemployment and tight la- bor markets,leading to rising wages.Yet some countries chose to import immigrant labor to moderate wages and labor-market shortages(for example,France,Germany, and Australia),whereas others discouraged immigration altogether(for example,Ja- pan)or began closing the door to labor-market immigrants well before the oil shock of 1973(for example,Great Britain,Switzerland,and Sweden).In light of similar labor market conditions,the variation in immigration control represents an interest- ing empirical puzzle. Moreover,the international flow of people has not yet been systematically drawn into the research agenda in international political economy.3 Although the determi- nants of global patterns of trade,production,and capital flows are now widely stud- ied,immigration remains predominantly in the domain of economists,demogra- phers,and sociologists.The literature on immigration policy formation that does exist tends to be country specific rather than comparative,making it difficult to sort between idiosyncratic factors and more generally applicable theories.As James F. Hollifield notes,"truly comparative works on immigration are few.In the field of migration studies,the tendency has been to collect national case studies,bind them together,and call the study comparative.Such compendia are useful sources of infor- mation,but they rarely yield theoretical insights."4 Comparative research on the sources of immigration policy and patterns will inform the research agenda of inter- national political economy by including an important flow of resources across na- tional boundaries. The article is divided into four parts.In the first section I briefly describe the distinctive patterns of immigration in selected OECD countries and the standard explanations advanced to account for these patterns.I explain in the second section 3.Haus1995 4.Hollifield 1992,17

In advanced market economy countriesin the late 1990s, the importance of under￾standing the politics of immigration control may seem self-evident. Immigration is a Janus-faced policy with both domestic and international consequences. Domesti￾cally, many politicians are confronting a politically powerful backlash against for￾eign residents of all types: undocumented and documented,foreign born and foreign by ethnicity, workers and dependents, voluntary migrants and involuntary refugees and asylum seekers. At the same time, employers continue to petition for greater access to migrant labor, both skilled and unskilled. Internationally, con ict arises over disparate nationalinterests,such asthe desire by developingcountriesfor greater emigration versusthe preference for limited immigration by advanced industrial coun￾tries. Alternatively, conict arises when one country implements policies that transfer its controlproblemsto other nations.The 1993German policy to reject asylumseekerstransit￾ing ‘‘safe’’ countries en route to Germany reduced Germany’s intake of asylum seek￾ers but only at the expense of an increase of immigrants in neighboring countries. Empirically, national patterns of immigration control vary widely, despite similar positions of advanced industrial countries in the global economy. From the mid- 1950sto the mid-1970s, for example, many OECD (Organization for Economic Co￾operation and Development) countries experienced low unemployment and tight la￾bor markets, leading to rising wages. Yet some countries chose to import immigrant labor to moderate wages and labor-market shortages(for example, France, Germany, and Australia), whereas others discouraged immigration altogether (for example, Ja￾pan) or began closing the door to labor-market immigrants well before the oil shock of 1973 (for example, Great Britain, Switzerland, and Sweden). In light of similar labor market conditions, the variation in immigration control represents an interest￾ing empirical puzzle. Moreover, the international  ow of people has not yet been systematically drawn into the research agenda in international political economy.3 Although the determi￾nants of global patterns of trade, production, and capital  ows are now widely stud￾ied, immigration remains predominantly in the domain of economists, demogra￾phers, and sociologists. The literature on immigration policy formation that does exist tends to be country speciŽ c rather than comparative, making it difficult to sort between idiosyncratic factors and more generally applicable theories. As James F. HolliŽ eld notes, ‘‘truly comparative works on immigration are few. In the Ž eld of migration studies, the tendency has been to collect national case studies, bind them together, and call the study comparative. Such compendia are usefulsources of infor￾mation, but they rarely yield theoretical insights.’’ 4 Comparative research on the sources of immigration policy and patterns will inform the research agenda of inter￾national political economy by including an important  ow of resources across na￾tional boundaries. The article is divided into four parts. In the Ž rst section I brie y describe the distinctive patterns of immigration in selected OECD countries and the standard explanations advanced to account for these patterns. I explain in the second section 3. Haus 1995. 4. HolliŽ eld 1992, 17. 686 International Organization

Political Geography of Immigration Control 687 how the geographic concentration of immigrant communities affects the domestic distribution of costs and benefits associated with immigration control policies.5 This geographic concentration provides the spatial context for the politics of immigration control.I then examine the local conditions that create support for and opposition to immigration by focusing on the business community and the general population.To explain the changing attitudes of the native population toward the immigrant commu- nity,I adopt a theory of native-immigrant competition over scarce resources.Busi- ness support varies with the tightness and flexibility of labor markets as well as with the potential for capital mobility,among other factors.I then connect the underlying structural conditions to the local and national political agendas through the dynamics of party competition.In the third section,I illustrate the model with an analysis of British immigration policy.In the fourth section I compare the analysis with the British literature on immigration control to evaluate the power of the general frame- work against country-specific explanations.Finally,I return to the cross-national variation in immigration control and suggest areas for further theoretical and empiri- cal research. Although standard explanations of immigration policy favor domestic political factors,my approach is distinctive in three ways.It focuses on the local community as the primary unit of political analysis;it provides a theory to explain changing local preferences for and against immigration;and it systematically incorporates institu- tional aspects of the political system through which societal demands are funneled. Immigration Patterns and Prevailing Explanations Trends in Immigration Industrial democracies vary greatly in the degree to which they discourage immigra- tion.Because countries employ different labels for different types of individuals who cross national borders,it is impossible to provide a concise picture of cross-national variation in resident alien intake(defined as the level of aliens permitted to enter the host country and take up residence for at least twelve months).Nonetheless,a variety of indicators suggests that tolerance for resident aliens varies substantially among advanced industrial countries.One such indicator is the average annual gross flow of legal resident aliens per capita.As depicted in Table 1,the intake range is broad.At the low end,Japan permitted entrance to only 3 aliens per ten thousand national population per year on average between 1962 and 1991.Australia and New Zealand are at the high end of the spectrum,allowing the entrance of 81 and 136 resident aliens,respectively,per ten thousand national population.Belgium and Canada are 5.Immigration policy is generally understood to be composed of two components-immigration control and immigrant integration.My research question specifically addresses the former rather than the latter and,as such,may not be generalizable to this second aspect of immigration policy. 6.A second indicator is the"stock"of resident aliens as a proportion of total population:see SOPEMI 1993 for current and historical data

how the geographic concentration of immigrant communities affects the domestic distribution of costs and beneŽ ts associated with immigration control policies.5 This geographic concentration providesthe spatial context for the politics of immigration control. I then examine the local conditionsthat create support for and opposition to immigration by focusing on the business community and the general population. To explain the changing attitudesof the native populationtoward the immigrant commu￾nity, I adopt a theory of native–immigrant competition over scarce resources. Busi￾ness support varies with the tightness and  exibility of labor markets as well as with the potential for capital mobility, among other factors. I then connect the underlying structural conditionsto the local and national political agendasthrough the dynamics of party competition. In the third section, I illustrate the model with an analysis of British immigration policy. In the fourth section I compare the analysis with the British literature on immigration control to evaluate the power of the general frame￾work against country-speciŽ c explanations. Finally, I return to the cross-national variation in immigration control and suggest areas for further theoretical and empiri￾cal research. Although standard explanations of immigration policy favor domestic political factors, my approach is distinctive in three ways. It focuses on the local community asthe primary unit of political analysis; it provides a theory to explain changing local preferences for and against immigration; and it systematically incorporates institu￾tional aspects of the politicalsystem through which societal demands are funneled. Immigration Patterns and Prevailing Explanations Trends in Immigration Industrial democracies vary greatly in the degree to which they discourage immigra￾tion. Because countries employ different labelsfor different types of individualswho cross national borders, it is impossible to provide a concise picture of cross-national variation in resident alien intake (deŽ ned as the level of aliens permitted to enter the host country and take up residence for at least twelve months). Nonetheless, a variety of indicators suggests that tolerance for resident aliens varies substantially among advanced industrial countries. One such indicator is the average annual gross  ow of legal resident aliens per capita.6 As depicted in Table 1, the intake range is broad. At the low end, Japan permitted entrance to only 3 aliens per ten thousand national population per year on average between 1962 and 1991. Australia and New Zealand are at the high end of the spectrum, allowing the entrance of 81 and 136 resident aliens, respectively, per ten thousand national population. Belgium and Canada are 5. Immigration policy is generally understood to be composed of two components—immigration control and immigrant integration. My research question speciŽ cally addresses the former rather than the latter and, as such, may not be generalizable to thissecond aspect of immigration policy. 6. A second indicator is the ‘‘stock’’ of resident aliens as a proportion of total population;see SOPEMI 1993 for current and historical data. Political Geography of Immigration Control 687

688 International Organization TABLE 1.Average legal resident alien intakes in fourteen OECD countries, 1962-91(annual flow of legal alien residents per ten thousand) Country Mean Standard deviation Number ofyears Japan 3 23 France 23 14 29 United States Finland 91 United Kingdom 38 Netherlands 670 器 Norway 47 Sweden 29 3 Belgium 1 28 Canada 62 21 Denmark 6 Australia 81 29 Germany (West) 122 93082 New Zealand 136 2 Total 58 41 6 Source:Money 1996. closest to the unweighted annual average of the data set,with 58 per ten thousand per year. Some surprising facts emerge from these data.For example,the United States is widely perceived as a country of immigrants with a tradition of openness to immigra- tion.Yet,although the United States admitted large absolute numbers of immigrants, on a per capita basis it is located toward the low end of the scale,with 24 aliens per year per ten thousand on average for the period under review.Another anomaly is Germany.Even though it proclaims that it is not a country of immigration,(West) Germany admitted relatively large flows of aliens on a per capita basis,even when excluding the ethnic Germans who migrated from eastern Europe and the former Soviet Union,as these figures do. Explanation of Trends Many explanations have been advanced to account for variation in immigration con- trol policies among advanced industrial countries.Most point to aspects of national identity,to economic factors,or to the interaction between economic and cultural dimensions of immigration. One group of analysts focuses on conceptions of ethnicity,citizenship,and/or national identity to predict immigration policy.?Doris Meissner,for example,juxta- poses European nations in which citizenshipis tied to"shared ethnicity and national- 7.See,for example,Foot 1965 on Britain;Higham 1963 on the United States;and Meissner 1992 for a comparative analysis

closest to the unweighted annual average of the data set, with 58 per ten thousand per year. Some surprising facts emerge from these data. For example, the United States is widely perceived as a country of immigrants with a tradition of opennessto immigra￾tion. Yet, although the United States admitted large absolute numbers of immigrants, on a per capita basis it is located toward the low end of the scale, with 24 aliens per year per ten thousand on average for the period under review. Another anomaly is Germany. Even though it proclaims that it is not a country of immigration, (West) Germany admitted relatively large  ows of aliens on a per capita basis, even when excluding the ethnic Germans who migrated from eastern Europe and the former Soviet Union, as these Ž gures do. Explanation of Trends Many explanationshave been advanced to account for variation in immigration con￾trol policies among advanced industrial countries. Most point to aspects of national identity, to economic factors, or to the interaction between economic and cultural dimensions of immigration. One group of analysts focuses on conceptions of ethnicity, citizenship, and/or national identity to predict immigration policy.7 Doris Meissner, for example, juxta￾poses European nationsin which citizenship istied to ‘‘shared ethnicity and national- 7. See, for example, Foot 1965 on Britain; Higham 1963 on the United States; and Meissner 1992 for a comparative analysis. TABLE 1. Average legal resident alien intakesin fourteen OECD countries, 1962–91(annual  ow of legal alien residents per ten thousand) Country Mean Standard deviation Number of years Japan 3 1 23 France 23 14 29 United States 24 13 29 Finland 26 6 11 United Kingdom 38 7 28 Netherlands 40 10 14 Norway 47 9 29 Sweden 51 15 30 Belgium 53 13 28 Canada 62 21 28 Denmark 66 9 30 Australia 81 30 29 Germany (West) 122 48 29 New Zealand 136 26 28 Total 58 41 365 Source: Money 1996. 688 International Organization

Political Geography of Immigration Control 689 ity"to attitudes in the"traditional settler nations,"such as Australia,Canada,and the United States,"where nation building through immigration led to ideas of member- ship based on civic participation and a generally shared commitment to democratic values."8 From this perspective,"settler"states are those where nationalidentity and citizenship are based on civic values of participation;immigrants are viewed as po- tential citizens,and the state is relatively open to immigration.In contrast,"nonim- migrant"nations are those where citizenship and national identity are based on ethnicity; where these cultural values hold,immigrants are viewed as aliens or foreigners,and low levels of immigration are anticipated.Because national identity is fixed,in order to account for the variation over time the analyses introduce additional exogenous variables. From the "identity"perspective,elites,institutional structures,or extremist groups, for example,can damp down or kindle pressures to control alien entry.10 Economic interests of domestic political actors are a second major explanatory scheme.One group of analysts favors employer interests in explaining levels of openness.According to Marxian analyses,employer interests are dominant because of their privileged access to the capitalist state;for public choice analysts,it is because employers face fewer collective action problems than do employees or the public at large:12 for labor market analysts,employer interests are a reasonable proxy for other societal interests because migrant labor is concentrated in the secondary labor market and therefore complements and enhances the returns to the native labor force deployed in the primary labor market.3 Alternatively,domestic political actors are believed to have competing economic interests,forcing them to vie in the politi- cal arena for different policy outcomes.That is,"different social classes within the national territory will have conflicting interests which can result in opposing posi- tions on immigration."14 In other words,employers are one important class of politi- cal actors in favor of immigration,but now unions(or other interest groups)enter the political equation as actors who"might be opposed on the ground that this will harm workers'wages and working conditions."From the "economic"perspective,varia- tion is explained either in terms of employer demand for labor and hence for mi- grants or in terms of power balances among societal actors,some who support immi- gration,others who oppose it. Finally,many analysts point to the tension between economic benefits that accrue from immigration and the cultural costs that openness entails.One method of model- ing the relationship between economic and cultural variables and political outcomes focuses on the competition for scarce resources.5 National-ethnic identity is defined 8.Meissner 1992,70. 9.Altematively,"this view assumes that ethnic and racial strife is inevitable and invariant"and cannot account for variation over time;see Olzak 1992. 10.See,for example,Roeder 1994. 11.See Castells 1975;Castles and Kosack 1973;and Petras 1981. 12.Freeman 1995. 13.Piore1979. 14.Leitner1995,262. 15.For alternative ways of modeling the relationship between the economic and cultural variables,see Hollifield 1992;Freeman 1995;Zolberg 1983;and Leitner 1995

ity’’to attitudesin the ‘‘traditionalsettler nations,’’such asAustralia, Canada, and the United States, ‘‘where nation building through immigration led to ideas of member￾ship based on civic participation and a generally shared commitment to democratic values.’’ 8 From this perspective,‘‘settler’’states are those where national identity and citizenship are based on civic values of participation; immigrants are viewed as po￾tential citizens, and the state is relatively open to immigration. In contrast, ‘‘nonim￾migrant’’ nations are those where citizenship and national identity are based on ethnicity; where these cultural values hold, immigrants are viewed as aliens or foreigners, and low levels of immigration are anticipated.Because nationalidentityisŽ xed,in order to account for the variation over time the analyses introduce additional exogenous variables.9 From the ‘‘identity’’ perspective, elites, institutionalstructures, or extremist groups, for example, can damp down or kindle pressures to control alien entry.10 Economic interests of domestic political actors are a second major explanatory scheme. One group of analysts favors employer interests in explaining levels of openness. According to Marxian analyses, employer interests are dominant because of their privileged access to the capitalist state;11 for public choice analysts, it is because employers face fewer collective action problems than do employees or the public at large;12 for labor market analysts, employer interests are a reasonable proxy for other societal interests because migrant labor is concentrated in the secondary labor market and therefore complements and enhancesthe returns to the native labor force deployed in the primary labor market.13 Alternatively, domestic political actors are believed to have competing economic interests, forcing them to vie in the politi￾cal arena for different policy outcomes. That is, ‘‘different social classes within the national territory will have con icting interests which can result in opposing posi￾tions on immigration.’’ 14 In other words, employers are one important class of politi￾cal actorsin favor of immigration, but now unions(or other interest groups) enter the political equation as actors who ‘‘might be opposed on the ground that this will harm workers’ wages and working conditions.’’ From the ‘‘economic’’ perspective, varia￾tion is explained either in terms of employer demand for labor and hence for mi￾grants or in terms of power balances among societal actors,some who support immi￾gration, others who oppose it. Finally, many analysts point to the tension between economic beneŽ ts that accrue from immigration and the cultural coststhat openness entails. One method of model￾ing the relationship between economic and cultural variables and political outcomes focuses on the competition for scarce resources.15 National–ethnic identity is deŽ ned 8. Meissner 1992, 70. 9. Alternatively, ‘‘this view assumes that ethnic and racial strife is inevitable and invariant’’ and cannot account for variation over time; see Olzak 1992. 10. See, for example, Roeder 1994. 11. See Castells 1975; Castles and Kosack 1973; and Petras 1981. 12. Freeman 1995. 13. Piore 1979. 14. Leitner 1995, 262. 15. For alternative ways of modeling the relationship between the economic and cultural variables, see HolliŽ eld 1992; Freeman 1995; Zolberg 1983; and Leitner 1995. Political Geography of Immigration Control 689

690 International Organization as a contested social boundary that becomes politically significant only when compe- tition for scarce resources arises between the native and the immigrant populations. Susan Olzak,for example,argues that"competition potentially occurs when two or more groups come to exploit the same realized niches."6"Niche overlap"triggers ethnic competition and,with it,anti-immigrant sentiment.From the"competition" perspective,anti-immigrant sentiment and hence political opposition to immigration are not constant but change in light of changing circumstances:the size of the immi- grant community,its rate of growth,and other socioeconomic factors. Each of these alternative hypotheses has been criticized on both logical and empiri- cal grounds.7 However,rather than discard the hypotheses,I suggest that because immigrants are geographically concentrated,the analysis must be disaggregated to the level of the salient local political units.Then,each of these perspectives may provide some insight into the dynamics of immigration policy.Those who emphasize national identity suggest that the analysis must also be sensitive to factors that mediate between public opinion and policy outputs,that is,the political institutions of the nation.Those who emphasize economic variables suggest that important political actors are not limited to those concerned with national identity. Societies have political actors who actively promote immigration as well as those who oppose it.Those who focus on the interaction of cultural and economic variables suggest that the intensity and hence the political significance of national- ethnic identity may vary,and that explanations should incorporate variables that affect the intensity of public opinion.In this sense,the three types of analyses are incomplete but complement each other.The explanation I offer builds on these in- sights while emphasizing a fourth element,the geographic concentration of the immi- grant community. The Political Geography of Immigration Control The Geographic Concentration of Migrants Among demographers and sociologists,the geographic specificity of migratory pat- terns is well known,on both the sending and receiving end."Migrants arriving in a particular country do not spread out randomly throughout all possible destina- tions."18 Examples abound.In the United States,six states of fifty-California,Texas, Florida,New York,New Jersey,and Illinois-host 73 percent of the immigrant popu- lation;33 percent of the foreign-born population reside in California alone.In 1990, 8 percent of the U.S.population were foreign born,but they comprised 22 percent of California's population and 16 percent of New York's.This pattern holds in Australia as well.There,in 1991,22 percent of the population were foreign born;this popula- tion is concentrated in the cities and suburbs of Sydney and Melbourne.Of the 147 16.01zak1992. 17.In conformity with academic fashion,all of the authors cited provide a critique of the altemative theoretical frameworks while making the case for their own. 18.White1993b,52

as a contested social boundary that becomes politicallysigniŽ cant only when compe￾tition for scarce resources arises between the native and the immigrant populations. Susan Olzak, for example, argues that ‘‘competition potentially occurs when two or more groups come to exploit the same realized niches.’’ 16 ‘‘Niche overlap’’ triggers ethnic competition and, with it, anti-immigrant sentiment. From the ‘‘competition’’ perspective, anti-immigrant sentiment and hence political opposition to immigration are not constant but change in light of changing circumstances: the size of the immi￾grant community, its rate of growth, and other socioeconomic factors. Each of these alternativehypotheseshas been criticized on both logical and empiri￾cal grounds.17 However, rather than discard the hypotheses, I suggest that because immigrants are geographically concentrated, the analysis must be disaggregated to the level of the salient local political units. Then, each of these perspectives may provide some insight into the dynamics of immigration policy. Those who emphasize national identity suggest that the analysis must also be sensitive to factors that mediate between public opinion and policy outputs, that is, the political institutions of the nation. Those who emphasize economic variables suggest that important political actors are not limited to those concerned with national identity. Societies have political actors who actively promote immigration as well as those who oppose it. Those who focus on the interaction of cultural and economic variables suggest that the intensity and hence the political signiŽ cance of national– ethnic identity may vary, and that explanations should incorporate variables that affect the intensity of public opinion. In this sense, the three types of analyses are incomplete but complement each other. The explanation I offer builds on these in￾sights while emphasizing a fourth element, the geographic concentrationof the immi￾grant community. The Political Geography of Immigration Control The Geographic Concentration of Migrants Among demographers and sociologists, the geographic speciŽ city of migratory pat￾terns is well known, on both the sending and receiving end. ‘‘Migrants arriving in a particular country do not spread out randomly throughout all possible destina￾tions.’’ 18 Examples abound.In the United States,six states of Ž fty—California,Texas, Florida, New York, New Jersey, and Illinois—host 73 percent of the immigrant popu￾lation; 33 percent of the foreign-born population reside in California alone. In 1990, 8 percent of the U.S. populationwere foreign born, but they comprised 22 percent of California’s population and 16 percent of New York’s. This pattern holdsin Australia as well. There, in 1991, 22 percent of the population were foreign born; this popula￾tion is concentrated in the cities and suburbs of Sydney and Melbourne. Of the 147 16. Olzak 1992. 17. In conformity with academic fashion, all of the authors cited provide a critique of the alternative theoretical frameworks while making the case for their own. 18. White 1993b, 52. 690 International Organization

Political Geography of Immigration Control 691 parliamentary constituencies with approximately equal population,the smallest per- centage of foreign born was 4.8 percent,and the largest was ten times that level,49 percent.Thirty of the 147 constituencieshad foreign-born populationsof at least 30 percent;all but seven of these are located in the Melbourne and Sydney metropolitan areas.A similar concentration of immigrants is found in European states.In Britain, for example,58 percent of the Afro-Caribbeans live in Greater London as do 80 percent of the black African immigrants.In France,40 percent of the foreign-born population can be found in the Parisian basin,with the remainder of the immigrants concentratedin the industrial regions surrounding Lyon,Marseilles,and Strasbourg.1 This spatial concentration of migrants is attributed to several factors.Paul White enumerates the economic,social,political,and geographic forces that affect migrant destinations in the host countries,the most important of which are the initial eco- nomic pull of labor markets and the subsequent reinforcement of migrant concentra- tion through migrant networks or"chain migration."20 This brief overview indicates that the spatial concentration of immigrants in host countries is well documented empirically and well understood theoretically.The po- litical significance of this concentration,however,has been widely overlooked.One factor that may diminish the significance of the immigrant community's geographic concentration-and may account for why this aspect of immigration has been ig- nored in political analyses of immigration policy-is internal migration.21 Internal migration tends to distribute the economic costs and benefits more evenly among the indigenous population.Furthermore,the social impact of immigration is reduced by the self-selected out-migration of individuals with the least tolerance for immigrants. There are two reasons,however,to believe that internal migration does not mitigate entirely the consequences of immigrants'geographic concentration.First,internal migration is never so complete that it entirely separates the indigenous and immi- grant communities.Therefore,for better or worse,interaction between the two popu- lations is always higher in some areas than in others.Second,what appears to be a recent trend in internal out-migration may,in fact,intensify relations between the two communities by increasing the ratio of immigrants to the remaining indigenous population. The argument is straightforward.Because immigrant communities are spatially concentrated,the impact of the immigrant community is not evenly spread across the indigenous population.For better or worse,interaction between the host and immi- grant populations is higher where they are concentrated.Internal migration may di- minish the impact of spatial concentration but does not erase it.So we need to exam- ine the spatial distribution of both the benefits and costs of immigration in order to understand the politics of immigration control.It is this interaction that I attempt to model in the next section,followed by an analysis of how local preferences in sup- 19.See Martin and Midgley 1994 for the United States;Kopras 1993 for Australia;Cross 1993,124, for Britain;and King 1993 for France. 20.See White 1993a:King 1993;and Massey 1987,1989. 21.See Borjas 1994;and Frey 1994 for the United States;for Britain,see Cross 1993:and for Europe more broadly.see White 1993b

parliamentary constituencieswith approximately equal population, the smallest per￾centage of foreign born was 4.8 percent, and the largest was ten times that level, 49 percent. Thirty of the 147 constituencies had foreign-born populations of at least 30 percent; all but seven of these are located in the Melbourne and Sydney metropolitan areas. A similar concentration of immigrants is found in European states. In Britain, for example, 58 percent of the Afro-Caribbeans live in Greater London as do 80 percent of the black African immigrants. In France, 40 percent of the foreign-born population can be found in the Parisian basin, with the remainder of the immigrants concentrated in the industrialregionssurroundingLyon, Marseilles, and Strasbourg.19 This spatial concentration of migrants is attributed to several factors. Paul White enumerates the economic,social, political, and geographic forces that affect migrant destinations in the host countries, the most important of which are the initial eco￾nomic pull of labor markets and the subsequent reinforcement of migrant concentra￾tion through migrant networks or ‘‘chain migration.’’ 20 This brief overview indicates that the spatial concentration of immigrants in host countriesis well documented empirically and well understood theoretically.The po￾litical signiŽ cance of this concentration, however, has been widely overlooked. One factor that may diminish the signiŽ cance of the immigrant community’s geographic concentration—and may account for why this aspect of immigration has been ig￾nored in political analyses of immigration policy—is internal migration.21 Internal migration tendsto distribute the economic costs and beneŽ ts more evenly among the indigenous population. Furthermore, the social impact of immigration is reduced by the self-selected out-migration of individualswith the least tolerance for immigrants. There are two reasons, however, to believe that internal migration does not mitigate entirely the consequences of immigrants’ geographic concentration. First, internal migration is never so complete that it entirely separates the indigenous and immi￾grant communities.Therefore, for better or worse, interaction between the two popu￾lations is always higher in some areas than in others. Second, what appears to be a recent trend in internal out-migration may, in fact, intensify relations between the two communities by increasing the ratio of immigrants to the remaining indigenous population. The argument is straightforward. Because immigrant communities are spatially concentrated, the impact of the immigrant community is not evenly spread across the indigenous population. For better or worse, interaction between the host and immi￾grant populationsis higher where they are concentrated. Internal migration may di￾minish the impact of spatial concentration but does not erase it. So we need to exam￾ine the spatial distribution of both the beneŽ ts and costs of immigration in order to understand the politics of immigration control. It is this interaction that I attempt to model in the next section, followed by an analysis of how local preferences in sup- 19. See Martin and Midgley 1994 for the United States; Kopras 1993 for Australia; Cross 1993, 124, for Britain; and King 1993 for France. 20. See White 1993a; King 1993; and Massey 1987, 1989. 21. See Borjas 1994; and Frey 1994 for the United States; for Britain, see Cross 1993; and for Europe more broadly,see White 1993b. Political Geography of Immigration Control 691

692 International Organization port of and in opposition to immigration are filtered through specific national politi- cal institutions to affect political outcomes. I begin with a model of immigration policy as a function of domestic political interests.I assume,following the "regulation school"of politics,that politicians maximize the likelihood of(re)election by promising and producing policies that maximize political support while minimizing political opposition.2 Immigration con- trol can be modeled as a function of the support for and opposition to immigrant flows,arising from distinct societal sources.However,I focus on the local support for and opposition to immigration,the politicization of immigration in local politics, and its translation onto the national political agenda.I draw on the literature on immigration control but apply it at the local level. Support for Immigration On the demand side,firms are the primary actors with an interest in immigrant labor and an incentive to lobby government regarding immigration issues.I argue that local support for immigration varies over time in response to employers'labor mar- ket needs.Local support is strongest in periods of low unemployment;however, support will be mitigated by flexible labor markets and high capital mobility. Other factors being equal,low levels of local unemployment put pressure on local wage levels.Local labor market conditions,however,may be insufficiently attractive to generate internal migration,giving rise to geographically segmented labor markets and geographically specific corporate demands for immigrant labor.Regardless of conditions in the economy as a whole,local labor market conditions may intensify the demand for immigrant labor. Other factors are not always equal however.Firms with high capital mobility,such as manufacturing firms with standardized technology,can choose to export capital to sites of cheap foreign labor,whereas other firms have no alternative but to petition for the importation of labor.Natural resource-based firms,agricultural producers, and service firms (in-person delivery of services)are spatially fixed and therefore unable to reduce labor costs through capital exports.Firms in declining sectors of the economy also lack capital mobility.Given equal levels of unemployment,support for immigration will be strongest among firms with limited capital mobility.Therefore, local demand for immigrant labor will be strongest where firms with limited capital mobility dominate the local economy. Firms also confront variation in labor market flexibility.Numerous studies demon- strate that migrant labor provides desired flexibility by participating in the secondary labor market.23 That is,immigrants permit the expansion of dual labor markets and increase flexibility of production.Firms confronting inflexible labor markets will prefer higher levels of immigration than firms confronting flexible labor markets. Given equal levels of unemployment and similar levels of capital mobility,local 22.Peltzman 1976. 23.Piore1979

port of and in opposition to immigration are Ž ltered through speciŽ c national politi￾cal institutionsto affect political outcomes. I begin with a model of immigration policy as a function of domestic political interests. I assume, following the ‘‘regulation school’’ of politics, that politicians maximize the likelihood of (re)election by promising and producing policies that maximize politicalsupport while minimizing politicalopposition.22 Immigration con￾trol can be modeled as a function of the support for and opposition to immigrant  ows, arising from distinct societal sources. However, I focus on the local support for and opposition to immigration, the politicization of immigration in local politics, and its translation onto the national political agenda. I draw on the literature on immigration control but apply it at the local level. Support for Immigration On the demand side, Ž rms are the primary actors with an interest in immigrant labor and an incentive to lobby government regarding immigration issues. I argue that local support for immigration varies over time in response to employers’labor mar￾ket needs. Local support is strongest in periods of low unemployment; however, support will be mitigated by  exible labor markets and high capital mobility. Other factors being equal, low levels of local unemployment put pressure on local wage levels. Local labor market conditions,however, may be insufficiently attractive to generate internal migration, giving rise to geographically segmented labor markets and geographically speciŽ c corporate demands for immigrant labor. Regardless of conditions in the economy as a whole, local labor market conditions may intensify the demand for immigrant labor. Other factors are not always equal however. Firms with high capital mobility,such as manufacturing Ž rms with standardized technology, can choose to export capital to sites of cheap foreign labor, whereas other Ž rms have no alternative but to petition for the importation of labor. Natural resource–based Ž rms, agricultural producers, and service Ž rms (in-person delivery of services) are spatially Ž xed and therefore unable to reduce labor coststhrough capital exports. Firms in declining sectors of the economy also lack capital mobility.Given equal levels of unemployment,support for immigration will be strongest among Ž rms with limited capital mobility. Therefore, local demand for immigrant labor will be strongest where Ž rms with limited capital mobility dominate the local economy. Firms also confront variation in labor market  exibility.Numerousstudies demon￾strate that migrant labor provides desired  exibility by participating in the secondary labor market.23 That is, immigrants permit the expansion of dual labor markets and increase  exibility of production. Firms confronting in exible labor markets will prefer higher levels of immigration than Ž rms confronting  exible labor markets. Given equal levels of unemployment and similar levels of capital mobility, local 22. Peltzman 1976. 23. Piore 1979. 692 International Organization

Political Geography of Immigration Control 693 demand for immigrant labor will be strongest where employers experience inflexible labor markets. Opposition to Immigration I argue that local conditions trigger anti-immigrant sentiment through the level of native-immigrant competition.Although this thesis is not original,I clarify the con- ditions under which competition increases and diminishes,thereby isolating those factors that intensify opposition to immigration.This competition contains at least three dimensions.The first dimension,labor market competition,is triggered by economic recession.The second dimension,competition over state resources,is trig- gered by economic recession and the rate of growth of the immigrant community,as well as the level of immigrant access to publicly provided goods.The third dimen- sion,competition over societal identity,is triggered by the size of the immigrant community but is offset by assimilation of the immigrant community into the native population.The argument is conjunctural;that is,opposition to immigration that becomes politically important is triggered by the presence of an immigrant commu- nity in conjunction with economic recession.It is aggravated by the degree to which the migrant community challenges the preeminence of the native community. Competition over market-based resources.A considerable amount of controversy exists regarding the position of the migrant vis-a-vis the native workforce.24 Some economists argue that migrants complement the native workforce and actually en- hance the returns of the native population by increasing their productivity and,hence, their wages.Others argue that the immigrant labor force substitutes for the native workforce;in this view,immigrants obtain employment at the expense of the native population.For example,where indigenous labor is skilled and immigrant labor un- skilled,immigrants may be employed to increase the productivity of the native work- force,thereby increasing the wages of those workers.In contrast,unskilled immi- grant workers may compete with and displace native unskilled workers,thereby reducing the wages of those workers.Because the labor market is never completely segmented,undoubtedly there is some truth to both propositions.I focus on the variation in labor market segmentation during periods of recession and economic prosperity that modifies the level of competition between the native and immigrant workforce. Economic recession places the native and immigrant labor forces in more direct competition than in periods of economic prosperity.The dynamic proceeds because workers are often willing to take otherwise unacceptableemployment during periods of economic downturn.To be sure,this willingness is mitigated by the presence of a "reservation wage,"the remuneration available from nonwork sources,such as un- employment benefits and family allowances.To the extent that employment exists at greater than the reservation wage,unemployed workers accept employment that is 24.See Borjas 1994 for an overview of the debate as well as extensive citations of the literature

demand for immigrant labor will be strongest where employers experience in exible labor markets. Opposition to Immigration I argue that local conditions trigger anti-immigrant sentiment through the level of native–immigrant competition.Although this thesis is not original, I clarify the con￾ditions under which competition increases and diminishes, thereby isolating those factors that intensify opposition to immigration. This competition contains at least three dimensions. The Ž rst dimension, labor market competition, is triggered by economic recession. The second dimension, competition over state resources, is trig￾gered by economic recession and the rate of growth of the immigrant community, as well as the level of immigrant access to publicly provided goods. The third dimen￾sion, competition over societal identity, is triggered by the size of the immigrant community but is offset by assimilation of the immigrant community into the native population. The argument is conjunctural; that is, opposition to immigration that becomes politically important is triggered by the presence of an immigrant commu￾nity in conjunction with economic recession. It is aggravated by the degree to which the migrant community challengesthe preeminence of the native community. Competition over market-based resources. A considerable amount of controversy exists regarding the position of the migrant vis-a`-vis the native workforce.24 Some economists argue that migrants complement the native workforce and actually en￾hance the returns of the native populationby increasing their productivityand, hence, their wages. Others argue that the immigrant labor force substitutes for the native workforce; in this view, immigrants obtain employment at the expense of the native population. For example, where indigenouslabor is skilled and immigrant labor un￾skilled, immigrantsmay be employed to increase the productivity of the nativework￾force, thereby increasing the wages of those workers. In contrast, unskilled immi￾grant workers may compete with and displace native unskilled workers, thereby reducing the wages of those workers. Because the labor market is never completely segmented, undoubtedly there is some truth to both propositions. I focus on the variation in labor market segmentation during periods of recession and economic prosperity that modiŽ es the level of competition between the native and immigrant workforce. Economic recession places the native and immigrant labor forces in more direct competition than in periods of economic prosperity. The dynamic proceeds because workers are often willing to take otherwise unacceptable employment during periods of economic downturn. To be sure, this willingnessis mitigated by the presence of a ‘‘reservation wage,’’ the remuneration available from nonwork sources, such as un￾employment beneŽ ts and family allowances. To the extent that employment exists at greater than the reservation wage, unemployed workers accept employment that is 24. See Borjas 1994 for an overview of the debate as well as extensive citations of the literature. Political Geography of Immigration Control 693