Date post: | 11-Apr-2023 |
Category: |
Documents |
Upload: | independent |
View: | 0 times |
Download: | 0 times |
PIE
TE
R B
EV
EL
AN
DE
R &
DO
N J. D
EV
OR
ET
Z (E
DS
.) m
AL
mö
UN
IVE
RS
ITY
• m
Im 2
00
8T
hE
Ec
ON
Om
IcS
Of
cIT
IZE
NS
hIP
TheEconomics of CitizenshipEditors: Pieter Bevelander & Don J. DeVoretz
mALmö UNIVERSITY
SE-205 06 MalmöSweden
tel: +46 46 665 70 00www.mah.se
© Malmö University (MIM) and the authors
Cover: Diane Coulombe
Printed in Sweden by Holmbergs, Malmö 2008
ISBN 978-91-7104-079-4 / Online publication
www.bit.mah.se/MUEP
Malmö University
Malmö Institute for Studies of Migration, Diversity and Welfare (MIM)
SE-205 06 Malmö
Sweden
www.mah.se/mim
conTenTs
PREFACE ....................................................................... 7AboUT ThE AUThoRs .................................................... 9
INTRoDUCTIoN .......................................................... 13Irene Bloemraad
A Canadian Narrative 1. ThE ECoNoMIC DETERMINANTs AND CoNsEQUENCEs oF CANADIAN CITIZENshIP AsCENsIoN .................. 21 Don J. DeVoretz & Sergiy Pivnenko
A Dutch Narrative 2. NATURALIZATIoN AND soCIoECoNoMIC INTEGRATIoN: ThE CAsE oF ThE NEThERLANDs ........ 63 Pieter Bevelander & Justus Veenman
A Norwegian Narrative 3. ThE ECoNoMICs oF NoRWEGIAN CITIZENshIP ....... 89 John E. Hayfron
A swedish Narrative 4. ThE ECoNoMICs oF CITIZENshIP: Is ThERE A NATURALIZATIoN EFFECT? ...................................105 Kirk Scott
An American Narrative 5. IMMIGRANT NATURALIZATIoN AND ITs IMPACTs oN IMMIGRANT LAboUR MARKET PERFoRMANCE AND TREAsURY .....................................................127 Ather H. Akbari
6. ThE ECoNoMICs oF CITIZENshIP: A sYNThEsIs ......155 Pieter Bevelander & Don J. DeVoretz
REFERENCEs ..............................................................169
7
PREFACE
The genesis of this volume occurred at a workshop on immigrant citi-zenship acquisition held in 2004 in Malmo, Sweden under the framework of the Willy Brandt Professorship. This inter-disciplinary workshop ultima-tely reflected a deep rift between economists and other social scientists, which in turn led to this economically focused analysis. We are both grate-ful to the workshop participants and Malmo University for supporting this initial effort and shedding light on future research. As scholarly and public interest grew over citizenship issues in North America and Europe, both the Metropolis Project (Vancouver) and the Volkswagen Foundation (IZA) provided funds to support the research for this book. The overarching aim of the editors was to see if economic modeling could provide a general framework to explore the conditioners to immigrant citizenship ascension in both worlds. To the extent that we have succeeded in this task, a large measure of thanks goes to the individual country authors who sought out unique data sets, wrestled with understanding varying concepts of citizen-ship, and had the patience to employ and test the economic model offered in this book. Moreover we would like to thank all anonymous reviewers who read and commented on the initial manuscripts. We also offer many thanks to all authors for the country-specific narratives on citizenship issues preceding each chapter. Last but not least we thank Dr. Diane Coulombe for her careful copyediting and her help in keeping this project moving forward.
The Editors
9
ABOUT THE AUTHORS
Dr. Ather H. Akbari has been with the Economics Department at Saint Mary’s University since 1989. His research has been published in several economics and interdisciplinary journals and has also been quoted in top international and national newspapers and magazines such as The Wall Street Journal and The Economist magazine. Dr. Akbari also leads the eco-nomics domain of the Atlantic Metropolis, a Canadian centre for excel-lence in research on immigration. Dr. Akbari’s extensive past work on immigration includes studies on the public-finance impact of Canadian immigrants and work on their earnings functions and human capital stock.
Pieter Bevelander is an Associate Professor and the current Willy Brandt Research fellow at the Malmö Research Institute for Migration, Diversity and Welfare. He is also a senior lecturer at the School of International Mig-ration and Ethnic Relations, Malmö University, Sweden and a research fellow at IZA, the Institute for the Study of Labour, Bonn, Germany. His main field of research is international migration focussing on various aspects of immigrant integration. His latest research topics include a com-parison of the ethnic social capital in Canada and the Netherlands, social capital and electoral participation of immigrants and minorities in Canada, and the attitudes of the Swedish-born towards Muslims in Sweden. He has published in the International Migration Review, the Journal of Popula-tions Economics, the Journal of Ethnic and Migrations Studies and the Journal of International Migration and Integration.
Irene Bloemraad, Assistant Professor in Sociology at the University of Cali-fornia, Berkeley, studies immigration, political mobilization and citizen-ship, placing the U.S. experience in international context. Her recent book, Becoming a Citizen: Incorporating Immigrants and Refugees in the United States and Canada (University of California Press, 2006), argues that the United States’ lack of general integration policies has led to lower levels of citizenship among immigrants in the United States compared to Canada,
10
and poorer outcomes in political participation. Bloemraad has published articles on naturalization, dual citizenship, immigrant community organi-zations and ethnic leadership in academic journals such as Social Forces, the International Migration Review, Social Science Quarterly, and the Journal of Ethnic and Migration Studies. Her current projects, funded by the Russell Sage Foundation, examine the political and civic socialization of mixed status Mexican American families and the role of organizations in facilitating immigrants’ civic and political participation. She regularly talks about immigration to community groups and the media and belongs to IZA, a German institute devoted to studying immigration issues.
Dr. Don J. DeVoretz is a Professor of Economics at Simon Fraser Univer-sity where he was the Co-Director of RIIM, Vancouver’s Centre of Excel-lence on Immigration Studies from 1996 to 2007. Dr. DeVoretz has held visiting appointments at Duke University, University of Ibadan (Nigeria), University of the Philippines, University of Wisconsin, and the Norwegian School of Economics. He was the Willy Brandt Guest Professor at IMER, Malmo University in 2004. He is a Research Fellow with IZA (Germany), the Migration Research Group (Germany), and the Asia Pacific Founda-tion of Canada. Dr. DeVoretz was named a British Columbia Scholar to China in 2000. His current research interests include the economics of immigration with special emphasis on “brain circulation” and citizenship issues. His research findings have been reported in professional journals and major print and electronic media. John E. Hayfron received his PhD in economics from the University of Bergen in Norway in 1999. He is currently a Visiting Assistant Professor at Western Washington University. He has taught economics courses at various colleges and universities in Canada and the United States. His research is in the area of labour economics with special interest in the eco-nomics of immigration and the economics of gender and race. His publica-tions have appeared in the Journal of Population Economics, Applied Eco-nomics, and in edited books. Sergiy Pivnenko was born in Ukraine where he received degrees in physics and economics. In Canada he obtained a graduate degree in economics and worked as a research consultant. In 2005 Pivnenko completed work for the Government of British Columbia on the economic profiles of immigrants. His research contributions cover the labour market and the public-finance performance of immigrants and have appeared in refereed journals and working papers.
ABOUT THE AUTHORS
11
Kirk Scott in an Associate Professor at the Centre for Economic Demo-graphy and the Department of Economic History at Lund University. His research has dealt primarily with broadly defined issues of immigrant inte-gration. Topics such as immigrant income and employment outcomes were of initial interest, while more recently the focus of his research has shifted towards other measures of integration such as health, mortality, and ferti-lity. His recent publications have appeared in Population Studies, the Inter-national Migration Review, Demographic Research, the Journal of Socio-Economics, and several anthologies.
Justus Veenman is a Professor of Economic Sociology at the Economics Department, Erasmus University Rotterdam. He studied Economics in Rot-terdam and obtained his PhD from Utrecht University. His research focuses on social inequality, in particular on ethnic minorities in education and on the labour market. He has published several books and numerous articles (see www.uu.nl/uupublish/onderzoek/onderzoekcentra/ercomer/). From 1986 until 2005 he was the Director of the Institute for Sociological and Economic Research of Erasmus University Rotterdam.
ABOUT THE AUTHORS
13
INTRODUCTION
Irene Bloemraad
Debating Immigrant CitizenshipWhy would someone born and raised in one country decide to become a citizen of another? Does acquiring a new citizenship change one’s life? Does it matter?
More and more, countries of immigration imply that citizenship matters as an issue of national identity. Through much of the 1980s and early 1990s, Western countries extended rights and public services to non-citizen immigrants. The steady erosion of distinctions between citizens and non-citizens, coupled with seemingly greater tolerance for diversity and multiculturalism, led some to bemoan the devaluation of citizenship (Schuck 1998) while others heralded a new postnational age of member-ship that would make state-centered citizenship obsolete (Bauböck 1994; Soysal 1994).
Today, the celebration of citizenship and integration has replaced talk of multiculturalism from Australia to the Netherlands. Countries across continental Europe are introducing integration and language classes to spur immigrants’ incorporation. The United States changed the test required of would-be Americans to make it more “meaningful,” while Great Britain instituted its first-ever citizenship ceremony in 2004, presided over by Prince Charles, to underscore the significance of British citizenship. Impli-citly or explicitly, citizenship is portrayed as a marker of, or perhaps a way-station to, full social and political integration.
This book takes a different tact. It puts the focus on the economic aspects of citizenship, asking three questions: What are the economic deter-minants animating an immigrant’s choice to acquire citizenship? What are the economic consequences of choosing citizenship for the foreign-born worker? What are the economic consequences of immigrants’ citizenship for the country of reception? In asking these questions, the volume joins an on-going conversation of philosophers, sociologists and political scientists (not to mention politi-
14
cians, policy-makers and pundits). It takes the citizenship discussion in a new direction by bringing together, for the first time, an international group of economists who try to outline an economics of citizenship applicable to both traditional immigrant-receiving countries and the new countries of immigration in Europe. Do we need more voices in this already cacophonous debate over citi-zenship? The contributions in this book suggest we do. The volume asks questions in new ways and is undergirded by assumptions not frequently articulated in existing discussions. It has the singular advantage of asking similar questions across different host societies and national origin groups. Certain chapters rely on unique data sources that offer greater possibilities for modeling citizenship than before. The sum of these parts takes a big stride in bringing economists and economic approaches into the study of immigrant citizenship. While not everyone will agree with the economic approach adopted here, there is much to learn.
To Think Like an Economist: The Promise and Pitfalls of an Economic ApproachMany in the social sciences worry about the “imperialism” of economists, who have progressively used economic assumptions and thinking to tackle problems outside the traditional purview of economics.1 To be frank, I’m not immune to these worries. As a political sociologist with a long-standing interest in immigrant citizenship, I felt some trepidation when asked to write the introduction to a book entitled The Economics of Citizenship. Can citizenship be reduced to economic calculations?
Economic assumptions and models are seductive for their simplicity and clarity. They might consequently help us cut through the tangled terrain of existing citizenship studies. For example, despite countries’ growing focus on citizenship as a tool or badge of national identity, a number of the vignettes that begin each chapter distinguish between self-identity and legal status. In the words of one migrant from the Middle East, currently a Swedish citizen, “I have never felt ‘Swedish,’ and I do not believe that one can become a Swede through naturalization. The most one can expect is to become a citizen of Swedish society.” Such sentiments hint that a more instrumental approach to citizenship may not be amiss. What do we learn from taking an economic approach to immigrant citizenship? And what do we miss?
1 See, for example, non-economists’ reaction to the work of Gary Becker or to the success of Freakonomics (2005) by Steven Levitt and Stephen Dubner.
IRENE BLOEMRAAD
15
Economic Assumptions
The contributions in this volume offer, in line with mainstream economics, an elegant rational actor model of citizenship. Immigrants make a rational choice over whether or not they will acquire citizenship by weighing the benefits of citizenship acquisition against the costs of naturalization. When benefits outweigh costs, immigrants take up citizenship. When the ledger weighs more heavily on the negative, immigrants opt to remain non-citi-zens.
This approach has many benefits. It is simple. It can be readily applied in different contexts and across immigrant groups, one of the key contribu-tions of this volume. Thus, we learn that while Canada and Sweden allow dual citizenship (since 1977 and 2001, respectively), the Netherlands and Norway do not. Surely one of the biggest costs of naturalization is losing one’s prior citizenship. We would imagine, ceteris paribus, that immigrants would be more likely to take up citizenship in places where they are not required to renounce a prior nationality.
Yet the evidence presented across the country cases does not support such a simple conclusion. Citizenship levels and rates are very high in Canada, a country with a longstanding acceptance of dual citizenship, but they are also remarkably high in the Netherlands, which only allows dual citizenship in very particular cases. Further, the Swedish data, analyzed by Kirk Scott, finds that the 2001 law allowing multiple citizenships had no appreciable effect on citizenship trends in that country.
The benefits of citizenship also vary across countries. As Ather Akbari outlines in the chapter on the United States, since 1996 the advantages of U.S. citizenship have become more pronounced. Non-citizens now face more restricted access to public benefits and enjoy less recourse against deportation in the event they are convicted of minor crimes. The benefits of citizenship and costs of non-citizenship appear less stark in Canada and the Netherlands. In the Netherlands, a country with a strong social security net, access to benefits is not restricted by citizenship, a situation similar to that in Canada where only pension benefits for those living outside the country involve a citizenship component. Yet American levels of naturali-zation are below those of the other two countries, and appear to remain lower even when we try to adjust for the large undocumented population in the United States, as Akbari does.
Maximizing What? In Comparison to What?
This quick, somewhat unsuccessful, application of a cost/benefit approach suggests that the simple model needs to be unpacked. One key question arises: What, exactly, are immigrants trying maximize when they weigh the
INTRODUCTION
16
costs and benefits of citizenship? Most contributors to this volume would agree that immigrants seek something we might call “quality of life”, an assumption that this sociologist would find unproblematic. Of course, the devil is in the details. What constitutes “quality of life” and how might this vary for different immigrants in different countries of immigration?
One of the most consistent, and consequently important, findings across the five substantive cases in this book is the identification of diverse patterns of citizenship acquisition among refugees, migrants who come through family reunification, economic migrants from less developed countries (whether low, semi or high-skilled) and high skilled migrants from other OECD countries. Almost every chapter reports that refugees (or, more specifically, people from countries that have produced substantial refugee flows) rush to acquire citizenship quickly and at high levels. In contrast, highly skilled immigrants from North America or Europe who move to another highly developed country drag their feet, only naturalizing after long periods of residence. In between, we find those from the develo-ping world or Eastern Europe who arrive through family reunification pro-visions or as economic migrants.
What can we make of these patterns, quite stable across all five countries? First, they demand that we add a reference point to any discus-sion of ‘quality of life’ and consider the distance separating immigrants’ current situation and their reference point, likely the country they left. If a German leaves Europe for the United States, the difference in living condi-tions, and the attractiveness of return, will be quite different compared to a Burmese refugee fleeing political persecution and economic misery in her homeland. All the chapters in this volume suggest that the citizenship cal-culations of migrants from highly developed countries differ from others. This finding has important policy implications since immigrant-receiving countries appear to be increasingly in competition with each other in trying to attract high-skilled migration. If host countries want these high skilled immigrants to become citizens, then they should target elite migrants from the developing world, something that might contribute to the brain drain from South to North. If they instead entice skilled workers from other highly developed countries, they run the risk of more limited civic or politi-cal incorporation, given this group’s reluctance to change passports.
A second challenge comes in defining and measuring “quality of life,” assuming that immigrants do indeed base citizenship decisions on how long-term improvements in life conditions off-set the short term bureaucra-tic hassle and financial cost of citizenship (not to mention feelings of dis-loyalty to the home community). Many of the naturalization vignettes embrace a capacious view of quality of life. The Canadian citizen of
IRENE BLOEMRAAD
17
Chinese birth speaks warmly of the Canadian lifestyle and norms, inclu-ding respect, equality, freedom, the rule of law, a cleaner environment and better childhood education. The Swedish citizen who differentiates Swedish ethnicity from citizenship appreciates the values of democracy and fairness that inform the Swedish social institutions. The Vietnamese refugee who has become a U.S. citizen provides a poignant story giving substance to the statistical finding that refugees, in particular, take up citizenship quickly. Few are likely surprised to learn that, given persecution in Vietnam, the security and stability of legal membership in the United States become attractive.
Unfortunately, the analyses in this volume are limited by available data in their conceptualization and modeling of costs and benefits as well as quality of life. As a few authors point out, many statistical datasets do not include enough immigrants for complete models of citizenship decision-making. In the North American case, researchers use the U.S. and Cana-dian censuses, which offer excellent coverage of immigrant populations, but a relatively limited selection of demographic and socio-economic vari-ables. As a result, quality of life, or economic considerations more specifi-cally, get reduced to labour market participation and income. Immigrant citizens are compared to non-citizens to see whether the former are more likely to be employed or have higher income than the latter. The vignettes suggest that, from the perspective of the immigrant, employment and income are minor considerations. This might help to explain why naturali-zed citizens in Sweden or Norway do not appear to enjoy an employment benefit.
Moving Beyond the Atomized Rational Actor
Perhaps part of the problem lies in seeing the citizenship decision as one of individual choice rather than as one nested in social networks involving family, friends and others in the community (Bloemraad 2006). As Don DeVoretz and Sergiy Pivnenko note in passing, some Chinese migrants in Canada might view citizenship as a family strategy, where one spouse keeps Chinese citizenship while the other takes on Canadian citizenship, ensuring that the family has two options. Conversely, the vignettes of immigrant citizens in Sweden and Norway both mention relationships to nationals of those countries as part of the citizenship process. These stories suggest a pattern of citizenship homophily rather than citizenship splitting. In either case, it raises the question of how families and households, rather than just individuals, make decisions about citizenship.2
2 This argument can, of course, be extended to children. If children born in the country of residence acquire that county’s citizenship at birth, parents might use the children’s
INTRODUCTION
18
The Mechanism of the Citizenship Benefit
While the findings on the economic consequences of citizenship reported here are mixed likely beneficial to the society as a whole, possibly but not invariably helpful to the individual migrant the analyses do find a positive association between citizenship and economic outcomes in some cases. Why might this be so? Correlation analyses cannot tell us exactly what mechanisms link citizenship and economic outcomes, but the authors provide some helpful suggestions for thinking through this problem.
The most straightforward answer, laid out in all chapters, centers on restricted access to certain jobs. In every country, specific jobs (be they police work, as in the Netherlands, top banking positions, as in Norway, or high-level civil service or military positions in all countries) are only open to citizens. It is not clear, however, whether access to such jobs drives a significant proportion of citizenship decisions. Ideally, we would need panel data over a relatively long time period, non-existent in many cases, to see whether we can identify occupational change before and after natu-ralization.
Other explanations for why citizenship might improve economic out-comes are more subtle. Various authors point out that citizenship may fun-ction as a signaling mechanism, perhaps convincing a native-born employer to take on an immigrant worker. DeVoretz and Pivnenko suggest that citi-zenship might communicate greater attachment to Canada, and perhaps a longer commitment to a particular company or job, and Akbari hypothesi-zes that citizenship could signal to employers “greater knowledge of local customs and traditions that is essential for a firm’s success.” If this is true, it is especially noteworthy that those from highly developed countries, likely to be light-skinned, are less likely to naturalize compared to those from the developing world, individuals probably more likely to have darker colored skin. If signaling is part of the benefit of citizenship, does this mean that light-skinned immigrants face less discrimination and do not need signals as much as darker-skinned migrants who face a greater employment or promotion hurdle? The signaling hypothesis also begs the question of whether employers even know or ask about citizenship status in making hiring and salary decisions.
citizenship as “insurance” and not naturalize themselves, or they may decide to become a citizen in order to share the same nationality as their children. In this light, Akbari’s finding that having children under 15 in the household decreases the odds of citizenship would appear to support the first hypothesis rather than the latter. It is possible, however, that the finding is a statistical artifact, since the comparison groups become both families without any children and families with children over 15 who might or might not live in the household.
IRENE BLOEMRAAD
19
Citizenship might have an economic influence in a different way. Pos-sessing citizenship might increase the immigrant’s sense of self-efficacy or change time horizons in such a way that he or she demands higher wages or goes after better jobs. We can imagine, for example, a situation where an immigrant assumes that he will return home after a number of years making money in a particular country. However, with time, the immigrant becomes integrated, builds social networks and considers staying longer. Eventually the person decides to become a citizen and, once the decision is made, starts searching for a new job with greater chances for mobility, since the desire to return to the country of origin has been put aside or delayed until retirement.
Self-Selection Biases and Omitted Variables
This last scenario (of the immigrant who finally decides to stay, and becomes a citizen in consequence) gets at the heart of a problem confronting almost all statistical analyses of citizenship. Does citizenship, on its own, drive eco-nomic or other outcomes, or is the independent effect of citizenship insignifi-cant once we appropriately model the differences in characteristics, motiva-tions and situations of those who do and do not naturalize? Put another way, are there unobserved variables that both increase the probability of naturali-zation and affect economic outcomes measures? Almost all the authors note the difficulty of untangling causality. The longitudinal panel data analyzed in the Norwegian chapter by John Hayfron and in the Swedish chapter by Kirk Scott consequently deserve special note as an important advance in our ability to track immigrants before and after naturalization.
Policy Implications and Future ResearchReaders will want to delve into the individual chapters, and read across them, for the useful overviews of citizenship laws and benefits as well as the specific findings for each country. A few policy implications deserve mention in closing.
First, as Don DeVoretz and Pieter Bevelander outline in the conclusion, there does not seem to be any simple relationship between the openness or difficulty of national citizenship regulations and aggregate naturalization patterns. This suggests that immigrants’ citizenship decisions go well beyond a simple cost/benefit analysis based solely on laws. This finding includes the case of dual citizenship, a topic of public debate in countries such as Canada, the Netherlands and Sweden. Perhaps the citizenship laws of the country of reception matter much less than the laws in the country of origin. More generally, economic, social, cultural or political contexts must affect citizenship decisions more than any one law.
INTRODUCTION
20
Second, current interest in high-skilled, elite immigrants in countries such as Germany, as well as the long-standing immigration policy favoring skilled economic migrants in Canada and Australia, might have to be re-thought if countries truly wish to maximize national identity over human capital gains. A number of chapters suggest that foreign-born citizens provide greater fiscal gains to the national treasury than non-citizens, but high-skilled immigrants from developed countries are reluctant to take on a new passport. DeVoretz and Pivnenko suggest that high-skilled immi-grants, predominantly from non-European countries, might acquire Cana-dian citizenship only as a human capital enhancement that allows them to move on to a third country to pursue further economic success. Thus, mig-rants to Canada might quickly acquire Canadian citizenship in order to more easily move to the United States, while immigrants may see Sweden as a way-station to mobility within the European Union. Either scenario does little to enhance national identity or civic cohesion in the short or medium term.
In line with this, the analysis of Dutch citizenship by Pieter Bevelander and Justus Veenman is particular noteworthy as they are the only scholars who are able to do a direct evaluation of Dutch integration classes. Such classes are part of the turn toward national identity across Europe, yet the authors find no effect on citizenship among the refugees that they study. These results raise the question of whether integration programs really help immigrant incorporation, or whether they are largely publicity stunts to mollify a native-born public worried about immigration.
Whatever one’s thoughts about the goals of citizenship, be it for immi-grants or the host society, this initial foray into an economics of citizenship will provide much food for thought.
IRENE BLOEMRAAD
21
A Canadian Narrative
I first came to Canada in 1999 as a visiting scholar under the Canada-China Scholars’ Exchange Program. This visit marked the beginning of a new chapter in my life. During my visit, I decided to apply for immigration to Canada. In the following year, my application was approved, and my family, including my spouse and my then 7 year old son, landed in Canada. After three years of continuous residence in Canada, we all became eligible for Canadian citizenship. On January 13th, 2005, just two days before my family moved into a newly purchased home in Vancouver, we all took our oaths and became Canadian citizens. Many immigration stories can be explained by the push-pull theory. My story lies more on the pull side. After 16 years of schooling in China and 2 years in Europe, I was endowed with rich human capital and sought an outlet for it. I considered Europe when I was completing my post- graduate education. I also thought about Australia after returning to Shanghai from Europe. Finally, Canada came into my sight. The choice of Canada as a destination was optimal for my family and I when we were thinking of where to move to. My education and profes-sional background gave me enough points to pass the so-called point system. The visiting scholarship provided me with experience in the Cana-dian workplace. Most importantly, the contacts and networks I was able to establish during my initial visit helped me to acquire a job immediately after I landed in Canada. Needless to say, Canada’s cleaner environment, better childhood education, and more relaxed and convenient lifestyle all entered into the equation in a positive way. Canadian citizenship is regarded as something of convenience by some. To me, it is the outcome of a three-year term experiencing real life in Canada. It is about the lifestyle that my family and I enjoy. It is about values, namely: respect, equality, freedom and rule of law, which my family and I appreciate. Some people suggest that citizenship is a sense of belong-ing. Yes, I believe that I belong in both Canada and China, although China does not recognize dual citizenship. Even though I fully committed to a new homeland, it is morally unacceptable to disregard a place where my Mom gave birth to me. Now, like many other Canadians, I pay taxes, buy retirement funds, pay my mortgage, and sometimes comment on Canadian policies and social problems. This is the life I have chosen.
23
Chapter 1
the eCONOMIC DeterMINaNtS aND CONSeQUeNCeS OF CaNaDIaN CItIZeNShIp aSCeNSION
Don J. DeVoretz & Sergiy Pivnenko
Introduction: Issues and Stylized FactsCanada’s original 1947 citizenship act was born out of the sacrifices endured in World War two and was an explicit statement of independence from its heretofore citizenship ties to great Britain. However, by the 1970s Canada was witnessing a transformation in its population with annual immigration flows of over 250,000 or one per cent of its population base. Moreover, this immigrant population was no longer european-based and immigrant integration in a multi-cultural framework became a paramount policy objective. to this end, one of the long-term goals of Canadian immi-gration policy was to insure that the majority of its foreign-born arrivals became citizens. the Canadian ministry of immigration was often charged to perform both immigrant selection and citizenship functions.1
Under the 1977 Citizenship Act, the ease of ascension under the mixed principles of jus sanguinis and jus soli can be seen by the relatively modest requirements for naturalization (see Appendix A). in sum, the basic requi-rements to attain citizenship in Canada include: 3 years in residency, being 18 years of age or older, having a knowledge of one official language and an adequate knowledge of citizenship responsibilities.2 in addition, dual citizenship has been permitted since 1977; this resulted in a substantial number of naturalized Canadians living outside of Canada conferring citi-zenship on subsequent foreign-born generations.3
1 the title of the ministry is Citizenship and immigration Canada. in the past, the immigration ministry was merged with the Ministry of Justice, and, prior to that, with the Ministry of Manpower and immigration. each re-organization of the immigration ministry reflected the perspective of successive governments on issues surrounding immigration.2 See Appendix A for details. 3 the entire Canadian diaspora is estimated to be around 3 million or 10% of Canada´s population (zhang 2006).
Don J. DeVoretz & Sergiy PiVnenko
24
Citizenship Issues
this cursory description of the ascension process belies the degree of con-troversy that has arisen in Canada with respect to citizenship acquisition. For example, some immigrant groups have recently argued that the three-year waiting period is too long and presents a legal employment barrier. the plaintiffs in this legal case argued that both job and earnings discrimi-nation arose under this requirement, since immigrants without citizenship were unable to practice their profession and enjoy the relatively high ear-nings derived from a federal position. in 2003 the Canadian Supreme Court upheld the citizenship requirement for an array of federal government jobs, and ruled against an immigrant class action suit to recover damages from this alleged discrimination.4
Another issue has arisen as a byproduct of linking dual citizenship pro-visions with the growth in return migration of erstwhile Canadian immi-grants. it has been observed that over 25% of the post-1986 Chinese immi-grants to Canada had returned to Hong kong or China by 2004, many with Canadian citizenship (DeVoretz & Ma 2002). Canadian policymakers have made ambivalent pronouncements over the economic impact of this phenomenon. Some policymakers consider the returning erstwhile Cana-dian immigrants a Canadian asset which will increase trade and invest-ment. other observers are less sanguine and feel that these Chinese-Cana-dian emigrants represent a potential future liability, especially if they return to retire, thus putting economic pressure on the social system.5 in addition, the dramatic rescue of 12,000 naturalized Canadian citizens in Lebanon in 2006 highlighted the issue of the potential political and economic costs inherent in a dual citizenship policy predicated on the notion of a capstone integration tool. these two diaspora related policy issues have put into question the efficacy of dual citizenship, the relative ease of citizenship ascension and the portability of the rights and obligations of Canadian citi-zens living outside of Canada.
Stylized Facts
one long-term goal of Canadian immigration policy is to insure that the majority of its foreign-born arrivals become citizens. to this end, the current Canadian Ministry of Citizenship and immigration performs both
4 the Court argued in the majority that, since there was no barrier to becoming a Cana-dian citizen, then inherently immigrants did not face discrimination, but just a waiting period which applied to all immigrants.5 of course, there are many non-economic objections to returning immigrants, including an alleged lack of patriotism or failure to integrate into the Canadian economy.
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
25
immigrant and citizenship selection functions.6 the process of citizenship acquisition is straightforward. in fact, the majority of foreign-born perma-nent immigrants to Canada are entitled to apply for citizenship after a three-year period of residency; thus, the 2001 Census of Canada reports that 74.6% of Canada’s foreign-born residents are citizens. in 2001, 57% of immigrants who had been residents for four to five years had become Canadian citizens. Moreover, among those who had lived in Canada for 6 to 10 years, 79% were citizens and among those who had been in the country for 30 years or more, 90% were citizens.
Census data also show that recent groups of immigrants are taking less time to become Canadian citizens than their previous counterparts. the 1991 Census showed that just over half (51%) of immigrants who had been residents for four to five years had become citizens. the proportion in 1981 was only 42%.
nonetheless, variations in acquisition of citizenship appear. Figure 1 portrays the important observation that there are differential rates of citi-zenship ascension by immigrants for each year in residence in Canada and by their country of origin.
Figure 1. Cumulative percentage of naturalizations among permanent immigrants from high income countries (UK, USa, Germany, Italy, Netherlands) and low income countries (China and India)
in fact, Figure 1 reports that over one-third of the newer immigrant flows from China and india ascend to citizenship after 15 years in residence.7 However, by their 35th year in residence 60% of the Chinese and indian stock of residents have acquired citizenship. A dramatically different
6 Until 2002 the Ministry also was in charge of border security and had the power of deportation at its disposal for non-citizens.7 the Census of Canada does not provide any information on the year of citizenship acquisition.
Source: Authors’ tabulations from 2001 Census of Canada, Statstics Canada PUMF.
cum
ulat
ive
%
Don J. DeVoretz & Sergiy PiVnenko
26
picture emerges for Canada’s traditional immigrant source countries of Western europe and the United States. Here only 10% of the immigrant stock naturalize in their first 10 years in Canada and one third after 35 years in residence.
As shown in Figure 2, the level of development of the immigrant’s source country affects citizenship ascension rates. in fact after 10 years in residence, about 81% of immigrants from non-oeCD countries become Canadian citizens. After the 25th year in residence, the process ends as the remaining stock of residents from developing countries (non-oeCD) have largely acquired citizenship.8 Similar but less dramatic picture emerges for immigrants from the oeCD – Canada’s traditional source countries in Western europe and the United States. Here only about 61% of those who have been in Canada for 6-10 years ascend to citizenship. this is about 10% higher than a similar observation made in 1996 Census. Figure 2 shows a nearly 20% gap in ascension rates between oeCD and non-oeCD immigrants which persists until 25-30 years in residence.
in sum, even given Canada’s relatively high rate of citizenship ascen-sion substantial variations in acquisition rates occur across time and place which we will attempt to explain below.
Figure 2. ascension to Canadian citizenship by immigrants from OeCD and non-OeCD countries
Literature reviewthe economic literature on citizenship primarily consists of two separate views. one view attempts to rationalize an immigrant’s decision to acquire citizenship, and the other investigates the economic consequences of such a decision. the evidence on the determinants of acquiring citizenship remains
8 An unknown number of the original entry cohort could have disappeared after 25 years and this would produce an upward bias in the rate of citizenship acquisition.
Arrival cohorts
Perc
enta
ge
of ci
tize
ns
in a
rriv
al c
oh
ort
s
Source: Authors’ tabulations from 1996 and 2001 Census of Canada 2,7% PUMF.
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
27
highly controversial, largely due to the specifics of the populations studied and the varying nature of the data used. While some authors (kelley and McAllister 1982; Portes & Mozo 1985) insist on the importance of socio-economic variables such as education, occupation and income, others (Bernard 1936; Barkan & khokhlov 1980; Portes & Curtis 1987) put forward cultural assimilation and demographic arguments as the major determinants of an immigrant’s naturalization decision. With the aid of 1980 U.S. Census micro data, yang (1994) was the first to apply a cost-benefit framework to investigate the effects of individual characteristics and the socio-economic conditions of the immigrant’s home and host country on the immigrant’s citizenship decision. yang’s findings indicate that cultural integration plays a more important role than economic inte-gration in the U.S. immigrant’s naturalization decision. Age at immigra-tion, marital status and the presence of children were among the demo-graphic factors that increased the odds of an immigrant becoming a citizen. While the home country level of development proved to be a significant predictor of immigrant’s naturalization decision, the availability of dual citizenship did not obtain the expected effect.
Another stream of studies ignored the economic rationale for beco-ming a citizen and addressed only the possible economic impacts derived from the immigrant ascending to citizenship. While Bratsberg et al. (2002) ignore the economic rationale for becoming a citizen, they do address the possible economic impact of immigrant citizenship in the United States labour market. Using a youth panel study, they find that immigrant ascen-sion to citizenship alters the immigrants’ occupational distribution and raises their earnings. Moreover, they argue that these effects are greater for immigrants from less developed countries.
other economic studies of citizenship are even more limited in scope since they mostly incorporate the citizenship effect in an ad hoc manner or as addendum to a larger study. Pivnenko and DeVoretz (2003) found a strong citizenship effect on Ukrainian immigrant earnings in Canada. Mata (1999) reports no evidence on the economic impact of Canadian citizen-ship on immigrant earnings after conducting a principal components ana-lysis with 1996 Canadian data. in reviewing the economic outcomes of Chinese-Canadian citizens who returned to Hong kong, DeVoretz and zhang (2004) found that citizens earned higher incomes than any other resident group in Hong kong. For his part, Bevelander (2000) reports that the log-odds of obtaining employment improved for those immigrants to Sweden who obtained citizenship in 1990.
We conclude from this brief literature survey that studies of citizenship ascension and its economic impact are fragmented and limited in scope. in
Don J. DeVoretz & Sergiy PiVnenko
28
addition, this literature review suggests that economic modeling appears particularly difficult because of demanding data requirements and a need to model strong institutional components affecting the labour market out-comes.
DataBoth the legal process to obtain Canadian citizenship and our model design dictate data selection and variable definitions. in this study, we select a population of immigrants from the 2001 Census of Canada Public Use Microdata Files (PUMF). in order to ensure that potentially-naturalized citizens have met the time requirements to apply for Canadian citizenship, we restrict our sample to immigrants who landed prior to 1998.9 Since our model focuses on the wage effect that may arise from citizenship acquisi-tion by employed foreign-born workers, we would ideally like to have wage rate data. However, since our data source does not provide informa-tion on hourly wage rates, our regression analysis must be performed using the individual’s annual wage or salary earned in 2000, controlled by weeks worked in that year.10 Also, any individual records reporting inconsistent observations on wage, salary income, or weeks worked are excluded from our data set.11 Moreover, since we focus our analysis on employed, wor-king-age immigrants, individuals over 64 and under 25 years are excluded from our data set.
our model also dictates that the majority of the data used as explana-tory variables are recoded as zero-one dummy variables. Marital status (MArrieD) is coded as a 1- for legally married, or 0- otherwise. next, we recoded the educational variable ‘highest degree earned’ into a string of dummies indicating the acquisition of either a trades certificate or diploma (DiPL), bachelor degree (BACH), above bachelor (BACHPL, i.e. masters or medical degrees) or an earned doctorate (PHD). Similarly, we transfor-med the occupational variable of the immigrant into a series of dummy variables representing professional (ProF), managerial (MAng) or a skilled (SkL) classification.
9 Landed immigrants must live in Canada for at least three of the four years preceding their citizenship application. 10 given the annual nature of our data, we must ignore the possibility that our wage data may contain earnings derived from overtime premiums, or that foreign-born resi-dents may work at more than one job.11 that is, positive earnings for zero weeks or zero earnings for positive weeks. in addi-tion, by restricting wage earnings to a minimum of $1,000 we further cleaned our sample of those who reported hourly or weekly wages instead of annual earnings.
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
29
table 1. Descriptive statistics for selected immigrant groups by country of origin and citizenship status
Non-OECD OECDNon- Citizen
Citizen Non- Citizen
Citizen
n=3,271 n=21,867 n=4,196 n=16,291Age Composition 25-34 33.5% 23.4% 19.2% 11.9% 35-44 36.4% 34.1% 31.0% 24.4% 45-54 21.5% 29.8% 31.2% 37.8% 55-64 8.6% 12.7% 18.5% 25.8%Composition of Arrival Cohorts (years Since immigration) 0-10 24.8% 75.2% 43.0% 57.0% 11-20 9.7% 90.3% 26.8% 73.2% 21-30 5.4% 94.6% 25.2% 74.8% 31-40 3.8% 96.2% 16.2% 83.8% 41+ 2.4% 97.6% 7.6% 92.4%Composition of Citizenship groupsMarital Status Legally Married 74.4% 73.2% 73.8% 74.9% other 25.6% 26.8% 26.2% 25.1%educational Attainment High school or less 51.8% 40.1% 43.4% 42.5% Diploma or certificate above high school
26.6% 32.4% 35.6% 36.5%
University: bachelor degree
13.7% 18.3% 11.0% 11.1%
University: above bache- lor, master’s, or medical
6.9% 8.0% 7.7% 8.1%
University: earned doctorate
1.0% 1.3% 2.2% 1.8%
occupational group Managerial 5.0% 8.1% 12.3% 12.6% Professional 16.1% 24.9% 24.5% 25.2% Skilled 18.5% 19.1% 24.7% 24.5% Low skill 60.4% 47.8% 38.4% 37.7%Weeks Worked 0-25 12.0% 8.4% 7.5% 6.9% 26-40 13.4% 10.3% 9.9% 9.4% 41-52 74.6% 81.2% 82.6% 83.7%Full/Part time Weeks Mainly part-time 12.4% 11.1% 15.8% 13.0% Mainly full-time 87.6% 88.9% 84.2% 87.0%Mean Wage earnings All $25,441 $33,788 $38,762 $41,665 4-5 yrs in Canada $24,969 $28,693 $37,532 $39,799 16-20 yrs in Canada $28,494 $33,400 $37,802 $42,287
Source: Authors’ calculations.
Don J. DeVoretz & Sergiy PiVnenko
30
Based on the above definitions, table 1 provides some stylized facts by immigrant source country and citizenship status.12 Since our earlier work indicated that citizenship acquisition might be a by-product of the level of development of the foreign-born Canadians’ country of origin, we further divided our data into immigrants from oeCD and non-oeCD countries.
given that we will employ a human capital model of earnings, we focus our analysis on key socio-economic variables and observe, except for marital status, substantial differences across immigrant groups, as defined by citizenship and source country. For example, in the 25-44 age group about 57.5% of non-oeCD immigrants and 36.3% of oeCD immigrants reported Canadian citizenship. Also, naturalized citizens from non-oeCD countries are better educated than non-citizens (27.6% vs. 21.6% with a post-secondary degree) and fewer of them are in unskilled occupations (47.8% vs. 60.4% for non-citizens). in contrast, oeCD immigrants do not show variations in educational attainment or skill level across citizenship status. Finally, citizens from of non-oeCD origins work more weeks than non-citizens, while slightly more citizens of oeCD origins are employed full-time than non-citizens.
this combination of full-time employment and higher skill levels of foreign-born citizens contributes to greater annual wage earnings for citi-zens than non-citizens. Among non-oeCD origins, the average citizen-non-citizen annual earnings differential in 2000 was about $8,300, while a similarly defined differential for oeCD immigrants was about $3,000.
this brief overview indicates that citizenship status is correlated with larger human capital endowments and a more robust earnings performance for naturalized Canadian immigrants.
ModelCitizenship ascension
in the first stage or the decision of the immigrant to ascend into citizenship, we use a well-known economic citizenship model owing to DeVoretz and Pivnenko (2006). the basic argument is as follows: the economic problem that immigrants face is to choose a state -- citizenship or non-citizenship -- which maximizes their income net of citizenship ascension cost given their human capital stock. DeVoretz and Ma (2002) imbedded the citizenship
12 Since the available micro data file represented a 2.7% censored sample of the 2001 Census of Canada, we were able to identify a limited number of source countries. the identified oeCD countries include France, germany, greece, italy, the netherlands, Por-tugal, Spain, the United kingdom, other Western europe, northern europe and the United States. our sampled non-oeCD origins include Arab-African origins, Latin America, China (PrC), india, Latin America, Lebanon, the Philippines, Poland, the USSr (european part), Vietnam, and yugoslavia.
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
31
decision inside a more general model of moving and staying. each stage of this journey involves a decision to move or stay, and this decision is, in turn, conditioned by possible ascension to citizenship.
Figure 3. Immigration triangle
to highlight the citizenship decision, we focus on the movement path between the Sender country A and the entrepot country (B or Canada). initially, the immigrant resides in country A and decides to move to country B. this movement is asserted to be motivated by the prospect of higher earnings and the opportunity to acquire subsidized human capital and two public goods including a passport and citizenship.
Both the acquisition of subsidized human capital and the prospect of receiving public goods (citizenship and a passport) now increase the proba-bility that this immigrant will ascend to citizenship, if the expected ear-nings stream in country B net of costs of citizenship acquisition exceeds the option of returning home. However, if country A does not recognize dual citizenship, this will raise the cost of possible return migration and reduce the probability of ascending to citizenship in country B.13 But will the newly ascended citizen of country B stay in country B? only if the net income
13 given the lack of dual citizenship, one apparent strategy for Chinese immigrants is for one of the two spouses to ascend to Canadian citizenship, while the other spouse remains Chinese. this ensures access to China for the spouse who is not a Canadian citizen.
eNtrepOtCountry B
SeNDerCountry A
rOWCountry C
Human Capital triangle
Don J. DeVoretz & Sergiy PiVnenko
32
gains from staying as a citizen in country B exceed the income gains from a citizen of country B moving to country C in the rest of the world (roW) will the immigrant remain in country B. if the cost of movement to country C is lowered via the attainment of citizenship and a passport in country B and the income in country C exceeds that earned in country B, the natura-lized citizen of country B will move and become a citizen of country C. in sum, the optimization problem for the immigrant is to choose a mobility path and citizenship status in one or more of the destinations which maxi-mizes net income given his/her human capital endowment, and the transac-tion costs of movement and obtaining citizenship.
the array of costs implied by citizenship acquisition differs by country of origin and the individual immigrant. First, in the absence of mutual recognition of dual citizenship by both Canada and the sending country, the major cost of ascending to Canadian citizenship is the loss of home country citizenship. this loss in turn implies limited access to the home country’s labour market and loss of public services including social insu-rance and subsidized education. in addition, application fees and any fore-gone income arising from continued residence in Canada to fulfill citizen-ship residency requirements add to the costs of ascending to Canadian citizenship.
on the other hand, the benefits derived from Canadian citizenship include possible greater job access and any wage premium paid by private and public Canadian employers to citizens, potential access to the United States labour market with a nAFtA visa, and possible visa wavers by third countries given a Canadian passport.
given these observations and our economic model, we argue that rates of ascension to citizenship will be a positive function of the immigrant’s various demographic and socio-economic characteristics that affect mone-tary or non-monetary benefits associated with Canadian citizenship. For example, the immigrant’s age determines his/her remaining years in the Canadian labour force and thus the size of any lifetime premium associated with citizenship. Consequently the older the immigrant is at immigration, the less likely he/she will become a citizen given the shorter payoff period. Similarly, with a greater level of educational attainment, an immigrant will potentially reap more income from citizenship and hence may have a greater propensity to acquire citizenship.
Successful economic integration as indicated by the immigrant’s income, occupation and home ownership status increases the benefits of remaining in Canada and adds to the cost of moving back to the immigrant’s country of origin. the degree of cultural assimilation, which could be proxied by years since immigration, is expected to demonstrate a strong
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
33
positive effect on citizenship acquisition rates. Correspondingly, immi-grants who have lived in Canada for a long period will have a lower opp-ortunity cost when acquiring citizenship since they are less likely to suc-cessfully re-integrate into their home country labour market. Moreover, living in an urban environment with a greater concentration of immigrants and having better access to labour market information can make a diffe-rence to an immigrant’s perception of the benefits derived from citizenship and raise the odds of citizenship acquisition. gender may play a role in the naturalization decision as well. yang (1994) suggested that men have a greater inclination to seek citizenship than women since they are more likely to seek jobs where citizenship may be an advantage.
Costs and benefits of acquiring Canadian citizenship are also affected by the institutional environments in both the sending and receiving countries. on the one hand, the recognition of dual citizenship by the home country plays an important role in an immigrant’s decision to naturalize. in cases where dual citizenship is not recognized by the sending country the costs of Canadian citizenship acquisition rise and the odds of citizenship acquisition are reduced. on the other hand, changes in Canada’s immigra-tion legislation may introduce a period effect that will influence rates of naturalization. in particular, the 1976 Citizenship and immigration Act relaxed entry requirements for immigrants from developing countries and this will possibly affect citizenship acquisition rates during the post-1980 period. these developing countries are all characterized by low opportu-nity costs derived from acquiring Canadian citizenship. to capture this implied differential costs for citizenship acquisition by level of develop-ment, we will divide our sample into oeCD and non-oeCD immigrant source regions.
thus, the immigrant’s demographic and human capital characteristics, plus both Canada’s and source country’s socio-economic contexts all will be incorporated in an economic model of citizenship acquisition. 14
the above arguments lead us to a specification of an ascension model
in a logistic function form or where is a
probability of observing a citizen in our immigrant sample conditioned on vector of explanatory variables . these variables include individual attri-butes and the socio-economic context variables discussed earlier which may influence the naturalization decision. the vector of parameters is esti-mated by the Maximum Likelihood Method.
14 one outstanding modeling problem must be noted. it is quite possible that the deci-sion to become a citizen and the resulting rise in earnings may be jointly determined. in other words, does citizenship lead to greater earnings, or do higher earnings give a greater incentive to become a citizen? We will address this issue and its possible resolution in the next section of the paper.
)exp(1)exp(
)|1(�
�
i
iii X
XXYP
+== )|1( ii XYP =
iX
Don J. DeVoretz & Sergiy PiVnenko
34
Since we also feel that citizenship may vary by gender, we further dis-aggregate our results by gender. in table 2 we report the results for our model of citizenship ascension for a sample drawn from all the immigrant-sending countries as listed above. the maximum likelihood estimates of the logistic model yield a curvilinear relationship between age and naturali-zation rate. the fact is that the rate of ascension is increasing in age but, at decreasing rate, is consistent with our human capital view on the naturali-zation decision. in other words, the younger in age at naturalization, the greater lifetime benefits an immigrant can expect to accrue from the new citizenship status.
table 2. Model of probability of acquiring Canadian citizenship: immigrants from all countries
Coeff. b/St.Er. P[|Z|>z] Mean of X Elasticity
Constant 0.019187 0.164 0.8699
AgeP 0.007346 1.399 0.1617 45.88071 0.055009
AgeSQ -0.00011 -1.843 0.0654 2222.033 -0.03937
ySiM 0.080457 74.192 0 24.54317 0.322278
yiPoSt75 0.01155 0.295 0.7682 0.347505 0.000654
P75_ySiM 0.021916 9.573 0 4.296566 0.015368
tyS -0.00023 -7.85 0 -57.6105 0.002147
FeMALe -0.10292 -7.272 0 0.510275 -0.00857
Pro 0.279808 14.964 0 0.220901 0.00957
SkL 0.1378 7.978 0 0.244092 0.005361
LntinC 0.00012 4.106 0 -42.07 -0.00082
HoWn 0.192035 12.01 0 0.777668 0.02526
DUAL -0.19443 -9.606 0 0.601698 -0.01885
CMA 0.211616 11.696 0 0.834479 0.030215
oeCD -1.25681 -52.936 0 0.647082 -0.11868
number of
observations154458 Log likelihood function -68474.07
Chi squared 15186.62 restricted log likelihood -76067.38
note: Logistic regression: dependent variable Ctzn
Source: Authors’ calculations from 1991, 1996 and 2001 Censuses of Canada.
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
35
in table 2 for male immigrants most of the life-cycle variables obtain the predicted sign and are significant. the effect of the income variable (LnWDiF) that measures the predicted logarithmic differences of citizen versus non-citizens wages is relatively small and negative.15 years since immigration (ySM) positively and significantly influenced the log odds of ascending to citizenship. As we expected, the period dummy (yiPoSt75) which reflects a change in Canada’s immigrant source regions had a posi-tive but statistically insignificant effect on naturalization rates. in addition, we tested for the possible effect of the 1977 citizenship law on the speed of naturalization by using an interaction variable. A positive and statistically significant coefficient on P75_ySM suggests that years since immigration became a more important factor in determining naturalization after the amended changes to Canadian Citizenship act.
Contrary to our expectations the immigrant’s total years of schooling (tyS) had a small and negatively signed effect on the immigrant’s propen-sity to naturalize. the significantly negative coefficient for the gender dummy (FeMALe) suggests that males are more likely to ascend to Cana-dian citizenship which supports yang’s (1994) findings.
our estimates also illustrate the role of economic assimilation in the naturalization decision. Home ownership (HoWn) and the logarithm of total income (LntinC) are significant conditioners and yield the predicted positive signs. Also, a higher occupational status (Pro, SkL) yields a strong positive relationship with the rate of naturalization.
the characteristic of Canada’s socio-economic context, the Census metropolitan area indicator (CMA), is strong and positively signed which supports the idea that living in urban environment fosters immigrant natu-ralization. the significant negative coefficient for the oeCD dummy indi-cates that the immigrant’s source country level of development is an important determinant of citizenship ascension. Contrary to our expecta-tions the coefficient on the dual citizenship dummy (DUAL) which indica-tes the effect of source country citizenship regime on immigrant’s naturali-zation decision is strongly negative. this result could be explained by our data limitations. our available list of the source countries in the oeCD group (except germany and greece) includes only countries with dual citi-zenship regime, whereas in non-oeCD group only Poland recognizes dual
15 For non-citizens this variable is calculated as
€
LNWDIF i = ˆ β C X i − LNWAGE i , for citizens
€
LNWDIF i = LNWAGE i − ˆ β NC X i ; where LnWAgei – logarithm of individual’s annual wage earnings, Xi – vector of individual’s characteristics,
€
ˆ β NC and
€
ˆ β C are vectors of oLS coef-ficients estimated from log-linear earnings equations for non-citizens and citizens respec-tively. this variable equals the mean income difference between a 25-year old immigrant with and without Canadian citizenship, from the particular country of origin, for the sampled observation. Figures 7 and 8 illustrate how this was computed.
Don J. DeVoretz & Sergiy PiVnenko
36
citizenship. thus, the DUAL dummy variable essentially becomes an indi-cator of the level of development.16
Wage equation
DeVoretz and Pivnenko (2004) have already demonstrated that a signifi-cantly positive earnings effect derives from ascension to Canadian citizen-ship. in addition, they acknowledged the possibility that reverse causality can occur if the higher earnings observed among naturalized citizens influ-ence the immigrant’s decision to acquire Canadian citizenship.17 in other words, citizenship status (C), and the natural logarithm of citizen and non-
citizen gross annual wages may be determined simultaneously. thus, citi-zenship status may also be a function of the expected citizen/non-citizen wage differential, since immigrants may incorporate the potential wage premium associated with citizenship status in their decision to become citi-zens. Following Heckman (1976) and Lee (1978) we estimate the outlined empirical model in order to account for this potential selection bias and the implied simultaneity:
€
C i = α0 + α1X i + α2Yi + α3LW ˆ D IF i + εi (1)
€
lnWi = β0 + β1X i + β2Zi + β3λi + ν i , if Ci =1 (2)
€
lnWi = γ 0 + γ1X i + γ 2Zi + γ 3λi + τ i , if Ci =0 (3)
where
Ci – binary variable indicating the immigrant’s choice of citizenship status
(1- citizen, 0- non-citizen);
Xi – vector to represent the immigrant’s human capital characteristics;
yi – other determinants of citizenship status;
zi – control variables in wage equation (such as occupational choice and
weeks worked).
€
λi =φ( ˆ C i
*)Φ( ˆ C i
*) if Ci =1, or
€
λi =−φ( ˆ C i
*)(1− Φ( ˆ C i
*)) if Ci =0 – selectivity variable (inverse
Mill’s ratios for citizens and non-citizens respectively).
16 We believe that by using the 20% uncensored census samples the results will improve as they did for Bloemraad (2002).17 robinson and tomes (1982) were the first to apply the Heckman correction in the migration context.
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
37
€
LW ˆ D IF i - simulated citizen-non-citizen wage differential which equals the
difference between the logarithms of: i) observed and opportunity wages
€
lnWi − ( ˆ γ 0 + ˆ γ 1X i + ˆ γ 2Zi + ˆ γ 3λi ) for citizens, or ii) opportunity and observed wages
€
( ˆ β 0 + ˆ β 1X i + ˆ β 2Zi + ˆ β 3λi ) − lnWi for non-citizens 18.
in the first stage the selection equation is estimated by a maximum likelihood technique as an independent ProBit model to determine the decision to acquire Canadian citizenship. A vector of inverse Mills ratios (iMrs), estimated expected error, is then generated from the parameter estimates of the selection equation. the citizen’s wage is observed only when the selection equation equals 1 (i.e., an immigrant acquires citizen-ship) and its logarithm is then regressed on the explanatory variables and the vector of iMrs from the selection equation by ordinary least squares (oLS) method. Similarly we obtain oLS coefficients in the non-citizen wage equation. Hence, in the second stage we rerun the regression with the estimated expected error included as an extra explanatory variable, remo-ving the part of the error term correlated with the explanatory variable, and thus avoiding the suspected selection bias.
in order to generate the iMrs we use the reduced form of equation (1) that excludes wage differentials for the citizen/non-citizen workers. then we incorporate the estimated lambdas (or iMrs) into wage equations (2) and (3) and run an oLS procedure to estimate the selection bias corrected regressions coefficients. next, we use the estimated coefficients for citizens (non-citizens) to forecast the opportunity wages for non-citizens (citizens). Finally, we estimate a ProBit equation (1) using the simulated citizen/non-citizen wage differentials.
the results for the earnings equations are presented in table 2. At this point we must keep in mind that some of the model’s estimated coefficients do not lend themselves to a straightforward interpretation. if a variable appears only in the wage equation, its coefficient can be simply interpreted as the marginal effect of a one-unit change in the variable that appears in this one equation. if, on the other hand, the variable appears in both the selection and wage equations, the coefficient in the outcome equation is affected by its presence in the selection equation as well. table 3 shows the estimated coefficients derived from the earnings equations, implying that the indirect effects of age and education on the logarithm of wages are not shown in this table.
18 Here we rationalize our assumption that immigrants form their citizenship premium expectations based on the observed performances of their counterparts with similar back-ground but opposite citizenship status.
Don J. DeVoretz & Sergiy PiVnenko
38
table 3. estimation of log-wage equations for citizens and non-citizens: employed immigrants in 2001
All immigrants OECD immigrants Non-OECD immigrants
CitizensNon-
citizensCitizens
Non- citizens
CitizensNon-
citizens
Constant4.3794
(8.92)
7.5299
(11.98)
4.4912
(12.17)
6.5664
(17.32)
5.0640
(19.32)
6.4401
(15.51)
AgeP0.0562
(7.49)
0.0470
(5.48)
0.0690
(9.55)
0.0483
(4.99)
0.0472
(7.90)
0.0501
(4.52)
AgeSQ-0.0004
(-6.74)
-0.0003
(-3.58)
-0.0006
(-9.19)
-0.0004
-3.57 ()
-0.0004
(-6.69)
-0.0005
(-4.00)
FeMALe-0.2634
(-20.22)
-0.2913
(-17.41)
-0.3214
(-24.60)
-0.3359
(-15.40)
-0.2173
(-18.91)
-0.2515
(-10.18)
HLAng0.1804
(13.71)
0.1966
(11.11)
0.1256
(7.76)
0.1017
(3.63)
0.1469
(12.85)
0.0793
(3.13)
DiPL0.2223
(4.97)
0.2236
(3.43)
0.1160
(5.51)
0.0700
(2.27)
0.2491
(6.47)
0.2260
(2.92)
BACH0.4434
(5.63)
0.4400
(3.68)
0.3171
(9.46)
0.2639
(5.41)
0.3953
(8.08)
0.2576
(2.58)
BACHPL0.4252
(7.94)
0.3633
(4.73)
0.3418
(9.81)
0.2744
(5.34)
0.4311
(10.34)
0.2737
(3.35)
PHD0.3718
(4.92)
0.3391
(3.46)
0.3097
(4.27)
0.3252
(3.29)
0.5185
(7.31)
0.3572
(2.52)
MAng0.4224
(18.41)
0.5015
(16.71)
0.4400
(20.53)
0.4771
(13.41)
0.3688
()16.91
0.3666
(6.48)
ProF0.3229
(17.83)
0.3517
(14.76)
0.2895
(15.56)
0.3117
(10.10)
0.3365
(21.29)
0.3293
(8.88)
SkiL0.1577
(9.37)
0.1821
(8.57)
0.1519
(9.18)
0.1858
(6.79)
0.1444
(9.49)
0.1069
(3.31)
LnWeekS0.7015
(43.35)
0.6958
(36.29)
0.7034
(40.99)
0.6966
(26.02)
0.6990
(50.75)
0.6781
(25.97)
FtW0.6991
(34.29)
0.5830
(24.21)
0.7459
(37.85)
0.6392
(21.09)
0.6586
(35.46)
0.5070
(13.43)
LAMBDA2.3007
(2.76)
1.5262
(2.58)
1.2920
(2.88)
0.7346
(2.06)
1.3316
(3.24)
0.5043*
(1.45)
Adj. r2 0.42 0.39 0.39 0.39 0.45 0.40Sample size 38,158 7,467 21,867 3,271 16,291 4,196* insignificant at 5%Source: Authors’ calculations.
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
39
Although there are sizable differences in the coefficients across non-citizens and naturalized citizens by place of origin, most variables behave similarly and yield the expected signs as predicted by a human capital model. this outcome can be explained by the fact that the majority of oeCD immi-grants in the selected population come from either english- or French-spea-king countries. the impacts of various educational attainment variables and the immigrant’s occupational choice as measured by their estimated coefficients are similar for naturalized citizens and non-citizens in both the oeCD and non-oeCD groups. the educational coefficients, however, show a slightly greater effect on naturalized citizens’ earnings regardless of the immigrant’s source country.
the coefficient of the inverse Mill’s ratio (iMr) or lambda represents the product of the standard deviation of the errors in the wage equation and the correlation between the wage equation error term and the selection equation error term. Lambda (iMr) is positive and significant for citizens in both groups, and citizens in non-oeCD group; positive and insignificant for non-citizens in non-oeCD group. the significance of the coefficients for the iMr in the citizens’ equation indicates the importance of control-ling for self-selection bias. this suggests that the results reported by DeVo-retz and Pivnenko (2004) were biased since they ignored self-selection.
Adjusted for self-selection bias, our results now suggest that the unob-served element in citizenship choice plays an important role in the earnings regressions. in fact, a positive lambda indicates a positive correlation between the error terms in the selection and wage equations. thus, in our specification there are unobserved variables that both increase the proba-bility of ascension to citizenship and lead to higher-than-average values for the dependent variable in the earnings equation.
As noted above, since lambda is the correlation between the errors in the selection and wage equations, it is extremely sensitive to model specifi-cation, and therefore its interpretation will necessarily carry a certain degree of ambiguity. in tables B-1 and B-3 (see Appendix B) the estimates for sub-samples of male and female immigrants display similar results, alt-hough with a few exceptions. First, the coefficient on the reported lambdas for male immigrants from oeCD countries and female non-citizens from non-oeCD countries are insignificant and negative. Second, educational variables are insignificant for oeCD male citizens and non-oeCD female non-citizens.
Don J. DeVoretz & Sergiy PiVnenko
40
Selection equation
in our earlier study we argued that the decision to ascend to citizenship is conditioned on the costs and benefits associated with citizenship status. We included age, the expected wage premium, as well as educational attain-ment, marital status and married with children dummies in the selection equation as factors that may affect the costs and benefits of citizenship.
table 4 shows the maximum likelihood estimates of ProBit coeffi-cients derived from the selection equation. in general, our model better predicts the citizenship decision for immigrants from non-oeCD countries. For both source country groups the age variable has a positive effect on immigrant’s decision to acquire Canadian citizenship, as older people demonstrate a stronger commitment to stay in Canada. Marital status dis-plays a significant and positive effect only for the oeCD group, while a negative and significant coefficient for the interaction dummy in the same group indicates that the presence of children decreases the likelihood of acquiring citizenship.
table 4. prOBIt estimation of regression coefficients in citizenship equation (t-ratios in brackets)
All immigrants oeCD immigrants non-oeCD immigrants
estimated
Coefficient
Marginal
effect
estimated
Coefficient
Marginal
effect
estimated
Coefficient
Marginal
effect
Constant
-3.3974
(-25.68) -0.8130
-2.3542
(-12.15) -0.6576
-4.2481
(-21.68) -0.8414
AgeP
0.0093
(11.29) 0.0022
0.0145
(12.40) 0.0041
0.0160
(13.33) 0.0032
DiPL
0.0405
(2.43) 0.0096
0.0059
(0.26) 0.0016
0.1490
(6.04) 0.0286
BACH
0.0857
(3.73) 0.0199
-0.0220
(-0.63) -0.0062
0.1513
(4.84) 0.0283
BACHPL
-0.0753
(-2.61) -0.0186
-0.0650
(-1.62) -0.0186
-0.0228
(-0.54) -0.0046
PHD
-0.3860
(-6.62) -0.1094
-0.3872
(-5.22) -0.1240
-0.1612
(-1.60) -0.0349
MArSt
-0.0576
(-2.95) -0.0136
0.0115
(0.43) 0.0032
-0.1326
(-4.53) -0.0253
MAr_CH15
-0.0402
(-2.19) -0.0097
-0.1259
(-4.68) -0.0360
-0.0058
(-0.22) -0.0011
LWDiFF
0.3931
(29. 51) 0.0941
0.2475
(13.13) 0.0691
0.4732
(23. 98) 0.0937eStreLLA
r-SQUAre0.049 0.024 0.040
Source: Authors’ calculations.
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
41
Coefficients on the family characteristics mentioned above are both nega-tive in the non-oeCD group. the attainment of a bachelor degree positi-vely affects the immigrant’s citizenship decision, while obtaining a more advanced degree produces a negative effect. Finally, as expected, the simu-lated citizen/non-citizen wage differential has a strong positive effect on the citizenship decision across both source country groups.
While results for sub-samples of males (see table C-2 in Appendix B) display a similar pattern as reported for the aggregate results in table 3, results for females (table C-4) differ in two respects. First, all educational qualifications for oeCD females have a negative effect on citizenship acquisition. Second, non-oeCD female immigrants derive a positive effect from the presence of children, with a strong negative coefficient on marital status.
Decomposition analysis
the wage gap between the two immigrant groups may be owing to diffe-rences in individual characteristics and differences in returns to these cha-racteristics. the oaxaca-Blinder decomposition method (oaxaca 1973; Blinder 1974) has become a routine method in labour market discrimina-tion studies to explain segmented group wage differences. in our case the citizenship status of an immigrant segments the labour market, and thus we employ a decomposition analysis. in its simplest version the idea is to isolate the fraction of wage differentials unexplained by human capital endowments usually attributed to labour market discrimination. Accor-dingly, we have to adopt one of the estimated wage structures as the nondi-scriminatory norm for the group believed to be dominant in the labour market (citizens) relative to the comparison group (non-citizens).
the human capital portion of the overall wage differential is obtained as a sum of the differences in the mean characteristics of the two groups weighted by the estimated coefficients for the nondiscriminatory wage standard. the portion of the overall wage differential owing to discrimina-tion will then be the residual left over after netting out for the human capital portion.19 in our study we adapt this decomposition methodology to explain citizen/non-citizen immigrant wage differentials. Further we treat non-citizens as a disadvantaged group since non-citizens are explicitly discriminated against in the public sector by blocking job access. Moreover, in the private sector we argue that citizenship status serves as a signal: first, it indicates an attachment to Canada, and, second, it may indicate greater cultural integration.
19 this could also be directly calculated as a sum of the difference in estimated coeffi-cients between the two groups weighted by the mean characteristics of the discriminated group.
Don J. DeVoretz & Sergiy PiVnenko
42
neuman and oaxaca (2003) acknowledged that when you introduce a correction for selectivity bias this in turn introduces some fundamental ambiguities in the context of wage decompositions. thus we follow one of their suggested decomposition modifications we are using below. 20
Based on the estimates obtained earlier, we define the citizen/non- citizen wage differential in matrix notation as (4)
where the first term on the right hand side represents the effect of the dif-ferences in mean characteristics, and the second term depicts the effect of differential returns to these characteristics, while the third term represents the selectivity residual.21
the decomposition results presented in table 5 suggest that the diffe-rences in immigrants’ human capital endowments and returns to these endowments are equally important determinants of the observed wage dif-ferentials between citizens and non-citizens.
table 5. Decomposition of wage differentials between naturalized citizens and permanent residents of Canada: population of foreign-born employees 25-65 years old
naturalized citizens
permanent residents
wage differential
Difference
unexplained by
human capital
endowments
(structural differences
and selectivity)
Differences due to
the human capital
endowments
All 15.3% 7.1% 8.2% Males 14.2% 6.0% 8.2% Females 15.2% 8.2% 7.0%oeCD 9.8% 3.5% 6.3% Males 7.6% 3.0% 4.6% Females 7.7% 4.0% 3.7%non-oeCD 28.9% 14.6% 14.3% Males 29.3% 13.4% 15.8% Females 29.5% 15.8% 13.7% Source: Authors’ calculations.
20 in our case, the selectivity component will be netted out of the endowment effect and, instead, combined with structural differences effect. these two effects will in turn be treated as a part of the wage differential unexplained by human capital characteristics.21 Vectors of the estimated coefficients from citizens and non-citizens wage equations ( and ) exclude and respectively.
€
lnWC − lnWN = (X C − X N )T ˆ γ + X NT ( ˆ β − ˆ γ ) + ( ˆ γ 3λ C − ˆ β 3λ N )
€
ˆ β
€
ˆ γ
€
ˆ β 3
€
ˆ γ 3
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
43
For our general sample of employed immigrants, we detected a wage diffe-rential of 15% between citizens and non-citizens. Moreover, as shown in table 4, human capital endowments are relatively more important for males than females.
in general, immigrant citizens from non-oeCD countries enjoyed a greater wage advantage than non-citizens from oeCD countries (28.9% vs. 9.8%).
public Finance Implications of Citizenship A crucial and reoccurring question in the minds of immigration critics is: “Do the foreign-born make a net positive contribution to the treasury?” We expand this question here by asking if the naturalized foreign-born contribute more or less than non-citizens to the treasury. if citizen status increases income opportunities, then, under a progressive tax system, natu-ralized citizens should contribute more to the treasury. it is however pos-sible that ascension to citizenship increases public entitlements. Hence, the impact of citizenship on public finance transfers is ambiguous.
Simon (1984) provides us with a theoretical framework to answer the question formulated above by suggesting that income and, therefore, tax payments, are concave in age, while public transfers consumption is convex in age, as shown in Figure 4.
Figure 4. age-Consumption tax profiles by birth status: optimistic case
$
Cn
Age
foreign-born taxes
Cn
Ci
Ti
Tn
Canadian taxes
Canadian-born consumption
foreign-born consumption
X
CiTn Ti
18 25 78
Don J. DeVoretz & Sergiy PiVnenko
44
in addition, DeVoretz and Pivnenko (2004) suggest that many Canadian naturalized citizens may be selected three times: first, by themselves when they choose to immigrate; secondly, by Canada via the point system; and then by self-selection into the naturalization process. thus, the hypothesi-zed age transfer patterns may vary by birth and citizenship status. the effect of this triple selection is observed in many dimensions. First, Cana-dian-born residents start consuming public finance transfers (denoted Cn-Cn
in Figure 4) for health services at birth, and then, at the age of six, receive educational subsidies. next, between ages 19 and 22, the Canadian-born population simultaneously exits the education sector and enters the labour force. this reduces their consumption of subsidized education and associa-ted government transfers. At this point in the life cycle, the Canadian-born begin to pay taxes (tn-tn). in contrast to this life-cycle pattern, an immigrant’s public good consumption curve (Ci-Ci) begins later, after entry to Canada (i.e. age 26) and may lie everywhere below (or above) the public good consumption curve of its Canadian-born cohort, depending upon economic and demographic circumstances.
the ambiguity in the level of consumption of public transfers by the foreign-born, naturalized or not, at every age, results from two countervai-ling forces. on the one hand, the immigrant’s initial lack of eligibility to consume some subsidized services, such as health, is tied to residency requi-rements. other programs, such as (Un)employment insurance, are income-contingent and thus may reduce immigrant access to public transfers. on the other hand, the initial risk faced by the immigrants in the labour market could lead to their greater use of entitlement programs, such as social assis-tance, upon arrival. Finally, it is hypothesized that initially low foreign-born tax payments (ti-ti) accelerate to a “‘crossover” point (at X), where tax payments by the foreign-born exceed those of the Canadian-born as a result of higher foreign-born earnings because of citizenship acquisition and of the progressive nature of Canada’s tax system.
Figure 4 best represents the optimistic case. A pessimistic case presen-ted in Figure 5 would have the foreign-born earn less and never “catch up” to the Canadian-born where earnings are concerned, and this would result in lower tax payments. it would also delay or completely forestall their tax “crossover.” in this case, given a low household income, the foreign-born consumption of government transfers would lie above that predicted in Figure 5.
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
45
Figure 5. age-Consumption tax profiles by birth status: pessimistic case
total federal government net treasury payments in 1995 by birthplace and citizenship status of the head of household are reported in Figure 6.22 these government transfers to the federal treasury are a result of calculating taxes paid minus the receipt of pensions, child tax credits, employment benefits and other money transfers by individual households.23
22 the life-cycle transfer model described in Figures 4 and 5 was estimated in the Cana-dian context by DeVoretz and Pivnenko (2004), and represents a flexible theoretical tool. 23 Federal taxes consist of the income tax and a consumption tax (i.e., the gSt) paid per household. Calculation of a household income tax rate was based on the reported hous-ehold income and the applicable average and marginal tax rates after appropriate house-hold deductions were made. the gSt contribution was derived by taking the average propensity to consume, as reported in FAMeX 1995, for a similar household, multiplied by the appropriate gSt or HSt in the relevant province. See Appendix C for details.
$
Cn
Age
foreign-born taxes
CnCi
TiTn
Canadian taxes
Canadian-born consumption
foreign-born consumption
X
Ci
Tn Ti
18 25 78
Don J. DeVoretz & Sergiy PiVnenko
46
Figure 6. Net treasury transfers by Canadian-born and foreign-born by citizenship status, 1995 (5-year moving average)
Source: Authors’ calculations.
Foreign birth status and citizenship status have two separate effects on life-cycle transfer payments. Being a non-citizen in Canada substantially reduces net treasury payments; however, treasury transfers rise when the immigrant ascends to citizenship, and these transfers exceed the native-born transfers after age 50. in fact, a naturalized citizen transfers a dis-counted value of $67,986 to the federal treasury, an amount almost equal to the Canadian-born contribution of $72,208, but greatly superior to the non-citizen lifetime discounted transfer of $35,164.24
We can expand this analysis to complement our earlier work by esti-mating net treasury transfers for oeCD and non-oeCD immigrants by citizenship status as reported in Figures 7 and 8.25
24 the net present value = where r=.05, t=1 when i=27 all in 1995 dollars.25 oeCD countries include: France, germany, greece, italy, the netherlands, Portugal, Spain, the United kingdom, and the United States. non-oeCD countries include: China (PrC), india, Lebanon, the Philippines, Poland, former USSr (european), Vietnam, and former yugoslavia.
-$8 000
-$6 000
-$4 000
-$2 000
$0
$2 000
$4 000
$6 000
$8 000
$10 000
27 37 47 57 67 77
Age
Canadian $
CB All FB cit All FB ncit
Net Present Value at 5%: $72,208(CB), $67,986 (FBcit), $35,164(FBncit)
€
NPV = Σi = 27
i = 75 (TotTaxi
− GovTransi)
(1 + r )t
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
47
Figure 7. Net treasury transfers for Canadian-born and OeCD immigrants by citizenship status, 1995 (5-year moving average)
Source: Authors’ calculations.
Figure 7 reveals that naturalized citizens from the oeCD transfer more to the Canadian federal treasury over their entire lifetime than either Cana-dian-born citizens or non-citizens from the oeCD. Figure 8, however, portrays the dramatic effect of citizenship status on the size of public trans-fers for immigrants from non-oeCD countries: Canadian citizens generally transfer twice as much as non-citizens from non-oeCD countries each year in the lifecycle.
Figure 8. Net treasury transfers for Canadian-born and non-OeCD immigrants by citizenship status, 1995 (5-year moving average)
Source: Authors’ calculations.
-$8 000-$6 000
-$4 000-$2 000
$0
$2 000$4 000$6 000
$8 000$10 000
27 37 47 57 67 77
Age
1995 Canadian $
CB OECDcit OECDncit
Net Present Value at 5%: $72,208(CB), $86,417 (OECDcit), $71,491(OECDncit)
-$8 000
-$6 000
-$4 000
-$2 000
$0
$2 000
$4 000
$6 000
$8 000
27 37 47 57 67 77
Age
CB NOECDcitNOECDncit
Net Present Value at 5%: $72,208(CB), $59,992 (NOECDcit), $18,548(NOECDncit)
-$8 000-$6 000
-$4 000-$2 000
$0
$2 000$4 000$6 000
$8 000$10 000
27 37 47 57 67 77
Age
1995 Canadian $
CB OECDcit OECDncit
Net Present Value at 5%: $72,208(CB), $86,417 (OECDcit), $71,491(OECDncit)
CB NOECDncitNOECDcit
Don J. DeVoretz & Sergiy PiVnenko
48
Figure 10. Net Treasury Transfers Profiles for immigrants from OECD
and non-OECD countries by Canadian citizenship status
(5-year moving average)
$0
$2 000
$4 000
$6 000
$8 000
$10 000
$12 000
15 25 35 45 55 65 75 85
Age
19
95
Ca
na
dia
n d
oll
ars
OECDcit OECDncit NOECDcit NOCDncit
Source: Authors' calculations
these treasury transfers are summarized across citizenship and place of birth status in table 6. By place of birth, oeCD citizens transfer the largest amount, followed by the transfers contributed by Canadian-born and all naturalized citizens. Moreover, citizenship status for all foreign-born groups has a greater impact on public finance contributions than birth status, except for immigrants from the oeCD.
table 6. Net present value of public finance transfers: 1995 dollars, 5% rate
Canadian-bornAll
Foreign-born
OECD
immigrants
Non-OECD
immigrantsCitizens $72,208 $67,986 $86,417 $59,992non-Citizens $35,164 $71,491 $18,548
Are these differential treasury transfers by citizenship status a result of tax payments, use of federal services, or both? Figure 9 illustrates the large impact on tax payments owing to citizenship for non-oeCD immigrants, and a much more moderate citizenship impact for the oeCD immigrant group.
Figure 9. tax payment for immigrants from OeCD and non-OeCD countries by citizenship status (5-year moving average)
Source: Authors’ calculations.
Figure 10. Net Treasury Transfers Profiles for immigrants from OECD
and non-OECD countries by Canadian citizenship status
(5-year moving average)
$0
$2 000
$4 000
$6 000
$8 000
$10 000
$12 000
15 25 35 45 55 65 75 85
Age
19
95
Ca
na
dia
n d
oll
ars
OECDcit OECDncit NOECDcit NOCDncit
Source: Authors' calculations
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
49
Figure 10 portrays the impact of citizenship and birth status on the con-sumption of federally financed services. there is no citizenship effect by place of birth except after retirement (65) when the lack of consumption of pensions by non-citizens from non-oeCD countries (noeCDncit) res-trains their use of public service.
Figure 10. Consumption to federal transfers for immigrants from OeCD and non-OeCD countries by Canadian citizenship status (5-year moving average)
Source: Authors’ calculations.
Figure 11 documents this effect in more detail. After age 65, federal pension payments to all naturalized citizens and non-citizen oeCD immigrants grow but there is little growth in pension payments for non-oeCD immi-grants who have not ascended to citizenship. this corresponds to the fin-dings of Shamsuddin and DeVoretz (1999) who noted that the newest wave of immigrants are less eligible for government-financed pensions.
Figure 10. Net Treasury Transfers Profiles for immigrants from OECD
and non-OECD countries by Canadian citizenship status
(5-year moving average)
$0
$2 000
$4 000
$6 000
$8 000
$10 000
$12 000
15 25 35 45 55 65 75 85
Age
19
95
Ca
na
dia
n d
oll
ars
OECDcit OECDncit NOECDcit NOCDncit
Source: Authors' calculations
Figure 10. Net Treasury Transfers Profiles for immigrants from OECD
and non-OECD countries by Canadian citizenship status
(5-year moving average)
$0
$2 000
$4 000
$6 000
$8 000
$10 000
$12 000
15 25 35 45 55 65 75 85
Age
19
95
Ca
na
dia
n d
oll
ars
OECDcit OECDncit NOECDcit NOCDncit
Source: Authors' calculations
Don J. DeVoretz & Sergiy PiVnenko
50
Figure 11. pension payments to immigrants from OeCD and non-OeCD countries by Canadian citizenship status (3-year moving average)
Source: Authors’ calculations.
ConclusionsAt the micro level, our analysis of the economic implications of Canadian citizenship focused on the individual determinants of the naturalization decision and on the subsequent earnings performance of immigrants. Despite specification difficulties and demanding data requirements, our model based on Heckman’s selectivity correction procedure generated some important conclusions.
First, possible selectivity bias has to be taken into account in the analy-sis of immigrants’ earnings performance across their citizenship status. our results showed that the significance and the magnitude of the selectivity effects varied by gender and source country groups. Furthermore, the oaxaca decomposition results derived from the selectivity corrected ear-nings equations demonstrated the importance of the endowments compo-nent as compared to the estimates made in the presence of selection bias.
Second, we confirmed our earlier findings that the naturalization deci-sion is conditioned by the expected wage gain, level of education, marital status, age and presence of children. Statistically stronger results were obtained with the non-oeCD immigrant group.
At the macro level, our study focused on the implications of Canadian citizenship for the lifetime public finance contributions of immigrants. We found evidence that all immigrants, regardless of their source country group and citizenship status, make positive contributions to Canada’s tre-asury over their life cycle. naturalized citizens from oeCD countries trans-
$0
$2 000
$4 000
$6 000
$8 000
61 66 71 76 81
Age
OECDcit OECDncit NOECDcit NOCDncit
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
51
ferred the most to the federal treasury with their contributions exceeding the corresponding value for the Canadian-born by more than $14,000. As expected from our triple selection arguments, our estimates indicate that naturalized citizens made greater net contributions to the federal treasury than their non-citizen counterparts regardless of source country. this dif-ference was smaller among the immigrants from oeCD countries ($86,417 vs. $71,491), and substantially greater in the non-oeCD immigrant group ($59,992 vs. $18,548). the relatively poor public finance performance of non-citizens in the latter group was primarily explained by their low income and their low level of tax payments over the entire lifespan. in contrast to the other three groups that displayed almost identical pension and transfer acquisition profiles, the non-citizens from non-oeCD countries received considerably smaller government transfers and lower pensions after the age of 55. in sum, citizenship status has substantial micro implica-tions on immigrant earnings and macro economic implications on Canada’s treasury.
Don J. DeVoretz & Sergiy PiVnenko
52
appendix a
excerpts from Canada’s Citizenship actPart i: the right to Citizenship(1) the Minister shall grant citizenship to any person who(a) makes application for citizenship;(b) is eighteen years of age or over;(c) is a permanent resident within the meaning of subsection 2(1) of the Immigration and Refugee Protection Act, and has, within the four years immediately preceding the date of his or her application, accumulated at least three years of residence in Canada calculated in the following manner:
(i) for every day during which the person was resident in Canada before his lawful admission to Canada for permanent residence the person shall be deemed to have accumulated one-half of a day of resi-dence, and(ii) for every day during which the person was resident in Canada after his lawful admission to Canada for permanent residence the person shall be deemed to have accumulated one day of residence;
(d) has an adequate knowledge of one of the official languages of Canada;(e) has an adequate knowledge of Canada and of the responsibilities and privileges of citizenship; and(f) is not under a removal order and is not the subject of a declaration by the governor in Council made pursuant to section 20.
Source: Department of Justice Canada, http://laws.justice.gc.ca/en/C-29/34586.html
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
53
appendix B
Male and Female Wage and Citizenship Models
table B-1. estimation of log-wage equations for citizens and non-citizens: employed male immigrants in 2001
OECD Non-OECD
CitizensNon-
CitizensCitizens
Non- Citizens
Constant7.2583
(3.48)
3.9108
(2.40)
4.2140
(7.09)
8.1689
(8.03)
AgeP0.0402*
(1.28)
0.0328*
(1.19)
0.0529
(4.26)
0.0557
(2.89)
AgeSQ-0.0006
(-3.15)
-0.0006
(-3.09)
-0.0004
(-3.32)
-0.0004
(-2.00)
HLAng0.1405
(3.50)
0.1050
(2.10)
0.1535
(6.47)
0.0544*
(1.35)
DiPL-0.0132*
(-0.11)
-0.1093*
(-0.65)
0.3610
(3.33)
0.5144
(2.28)
BACH0.0875*
(0.50)
-0.0900
(-0.37)
0.5365
(3.99)
0.6510
()2.25
BACHPL0.1530*
(0.93)
0.0081*
(0.04)
0.5767
(4.81)
0.6315
(2.51)
PHD0.4887
(3.22)
0.4379
(2.24)
0.6264
(4.14)
0.7607
(2.41)
MAng0.4216
(8.36)
0.4844
(7.79)
0.3752
(9.04)
0.3408
(4.15)
ProF0.2483
(5.09)
0.3133
(5.21)
0.3363
(9.83)
0.3157
(5.09)
SkiL0.1398
(3.50)
0.1720
(3.58)
0.1531
(5.09)
0.0639*
(1.37)
LnWeekS0.6631
(13.96)
0.7259
(13.69)
0.7516
(23.45)
0.6242
(13.98)
FtW0.9282
(12.46)
0.8287
(9.92)
0.7726
(14.77)
0.4843
(6.47)
LAMBDA-2.7163*
(-1.02)
-1.8582*
(-0.95)
2.2851
(2.29)
1.7849
(2.00)
Adj. r2 0.35 0.30 0.35 0.24
Sample size 8,557 2,032 10,693 1,643
* insignificant at 5%
Don J. DeVoretz & Sergiy PiVnenko
54
table B-2. prOBIt estimation of citizenship equation: male immigrants in 2001
OECD Non-OECD
Estimated
Coefficient
Marginal
Effect
Estimated
Coefficient
Marginal
Effect
Constant-1.5459
(-4.21) -0.4165
-5.6320
(-17.90) -1.1188
AgeP0.0157
(9.61) 0.0042
0.0140
(8.15) 0.0028
DiPL0.0711
(2.18) 0.0190
0.1693
(4.72) 0.0323
BACH0.0800
(1.57) 0.0209
0.1574
(3.48) 0.0294
BACHPL0.0591
(1.01) 0.0156
0.0034
(0.06) 0.0007
PHD-0.1638
(-1.82) -0.0471
-0.1065
(-0.93) -0.0224
MArSt0.0176
(0.45) 0.0048
-0.1924
(-4.28) -0.0360
MAr_CH15-0.0453
(-1.22) -0.0123
-0.0259
(-0.71) -0.0052
LWDiF0.1564
(4.42) 0.0421
0.6105
(19.45) 0.1213eStreLLA
r-SQUAre0.017 0.049
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
55
table B-3. estimation of log-wage equations for citizens and non-citizens: employed female immigrants in 2001
OECD Non-OECDCitizens Non-citizens Citizens Non-citizens
Constant5.2866
(24.65)
6.0792
(15.61)
5.6437
(21.44)
5.2294
(9.85)
AgeP0.0484
(7.06)
0.0361
(2.64)
0.0383
(5.85)
0.0416
(2.55)
AgeSQ-0.0005
(-6.52)
-0.0003
(-2.16)
-0.0004
(-5.54)
-0.0005
(-2.94)
HLAng0.1190
(6.42)
0.1081
(2.71)
0.1378
(10.70)
0.1049
(2.93)
DiPL0.0756
(4.48)
0.0384*
(1.10)
0.1667
(5.44)
0.1037*
(1.15)
BACH0.2916
(11.05)
0.2923
(5.53)
0.2876
(7.14)
0.0765*
(0.65)
BACHPL0.3307
(11.08)
0.2788
(4.59)
0.3376
(10.31)
0.1621*
(1.85)
PHD0.3457
(3.41)
0.4386
(2.82)
0.5213
(5.28)
0.1130*
(0.45)
MAng0.4495
(16.83)
0.4622
(8.81)
0.3548
(12.90)
0.3981
(4.53)
ProF0.3237
(16.04)
0.3094
(7.62)
0.3388
(19.89)
0.3340
(6.79)
SkiL0.1478
(7.51)
0.1814
(4.46)
0.1332
(7.34)
0.1851
(3.52)
LnWeekS0.7246
(40.77)
0.6877
(19.74)
0.6600
(46.23)
0.7120
(20.40)
FtW0.6853
(38.45)
0.5903
(16.94)
0.6170
(35.39)
0.5076
(11.09)
LAMBDA0.0879*
(0.42)
0.2430*
(1.02)
0.3656*
(0.94)
-0.2935*
(-0.63)
Adj. r2 0.43 0.38 0.36 0.35Sample size 7,734 2,164 11,174 1,628
* insignificant at 5%
Don J. DeVoretz & Sergiy PiVnenko
56
table B-4. prOBIt estimation of citizenship equation: female immigrants in 2001
OECD Non-OECD
Estimated
Coefficient
Marginal
Effect
Estimated
Coefficient
Marginal
Effect
Constant-2.5608
(-9.49) -0.7417
-2.5453
(-8.93) -0.5081
AgeP0.0137
(8.03) 0.0040
0.0195
(11.48) 0.0039
DiPL-0.0237
(-0.73) -0.0069
0.1340
(3.91) 0.0261
BACH-0.0608
(-1.26) -0.0179
0.1643
(3.76) 0.0308
BACHPL-0.1197
(-2.18) -0.0360
-0.0154
(-0.25) -0.0031
PHD-0.6676
(-4.87) -0.2348
-0.2094
(-0.98) -0.0470
MArSt0.0008
(0.02) 0.0002
-0.1005
(-2.57) -0.0196
MAr_CH15-0.1273
(-3.23) -0.0378
0.0176
(0.47) 0.0035
LWDiF0.2731
(10.77) 0.0791
0.2927
(10.14) 0.0584eStreLLA
r-SQUAre0.030 0.026
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
57
appendix C
public Finance regressions for OeCD and Non-OeCD Immigrants
table C-1. OeCD Immigrants: total transfers
a) Dependent Variable: ttLgoVrtrAnSFrPAyMnt
Unstandardized
Coefficients
Standardized
Coefficients t Sig.
B Std. error Beta(Constant) -4581.730 105.690 -43.351 .000
Age 200.508 1.400 .580 143.219 .000
LMArrieD -2141.543 47.970 -.171 -44.643 .000
CITIZ -11.552 49.677 -.001 -.233 .816
totAL
inCoMe-.017 .001 -.078 -21.434 .000
HoUSeHoLD
Size106.152 16.585 .026 6.401 .000
r r Square Adjusted r SquareStd. error of the
estimate
.593(a) .351 .351 4540.359
Don J. DeVoretz & Sergiy PiVnenko
58
table C-2. OeCD Immigrants: pension transfers
b) Dependent Variable: oAS PenSion giS
Unstandardized
Coefficients
Standardized
Coefficients t Sig.
B Std. error Beta(Constant) -3708.680 43.479 -85.299 .000
Age 122.042 .576 .717 211.902 .000
LMArrieD -1298.885 19.734 -.211 -65.819 .000
CITIZ -129.230 20.436 -.020 -6.324 .000
totAL inCoMe -.012 .000 -.113 -37.055 .000
HoUSeHoLD
Size37.773 6.823 .019 5.536 .000
r r Square Adjusted r SquareStd. error of the
estimate
.739(a) .546 .546 1867.814
table C-3. OeCD Immigrants: Unemployment Benefits
c) Dependent Variable: Ui BeneFitS
Unstandardized
Coefficients
Standardized
Coefficients t Sig.
B Std. error Beta(Constant) 1080.376 41.210 26.217 .000
Age -12.850 .546 -.118 -23.540 .000
LMArrieD 98.440 18.704 .025 5.263 .000
CITIZ 17.021 19.370 .004 .879 .380
totAL inCoMe -.001 .000 -.013 -2.934 .003
HoUSeHoLD
Size-14.984 6.467 -.012 -2.317 .020
r r Square Adjusted r SquareStd. error of the
estimate
.113(a) .013 .013 1770.338
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
59
table C-4. Non-OeCD Immigrants: total transfers
a) Dependent Variable: ttLgoVrtrAnSFrPAyMnt
Unstandardized
Coefficients
Standardized
Coefficients t Sig.
B Std. error Beta(Constant) -1445.325 104.139 -13.879 .000
Age 135.107 1.423 .480 94.933 .000
LMArrieD -1276.878 52.603 -.118 -24.274 .000
CITIZ 517.295 54.205 .047 9.543 .000
totAL inCoMe -.003 .001 -.011 -2.347 .019
HoUSeHoLD
Size-219.868 14.358 -.076 -15.313 .000
r r Square Adjusted r SquareStd. error of the
estimate
.518(a) .269 .269 4285.853
table C-5. Non-OeCD Immigrants: pension transfers
b) Dependent Variable: oAS PenSion giS
Unstandardized
Coefficients
Standardized
Coefficients t Sig.
B Std. error Beta(Constant) -1997.320 45.129 -44.258 .000
Age 89.514 .617 .630 145.141 .000
LMArrieD -1108.122 22.796 -.203 -48.611 .000
CITIZ 352.108 23.490 .064 14.990 .000
totAL inCoMe -.010 .001 -.082 -19.531 .000
HoUSeHoLD
Size-112.868 6.222 -.078 -18.140 .000
r r Square Adjusted r SquareStd. error of the
estimate
.679(a) .461 .461 1857.293
Don J. DeVoretz & Sergiy PiVnenko
60
table C-6. Non-OeCD Immigrants: UI Benefits
c) Dependent Variable: Ui BeneFitS
Unstandardized
Coefficients
Standardized
Coefficients t Sig.
B Std. error Beta(Constant) 767.451 43.073 17.818 .000
Age -10.164 .589 -.101 -17.268 .000
LMArrieD 230.358 21.757 .060 10.588 .000
CITIZ -6.083 22.420 -.002 -.271 .786
totAL inCoMe .002 .000 .025 4.421 .000
HoUSeHoLD
Size-8.571 5.939 -.008 -1.443 .149
r r Square Adjusted r SquareStd. error of the
estimate
.111(a) .012 .012 1772.664
tHe eConoMiC DeterMinAntS AnD ConSeQUenCeS oF CAnADiAn CitizenSHiP ASCenSion
61
appendix D
Calculation of taxesage>14 and age<76
Calculation of income tax (Census of Canada 1996):if totincp<6,457 then inctax=0if totincp<29,591 and totincp>6,456 then inctax=.17*totincp-chdbnpif totincp>29,590 and totincp<59,181 then inctax=5,030+.26*(totincp-29,590)-chdbnpif totincp>59,180 then inctax= 12,724+.29*(totincp-59,180)-chdbnpwhere totincp – total income, chbnp – child benefits.
Calculation of average propensity to Consume (FaMeX 1995):APC=(total current consumption- rent- food from stores- health care- child care)/household income after taxes
Average Propensity to Consume
Meanage 15-24 .98
age 25-34 .85
age 35-44 .74
age 45-54 .70
age 55-64 .73
age 65-75 .63
Calculation of total taxes (Census of Canada 1996):Salestax= .14*APC*(totincp-inctax)gSt= .07*APC*(totincp-inctax)totfedtax=inctax+gSt
tottaxes=inctax+salestax.
63
A Dutch Narrative
Naturalization: Weighing the pros and consSometimes people have to decide on their own naturalization. In order to make up their minds, they have to carefully weigh the pros and cons. The decision is also dependent on their country of origin to determine whether naturalization will be advantageous or not. In this respect, the attractive-ness of a European passport is relative and a decision on naturalization is, therefore, not an easy one. The arguments that I will use in this short essay are relevant for someone like me, who is coming from a non-European country.
The cons of naturalization1) Naturalization means that one loses his original nationality, and such a loss inevitably provokes a variety of thoughts and sentiments. Apart from the emotional factors, loss of nationality can produce material drawbacks at certain moments or in specific situations, such as when visiting one’s country of origin.2) The feeling of betrayal of one’s own culture. Those who prefer another nationality may have good reasons for their choice; nonetheless, the feeling of disloyalty may prevail. Everything the immigrant always believed in has been traded for something that appears to be ‘better’. Persons who prefer not to change their nationality (family members, friends, those who stay in the country of origin) may also see naturalization as disloyalty. When it comes to choosing a European nationality, the superiority of such nationality may seem evident in the eyes of Europeans. They take into con-sideration issues such as the political system, religion, women’s position in society, the legal system, etc. Non-Europeans, on the other hand, may be convinced that their political system, religion, legal system and so forth are superior. This conviction is similarly undermined by their choosing a Euro-pean nationality.3) The naturalized person may have problems with his own identity. An Arab becoming a European citizen may be confronted with the question of who or what he actually is in the world. Is he still an Arab after naturaliza-tion, or has his identity changed to that of a European? The personal pro-blems arising from these feelings may be compared to the psychological problems of someone who has experienced a sex change. The comparison is not perfect, but those who change their citizenship probably have already lived in another country, with another culture and another language, for a long time. Although naturalization will be preceded by integration proces-ses, it goes without saying that those who naturalize will not forget their culture, language, and identity. In this sense, the transition is not without problems.
64
The pros of naturalizationAs stated before, most people who decide to change their citizenship will have good reasons to do so, depending on their country of origin. In my case, the choice was to accept the Dutch nationality. Why make this deci-sion if one has come as a refugee to the Netherlands? 1) Since it is a European passport, a Dutch passport expedites interna-tional travel. 2) Naturalization improves the labour market chances of those who are highly educated. Those with Dutch nationality have all the rights that Dutch natives are entitled to.3) Naturalization implies an accepted formal status. That is why a foreig-ner with a Dutch passport has a better status than someone without Dutch citizenship. Although discrimination will actually not diminish, Dutch citi-zenship makes it easier to use the justice system to fight discrimination. 4) Naturalization may have advantages for the (potential) family of immi-grants with Dutch citizenship. For example, if the person in question has children, those children will automatically be granted Dutch nationality. 5) Naturalization implies increased voting rights. Those with Dutch natio-nality can vote in every election (not only at the municipal level) and have the right to be political candidates.
ConclusionLooking at the pros and cons of the naturalization of immigrants who come from a non-European country to a European country, one gets the impression that the advantages outweigh the disadvantages. The home country’s characteristics, such as the political system, the dominant reli-gion, the culture, etc., as well as international relations affect the balance of pros and cons. The decision to naturalize may therefore be quite diffe-rent if one comes from the USA, Australia or another European country. Indeed, personal circumstances also affect the decision. That is why refu-gees, asylum-seekers and those who are persecuted in their own country will benefit from naturalization. Although there might be some disadvanta-ges, naturalization also has an important symbolic meaning. It shows the positive attitude of the immigrant towards the receiving country.
65
Chapter 2
NatUraLIZatION aND SOCIOeCONOMIC INteGratION: the CaSe OF the NetherLaNDS
Pieter Bevelander & Justus Veenman
Introduction: Issues and Stylized FactsSince the middle of the last century, the Netherlands has received an increa-sing number of immigrants, turning it into a net-immigration country. Over the last two decades, immigrants to the Netherlands were largely refugees. Since the number of new citizens increased more dramatically in the 1990s than in other EU-countries, the issue of naturalization was brought into the limelight. For instance, in 1997 360,000 people obtained another nationa-lity in one of the EU countries. After France (84,000) and Germany (83,000), the Netherlands had the greatest number of naturalizations (60,000), followed by Great Britain (28,000), Sweden (25,000) and Belgium (22,000). In fact, in 1997 France, Germany and the Netherlands accounted for two-thirds of the EU total number of naturalizations (Van der Erf 2000: 1-2). The largest proportion of inhabitants who could have opted for naturalization was found in the Netherlands (almost 9%) and Sweden (5.5%), with France (1.5%) and Germany (1.1%) with the smal-lest numbers (Van der Erf 2000: 2).
In general, the probability of immigrants ascending to citizenship depends on their legal status and their individual characteristics. The ability to obtain Dutch citizenship depends, among other things, on the immigrant’s individual characteristics such as age, duration of legal residence, conduct, and degree of integration into Dutch society. However, in particular, the presence or absence of dual citizenship rights has affected naturalization rates in the Netherlands. Unless dual citizenship was prevented by law in the country of origin (as in the case of Morocco), those who opted for Dutch citizenship lost their original nationality. This general requirement became more flexible after 1992 in anticipation of a proposed change in
PIETEr BEVElANDEr & JUSTUS VEENMAN
66
the citizenship law; the latter, however, was not enacted by the Dutch Par-liament, and naturalization rules reverted to the strict policy of single citi-zenship after October 1997. The effect of a more flexible application of the citizenship act shows clearly in the figures on naturalization in the beginning of the 1990s. After a remarkable rise, especially between 1994 and 1996, the number of natu-ralizations decreased.
Figure 1. Naturalization as percent of foreign population for selected european Countries
Source: SOPEMI, 2004.
At the end of the 1990s, the decrease in naturalization rates as a conse-quence of the stricter application of the rules on dual citizenship was offset by the increased naturalization rate of refugees who had arrived in the Netherlands in the early 1990s, in particular from the former Yugoslavia and Somalia. Although the number of refugees clearly affected naturaliza-tion rates and accounted for 45% of all naturalizations in the Netherlands in 1998, non-refugees, such as Turks (23%) and Moroccans (19%) still obtained the highest rates of naturalization by country of origin (Van der Erf 2000). This is why we include these immigrant groups along with refu-gees in our analysis.
0
2
4
6
8
10
12
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
NetherlandsGermanySwedenBelgiumUnited KingdomNorway
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
67
Dutch Naturalization Debate
From the Dutch viewpoint, naturalization reflects a change in the immigrant’s identity and ability to adapt to the Dutch culture, and this commitment, in turn increases their socioeconomic integration. In this case, the individual immigrant in the Dutch context is signaling a long-term commitment to the Netherlands. In September 2004, the Dutch Immigra-tion Minister, Mrs. rita Verdonk, expressed the symbolic meaning of natu-ralization in Parliament by saying that giving up nationality of the home country is ‘a sign of loyalty to the host country’ (newspaper De Volkskrant, September 3, 2004). The Minister’s statement actually implied that holding dual citizenship was not desirable.
This negative view of dual citizenship became a real issue in the Dutch Parliament when the new Christen/Social Democrat government came into power in 2007 since two State Secretaries had dual citizenship (both Dutch and Moroccan/Turkish respectively). The leader of the right-wing political party PVV, Geert Wilders, wrote on the web log Geenstijl.nl that ‘a blind man can see that dual nationality implies dual loyalty.’ In his opinion, foreign governments infiltrate ‘the heart of the Dutch power system’ if members of the Dutch cabinet have a non-Dutch nationality. Since he feared a conflict of interests, Wilders proposed a vote of no-confidence based on the two State Secretaries’ dual citizenship status. The Speaker of Parliament interfered and advised against the motion. In later debates it became clear that some other political parties also wanted to establish stricter rules on dual nationality, as the number of those with dual nationa-lity rose rapidly in the Netherlands (from 400.000 in 1995 to 1 million in 2007). The current debate has led the Minister of Justice to prepare a change in the naturalization law to force children with two passports who had been living in the Netherlands for more than five years to give up one of the nationalities.
Stylized Facts
The charged political environment following the Van Gogh murder resul-ted in the Dutch government adopting both more restrictive immigration and naturalization policies at the beginning of the 21st century. Currently one can obtain Dutch nationality in two ways, other than by birth. The first is the option procedure for adults 18 years of age and older (a) born in the Netherlands, the Dutch Antilles or Aruba, who have lived there conti-nuously since birth, or for at least three years for stateless individuals, or (b) who have legally lived in the Netherlands, the Dutch Antilles or Aruba since age 4, or (c) who are former Dutch citizens and have lived in the Netherlands, the Dutch Antilles or Aruba for at least one year on a perma-
PIETEr BEVElANDEr & JUSTUS VEENMAN
68
nent residence permit or a residence permit for non-temporary residents, or (d) who have been married to a Dutch national for at least three years and have legally lived in the Netherlands for an uninterrupted period of at least 15 years, or (e) who are 65 years of age or over and have legally lived in the Netherlands for an uninterrupted period of at least 15 years. The option procedure is also applicable to minors in the custody of a Dutch citizen and who have been cared for and raised by this Dutch citizen for an uninter-rupted period of at least three years.1
People are eligible for naturalization, the second avenue to obtain Dutch citizenship, if they have a residence permit and in addition meet all the following conditions: they must (a) be of age, (b) have lived in the Netherlands, the Dutch Antilles or Aruba for an uninterrupted period of 5 years with a valid residence permit, (c) have integrated into Dutch society and be able to read, write, speak and understand Dutch to the standard of the naturalization test2, (d) have no record of a custodial sentence or a sub-stantial monetary fine, and (e) be prepared to renounce their other nationa-lity. If immigrants do not give up their other nationality, the Dutch natio-nality is not granted. There are, however, some exceptions to giving up one’s nationality due to the nationality law in the immigrants’ country of origin, or to the immigrants’ inability to contact the authorities of the country of which they are nationals.3
ImmigrationWith the exception of 1967, the Netherlands has been a net-immigration country since the 1960s. From that time until the 1990s, immigration was dominated by four groups; the largest (about 70 percent of the immigrants and their descendents) came from Turkey and Morocco, and Suriname and the Dutch Antilles. The latter still are Dutch colonies and, therefore, most Surinamese immigrants and all Antilleans have Dutch nationality. For this reason we exclude the latter and focus on the migration history and natu-ralization rates of Turks and Moroccans. In the 1990s, the immigration patterns changed as refugee flows increased. Given this switch in entry categories we will also describe the entry and naturalization rates of the
1 The same condition applies to minors under the joint custody of one Dutch and one non-Dutch parent.2 An exception will be made for those who have successfully completed an integration course (at NT2-level 2) or have attained another diploma. As of 15 March 2006 a civic integration examination abroad was introduced. Since then migrants wishing to settle in the Netherlands for, in particular, the purposes of marrying or forming a relationship, have been required to take the examination. In addition, their partner has to make at least 120% of the minimum income.3 Other exceptions apply when the immigrant is a national not recognized by the Nether-lands, or the immigrant is originally an Italian or French national and part of one of the target groups of the so-called Second Protocol.
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
69
five largest refugee groups in the Netherlands: Afghans, Iraqis, Iranians, former Yugoslavs and Somalis.4
The first immigrants from Turkey arrived in the beginning of the 1960s via Germany and Belgium, relieving the unskilled and low-skilled labour shortages in the Netherlands. The migration movement became institutio-nalized in 1964 when a labour recruitment agreement between the Nether-lands and Turkey was signed. The number of Turkish labour migrants, mainly target earners, increased rapidly and peaked in the early 1970s. At the end of 1973, labour market recruitment came to an end when Turks were no longer admitted as labour migrants. This, however, did not dimi-nish the number of immigrants from Turkey to the Netherlands, since increased family reunification and subsequent family formation continued unabated. The rise in Turkish immigration since the mid-1980s was caused by the children of the labour migrants marrying spouses from Turkey. Furthermore, since the 1970s the number of Turkish refugees to the Nether-lands has risen. By 2004 the number of Turks in the Netherlands was approximately 351,000 of which 45 percent were born in the Netherlands.
In the 1960s, the immigrant flow from Morocco to the Netherlands numbered 3,000 yearly until the first oil crisis in 1973. These Moroccans were mostly men with little education recruited for unskilled work in clea-ning companies and in the metal, timber, and food industries. The Dutch hoped their presence would only be temporary, and expected that many would return and start their own business at home with the savings derived from working in the Netherlands. This did not occur and, after 1973, labour recruitment was stopped by the Dutch government and family reu-nification of women and children became the primary form of Moroccan immigration. The majority of the Moroccan immigrants came from the rif area, one of the more traditional parts of the country. This explains their low educational level and the fact that they were not prepared to partici-pate in the Dutch society. Family reunification was supplanted by marriage migration in the first half of the 1980s and this group had a higher educa-tional attainment than earlier Moroccan immigrants. As a consequence of all these flows, approximately 306,000 Moroccans lived in the Netherlands circa 2004, of which almost 46 percent are second-generation Moroccans.
Afghan refugees to the Netherlands came mainly during the second half of the 1990s. From 1994 to 2002, asylum-seekers from Afghanistan were given residence permits without further scrutiny. Since 2002, refugee applications by persons from Afghanistan have been individually exami-ned. The Afghan population in the Netherlands is relatively young and less
4 Clearly refugees have a smaller incentive to return and perhaps a greater incentive to naturalize.
PIETEr BEVElANDEr & JUSTUS VEENMAN
70
educated than the Dutch population. The low degree of economic integra-tion in this group is also due to their relatively short stay in the Nether-lands. In 2004, 36,043 individuals from Afghanistan and their children lived in the country. The so-called ‘second generation’ constituted 11 per cent of the total Afghan population in the Netherlands.
refugees from Iraq came mainly in the second half of the 1990s, partly as quota refugees sent by UNHCr and partly as spontaneous refugee arri-vals and reunified family members. The Iraqi population in the Nether-lands increased sevenfold between 1996 and 2004. It is a young refugee group, like the Afghan population. The Iraqis are overrepresented in lower educational categories compared to Dutch natives, and only 45 percent of the Iraqi population aged 15 to 64 is gainfully employed. In 2004, 42,931 persons with an Iraqi background were reported in the Dutch population register, with 16 percent second-generation Iraqi residents.
Iranians started to arrive after the ‘Islamic revolution’ of 1979. The war with Iraq also increased the number that fled Iran. Those who were not accepted as refugees could often stay on conditional residence permits. Since 1995 a more restrictive asylum policy has led to a decrease in asylum-seekers from Iran. Compared to Dutch natives, Iranians have a high educa-tional profile with many holding university-level degrees. The total number of Iranians and their descendants in the Netherlands was 28,438 in 2004, with the second generation accounting for 16 percent of this number.
The migration background of former Yugoslavs in the Netherlands is diverse. This group consists of both earlier labour migrants who came to the Netherlands in the 1960s and 1970s and refugees from different parts of the former Yugoslavian republic who arrived in the early 1990s. In 2004, 76,346 individuals with roots in the former Yugoslavia were living in the Netherlands, of whom 28 percent were second-generation. The edu-cational profile of the group is somewhat lower than for Dutch natives and the labour participation of both males and females is just below that of natives.
Somali immigration to the Netherlands grew in the 1990s due to the civil war in Somalia. The more restrictive Dutch asylum application proce-dure introduced in 1995 led to a decrease in the number of asylum requests in subsequent years. A relatively large portion of asylum-seekers are minors, thus the Somali population is relatively young. The total number of Somalis in the Netherlands was 25,001 in 2004, of which 31 percent belonged to the second generation. The Somali group in the Netherlands has a lower educational level than natives and labour force participation of only 41 per cent.
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
71
In Figure 2, we plot both the number of immigrants to the Netherlands and the number of naturalizations per year since 1970. Figure 2 shows that the number of naturalizations increased until 1997 and then dropped, but naturalizations still remained greater than before the 1980s.
Figure 2. total number of naturalizations and number of immigrants, 1970-2004
Source: Statline, Statistics Netherlands.
Turkish and Moroccan immigrants of the 1970s and 1980s constituted a large share of the observed naturalizations. While in the 1980s more Moroccans than Turks naturalized, between 1992 and 1998 the opposite was true, as shown in see Figure 3. This temporary increase in naturaliza-tion among Turkish nationals was caused by the change in the conditions surrounding dual citizenship.
0
20000
40000
60000
80000
100000
120000
140000
160000
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
NaturalisationImmigration
PIETEr BEVElANDEr & JUSTUS VEENMAN
72
Figure 3. Number of naturalizations among turks and Moroccans, 1980-2003
Source: Statline, Statistics Netherlands.
Figure 4 reports the number of naturalizations among the five refugee groups from 1996 to 2003. It is clear that refugee naturalizations increased in the late 1990s and dropped after 2000. The naturalization increase among immigrants from the former Yugoslavia is similar to the one repor-ted for the Turks in Figure 3. This rise probably also relates to the more flexible dual citizenship policy applied between 1992 and 1997. In addi-tion, it reflects the long residency period of the majority of these refugees by the early 1990s.
0
5000
10000
15000
20000
25000
30000
35000
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
TurkeyMorocco
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
73
Figure 4. Number of naturalizations among refugees, 1996-2003
Source: Statline, Statistics Netherlands.
The high absolute numbers of naturalizations depicted in Figure 4 resulted in high rates of ascension to Dutch citizenship per immigrant group for 2002 (Turks and Moroccans) and 2003 (refugee groups), as reported by gender in Table 1.
table 1. Naturalization rates by gender per immigrant group, 2002-2003
Males Females
Afghanistan 64 67Somalia 80 83Iran 89 90Iraq 91 80Former Yugoslavia 72 75Turkey 56 50Morocco 66 57
Source: SPVA 2002 and 2003, ISEO/SCP/NIDI.
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
1996 1997 1998 1999 2000 2001 2002 2003
Former YugoslaviaSomaliaAfghanistanIran Iraq
PIETEr BEVElANDEr & JUSTUS VEENMAN
74
As shown in Table 1, refugee groups from Somalia, Iran and Iraq naturali-zed at a very high rate. Even though immigrants from Turkey and Morocco had lived longer in the Netherlands, they experienced lower rates of natu-ralization. Immigrants from the former Yugoslavia, who had mixed motives for migrating to the Netherlands, have a naturalization rate ‘in-between’ the high rates for refugees and the lower rates reported for the longer-stay-ing Mediterranean immigrants. For their part, both male and female refu-gees from Afghanistan have a lower naturalization rate than other refugee groups. This is due to the fact that most Afghan immigrants had not lived in the Netherlands for the required minimum of at least five years and, the-refore, were not eligible for Dutch citizenship. Across genders, Turks, Moroccans and Iraqi males naturalized at higher rates than females, whereas Iranians show the highest naturalization rate of all reported ethnic groups.
Context and LiteratureIf we assume that the decision to acquire another nationality is mainly based on rational calculations, one should know the net balance of advan-tages and disadvantages derived from naturalization to fully understand the phenomenon. While the immigrants’ loss of nationality is often consi-dered the primary cost of naturalization, this can be offset by the ability to obtain a Dutch passport and, therefore, to travel freely within the EU. The second positive consequence derived from naturalization is the right to vote in Dutch and EU general elections.5 Naturalization also opens up employ-ment opportunities in the civil service, the police, the justice system and the military, areas reserved exclusively for Dutch nationals.
Although a cost/benefit-analysis offers an interesting perspective to the naturalization decision, Yang (1994) argues that this is not the starting point of most studies on naturalization. research often focuses on the degree of immigrant integration as a determinant of naturalization. Yang thus distinguishes two scholarly traditions: one stresses the role of socio-economic achievements in the naturalization process, while the other emp-hasizes the importance of the immigrants’ cultural adaptation to the host society as well as their demographic characteristics. Yang notes that, despite their differences, both research traditions use immigrants’ characteristics as predictors of the probability of naturalization and both view naturalization as an outcome of the immigrants’ successful integration into the receiving country.
A shortcoming of both traditions is the omission of a cost/benefit-ana-lysis to predict naturalization. As stated previously, costs could be attribu-
5 Non-EU legal inhabitants of the Netherlands without Dutch nationality are only entit-led to vote and seek office in municipal elections.
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
75
ted to loss of citizenship status in the country of origin, which in turn implies the possible loss of a claim on inheritance or real estate in the country of origin. In addition, other costs are incurred including the time and money invested in Dutch language courses and tests before immigrants are allowed to start the naturalization procedure.
In sum, this chapter will shed light on the relationship between integra-tion and naturalization by answering the following questions: (1) Which causal factors influence the naturalization decision? and (2) Does naturali-zation affect employment opportunities and immigrant earnings and there-fore raise immigrants’ socioeconomic status?6 We should note that we do not have panel data or information on the moment of naturalization. Thus, given limitations on data availability, it is impossible to pinpoint exact causal relationships.
In the Netherlands, research on immigrant labour market integration is extensive, but the causes and consequences of naturalization have rarely been studied from an economic perspective. Only Bevelander and Veenman (2006) have examined the relationship between naturalization and the socioeconomic integration of Turks and Moroccans. Using a multivariate analysis they found only a weak relation between cultural integration and naturalization. Controlling for various individual characteristics, they reported that age, age at migration, and education influenced Turkish and Moroccan immigrants’ propensity to naturalize. In addition, gender affects the likelihood of obtaining Dutch citizenship since Turkish and Moroccan women who score high on a ‘modernization’ scale and Moroccan women who identify themselves as ‘Dutch’ were more likely to naturalize.
Method and DataAs noted previously, our research focuses on two aspects of naturalization: (a) the relationship between demographic factors and socioeconomic inte-gration and naturalization, and (b) the effect of naturalization on the employment opportunities and relative earnings of immigrants. Our analy-sis will therefore be twofold. We will first elaborate on the probability of obtaining Dutch citizenship depending on immigrant socioeconomic and demographic characteristics. Second, we will analyze employment oppor-tunities and wages as they relate to Dutch citizenship while controlling for some other individual characteristics. We insist that it will be difficult to
6 The extent to which rational calculations of costs and benefits determine the naturali-zation decision will vary across individuals. This variation can explain individual diffe-rences in the request to naturalize. Unfortunately, we do not have data that contain infor-mation on the perceived costs and benefits. We are therefore not able to study the effects of these factors directly, but will try to take these factors into account in our explanation of naturalization.
PIETEr BEVElANDEr & JUSTUS VEENMAN
76
discern causal relationships since panel data are unavailable and since we only have information on the year of naturalization for the refugee sample.
The operationalization of the three core variables is as follows: citizen-ship status is divided into having or not having obtained Dutch citizenship. Employment, the first dependent variable in the second part of the analy-sis, is divided into being unemployed and having any kind of employment. The income variable, the second dependent variable in this part of the ana-lysis, is the natural logarithm of monthly wages.
In keeping with Yang (1994), the analysis of Dutch citizenship ascen-sion includes as independent variables the educational level as well as demographic variables (age and gender), and immigrant-specific variables (country of birth, reason for migration, years in the host country, studies in the Netherlands or in the home country, and participation in an ‘integra-tion program’).7 We expect that large numbers of years of residence, high educational levels, and participation in an ‘integration program’ (all indi-cators of integration in Dutch society) will lead to higher rates of naturali-zation. Study programs completed in the Netherlands, not in the home country, should lead to further integration into Dutch society and the labour market. In turn, this should have a positive effect on the probability of ascending to Dutch citizenship.
As far as the variable ‘reason for migration’ is concerned, earlier studies have shown that refugees obtain the citizenship of their country of adop-tion to a larger extent if they have lost their home country citizenship or are not allowed to return (Bevelander 2000). This may indicate that those who migrated for political reasons will be more likely to obtain Dutch citi-zenship than other immigrants. In addition, one might expect that those who came to the Netherlands to work or to study, may not intend to stay permanently. Therefore, these two classes of immigrant, therefore, will have a lower probability of ascending to Dutch nationality.
In the employment and earnings analysis, age, gender, country of birth, years of residence, reason for migration, educational level, studies in the Netherlands, participation in an ‘integration program’, citizenship status, and years since acquiring Dutch citizenship are used as explanatory variab-les.8
7 Such programs have been compulsory since 1998, except for immigrants less than 15 years old or those who meet the language and knowledge requirements upon arrival in the Netherlands. This implies that participation in an integration program only repre-sents a true differential effect for those who participated in these programs before 1998. Unfortunately, we have no information on the year of participation. Correlation coeffi-cients of 0.05 and 0.14 were measured for the variables naturalization, participation in ‘integration’ program and employment, indicating no endogeneity between the variables.8 Weekly hours worked is included as a control variable in the monthly income regres-sions.
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
77
Again, in keeping with Yang (1994), the statistical analysis will use logistic regression techniques to predict the effect of various variables on the probability of obtaining Dutch nationality and the probability of being employed. Next, we use OlS regressions to estimate the log-linear earnings model. Both descriptive and earlier research on citizenship ascension, employment and earnings of these groups will be presented and used to analyze the results of the statistical analysis.
Our data source is the survey ‘Social Position and Use of Public Utili-ties by Immigrants’ (SPVA) for 2002 and 2003.9 These surveys carried out by the Institute for Sociological and Economic research of the Erasmus University in rotterdam, the Dutch Social and Cultural Planning Office and the Dutch Interdisciplinary Demographic Institute aim to describe and analyze the socioeconomic and cultural integration of the four largest immigrant groups and of the five largest refugee groups in the Netherlands. Given the geographical concentration of the migrant population, the SPVA consists of random samples of the population in thirteen cities, including the four largest cities in the Netherlands.10 In this paper we focus on seven groups: foreign-born (first generation) immigrants from Turkey, Morocco, Afghanistan, the former Yugoslavia, Iraq, Iran and Somalia. For our analy-sis we will use a weighted pooled sub-sample of the SPVA which comprises the population of these groups 18 to 64 years old and includes only those individuals who have been in the Netherlands for at least five years. This latter restriction arises from the waiting period inherent in the Dutch legis-lation to gain naturalization (see Stylized Facts).
resultsCitizenship acquisition
In Table 2, we present the results of four simple logistic regressions on the odds11 of obtaining Dutch nationality for immigrants and refugees. The first model with all immigrant groups included (first column) shows that a combination of higher educational attainment and having completed studies in the Netherlands significantly increases the probability of obtain-ing Dutch nationality. Contrary to what we expected, no significant effect is found for gender and years of residence (more than 5 years) variables. When we restrict our analysis to refugee groups (second column) and include the variables ‘integration program’ and migration reason, we find no significant effect derived from participation in an ‘integration program’.
9 See Appendix for sample values of selected variables.10 Further detailed information on the SPVA can be found in Groeneveld and Weijers-Martens (2003) and Van den Maagdenberg (2004).11 The odds are defined as p/(1-p), in which p= the probability of ascending to citizen-ship.
PIETEr BEVElANDEr & JUSTUS VEENMAN
78
refugees who have migrated because of war or for political reasons, as well as family-reunion migrants, though, have a significantly greater prob-ability of obtaining Dutch citizenship. Women from refugee countries also have a greater probability of ascending to Dutch citizenship. Finally, increasing years of residence is also positively and significantly related to Dutch citizenship acquisition for refugees.
Model 2 in Table 2 tests for the effect of differential entry group cate-gories on the probability of ascending to Dutch citizenship. The model is again subdivided into all countries (column 3) and into refugee groups and some extra explanatory variables (column 4). In the regression including all countries, we find that most variables noted above are significant inclu-ding years of residence. When we include country of birth, we find that refugees have a higher probability of obtaining Dutch citizenship than Mediterranean immigrants. This confirms the differences in the naturaliza-tion rates observed in Table 2. The analysis that only includes refugee groups shows once more that participation in an ‘integration program’ has no significant effect on the probability of acquiring Dutch citizenship.
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
79
Model 1 Model 2All groups Refugee
groupsAll groups Refugee
groupsAge 1.165*** 1.160*** 1.198*** 1.149***
Age squared 0.998*** 0.998*** 0.998*** 0.998***
SexMale 1 1 1 1
Female 0.978 1.306** 1.022 1.326***
Years of residence 0.994 1.111*** 1.054*** 1.107***
Educational levelPrimary Education 1 1 1 1
lower secondary Education
1.377*** 1.182 1.324** 1.224
Higher secondary Education
1.618*** 1.186 1.393*** 1.180
University education 2.020*** 1.413** 1.658*** 1.443**
Education home/ destination countryHome country 1 1 1 1
Netherlands 2.035*** 1.862*** 1.715*** 1.792***
Participation in ‘integration program’No participation 1 1
Participation 1.114 1.021
Reason for migrationlabour market/ Education
1 1
War/Political reasons 7.172*** 6.179***
Family reunion 5.244*** 4.368***
Home countriesAfghanistan 1 1
Somalia 2.627*** 2.144***
Iran 3.297*** 2.896***
Iraq 2.174*** 2.177***
Former Yugoslavia 1.144 1.191
Turkey 0.347***
Morocco 0.523***
Constant 0.088*** 0.004*** 0.033*** 0.004***
Number of observations
6795 3867 6795 3867
Degrees of freedom 8 11 14 19
-2 log Likelihood 7950,388 3399,824 7458,085 3313,361
table 2. Odds ratio of obtaining Dutch nationality on selected variables
Note: ***, ** indicate significance at 0.01 and 0.05 level respectively.
Source: SPVA 2002, ISEO/SCP/NIDI.
PIETEr BEVElANDEr & JUSTUS VEENMAN
80
In summary, our analysis reveals that the educational level of immigrants and studying in the Netherlands have a strong impact on the probability to natura-lize. Contrary to what we expected, participation in an ‘integration program’ does not increase the log odds to naturalize. Years of residence and being female have a significant and positive influence on the odds of obtaining Dutch citizenship. The finding that women are more likely to obtain Dutch citizen-ship contradicts Yang (1994) and other studies in this volume. A regression including the interaction between gender and immigrant groups showed that only women from Turkey have a significantly high probability of acquiring Dutch citizenship.12 Moreover, we found a highly significant probability to naturalize when the immigrant came to the Netherlands for political/war reasons or was motivated by family reunion. In fact, those who moved to the Netherlands for political, war and family reasons are more inclined to natura-lize than those who moved for labour market or educational reasons.
We expected this outcome, as those who migrate for labour or educa-tional reasons probably do not intend to stay permanently in the Nether-lands. Therefore, if they do stay 5 years or more, they are less apt to apply for naturalization to the same degree as other immigrants.
Citizenship and employment
We now explore if naturalization affects the labour market outcomes of immigrants. We therefore wish to answer the question of whether there are differences in employment opportunities between immigrants with non-Dutch nationality and immigrants with Dutch or dual nationality. Bevelan-der (2000) and others expected a positive relation between naturalization and employment integration. Table 3 shows the employment rate by gender and citizenship per immigrant group. A clear difference in employment rates arises between immigrants of all groups with Dutch nationality and those without. Dutch citizenship appears to be correlated with high employment rates.
Since individual characteristics could influence the results depicted in Table 3, we proceed with an analysis of employment opportunities for the various immigrant groups with different citizenship status while control-ling for individual characteristics. Table 4 shows that in all regressions the variables age and age squared have a significant effect. A high educational level coupled with studies completed in the Netherlands increase the log odds of gaining employment. According to the results in column 2, labour migrants have significantly higher log odds of obtaining employment in the Dutch labour market than immigrants who came because of war, political and family reasons.13 In column 3 we expand the analysis by including the
12 results available upon request.13 Yugoslavs who arrived in the 1960s and 1970s are included in the category of labour migrants.
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
81
variable ‘years since citizenship’. The effects of changing citizenship and years of residence are not significant in this model, whereas the variable ‘years since citizenship’ is.
table 3. employment rate by gender and citizenship, per immigrant group
Non-Dutch
nationality
Dutch
nationality
Non-Dutch
nationality
Dutch
nationality
Males FemalesAfghanistan 49 56 8 26
Somalia 36 53 0 16
Iran 25 67 14 51
Iraq 20 49 8 23
Former Yugoslavia 58 71 47 59
Turkey 59 64 21 36
Morocco 47 62 14 34
Source: SPVA 2002 and 2003, ISEO/SCP/NIDI.
In columns 4 to 6 we estimate the same models as in the first three columns but now include the various immigrant and refugee groups as variables. The results from these regressions show once more the important effect of education on the probability of being employed in the Nether-lands. Citizenship acquisition has a positive and significant effect on job opportunities. The results further show that immigrants from Somalia and Iraq are the least likely to be employed, while those from former Yugosla-via and Turkey the most likely.
Next we test for the effects of interaction variables between gender and citizenship, and immigrant group and citizenship. The results indicate that women with the Dutch nationality have a greater chance of being employed than women without Dutch citizenship. Individuals from Iran and Iraq with the Dutch nationality are more likely to be employed than those without and those who belong to other immigrant groups. In addition, separate country regressions show that the naturalization effect is positive for immigrants from all countries. Once more, the results derived from the interaction variables which capture the joint effects of gender and citizen-ship, and immigrant group and citizenship, indicate that women with the Dutch nationality, except for the Iraqis, have a greater chance of being employed than women without Dutch citizenship. For males, the regres-sion results show that only naturalized males from Iran and the former Yugoslavia have a higher probability to be employed than those without the Dutch citizenship.14
14 results available upon request.
PIETEr BEVElANDEr & JUSTUS VEENMAN
82
All
grou
ps
Ref
ugee
grou
ps
Ref
ugee
grou
ps
All
grou
ps
Ref
ugee
grou
ps
Ref
ugee
grou
ps
Age
1.14
8***
1.11
6***
1.13
7***
1.13
4***
1.11
6***
1.14
4***
Age
squ
ared
0.99
8***
0.99
8***
0.99
8***
0.99
8***
0.99
8***
0.99
8***
Sex
Mal
e1
11
11
1
Fem
ale
0.31
0***
0.39
2***
0.36
8***
0.27
6***
0.35
5***
0.33
2***
Yea
rs s
ince
mig
rati
on1.
042*
**1.
093*
**1.
024
1.02
1***
1.07
4***
1.00
7
Edu
cati
onal
leve
lPr
imar
y E
duca
tion
11
11
11
low
er s
econ
d. E
duc.
1.98
5***
1.89
6***
1.89
8***
1.73
0***
1.63
6***
1.63
8***
Hig
her
seco
nd. E
duc.
2.01
1***
1.88
8***
1.93
1***
1.84
5***
1.66
8***
1.68
6***
Uni
vers
ity
educ
atio
n2.
651*
**2.
521*
**2.
365*
**2.
649*
**2.
371*
**2.
176*
**
Edu
cati
on
Hom
e co
untr
y1
11
11
1
Net
herl
ands
1.65
4***
1.58
7***
1.49
7***
1.63
4***
1.53
9***
1.45
4***
Cit
izen
ship
Hom
e co
untr
y1
11
11
1
Dut
ch1.
468*
**1.
779*
**2.
021
1.64
9***
1.98
5***
3.15
7
Yea
rs s
ince
Cit
izen
ship
1.08
2***
1.08
5***
Part
icip
atio
n in
”in
burg
erin
gs”
prog
ram
s
No
part
icip
atio
n1
11
1
Part
icip
atio
n0.
993
1.04
31.
008
1.07
6
Rea
son
mig
rati
onl
abou
r m
arke
t/E
duc.
11
11
War
/Pol
itic
al r
easo
ns
0.73
3*0.
789
0.91
40.
963
Fam
ily r
euni
on
0.73
70.
821
0.80
50.
822
Cou
ntri
esA
fgha
nist
an1
11
Som
alia
0.66
8***
0.56
9***
0.60
7**
Iran
1.53
6***
1.17
81.
153
Iraq
0.63
1***
0.61
0***
0.66
6**
Form
er Y
ugos
lavi
a2.
968*
**2.
069*
**2.
002*
**
Turk
ey1.
810*
**
Mor
occo
1.49
6***
Con
stan
t0.
056*
**0.
021*
**0.
013*
0.07
6***
0.02
2***
0.00
6**
Obs
erva
tion
s67
9538
6738
6767
9538
6738
67
D. o
f fr
eedo
m9
1213
1516
17
-2 lo
g lik
elih
ood
7920
,392
4182
,306
2744
,60
7674
,372
4053
,586
2674
,188
tabl
e 4.
Odd
s ra
tio o
f gai
ning
em
ploy
men
t on
sele
cted
var
iabl
es
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
83
All
grou
ps
Ref
ugee
grou
ps
Ref
ugee
grou
ps
All
grou
ps
Ref
ugee
grou
ps
Ref
ugee
grou
ps
Age
1.14
8***
1.11
6***
1.13
7***
1.13
4***
1.11
6***
1.14
4***
Age
squ
ared
0.99
8***
0.99
8***
0.99
8***
0.99
8***
0.99
8***
0.99
8***
Sex
Mal
e1
11
11
1
Fem
ale
0.31
0***
0.39
2***
0.36
8***
0.27
6***
0.35
5***
0.33
2***
Yea
rs s
ince
mig
rati
on1.
042*
**1.
093*
**1.
024
1.02
1***
1.07
4***
1.00
7
Edu
cati
onal
leve
lPr
imar
y E
duca
tion
11
11
11
low
er s
econ
d. E
duc.
1.98
5***
1.89
6***
1.89
8***
1.73
0***
1.63
6***
1.63
8***
Hig
her
seco
nd. E
duc.
2.01
1***
1.88
8***
1.93
1***
1.84
5***
1.66
8***
1.68
6***
Uni
vers
ity
educ
atio
n2.
651*
**2.
521*
**2.
365*
**2.
649*
**2.
371*
**2.
176*
**
Edu
cati
on
Hom
e co
untr
y1
11
11
1
Net
herl
ands
1.65
4***
1.58
7***
1.49
7***
1.63
4***
1.53
9***
1.45
4***
Cit
izen
ship
Hom
e co
untr
y1
11
11
1
Dut
ch1.
468*
**1.
779*
**2.
021
1.64
9***
1.98
5***
3.15
7
Yea
rs s
ince
Cit
izen
ship
1.08
2***
1.08
5***
Part
icip
atio
n in
”in
burg
erin
gs”
prog
ram
s
No
part
icip
atio
n1
11
1
Part
icip
atio
n0.
993
1.04
31.
008
1.07
6
Rea
son
mig
rati
onl
abou
r m
arke
t/E
duc.
11
11
War
/Pol
itic
al r
easo
ns
0.73
3*0.
789
0.91
40.
963
Fam
ily r
euni
on
0.73
70.
821
0.80
50.
822
Cou
ntri
esA
fgha
nist
an1
11
Som
alia
0.66
8***
0.56
9***
0.60
7**
Iran
1.53
6***
1.17
81.
153
Iraq
0.63
1***
0.61
0***
0.66
6**
Form
er Y
ugos
lavi
a2.
968*
**2.
069*
**2.
002*
**
Turk
ey1.
810*
**
Mor
occo
1.49
6***
Con
stan
t0.
056*
**0.
021*
**0.
013*
0.07
6***
0.02
2***
0.00
6**
Obs
erva
tion
s67
9538
6738
6767
9538
6738
67
D. o
f fr
eedo
m9
1213
1516
17
-2 lo
g lik
elih
ood
7920
,392
4182
,306
2744
,60
7674
,372
4053
,586
2674
,188
All
grou
ps
Ref
ugee
grou
ps
Ref
ugee
grou
ps
All
grou
ps
Ref
ugee
grou
ps
Ref
ugee
grou
ps
Rea
son
mig
rati
onl
abou
r m
arke
t/E
duc.
11
11
War
/Pol
itic
al r
easo
ns
0.73
3*0.
789
0.91
40.
963
Fam
ily r
euni
on
0.73
70.
821
0.80
50.
822
Cou
ntri
esA
fgha
nist
an1
11
Som
alia
0.66
8***
0.56
9***
0.60
7**
Iran
1.53
6***
1.17
81.
153
Iraq
0.63
1***
0.61
0***
0.66
6**
Form
er Y
ugos
lavi
a2.
968*
**2.
069*
**2.
002*
**
Turk
ey1.
810*
**
Mor
occo
1.49
6***
Con
stan
t0.
056*
**0.
021*
**0.
013*
0.07
6***
0.02
2***
0.00
6**
Obs
erva
tion
s67
9538
6738
6767
9538
6738
67
D. o
f fr
eedo
m9
1213
1516
17
-2 lo
g lik
elih
ood
7920
,392
4182
,306
2744
,60
7674
,372
4053
,586
2674
,188
Not
e: *
**, *
*, *
indi
cate
sig
nific
ance
at
0.01
, 0.0
5 an
d 0.
1 le
vel r
espe
ctiv
ely.
Sour
ce: S
PV
A 2
002,
ISE
O/S
CP
/NID
I.
PIETEr BEVElANDEr & JUSTUS VEENMAN
84
Citizenship and earnings
Our final analysis deals with the effect of naturalization on earnings, more precisely on monthly income derived from work. As discussed in other chapters of this book, naturalized immigrants are expected to have higher hourly wages and a larger monthly income than those who do not natura-lize. As shown in Table 5, the monthly income for men and women within the various immigrant groups is in most cases higher for those with the Dutch citizenship, except for Turkish males and Afghan females. Thus we conclude that, since monthly income is dependent on hours worked, we should control for this factor as in the multivariate analysis presented in Table 6.
table 5. Mean earnings (€) of employment for immigrants and refugees by gender and citizenship
Non-Dutch
nationality
Dutch
nationality
Non-Dutch
nationality
Dutch
nationality
Males FemalesAfghanistan 1002 1094 860 784Somalia 944 1165 - 840Iran 1282 1391 700 1080Iraq 1107 1284 1003 1007Former Yugoslavia 1346 1412 1072 1126
Turkey 1515 1510 899 925Morocco 1333 1405 748 1004
Source: SPVA 2002 and 2003, ISEO/SCP/NIDI.
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
85
table 6. OLS estimation of log-linear earnings model: citizenship effect on earnings
All groupsCoeff.
Refugee groupsCoeff.
All groupsCoeff.
Refugee groupsCoeff.
Age .057*** .066*** .057*** .066***Age squared -.001*** -.001*** -.001*** .001***Female -.124*** -.078*** -.144*** -.014***Years since migration
.009*** .015*** .009*** .002***
Educational levellow second. Educ. .066*** .032 .045** .009High second. Educ. .117*** .070** .099*** .055**University Educ. .244*** .186*** .234*** .180***Dutch Education .050*** .050* .048*** .049**Dutch Citizenship .009 .033 .005 .035Lnhours .746*** .752*** .744*** .749***Somalia .020 .007Iran .048* .035Iraq -.051* -.045Former Yugoslavia .119*** .102***Turkey .026Morocco .013Constant 3.101*** 2.863*** 3.101*** 2.905***
Observations 2410 1551 2410 1551Adjusted R Square 0.583 0.606 0.589 0,612F-Statistics 333,631 236,429 214,743 173,878
Note: ***, **, * indicate significance at 0.01, 0.05 and 0.1 level respectively.
Source: SPVA 2002 and 2003, ISEO/SCP/NIDI.
We will now explore the effects of naturalization on the log monthly income of the immigrant and refugee groups, while taking into account several important individual characteristics expected to influence earnings.
As shown in Table 6, the control variables age and age squared yield the expected significant coefficients; an increase in immigrants’ age is cong-ruent with an increase in monthly income. The educational level also stron-gly correlates with our dependent variables: a high educational level obtai-ned in the Netherlands is translated into greater monthly income. While highly significant, coefficients for the immigrant-specific variables ‘years since migration’ and ‘Dutch education’ are positive, as expected. The log hours variable shows the expected correlation of increased income with increased hours of work. The results also reveal that greater earnings accrue to males as compared to females.
PIETEr BEVElANDEr & JUSTUS VEENMAN
86
Most importantly, if we control for all demographic and labour market variables listed in Table 6, the effect of naturalization is weak and positive but not significant in the aggregate analysis with all entry groups. The same result is obtained when we restrict our analysis to the refugee groups. Inclu-ding dummies for individual groups has only a small effect on the various background variables and does not change the effects of the citizenship variables. Finally, when we interact citizenship with country of birth, a weak but significant citizenship premium can be found for the refugee groups from the former Yugoslavia, Iran and Iraq in the Dutch labour mar-ket.15
ConclusionsIn this paper we analyzed the relationship between socioeconomic integra-tion and naturalization in the Netherlands. We started with an analysis of various factors, both demographic and educational, that affect naturaliza-tion rates. In this case, naturalization is thus the dependent variable. We then analyzed the relationship between employment opportunities and ear-nings on the one hand and naturalization on the other hand (naturalization now being the independent variable). We used a sample of Afghan, Somali, Iranian, Iraqi, former Yugoslav, Turkish and Moroccan immigrants in the Netherlands, with cross-sectional data from the 2002 and 2003 versions of the SPVA survey. With regard to the naturalization decision, we found that the immigrant’s educational level and studies completed in the Netherlands had a strong positive impact on citizenship acquisition. These findings support the hypothesis that integration into the receiving country positi-vely affects naturalization acquisition. Additional support for the integra-tion hypothesis is in the finding that the length of stay positively affects naturalization in the Netherlands. A direct test of the integration hypothe-sis, however, fails, since participation in Dutch integration programs does not have a significant effect on naturalization, although this outcome may also reflect the effectiveness of the program itself.16 We furthermore esta-blished that those admitted as refugees have a greater propensity to natura-lize than Turks and Moroccans. Immigrants motivated by political reasons or family reunion also are more likely to acquire the Dutch citizenship than those who moved to work or study. We argued the latter did not initially intend to stay and hence would be less likely to naturalize if they stayed for the required five years.
15 results of regressions with interactions available upon request.16 The result may also be affected by the fact that integration programs have been mandatory in the Netherlands since 1998.
NATUrAlIZATION AND SOCIOECONOMIC INTEGrATION: THE CASE OF THE NETHErlANDS'
87
The impact of naturalization on the immigrants’ labour market status was measured by changes in employment rates and wage levels. As expec-ted, we found higher employment rates among male and female immigrants with Dutch citizenship. Controlling for educational and demographic cha-racteristics, positive and significant effects on the employment opportuni-ties of naturalized immigrants were derived for most variables. The cited variables which were intended to capture the motivation to migrate and subsequent participation in the Dutch integration programs did not produce a significant effect on naturalization rates. The latter result may again indicate the limited effectiveness of the said programs.
Moreover, we found that immigrants with the Dutch citizenship had higher wages, again for men and women alike, in most cases. However, naturalization did not yield a significant effect on wages when we aggrega-ted across all groups. When we interacted citizenship with country of birth, our analysis showed a significant positive effect of naturalization for some refugee groups.
Finally, our analyses must be considered with caution since we could only use cross-sectional data. Accordingly we reported correlations and tempered our conclusions with respect to causality between citizenship status and economic outcomes. In fact, high employment rates and good wages may induce naturalization. For example, high educational attain-ment may be a consequence of citizenship acquisition rather than its cause, given the lack of formal restrictions imposed on foreigners in Dutch educa-tional system. At this point we offer the possibility that a third set of factors influence citizenship acquisition in the Dutch context. Since we reported that immigrants’ educational level and place of education (i.e., the Nether-lands) both have a strong positive and significant effect on the employment rate as well as on wages, it is now possible that these two variables in turn affect naturalization. Thus, we must remain well aware of the possibility that the relationship between naturalization and the labour market indica-tors can be explained in terms of a third variable, be it educational level and/or studies completed in the Netherlands. To test these and other causal possibilities requires longitudinal data, especially in the form of panel data.
PIETEr BEVElANDEr & JUSTUS VEENMAN
88
appendixSample values of selected variables
Citizen (St.Dev.) Non-Citizen (St.Dev.)
Mean age 39.37(11.94) 37.35(10.04)Mean Years since
migration
16.77(10.02) 15.43(8.25)
Females 1004 2063Males 1157 2790Primary schooling 1215 1815Secondary lower
education
300 790
Secondary higher
education
387 1258
University education 192 863Afghanistan 279 513Iran 92 785Iraq 174 712Somalia 128 571former Yugoslavia 212 576Turkey 761 866Morocco 515 830Total number 2161 4853
Source: SPVA 2002/2003.
89
A Norwegian Narrative
I came to Norway as a foreign student in the early 1980s, with the inten-tion of doing graduate studies and then returning to my home country of Ghana. One cannot do much with the Bachelor’s degree in psychology which I had when I left Ghana, and for that reason I had a very strong urge to do graduate studies in psychology, and to put into practice what I had learnt.
The decision to come to Norway was not so much because I found Norway an attractive country to study in, but because Norway was the only country that was willing to provide me with a tuition-free education, as well as to support me financially with respect to my living expenses. I had the desire to go to an English-speaking country, preferably Canada or the United States, and to a lesser extent Australia/New Zealand, and the United Kingdom.
My initial plan was to use Norway as a stepping stone to any of the English speaking countries I just mentioned. With relatively good grades in psychology following my Bachelor’s degree, I could have waited a little bit longer in Ghana and hoped that some form of scholarship would come my way from any of the English-speaking countries. Unfortunately, the politi-cal climate in Ghana in the early 1980s was very unpredictable. One could not be sure what might happen the next day. University students had incur-red the wrath of the then-ruling military government. Even though I was not politically active, one could never tell when he or she would fall victim to military brutalities. Thus, when I had the opportunity to do graduate studies in Norway, I did not hesitate to accept the offer.
The first year in Norway was used in learning the language and the social life of the country. During that year, I came to learn much about the country and to appreciate its high standard of living, along with an atmosphere of peace and respect for human rights and equality. Gaining command of the Norwegian language, I also found Norwegians friendly and welcoming. Accordingly, when I had an offer to do graduate studies in psychology in Canada, after 4 months in Norway, I turned down the offer. I felt secure in Norway, and did not know what I might be heading into by going to another country. I found my studies interesting and did not want to terminate what I had started. I therefore opted to stay, then finish my graduate studies, i.e., professional degree in psychology, and then return to Ghana. However, I fell in love with a Norwegian girl during my third year in the country, and this made it difficult for me to leave the country when I finished my professional degree five years later.
90
My Norwegian wife was not very eager to migrate to another country, and I had been fortunate to secure a local fellowship to do a PhD. I obtai-ned a PhD after 10 years in Norway, and soon got a job as a post-doctoral research fellow at the university. Three years later, I obtained a job as a lecturer, and have since then risen through the ranks to become a full pro-fessor of cross-cultural psychology. While working on my PhD, I had the opportunity to travel a lot. But I also experienced the difficulties involved in travelling with a Ghanaian passport. Thus, I seriously considered secu-ring a Norwegian passport. Norway does not allow dual citizenship, meaning that I had to give up my Ghanaian citizenship to be issued a Nor-wegian passport. It took me at least a year to evaluate the pros and cons, and to be sure that this was what I really wanted. By that time my status had changed from foreign student to an immigrant by virtue of my marri-age to a Norwegian. I married after four years in Norway. I applied for Norwegian citizenship after 9 years in Norway and it was granted after 4 months. Even though I gave up my Ghanaian citizenship at that time, I still identify myself as a Ghanaian, even after 24 years in Norway.
91
Chapter 3
the eCONOMICS OF NOrWeGIaN CItIZeNShIp
John E. Hayfron
IntroductionIn a recent newspaper article entitled Immigrants okay, as long as they ‘become Norwegian’, Aftenposten (2005) reports the results from a survey which measures Norwegians’ tolerance toward immigrants and concludes that “a majority of Norwegians think ethnic and cultural diversity is fine, but they also think new immigrants must assimilate as soon as possible.” This article further suggests that citizenship acquisition is an important mode of integration.
Given the importance of citizenship acquisition as an integration tool in Norway it is ironic to note that there is no detailed statistical analysis of the causes and implications of immigrant naturalization in Norway. However, Statistics Norway reports some stylized facts on the number and the socio-demographic characteristics of immigrants who become Norwe-gian citizens each year. According to recent statistics, in the 12 years since 1992 the number of eligible immigrants who have become Norwegian citi-zens has risen from 5,132 to 8,154, or an increase of 58.9 percent. The statistics show that the number of immigrants from western countries who have acquired Norwegian citizenship each year has increased from 544 to 870 (an increase of 59.9 percent), compared to an increase from 4,578 to 7,183 (an increase of 56.9 percent) for immigrants from non-Western countries.1
Figure 1 shows the trends in naturalization in Norway between 1977 and 2004. The observed sharp increase in the number of naturalized citi-zens in the mid-1990s is due to the dramatic increase in immigration, parti-cularly asylum-seekers, in the mid-1980s.
1 Western countries include Denmark, Finland, Sweden, Western Europe, North America, Australia and New Zealand. Countries in Eastern Europe, Asia, Africa and South/latin America are defined as non-Western.
JOHN E. HAYFrON
92
Figure 1. trends in Naturalization in Norway, 1977-2004
Source: Statistics Norway data, 2005.
The official statistics are useful in showing the growing number of natura-lized citizens in Norway over time. However, they do not explain why immigrants decide to acquire Norwegian citizenship, and neither do they explain why the naturalization rate (i.e., the rate at which immigrants are granted Norwegian citizenship) varies across sending countries as well as over time.
To understand why some eligible immigrants decide to naturalize while others do not, we need to know what factors affect their decision to become Norwegian citizens. In this chapter I approach this problem in two ways. First, I provide a theoretical framework that explains how individual immi-grants’ labour market characteristics impact their propensity to naturalize. I then develop a set of equations which characterize the modelling exercise and provide a framework to evaluate the relative importance of each cha-racteristic in the decision to acquire Norwegian citizenship.
Although Norwegian citizenship studies are rare, several of them docu-ment immigrant earnings assimilation in Norway. These studies reveal that immigrants experience a rapid earnings growth relative to otherwise equally skilled Norwegians over time (Hayfron 1998; longva & raaum 2003). One reason cited is that immigrants invest in Norwegian language proficiency (Hayfron 2001). Other studies indicate that the earnings of
Figure 1. T rends in N aturalization in N orway. 1977-2004
200
1200
2200
3200
4200
5200
6200
7200
8200
9200
10200
11200
12200
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
ALL Immigrants Western Countries Non-western Countries
THE ECONOMICS OF NOrWEGIAN CITIZENSHIP
93
immigrants are more sensitive to changes in macroeconomic conditions than the earnings of Norwegians (Barth et al 2004). What is missing from these earnings studies is the effect of naturalization on immigrants’ ear-nings assimilation. This omission is addressed here by accounting for the effect derived from naturalization on immigrant wages in Norway. theoretical analysis
Immigrants may decide to naturalize because they expect to economically or otherwise benefit from Norwegian citizenship. The rational choice assumption implies that immigrants judge their satisfaction from naturali-zation by weighing the expected additional benefits against the expected additional costs (Yang 1994; Watsula 2005; DeVoretz & Pivnenko 2006). Under this argument immigrants will naturalize if the net benefit derived from naturalization is positive. Apart from the economic incentives, such as expected wage increase and job prospects, other evidence suggests that non-economic benefits influence immigrants’ decision to naturalize. For instance, in their 1986 study Jasso and rosenzweig report a key incentive for United States immigrants to naturalize: after naturalization immigrants have the right to sponsor their family members in an expedited fashion to the United States. For his part, Yang (1994) found that men were more likely than women to naturalize since U.S male immigrants were more likely to be engaged in the labour market. However, Yang cited other non-economic conditioners as key determinants of immigrants’ naturalization in the United States: the desire to possess a U.S. passport and vote, the level of English language skills, and the existence of a strong ethnic community.
In the case of Norway, the benefits derived from citizenship include access to top-level positions normally reserved for Norwegian nationals in the government and banking sectors, increased travel mobility afforded by a Norwegian passport, governmental protection at home and abroad, the ability to sponsor close family members (i.e., children under 18 years old, a spouse or registered partner, someone with whom the person has cohabited for at least two years), and the right to vote in local and national elections, and to stand as a candidate in a national election.
The costs of naturalization include any out-of-pocket expenses (such as application fees,2 legal fees, fees the individual would pay for future visits to the country of origin), and the cost of lost opportunities in the sending countries, especially in the absence of dual citizenship. In addition, Norway requires its male nationals to perform compulsory military service, and this is costly to the prospective citizen.
2 The application fee for naturalization is NOK 2500 (about US$354).
JOHN E. HAYFrON
94
Norway’s actual naturalization process may also influence an immigrant’s desire for naturalization. For this reason, I review the Norwe-gian naturalization policy in the next section. I will then discuss the data and the econometric technique I used to estimate representative immi-grants’ propensity to naturalize and the effects of naturalization on their wages over time. This is followed closely by the interpretation of the esti-mated coefficients from both the random-effects probit and wage regres-sions respectively. I conclude with a summary and a discussion of the policy implications of this study.
Norwegian Nationality actFor the purpose of this study, the discussion of Norway’s naturalization policy will focus on three questions whose answers will shed light on the propensity to naturalize: First, given that Norway is increasingly becoming a pluralistic society, who is considered a Norwegian? This question is important given the relatively large number of ethnic minorities in Norway, and the fact that more than 50 percent of Norwegian immigrants are of non-Western origin (Statistics Norway, 2005). Second, I ask who is eligible for Norwegian citizenship. Finally, does Norway recognize dual citizen-ship?
The Norwegian Nationality Act of 1950 defines Norwegian citizen-ship on the basis of birth (i.e., the jus sanguinis principle), rather than birthplace (i.e., the jus soli principle). This means that a Norwegian is a person born of two Norwegian nationals.3 This also implies that a child born in Norway of two foreign-born parents (second-generation immi-grants) is not guaranteed automatic Norwegian citizenship.
With regard to eligibility to ascend to Norwegian citizenship, para-graph §6 of the Norwegian Nationality Act indicates that “The King, or anyone he so empowers, may upon application grant nationality to foreign national provided the applicant:• isnotlessthan18yearsofage,• hasresidedintherealmduringthelastsevenyears,• hasshowngoodconduct,and• does not have substantial debt in connection with maintenance pay-
ments.” 4
3 As of September 1, 2006, a person may be considered a Norwegian if that person has one Norwegian parent (www.udi.no/upload/Publikasjoner/Statsborgerskap/NorskStats-borgerskap.pdf).4 The official Guide to Naturalization indicates that, “If you are under obligation to pay maintenance, i.e., you have to pay a fixed amount each month to your children or your former spouse, and you have not paid the amount, your application for Norwegian nationality will be rejected.” The amount is about NOK 20,000.
THE ECONOMICS OF NOrWEGIAN CITIZENSHIP
95
The Norwegian nationality act forbids Norwegian nationals from holding two citizenships. For this reason, foreign nationals must relinquish their previous nationality if they are granted Norwegian citizenship. There are a few exceptions to this law. A child with at least one Norwegian parent may be able to hold dual citizenship. Immigrants from countries with laws that make renouncing their citizenship difficult even if they acquire the citi-zenship of another country, may be able to hold dual citizenship. Children born of two Norwegian parents in countries where the definition of natio-nality is based on the jus soli principle may also be able to hold dual citi-zenship.
Other exceptions apply to immigrants who arrive in Norway as child-ren. For instance, immigrant children who arrive in Norway before they are 12 years old, or those born in Norway can acquire Norwegian citizen-ship without any residency requirement. However, immigrant children between the ages of 12 and 13 upon arrival in Norway require only one year of residency, while those between 13 and 14 years of age require two years of residency, and children between 14 and 15 years of age require three years of residency to be eligible for Norwegian citizenship.
In sum, the answers to the questions above illustrate the complex and varied environment in which Norwegian immigrants decide to acquire citi-zenship.
Model Specification and DataAlthough an immigrant’s decision to acquire Norwegian citizenship at a particular point in time is not observed directly, it is possible to predict the outcome of such a decision-making process using a binary choice model. In this case, an immigrant’s propensity to acquire Norwegian citizenship is modeled in a series of probit equations as a function of the individual and time-specific characteristics:
(1)
€
Pr (NCitzit =1) = Φ(β0 + β1YrS it + β2YSM it + β3YSM it2 + β4Git + β5lnWit +ψZit ),
where the dependent variable, NCitzit (0, 1) denotes an individual i’s pro-pensity to naturalize in period t. YSMit denotes years of residence in Norway, InWit and denote the log of wages and employment in the public and banking sector jobs respectively. Z it is a set of other control variables, and Φ denotes the cumulative standard normal distribution function.
According to the official Guide to Naturalization, “certain positions are reserved for Norwegian nationals. This applies to a number of posi-tions in the police and the Norwegian legal system, as well as top adminis-trative positions in other public sectors.” For this reason, I test the hypo-
JOHN E. HAYFrON
96
thesis that access to these protected jobs increases an immigrant’s propensity to naturalize. That is β4 > 0. Since most of these top level jobs pay higher wages, I hypothesize that β5 > 0. Apart from the potential job prospects and the wage incentives, the assimilation-related variables, such as years of schooling, years since migration, age at migration, and marital status are expected to have significant impacts on immigrants’ citizenship ascension (Portes & Mozo 1985; Jasso & rosenzweig 1986; Yang 1994; liang 1994).
Location-specific characteristics, such as the economic conditions in the country of origin and the mode of entry into Norway, can impact an immigrant’s propensity to naturalize. Norway has a wide array of immi-grant entry categories, from asylum seekers to employer-sponsored immi-grants, which have varied in importance over time. This has resulted in a stock of immigrants with a wide array of individual characteristics. For their part, since refugees cannot usually return to their countries of origin, they have an induced permanent attachment to Norway. Therefore, I expect refugees to have a stronger propensity to naturalize than non-refugees. Similarly, given Norway’s generous social benefits5 and democratic system, non-refugees from developing countries with high rates of poverty, political instability, and unemployment will have greater incentives to naturalize, and eventually sponsor their immediate family members to Norway.
Norwegian panel Data
The six-wave panel data set used in this study is taken from the Norwegian Database “FD-Tygd Panel”. Data sources are national administrative and statistical registers. The Database contains a sample of 10 percent of the Norwegian population between 16 and 66 years of age. The data follow the individual over a period of eight years, 1992-2000. For each individual, periods with benefits from the National Insurance System are registered. The database contains information about labour market involvement, edu-cation, year of arrival, year of naturalization, country of birth, citizenship status, income history, and family size. Each year the sample is supplemen-ted with new 16 year olds and immigrants. Consistent with the minimum age and residency requirements for naturalization, the non-Nordic sample consists of individuals 18 years old or older (AGE ≥ 18) who have lived in Norway legally for at least seven consecutive years (YSM ≥ 7). The age at migration, the years prior to naturalization, and the years since naturaliza-tion variables are calculated as described in Table 1. All the variables with missing observations were dropped.
5 The Guide to Naturalization indicates that “the right to national insurance, social secu-rity and other benefits depend on how long you have lived in this country and is not dependent on your nationality status.”
THE ECONOMICS OF NOrWEGIAN CITIZENSHIP
97
table 1. Descriptive Statistics
Variable Definition Mean S.D
Citizenship =1 if individual is a naturalized citizen
.872 .334
Education Years of schooling 12.771 2.483
Age Census year minus year of birth
36.971 8.372
Years Since Migration Census year minus year of arrival
13.024 4.972
Age at Migration Age minus year of arrival 23.947 8.158
log (Wage) Natural logarithm of wages (1,000 Krone)
11.401 1.023
Male =1 if individual is a male .663 .473
Married =1 if individual is married .373 .484
Child <11 years old Number of children under 11 years of age
.579 .964
Oslo =1 if individual is a resident of Oslo
.222 .415
Fulltime =1 if individual works 30 hours a week
.753 .431
refugee =1 if individual originates from one of the ten top
refugee sending countries
.149 .356
Years Prior Naturalization Year of naturalization minus year of arrival
10.811 5.080
Years Since Naturalization Census year minus year of naturalization
2.500 1.990
Hotel and restaurant =1 if individual is employed in the hotel and restaurant
industry.
.040 .197
Public and Banking Sectors =1 if individual is employed in the public sector, Banking
sector
.088 .284
Educational sector =1 if individual is employed in the education and
research industry.
.091 .267
Health & Social Services =1 if individual is employed
in the health and social services sector
.149 .356
Source: Author’s calculations.
JOHN E. HAYFrON
98
According to the descriptive statistics in Table 1, 87.2 percent of the eligible immigrants are naturalized Norwegian citizens. An average immi-grant is nearly 37 years old and has about 13 years of schooling.6 Moreo-ver, average immigrant age at arrival is 24, and the average immigrant has resided in the country for 13 years. The mean log wage is NOK11.40/1.43 EUr. About 66 percent of immigrants are males and 37.3 percent of them are married. About 58 percent of them have children aged 10 or younger, and 22 percent of immigrants reside in the Oslo municipality.7 About 15 percent of immigrants came to Norway as asylum-seekers, and 75.3 percent of the immigrants are full-time workers. The average years prior to natura-lization and years since naturalization are 18.8 and 2.5 respectively. The distribution of immigrants across industrial sectors is as follows: hotel and restaurant, 4 percent; government and banking, 8.8 percent; education and research, 9.1 percent; and health and social services, 14.9 percent. All these variables will be useful for my analysis of the immigrants’ naturalization process in Norway.
Determinants of Immigrants’ acquisition of Norwegian CitizenshipThe estimated coefficients and their test statistics from the random-effect probit regressions are reported in Table 2. The negative coefficient on the age at migration variable can be interpreted to mean that immigrants who arrive in Norway as adults are less likely to naturalize than those who arrive as children. This is to be expected since the residency requirement for naturalization in Norway is lower for immigrants who arrive as child-ren than for those who arrive as adults. The results show that each additio-nal year spent in Norway as a legal resident increases an immigrant’s pro-pensity to naturalize by 37 percent. This is no surprise since longer residency in Norway is required to acquire proficiency in the Norwegian language (Hayfron 2001) and to gather better information about the net benefits associated with naturalization. 8
6 Norway has one of the highest percentages of well educated immigrants in the world. Four out of ten immigrants to Norway have received a higher education. According to the OECD report which focuses on Europe and the USA, only England and Ireland exceed Norway in terms of the numbers of highly educated immigrants. (http://www.aftenposten.no/english/local/article1260069.ece).7 About 21.1% of the inhabitants of the greater Oslo district have an immigrant back-ground and 80% of these come from non-Western countries. As of 2004, 92.4% of Oslo residents from non-Western countries, compared to 6.8% from Western countries, have acquired Norwegian citizenship (Statistics Norway 2005).8 The panel data set does not have information on individual immigrants’ Norwegian language proficiency. Currently, immigrants are not required to take Norwegian language test to be eligible for Norwegian citizenship. However, as of September 1, 2008, immi-grants would have to either document that they have received 300 hours of Norwegian language training, or take a language test that will measure their language proficiency. (www.aftenposten.no/english/local/article1260069.ece).
THE ECONOMICS OF NOrWEGIAN CITIZENSHIP
99
Coe
fft-
rati
oC
oeff
t-ra
tio
Coe
fft-
rati
o
Yea
rs o
f Sc
hool
ing
.033
0 (
1.10
0) .
0036
(.1
24)
.0
426
(
1.29
1)
Age
at
mig
rati
on-.
0800
(-8
.217
) -
.078
4 (
-8.2
35)
-.0
901
(-
7.95
1)
Yea
rs S
ince
Mig
rati
on .7
384
(10.
712)
.77
76
(11.
693)
.8
021
(11.
505)
Yea
rs S
ince
Mig
rati
on S
q.-.
0155
(-9
.161
) -
.016
4 (-
9.82
3) -
.016
6
(-9.
649)
Mal
e .4
052
(2.
351)
.46
27
(2.
819)
.4
034
(
2.75
4)
Mar
ried
.822
1 (
3.50
6) .
8960
(
4.89
8) 1
.153
5
(5.
570)
Chi
ld (
< 11
yea
rs o
ld)
-.10
78(-
1.52
4) -
.028
4 (
-.39
7) -
.099
4 (-
1.22
4)
Osl
o-1
.183
4(-
4.54
0) -
.890
0 (
-5.0
81)
-.8
428
(-
5.25
8)
log
(W
age)
.0
707
(1.
091)
.13
66
(2.
274)
.1
567
(
2.64
7)
GD
P pe
r ca
pita
× 1
0 -
.000
3 (
3.40
8) -
.000
2 (
2.76
6) -
.000
2 (-
2.44
4)
ref
ugee
1.69
70 (
3.73
0) 2
.593
0 (5
.865
)
ref
ugee
× G
DP
per
capi
ta-.
0003
(.
189)
-.0
003
(-2.
018)
Publ
ic s
ecto
r/B
ank
.7
625
(1.
860)
-.3
040
(1.
056)
-.
1687
(.62
7)
Hot
el &
res
taur
ant
1.1
000
(3.0
29)
Hea
lth
& S
ocia
l ser
vice
s
.405
0 (1
.729
)
Edu
cati
on s
ervi
ces
-.
2297
(-.
939)
Con
stan
t -
3.58
38 (
3.77
5)-4
.798
4 (
5.51
5) -
5.59
69(-
6.20
0)
tabl
e 2.
ran
dom
-eff
ects
pro
bit r
egre
ssio
n of
Imm
igra
nts’
Citi
zens
hip
prop
ensi
ties
in N
orw
ay
Sour
ce: A
utho
rs' c
alcu
lati
ons.
JOHN E. HAYFrON
100
Male immigrants are more likely than females to naturalize. This is consis-tent with the gender hypothesis which suggests that males are likely to naturalize more quickly than females (Jasso & rosenzweig 1990), since males will receive a greater economic reward from citizenship ascension. Similarly, married immigrants are more likely to naturalize than their coun-terparts who are not married. In addition, an immigrant residing in the Oslo municipality has a lower propensity to naturalize than do those living in other municipalities perhaps due to the existence of a strong ethnic com-munity.
The statistically significant coefficient on the wage variable indicates that immigrants who earn higher average wages have greater incentives to naturalize than those earning lower wages. Though positive (.7625), the estimated coefficient on the public sector dummy variable (column 1 of Table 2), is marginally significant (t = 1.860). However, the coefficient on the public sector dummy becomes statistically insignificant when other industry dummy variables are added to the probit equation. The results indicate that immigrants in the hotel and restaurant sector, as well as in the health and social services sector are more likely to naturalize than their counterparts in other sectors.
To capture the effect of location-specific characteristics on immigrants’ propensities to naturalize, I added a measure of economic activity in a source country, GDP per capita in the probit equations.9 The coefficient on the GDP per capita variable is negative (-.00002) and statistically signifi-cant (t = -2.444), implying that immigrants from countries with a higher GDP per capita are less likely to naturalize than immigrants from countries with a lower GDP per capita. More specifically, if the GDP per capita in a particular source country increases by $1,000, the likelihood that the country’s citizens will naturalize falls by .02 percent. The results imply that immigrants from poor refugee-sending countries are more likely to natura-lize than immigrants from non-refugee-sending countries.10
effects of Citizenship acquisition on Immigrants’ WagesTo be able to determine the returns to immigrants’ naturalization in Norway, I estimate a series of wage equations of the form:
(2)
€
lnWit = β0 + β1Ncitzit + β2YPN ,it + β3YSN ,it + β4YSN ,it2 + β5X it + α i + eit ,
9 The nominal GDP per capita for each of the 180 member countries of the Internatio-nal Monetary Fund is expressed in 2000 US dollars.10 As of 2000, russians, Croatians, Somalians, Iraqis, Ukrainians, Yugoslavians (Serbia and Montenegro), Bosnia and Herzegovina, Afghans and Mongolians constituted the ten largest groups of asylum-seekers in Norway (Kommunal- og regionaldepartementet 2003).
THE ECONOMICS OF NOrWEGIAN CITIZENSHIP
101
where ln Wit is the natural logarithm of wages for individual i in year t, β0 is the general intercept. Ncitzit is a dummy variable indicating whether the individual i is a naturalized citizen or non-citizen in year t. YPN, it denotes the years prior to naturalization, YSN, it represents years since naturaliza-tion, 11 while X it is a set of other individual and time specific characteristics. The αi represents unobserved individual characteristics, such as innate ability. Innate ability affects the individual wages and is probably correla-ted with the individual’s level of education. By definition innate ability does not change over time. Similarly, eit represents the unobserved factors that affect individual wages and do change over time. The wage equation (2) can be estimated using either a fixed or random effects model depending on the assumption about the distribution of αi. The standard assumption is that αi and eit are independent, that eit has a zero mean and is serially uncorrelated (Hsiao 1992). 12
Although I estimate both models, I interpret only the parameter esti-mates from the random-effects models. The estimated coefficients are reported in columns 1 and 3 in Table 3. regressing the log wage on only the naturalization dummy variable, the results in column 3 show that natu-ralized citizens earn (exp (.2744) - 1 = 0.3099) 31 percent more on average than non-citizens. This is consistent with the hypothesis that naturalization has an instantaneous effect on wages (Bratsberg et al 2002). However, when age, age squared, years prior to naturalization, years since naturali-zation and years since naturalization squared are added to the wage equa-tions, not only does the coefficient on the citizenship dummy variable change in value, it becomes statistically insignificant (columns 2 and 4).
11 The years of legal residence in Norway (YSM) can be divided into two parts: Years Prior to Naturalization and Years Since Naturalization. Let YA = year of arrival, YN = year of naturali-zation, YC = census or reference year, YPN = years prior to naturalization and YSN = years since naturalization. YSM = YC – YA. YPN = YN – YA while YSN = YC – YN. Since YSM = YPN + YSN, it follows that YC – YA = YN – YA + YC – YN. The two YN s cancel each other. Example: YA =1980, YC =2000, YN=1987. YSM = 2000 – 1980 = 20 years. YPN = 1987 – 1980 = 7 years. YSN = 2000 – 1987 = 13 years. YPN + YSN = 20 years.12 In this specification, I assume that αi = eit = 0. E(αi eit ) = 0 i, t, E(eit) = 0 i,t. The Hausman’s specification test conducted favours a fixed effects model. X2=2.9 F test that αi = eit = 0. F (386, 1979) = 4.04. However, due to the number of time-invariant variab-les such as years of schooling, and the large sample size over six periods, using the fixed effects model may yield inconsistent estimates of the parameters (see Maddala 1986).
A A
John E. hayfron
102
Table 3. Parameter Estimates of the Wage Function: Dependent Variable = Log (Wage)
note: The t-ratios are given in parentheses.
Source: Author's calculations.
The statistically significant coefficients on the age and age squared variab-les indicate that as an immigrant grows older each year his or her wage rises by 2.3 percent, evaluated at the sample mean.13 This is consistent with the institutionalized wage setting mechanism in norway, whereby workers are paid more as they grow older (hayfron 1998). The results indicate that the estimated return to years since naturalization ( itYSN ) is higher than the return to years prior to naturalization ( itYPN ). Extending the post-
13
Fixed Effects Random Effects
(1) (2) (3) (4) (5) (6)
Citizenship .3192 (4.94)
-.1225 (-1.24)
.2744 (4.56)
-.0363 (-0.54)
-.0361 (-0.54)
-.0366 (-0.54)
Years of Schooling .0081 (0.63)
Age .4134 (6.88)
.2311 (9.09)
.2311 (9.08)
.2146 (8.54)
Age Squared -.0040 (-7.00)
-.0028 (-8.43)
-.0028 (-8.42)
-.0026 (-7.88)
Years Prior to Naturalization .0231 (3.23)
.0230 (3.21)
.0214 (3.07)
Years Since Naturalization .0826 (1.59)
.1699 (5.58)
.1699 (5.58)
.1713 (5.64)
Years Since Naturalization Sq. -.0148 (-3.11)
-.0146 (-3.09)
-.0145 (-3.08)
-.0149 (-3.17)
Male .3366 (4.92)
Married -.1372 (-2.06)
Full time .0010 (0.01)
Oslo .0746 (0.95)
Hotel -.0126 (-0.07)
Education Services .1842 (1.61)
Public Sector & Banking -.0359 (-0.31)
.0154 (0.13)
Health & Social Services .1116 (1.18)
Refugee -.9070 (-3.32)
-.9090 (-3.32)
-.9709 (-3.57)
Refugee × Citizenship .6930 (2.60)
.6932 (2.60)
.6869 (2.59)
Constant 11.122 (189.42)
1.910 (1.18)
11.161 (176.16)
6.4369 (13.25)
6.4417 (13.24)
6.4400 (12.62)
Number of observations 2382 2382 2382 2382 2382 2382
R-squared .0035 .0618 .0035 .1040 .1041 .1164
€
∂ log (wage)
∂Age= .2146 - .0052(36.9) = .0227 or 2.3%.
THE ECONOMICS OF NOrWEGIAN CITIZENSHIP
103
naturalization period by one year increases a naturalized citizen’s wage by 9.7 percent, evaluated at the sample mean.14
In Figure 2, the curvilinear curve with the dotted lines represents the predicted wage profile for non-naturalized citizens with YSN it = 0. Simi-larly, the wage profiles with the dashed and thick lines were constructed for naturalized citizens by setting YSN it = 2 and 5 years respectively. Without any doubt, years after naturalization increase wages for naturalized citi-zens. This is indicated by a parallel shift in the wage profiles for naturalized citizens.
Figure 2. Wage profiles
Finally, the results indicate that male immigrants receive higher wages on average than female immigrants with similar characteristics, and that wages are lower for married males. refugees receive lower wages on average than non-refugees. The statistically significant positive coefficient on the interac-tion term, 0.6842 (Refugee×Citizenship), indicates that refugees who natu-ralize as Norwegian citizens at a particular point in time obtain a wage premium over their counterparts who are non-citizens. Stated in terms of
14
Figure 2. Wage Profiles
7
8
9
10
11
12
13
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Age
Lo
g (
Wag
e)
IMMIGRANT CITIZEN (YSN=2) CITIZEN (YSN=5)
€
∂ log (wage)
∂YSN= .1713 - .0298(2.5) = .0968 or 9.7%.
JOHN E. HAYFrON
104
wage disparities, refugees who naturalize (and those who do not) earn 28.4 percent (and 96.8 percent) more than non-refugees.15
Summary and Conclusions This chapter examines an immigrant’s propensity to naturalize in Norway. It also examines the effect of naturalization on an immigrant’s wage. The results tend to confirm the findings in most studies that assimila-tion-related variables, such as age at migration, years since migration, gender, marital status, wages, as well as location-specific characteristics, such as GDP per capita and refugee status are key determinants of an immigrant’s propensity to naturalize. results from the wage equations indicate that naturalization has an instantaneous effect on immigrants’ wages. For instance, naturalized citi-zens earn 31.6 percent more than non-citizens. Extending the post-natura-lization period by one year increases a naturalized immigrant’s wage by 9.7 percent, evaluated at the sample mean. The results show that refugees who naturalize have a higher wage premium than their counterparts who are non-citizens.
These findings are important because they not only help us to fully understand the economic impact of naturalization on immigrants’ wages, but they also might contain some relevant policy implications, especially for Norway’s integration efforts. For instance, if Norway considers rapid and complete ascension to Norwegian citizenship an important step in the process of immigrant integration, then a policy of dual citizenship, accom-panied by a reduction of the residency requirement and enhanced econo-mic opportunities for Norway’s Third World immigrants would hasten their ascension to citizenship.
15 For refugees who are non-citizens = -.9684 (refugee = 1) + .6842 (refugee = 1 × citi-zenship = 0) = -.9684. For refugees who are citizens = -.9684 (refugee = 1) + .6842 (citi-zenship = 1) = -.2842.
105
A Swedish Narrative
I became a Swedish citizen some 15 years ago. Prior to arrival in Sweden, I lived in another European country as a political refugee from a Middle Eastern country. I came to Sweden through my relationship with a Swedish woman, and was granted a residence permit on this basis. In due course I chose naturalization.
I have never felt ‘Swedish,’ and I do not believe one can become a Swede through naturalization. Given the peculiar history of the country, ‘being Swedish’ is associated with ethnicity (not only in Sweden but inter-nationally). The most one can expect is to become a citizen of the Swedish society. I am also aware that many naturalized immigrants do not even feel that they are full citizens.
Whether racism and xenophobia are to be blamed for the failed social integration of many immigrant groups is beside the point here. I have been lucky in this respect, perhaps because I live and work in a small, cosmopo-litan community with a large share of university employees and internatio-nal students. In this context I’ve found it natural to feel I belong to this society as a citizen who greatly appreciates the fundamental values of democracy and fairness that inform the social institutions of this society. In my experience, you do not have to feel ‘Swedish’ to feel that you are a citizen of your naturalized country.
107
Chapter 4
the eCONOMICS OF CItIZeNShIp:IS there a NatUraLIZatION eFFeCt?
Kirk Scott
IntroductionSwedish rules regarding naturalization have changed somewhat during the past 40 years, but the fundamental principle was established in a proposi-tion to the parliament in 1968 (Proposition 1968). This new and innova-tive alien policy was based on a concept of equality and the allocation of civil rights based on residence rather than nationality. The idea was that residence in Sweden should be adequate grounds for gaining rights and privileges, and full inclusion in society would no longer be the sole pro-perty of citizens. In the spirit of Sweden as a universalistic welfare state (Esping-Andersen 1990), this inclusion was expanded during the following decades to include practically all spheres with only a very few exceptions. One exception is the right to political participation. While foreign residents have been granted the right to participate in local and county elections, voting in the parliamentary elections is restricted to Swedish citizens. A second area of restriction is found in employment deemed vital to national security, with the military, police, and certain other occupations closed to non-citizens. A third, and perhaps non-trivial, right which is not granted to non-citizens is the right of stay in the country. The Swedish government may, given adequate grounds, expel non-citizens from the country, while naturalized immigrants are immune to deportation. Despite these excep-tions, a recent government study found that the changes during the past 40 years have led to a dramatic reduction in the importance of citizenship (SOU 1999). Becoming a citizen in Sweden today is a fairly straightforward issue. One basic demand is that the individual applying for citizenship be of legal age (18 years or older) and able to provide evidence of identity. Additio-nally, the person must have been granted permanent residence in the country prior to application. If these criteria are fulfilled and the person
KIrK SCOTT
108
has lived an “honourable” (free from convictions for major crimes) life in Sweden, then the only other obstacle is the amount of time lived in the country. Citizens of the Nordic countries must have lived in Sweden for 2 years prior to citizenship, individuals granted refugee status should have lived in the country 4 years, and all other foreign citizens must have lived in Sweden for at least 5 years.1
While acquisition of Swedish citizenship has become successively easier, there has existed one formal deterrent. Dual citizenship was forbid-den in Sweden until 1 July, 2001. Many individuals may see uptake of a citizenship as a fairly casual act, but renunciation of a citizenship is much more serious. Prior to 2001 dual citizenship was only allowed in those cases where individuals were citizens of countries which did not allow renunciation of citizenship, and in cases where children had parents of dif-ferent nationalities. With the changing of the law in 2001, no such demands were placed on applicants. Swedish citizens were free to acquire citizenship in another country and retain their Swedish citizenship as long as this was accepted in the second country. Immigrants were also free to acquire Swedish citizenship. This change in the law to allow dual, and even multi-ple, citizenship was far from universally approved of, with less than half of the electorate in favour, but the principle of dual citizenship received very little criticism after the fact (Gustafsson 2005).
Stylized Facts
In general, immigrants to Sweden have a fairly high naturalization rate, but this varies considerably by nationality. Country-specific details will be pro-vided below, and we will focus here on aggregate figures for the collective immigrant population. Figure 1 is interesting largely from the total lack of variation, and shows that the number of foreign citizens has remained roughly constant during the past decade, and the same holds for the number of naturaliza-tions. Foreign citizens have naturalized at the rate of between 30,000 and 40,000 per year. This breaks down to roughly 8% of the foreign citizens becoming Swedish citizens in any given year. This figure is based on the total number of citizens, and not on the number formally eligible for citi-zenship.
1 Foreign citizens married to Swedish citizens may naturalize after only 3 years in the country, provided that they have been married for at least 2 of those years.
THE ECONOMICS OF CITIZENSHIP:IS THERE A NATURALIZATIONEFFECT?
109
Figure 1. Number of foreign citizens and naturalizations in Sweden, 1999-2005
Source: Author's calculations based on Statistics Sweden register RTB.
One interesting point to note in Figure 1 is that the change in Swedish law allowing dual citizenship, which took effect July 1 2001, does not appear to have had any major impact on naturalization rates. The temporal aspects of naturalization for different nationalities are further discussed below.
BackgroundThis chapter examines the role of naturalization in the economic integra-tion of foreign-born2 individuals in Sweden. Many studies make use of the acquisition of citizenship as a tool to explain the economic success or failure of immigrants. These studies usually include citizenship as a binary explanatory variable under the assumption that citizenship should have some positive effect on employment, earnings, or both. Such studies rarely delve into the mechanisms behind this “citizenship premium,” however. Citizenship can affect economic integration in several ways, and the mechanism for the effects can vary. From a policy standpoint it is crucial to understand this mechanism, since it will determine the potential effects of various naturalization policies. If there is simply an unambiguous effect through which all individuals receive some sort of premium to citizenship acquisition, then liberal naturalization policies would be in order. Various
2 Throughout this chapter the terms “immigrant” and “foreign-born” will be used inter-changeably.
KIrK SCOTT
110
reasons can be found to explain this effect, such as restriction of various types of employment to citizens, or a lower propensity among natives to discriminate against immigrants who signal their “loyalty” to the host country through naturalization. There is also the distinct possibility that there is no naturalization premium, and the observed effect is simply an effect of selection. Those who opt to receive citizenship may simply be more positively self-selected than those who do not become citizens. This paper will delve into this question. This study will begin with a discussion of the pattern of Swedish citi-zenship acquisition for immigrants from 16 countries. It will then chart the effects of citizenship on employment and income, in the same way that many other studies have. Finally, it will make use of a longitudinal approach to attempt to isolate the naturalization effect from a selection effect.
previous Studies
There has not been an overwhelming amount of work done on the effects of naturalization, especially in light of the interest in immigrant and ethnic studies in general. It can be argued that there have been more studies on the determinants of naturalization (Alvarez 1987; DeSipio 1987; Jasso & rosenzweig 1986; Kelley & McAllister 1982; Portes & Curtis 1987; Yang 1994). In recent years this interest has, however, shifted and studies of the impact of naturalization have begun to appear in greater numbers. Bratsberg et al. (2002) examine the effects of naturalization on wage growth in the United States. They study a longitudinal sample of young male immigrants and find that there is a positive effect of naturalization on the economic integration of immigrants. They find evidence that wage growth accelerates after naturalization, and that the occupational structure shifts towards more white-collar and public-sector employment. Unlike Chiswick (1978), who is cited as a motivation for the study, they found a naturalization effect even after controlling for time spent in the United States. Studies of naturalization in Canada have come to mixed results, with DeVoretz and Pivnenko (2004) finding a positive impact of citizenship for Ukrainian immigrants to Canada. On the other hand, Mata (1999) finds no conclusive effect of naturalization on immigrants in general using the same cross-sectional data from the 1996 census. European studies are rare, but one study of the experiences of mig-rants from the former Yugoslavia in Sweden and Austria by Kogan (2003) finds that there could be multiple processes in the different European nations. Ex-Yugoslavs in Austria showed a positive effect of naturalization
THE ECONOMICS OF CITIZENSHIP:IS THErE A NATUrAlIZATIONEFFECT?
111
on labour market outcomes, while those in Sweden showed no real benefit. The explanation here is that the two countries have differing systems of rights for non-citizens, as discussed above. Another study examining Sweden specifically comes to another con-clusion. Using cross-sectional data from 1970 and 1990, Bevelander (2000) finds that naturalization status had negligible effects on probability of being employed in data from the 1970 census, but that naturalized immi-grants did have a higher propensity to be employed than foreign citizens in the 1990 census. Studies of the effects of citizenship all face one major obstacle. To acquire citizenship an immigrant must meet a number of criteria, one of which is almost universal: the demand that an individual live a certain number of years in the host country to be eligible for naturalization. This demand implies that there must be some sort of selection process at work, since some immigrants will return home or proceed to a third country prior to meeting the eligibility requirement. In addition, some immigrants choose to NOT become citizens in the destination country, as mentioned above, which adds another dimension of selection to the equation. The previous studies have come to varying conclusions as to the effects of naturalization, and this study will attempt to isolate some of them.
Immigrant Integration in General
Based on various sources of Swedish data, many studies have shown that labour force attachment among immigrants has been weakening over the past three decades, and the relative incomes earned by those immigrants actually in the labour force have been declining. Prior to 1970, immigrants exhibited economic performance similar to, if not better than, native-born Swedes with the same occupations. After 1970, there are indications that this shifting immigrant labour market performance is not merely a reflec-tion of shifting quality of immigrant cohorts, but also of shifting labour market conditions which adversely affect all migrants, even those from cohorts which were fairly successful in earlier years (Bengtsson & Scott 2006; Bevelander & Nielsen 2001; Ekberg & Gustafsson 1995; rooth 1999; Scott 1999; rosholm et al. 2006).
Citizenship acquisition in Sweden
Any study of the effect of citizenship must be based on an understanding of the extent of citizenship acquisition. To this end survivor plots have been calculated to show the percentage of the male immigrant population acqui-ring Swedish citizenship by years since migration. Figures 2.1-2.4 show the Kaplan-Meier estimates of the percentage of the immigrant stock without Swedish citizenship by years since migration for the period 1980-2001.
KIrK SCOTT
112
These figures show that the major immigrant nationalities in Sweden can be grouped into three categories of citizenship acquisition. Immigrants from Scandinavia, Germany, Italy and the United States all have relatively low rates of naturalization, and the naturalization process seems to be a fairly gradual one, with a relatively constant slope. These countries can basically be considered industrialized during the entire period and, with the exception of Finland, cannot be considered labour-surplus countries. The second category consists of the labour-surplus countries of Greece, Turkey and the former Yugoslavia. Here we can see high rates of citizen-ship acquisition, with a slightly faster rate in the period immediately follo-wing migration, and diminishing rates as time passes. The third category consists of the refugee-sending countries of Czechoslovakia, Poland, Iran, Iraq, Ethiopia, Chile, and Vietnam. Common for these nationalities is an extremely high rate of citizenship acquisition in the period immediately fol-lowing immigration. This can most likely be explained by the desire to gain the added protection afforded Swedish citizens.
Figure 2.1. Kaplan-Meier estimates of the survivor function. Scandinavian males, 1980-2001
Source: Author's calculations. Swedish longitudinal immigrant database.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40
Denmark Norway Finland
THE ECONOMICS OF CITIZENSHIP:IS THErE A NATUrAlIZATIONEFFECT?
113
Figure 2.2. Kaplan-Meier estimates of the survivor function. eastern and Northern european males, 1980-2001
Source: Author's calculations. Swedish longitudinal immigrant database.
Figure 2.3. Kaplan-Meier estimates of the survivor function. Southern european males, 1980-2001
Source: Author's calculations. Swedish longitudinal immigrant database.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40
Poland Czechoslovakia USA Germany
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
Greece Turkey Yugoslavia Italy
KirK Scott
114
Figure 2.4. Kaplan-Meier estimates of the survivor function. Non-European males, 1980-2001; Norway included for comparison
Source: Author's calculations. Swedish longitudinal immigrant database.
table 1 shows the share of each nationality reported as having acquired Swedish citizenship in the 1990 census. one interesting aspect to note here is that females of most nationalities naturalize at approximately the same or somewhat higher rates as their male counterparts.
Table 1. Share naturalized by gender and nationality, 1990
Male Femalechile 0.53 0.55
czech 0.94 0.89
Denmark 0.54 0.59
Ethiopia 0.77 0.74
Finland 0.59 0.60
Germany 0.67 0.79
Greece 0.71 0.76
iran 0.40 0.31
iraq 0.69 0.75
italy 0.48 0.62
Norway 0.43 0.45
Poland 0.80 0.73
turkey 0.42 0.27
USA 0.42 0.41
Yugoslavia 0.40 0.32
Source: Swedish Census 1990, contingent on having lived in Sweden 5+ years.
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40
Chile Ethiopia Iran Iraq Vietnam Norway
Figure 2.4 chapter 4
THE ECONOMICS OF CITIZENSHIP:IS THErE A NATUrAlIZATIONEFFECT?
115
recent changes in naturalization patterns are illustrated in Figure 3 by showing the number of foreign-born individuals residing in Sweden, and the number of those who have retained their foreign citizenships. What is clear is that there has been an increase in propensity to become a Swedish citizen, and that this increase has occurred largely between 1990 and 2000. In 1980 approximately 2/3 of all foreign-born living in Sweden had retai-ned their foreign citizenship, while by 2000 this figure had fallen to less than half. Note that this diagram does not account for years in Sweden, and as such includes individuals who are not yet eligible for citizenship. It does, however, show a tendency towards increasing uptake of citizenship during a period when dual-citizenship was not yet allowed.
Figure 3. total foreign-born population and number of foreign citizens residing in Sweden, 1980, 1990 and 2000
Source: Author’s calculations based upon Statistics Sweden Census of Housing and Popu-
lation, 1980, 1990 and 2000.
DataThis study uses two primary data sources. The first is the Swedish census of 1990, and the second is the Swedish longitudinal Immigrant database (SlI). The SlI is a longitudinal database created through the merging of a number of registers kept by Statistics Sweden, and was the result of coope-ration between Statistics Sweden, the Swedish Immigration Board, and the research Group in Population Economics at the Department of Economic History in lund. The database consists of a sample of immigrants from sixteen European and non-European countries. This sample begins in 1980
1200000
1000000
800000
600000
400000
200000
01980 1990 2000
Foreign Citizen Foreign-Born
KIrK SCOTT
116
and extends through 2001. The sample contains detailed continual infor-mation on approximately 550,000 individuals. The data is restricted to individuals aged 25 to 59 who have been in Sweden for at least five years. This is to ascertain that each immigrant is eligible for citizenship. This restriction implies that the first selection mechanism – remaining in the host country at least five years – has been ignored. Although the SlI contains continual information on many variables, the economic information is reported once per year. For simplicity the database has been recoded into a series of repeated cross-sections, since the estimation techniques will not require exact timing of citizenship acquisi-tion, allowing the use of yearly data.
analysisThe effects of naturalization are estimated through two processes. The first is the effect of naturalization on the probability of being gainfully employed, while the second is the effect of citizenship status on earnings for those who have some type of employment.
effects of Citizenship on employment
If there is a naturalization effect, one would expect that immigrants gaining Swedish citizenship should have a higher probability of being employed. Using the 1990 census, the probability of being employed full-time was estimated using a logistic regression. Table 2 shows the predicted proba-bility of having full-time employment calculated at Swedish mean values for the regressors. Since we are examining immigrants, the value for years since migration was set at ten. This table shows that those immigrants who have been naturalized tended to see a modest increase in the probability of having full-time employment in 1990. Some nationalities, particularly the Scandinavians, Chileans and Germans, saw almost no difference between the naturalized and non-naturalized groups. The Czechoslovakian men saw the greatest increase in probability of almost 24 percentage points, while American men and Greek women experienced considerable negative effects of citi-zenship.
THE ECONOMICS OF CITIZENSHIP:IS THErE A NATUrAlIZATIONEFFECT?
117
table 2. estimated probability of being full-time employed in 1990 using census data. Calculated at Swedish means, YSM=10. Logit contingent on YSM≥5, ages 25-59*
Men Women
Foreign Naturali-zed
“Naturali-zation Effect”
Foreign Naturali-zed
“Naturali-zation Effect”
Chile 0.75 0.75 0.00 0.61 0.59 -0.02
Czech 0.42 0.66 +0.24 0.42 0.48 +0.05
Denmark 0.77 0.74 -0.02 0.40 0.50 +0.10
Ethiopia 0.64 0.69 +0.06 0.55 0.52 -0.04
Finland 0.75 0.71 -0.04 0.56 0.56 0.00
Germany 0.78 0.76 -0.02 0.43 0.49 +0.07
Greece 0.59 0.65 +0.06 0.50 0.38 -0.11
Iran 0.61 0.64 +0.03 0.39 0.44 +0.05
Iraq 0.47 0.62 +0.15 0.40 0.49 +0.09
Italy 0.71 0.68 -0.03 0.47 0.43 -0.04
Norway 0.71 0.77 +0.05 0.39 0.52 +0.12
Poland 0.61 0.68 +0.07 0.44 0.53 +0.09
Turkey 0.56 0.54 -0.01 0.36 0.41 +0.06
USA 0.69 0.57 -0.12 0.34 0.30 -0.04
Yugo. 0.68 0.63 -0.05 0.54 0.53 +0.00
* Control for age, age squared, marital status, year since migration, years since migration squared, educational level. YSM = Years since migration.Source: Swedish Census 1990.
Many studies would take the results of Table 2 as evidence of some type of naturalization premium, but cross-sectional data such as this do not allow for any real examination of selection effects. To this end this study turns to the longitudinal data available in the SlI (see data section). Using data from 1980 to 2001 two separate analyses of employment were carried out. The dependent variable in these regressions is a dichotomous variable sig-nifying the acquisition of employment paying at least 3 base amounts3 in a given year. The lower income-bound set in this analysis was necessary, since we do not have information on hours worked, and can thus not identify full-time employment in the SlI. This income term should be considered more along the lines of an employment rate.
3 “Many of the benefits within social insurance are linked to the so-called base amount. The base amount is also used to calculate the pensionable income, pension points and maximum levels within social insurance. It is an index of price movements, which means that benefits follow price trends. Price trends are measured in the consumer price index.” National Social Insurance Board web page: http://www.rfv.se/english/social/base_k.htm
KIrK SCOTT
118
The data is then subjected to two analyses, one with all immigrants included, and one where the dataset is restricted to individuals who at some time acquire citizenship. Both datasets include only those having been in the country for 5 or more years. By only examining those that eventually become citizens we are violating a fundamental prohibition of estimating models conditioned on future events, but this approach can hopefully isolate some of the effects of selection into citizenship. Since we have data on individuals during multiple observation points the estimation technique should take account of unobservable individual characteristics as well as those which we can identify. As such this study employs a random effects logit as the estimator of choice.
Table 3 shows the results of the model, including both citizens and those who never become citizens. The probabilities presented are calcula-ted at Swedish mean values for the independent variables, with years since migration set at 10. Using longitudinal data the naturalization premium appears at first glance to be even more widespread than using the cross-sectional data. The only nationality which does not have a positive or neutral effect of naturalization is the USA.
This effect could still be one of selection, however. If the increased pro-bability of obtaining employment is a function of increased opportunities available to citizens or the decreased effect of discrimination on naturalized immigrants then we should expect to see a similar effect of citizenship in a sample selected on the prospective knowledge that the individuals will eventually become citizens. If, however, the naturalization effect is simply the result of naturalization serving as a variable identifying unobservable selection characteristics, then the effect of citizenship on employment should be lower than in the full sample. Table 4 shows the predicted probabilities for individuals in a sample restricted to those who at some point become citizens. This table shows that much of the naturalization effect noticeable in both the census data and an initial look at the longitudinal data is in fact a selection effect. When the sample is restricted to those who we know will become citizens the effect of naturalization becomes much smaller, and in roughly half of the cases actually becomes zero or negative.
THE ECONOMICS OF CITIZENSHIP:IS THErE A NATUrAlIZATIONEFFECT?
119
table 3. estimated probability of being full-time employed in 1990 using longitudinal data. Calculated at Swedish means, YSM=10. Logit contingent on YSM≥5, ages 25-59*
Men Women
Non-
Natura-
lized
Natura-
lized
“Naturali-
zation
Premium”
Non-
Natura-
lized
Natura-
lized
“Naturali-
zation
Premium”
Chile 0.92 0.93 +0.01 0.91 0.92 +0.01
Czech 0.84 0.90 +0.06 0.82 0.88 +0.06
Denmark 0.95 0.97 +0.01 0.89 0.90 0.00
Ethiopia 0.77 0.86 +0.09 0.78 0.88 +0.10
Finland 0.94 0.94 0.00 0.93 0.93 0.00
Germany 0.93 0.95 +0.03 0.82 0.85 +0.03
Greece 0.81 0.81 0.00 0.73 0.69 -0.05
Iran 0.68 0.78 +0.09 0.58 0.72 +0.13
Iraq 0.69 0.78 +0.09 0.67 0.78 +0.11
Italy 0.89 0.92 +0.03 0.77 0.74 -0.03
Norway 0.93 0.94 +0.02 0.89 0.93 +0.04
Poland 0.88 0.91 +0.03 0.86 0.89 +0.03
Turkey 0.77 0.80 +0.04 0.66 0.71 +0.06
USA 0.73 0.70 -0.03 0.52 0.67 +0.15
Yugo. 0.91 0.94 +0.02 0.90 0.91 0.00
*Control for years since migration, years since migration squared, age, age squared, citi-zenship, educational level, year of measurement. Full regression results available upon request. YSM = Years since migration.Source: SLI.
KIrK SCOTT
120
table 4. estimated probability of being full-time employed in 1990 using longitudinal data. Calculated at Swedish means, YSM=10. Logit contingent on YSM≥5 and Swedish citizenship at some point, ages 25-59*
Men Women
Non-
Natura-
lized
Natura-
lized
“Naturali-
zation
Premium”
Non-
Natura-
lized
Natura-
lized
“Naturali-
zation
Premium”
Chile 0.94 0.94 0.00 0.93 0.93 -0.01
Czech 0.88 0.92 +0.04 0.86 0.89 +0.03
Denmark 0.97 0.97 0.00 0.91 0.90 -0.02
Ethiopia 0.80 0.87 +0.07 0.86 0.92 +0.06
Finland 0.95 0.94 -0.01 0.94 0.92 -0.01
Germany 0.95 0.96 +0.01 0.90 0.88 -0.02
Greece 0.92 0.88 -0.04 0.89 0.80 -0.08
Iran 0.77 0.82 +0.05 0.67 0.76 +0.09
Iraq 0.74 0.80 +0.06 0.73 0.81 +0.08
Italy 0.94 0.92 -0.03 0.89 0.83 -0.06
Norway 0.98 0.98 -0.01 0.88 0.92 +0.04
Poland 0.92 0.92 0.00 0.90 0.91 0.00
Turkey 0.81 0.82 +0.01 0.65 0.68 +0.03
USA 0.89 0.73 -0.16 0.81 0.86 +0.05
Yugo. 0.94 0.94 +0.01 0.91 0.92 0.00
*Control for years since migration, years since migration squared, age, age squared, citi-zenship, educational level, year of measurement. Full regression results available upon request. YSM = Years since migration.Source: SLI.
effects of Citizenship on earnings
Once employment has been secured, it is possible that a naturalization effect could also exhibit itself in earnings. The argument here is similar to the argument for a naturalization premium in employment. Jobs with higher wages or better career opportunities may be limited to citizens, or employers may choose to hire those who have demonstrated a desire to stay in the country. If this is so, then we should see a positive effect of citi-zenship on earnings. Table 5 shows the effects of citizenship on earnings in the 1990 census. Many nationalities exhibit earnings premiums of the mag-nitude of only a few percent, while others see large increases in income for those with citizenship.
THE ECONOMICS OF CITIZENSHIP:IS THErE A NATUrAlIZATIONEFFECT?
121
table 5. effects of naturalization on income, ages 25-59, 1990, contingent on YSM≥5 and full-time employment
Male Female
Chile 0.07*** 0.06**
Czech 0.27*** 0.14**
Denmark 0.00 0.05***
Ethiopia 0.22*** 0.14*
Finland 0.01*** -0.02***
Germany 0.04*** 0.06***
Greece 0.05** 0.04
Iran 0.19*** 0.43***
Iraq 0.08* 0.28**
Italy 0.03 -0.01
Norway 0.00 0.07***
Poland 0.15*** 0.13***
Turkey 0.02 -0.04
USA 0.10** -0.04
Yugo. 0.07*** 0.00
Note: * - 10%, ** - 5%, *** - 1%. YSM = Years since migration.Source: Swedish Census 1990.
Again we have the question of if these increases are actually due to citizen-ship itself, or to the fact that those who eventually become citizens are positively selected and, therefore, some other characteristic is driving the results. As with the examination of employment, longitudinal data is needed to answer this question. Table 6 presents the results of a random effects generalized-least-squa-res regression using data from the SlI. The table presents the effect of natu-ralization on income for two different samples of men. The first sample consists of all men in Sweden more than 5 years, while the second sample is restricted to men who receive citizenship at some point between 1980 and 2001. The idea here is similar to the employment equations: to separate selection and citizenship effects. The difference between these two figures is also reported. Here we can see that the selection effect on wages seems somewhat ambiguous.
KIrK SCOTT
122
table 6. effects of naturalization on income, males, ages 25-59, 1980-2001, contingent on YSM≥5*
Men Women
Total
Sample
Naturalized
Sometime
“Naturali-
zation
Premium”
Total
Sample
Naturalized
Sometime
“Naturali-
zation
Premium”
Chile 0.02 -0.01 -0.03 0.01 -0.01 -0.02
Czech 0.03 0.09 +0.06 0.08 0.09 +0.01
Denmark 0.00 0.03 +0.03 -0.03 -0.04 -0.01
Ethiopia 0.04 0.06 +0.02 0.03 0.02 -0.01
Finland -0.01 -0.02 -0.01 0.00 0.00 0.00
Germany -0.01 0.03 +0.04 0.03 0.04 +0.01
Greece 0.02 -0.02 -0.04 -0.06 -0.08 -0.02
Iran 0.02 0.02 0.00 0.02 0.02 0.00
Iraq 0.03 0.04 +0.01 0.01 0.00 -0.01
Italy 0.05 0.03 -0.02 0.04 0.04 0.00
Norway -0.02 -0.04 -0.02 0.02 0.03 +0.01
Poland 0.01 0.01 0.00 0.01 0.00 -0.01
Turkey -0.01 0.00 +0.01 0.00 0.00 0.00
USA -0.08 -0.08 0.00 0.06 0.07 +0.01
Yugo. 0.01 0.00 -0.01 -0.02 -0.02 0.00
*Control for years since migration, years since migration squared, age, age squared, citi-zenship, educational level, year of measurement. Full regression results available upon request. YSM = Years since migration.Source: SLI.
While the employment analysis showed a clear difference in predicted pro-babilities of employment between the full and restricted samples, the income analysis shows very little difference. To gain a further understan-ding of the effect on naturalization we calculate earnings profiles for indivi-duals taking years since citizenship into account, following Bratsberg et al (2002). If citizenship itself is actually an important factor in income deve-lopment then we should expect to see an upward shift in the slope of the earnings profile at the point of citizenship. Tables 7 and 8 present the per-centage change in income from ages 25 to 45 in fixed prices for various standard individuals, having primary school education, being married, and entering Sweden at age 25. The calculations were then made with two vari-ations. One where the individual received Swedish citizenship after 10 years in Sweden, and one where the individual did not receive Swedish citi-zenship. These calculations were carried out on the entire sample available in the SlI, and on a restricted sample conditioned on having received Swedish citizenship at some time between 1980 and 2001. To clarify, a
THE ECONOMICS OF CITIZENSHIP:IS THErE A NATUrAlIZATIONEFFECT?
123
counterfactual probability is predicted for the restricted sample, showing their estimated earnings growth in the absence of naturalization.
table 7. percentage difference in income between ages 25-45. estimated at becoming citizens at YSM=10 & Never Becoming Citizen. Males*
Pooled Sample,
All Individuals
restricted Sample,
Become Citizens
Never
Citizens
Citizens
after 10
Years Difference
Never
Citizens
Citizens
after 10
Years Difference
Chile 1.22 1.35 0.14 1.26 1.31 0.05
Czech 1.54 1.55 0.01 1.60 1.56 -0.04
Denmark 1.26 1.18 -0.08 1.17 1.12 -0.05
Ethiopia 1.15 1.34 0.19 1.14 1.34 0.20
Finland 1.13 1.14 0.02 1.13 1.11 -0.02
Germany 1.42 1.36 -0.06 1.29 1.36 0.08
Greece 1.12 1.13 0.01 1.27 1.24 -0.03
Iran 1.20 1.81 0.61 1.25 1.85 0.60
Iraq 1.07 1.42 0.35 1.13 1.49 0.35
Italy 1.32 1.48 0.17 1.41 1.56 0.15
Norway 1.35 1.35 0.00 1.60 1.59 0.00
Poland 1.28 1.31 0.03 1.42 1.39 -0.03
Turkey 1.16 1.24 0.07 1.19 1.25 0.05
USA 1.50 1.63 0.13 1.63 1.79 0.16
Yugo. 1.09 1.13 0.05 1.07 1.13 0.06
*Control for years since migration, years since migration squared, age, age squared, citi-zenship, educational level, year of measurement. Full regression results available upon request. YSM = Years since migration.Source: SLI.
KIrK SCOTT
124
table 8. percentage difference in income between ages 25-45. estimated at becoming citizens at YSM=10 & Never Becoming Citizen. Females*
Pooled Sample,
All Individuals
restricted Sample,
Become Citizens
Never Citi-
zens
Citizens
after 10
Years Difference
Never Citi-
zens
Citizens
after 10
Years Difference
Chile 1.25 1.19 -0.06 1.33 1.15 -0.18
Czech 1.35 1.24 -0.10 1.57 1.24 -0.32
Denmark 1.20 0.83 -0.37 1.37 0.83 -0.54
Ethiopia 1.09 1.16 0.07 1.03 1.20 0.17
Finland 1.28 0.95 -0.33 1.28 0.92 -0.36
Germany 1.25 1.00 -0.25 1.24 1.02 -0.23
Greece 1.15 0.94 -0.21 1.25 1.05 -0.20
Iran 1.19 1.59 0.40 1.19 1.62 0.42
Iraq 1.14 1.24 0.10 1.15 1.27 0.11
Italy 1.16 1.28 0.12 1.19 1.25 0.07
Norway 1.28 0.91 -0.37 1.39 1.13 -0.26
Poland 1.31 1.10 -0.21 1.39 1.16 -0.23
Turkey 1.15 1.15 0.00 1.27 1.15 -0.11
USA 1.30 1.30 -0.01 0.72 1.44 0.72
Yugo. 1.19 1.00 -0.20 1.35 0.98 -0.37
*Control for years since migration, years since migration squared, age, age squared, citi-zenship, educational level, year of measurement. Full regression results available upon request. YSM = Years since migration.Source: SLI.
The results in Tables 7 and 8 show somewhat more than those in the reg-ressions using citizenship as a dummy. These tables can generally be inter-preted in the following manner: if the percentage point difference reported for the pooled sample is negative, then there is a negative selection into citizenship. Negative selection in this respect means that individuals attai-ning citizenship after 10 years will on average have a slower income pro-gression than those who never attain citizenship. The column showing the differences for the restricted sample then becomes important. If there is a negative effect of citizenship for the entire sample, how does this affect those who will eventually obtain citizenship? If this value is positive we can assume that for this negatively selected group, citizenship does seem to improve earnings. For the case where the difference is positive for the pooled sample, the effect in the restricted sample can give us some idea of the importance of citizenship versus selection. If the difference is lower in
THE ECONOMICS OF CITIZENSHIP:IS THErE A NATUrAlIZATIONEFFECT?
125
the selected sample than in the pooled sample, we can assume that there is a naturalization premium, but that this effect is blurred with selection effects in the pooled sample. An examination of the data is not easy. If one examines those nationa-lities with a positive effect of naturalization in the full sample, there appears to be a tendency for a lower effect in the restricted sample. This indicates that the citizenship “premium”, while not non-existent, may not be as strong as suspected. This pattern is strongest for men, and it also appears that men have the most positive effect of naturalization, with 12 nationali-ties seeing positive or neutral effects in the full sample, and 10 in the restric-ted sample. Women appear to have a different pattern, with the majority of the nationalities showing negative effects in both samples. In addition, there tends to be a process whereby those nationalities exhibiting negative effects of nationality in the full sample show even greater negative effects when the sample is restricted to those who will eventually naturalize.
Conclusion: Selection or Citizenship effect?After examining the tables above, it is time to confront the actual question: Is there a naturalization premium the form of labour market advantages for immigrants who acquire Swedish citizenship? The answer from the above must be a resounding “maybe.” It seems that there really is no citizenship premium on obtaining employment. In fact, when eliminating some selection effects by restricting the sample to those who do actually naturalize at some point, we can see that the effects of citizenship are negligible or even negative in many cases. The effect of citizenship that turns up in cross-sectional data appears to be caused by characteristics inherent in the group which naturalizes, and not in the state of citizenship itself. This could be caused by aspiring citizens investing more in country-specific human capital with the intent of remain-ing in the country. This investment seems to pay off regardless of actual attainment of citizenship, however, which points against hypotheses which state that citizenship may be used as a device to signal intent to stay in the host country. The fact that citizenship often shows a negative relationship to employment probability is something that requires future examination. The Swedish social system is constructed in such a manner that very few benefits require citizenship. Examples do exist, however, and they are often related to the ability to take Swedish benefits to another country. Certain portions of the Swedish pension system fell into this category during the examination period, as did disability insurance. It may be that the results in those countries exhibiting negative effects of citizenship are driven by
KIrK SCOTT
126
people desiring to take their benefits back to their home country. Speaking against this, however, is the fact that Sweden did not allow dual citizenship during this period, which could make return to the home country difficult. More research is needed on this topic, with emphasis on labour market status such as pension, sickness benefits, etc. The study of income development following citizenship, on the other hand, shows that there is indeed some premium associated with naturaliza-tion. This premium is, however, generally small in terms of differential income development, with those who naturalize showing earnings growth of only a few percentage points higher than non-naturalized immigrants over a twenty-year period. In addition, this effect seems to be largely confi-ned to male immigrants. The final conclusion of this paper is that the naturalization “premium” is largely a question of selection on the part of the individual, not a ques-tion of rights to employment or differential discrimination based on citi-zenship. This is clear from the fact the effects of “citizenship” are evident even prior to the actual naturalization itself. Going out on the policy implication limb, it seems that there are few incentives for the government to ease access to citizenship. Instead, indica-tors point to the need for increased access to training programs which provide the foreign-born with Swedish-specific human capital. This conclu-sion is in line with the results of most economic studies of immigrant inte-gration in Sweden.
127
An American Narrative1
I was born in Saigon in 1968, the sixth of nine children. Before 1975, I can describe my life as a good life, because my father was in the army, and he [was] one of the high-ranking officers in the military… And my family can be described as middle-class income. So I didn’t see much difficulty… we were very fortunate. It became hard right after 1975 when the Communists took over, because my father [was taken] away to prison, what they called “re-educa-tion.” And he passed away a year after…my family [has] to answer many questions to the government, and they wanted to isolate us to another, very wild, place. Also, with the way the Communists do it, especially for men, they wanted to destroy three generations. I [finally] left Vietnam in 1986, December of 1986 [after 19 tries]. Spent about eight days on the boat. We ended up in Thailand. I stay[ed] at first in a coconut farm for about three months, waiting for the American Embassy to get involved. After the paperwork is done, they send us to one of the refugee camps. …I stayed there a little bit over a year. You go through all the interview process, and because my father was in the army for the American and Vietnamese Republics, the American Embassy took a look into my application. And fortunately I was passed with all five different interviews. And [they] promised me to come to the United States. But before you go to the United States, the Americans will transfer you to another refugee camp, which is in the Philippines for about six months. We only meet with the INS at the end of the time. So they sit down, interview you, so they can figure out where you will be staying when you come to United States. I say, “I’m free. You know, my goal – the reason I left Vietnam—is for two reasons. First, for the freedom. I wanted to pursue freedom. And the second, important for me, is to continue with educa-tion.” …they come back, and they tell me, “How about Boston?” I say in answer, “Any state. Wherever [I can] pursue my goals.” And in March of 1988, I arrived to Logan Airport [in Boston].
…
By law, if you are [an] immigrant, you have to wait for five years to become a citizen. And I wait for every single day to become [one]. I mean, until five years. I became a citizen right away. It take me only three minutes to pass the test. Even though I studied for day and night to prepare for [the] hundred questions and learn about history. But most fortunately, because
1∗Adapted from an interview reported in Irene Bloemraad’s Becoming a Citizen: Incorporating Immigrants in the United States and Canada (UC Press, 2006). With author’s permission.
Ather h. AkBArI
128
I spend years in school, you learn about the American system of politics. So I came in, and I think they already pre-approved me. They opened up [my file], and they said, “Well, did you know that an amendment can be changed?” Before I answer, he said, “Yes, isn’t it?” I said, “Yes”, and he said, “Okay, you passed.”
…
I think, if you know about the American system, if you know about what is the life here, what is the benefit as [an] American citizen, you can’t wait to become [an] American citizen, because this is your right. You become like regular people. You pay tax, you go to vote – that’s your voice. And you live here. You will establish your life in here. This is America. …I see that my life will be probably pursued here for a long period of time, and I said, “Well, in order to get all the rights, the benefits, and the voice, you need to be a part of the community and society.”
129
Chapter 5
IMMIGraNt NatUraLIZatION aND ItS IMpaCtS ON IMMIGraNt LaBOUr MarKet perFOrMaNCe aND treaSUrY
Ather H. Akbari
Introductionthe United States of America has been a popular immigrant destination for centuries. According to the United Nations’ estimates, around 190 million people, or 3 percent of the world’s population, now reside in a country where they were not born. two-thirds of these people live in Australia, Canada and the United States, with the majority in the latter. Over the period 1990-2005, the percentage of foreign-born residents in the United States relative to the United States’ total population rose dramatically from 7.9 per cent to 12.1 per cent. this recent large-scale inflow of immigrants to the United States has generated a far-reaching public debate in the United States over immigrant economic and social integration. this chapter con-tributes to that debate through an analysis of one integration tool, citizen-ship acquisition.
A wide body of literature has emerged to analyze the economic perfor-mance of immigrants to the United States2 but one caveat in this literature is that it does not distinguish between the legal status of immigrants (which is defined in this paper as naturalized citizens), legal immigrants, legal non-immigrants, and undocumented migrants. As Massey and Bartley (2005) have noted, the economic performance of these status groups may differ from each other due to their differential treatment under United States law, which in turn may affect their economic opportunities. Since 1996, changes in United States immigration laws have resulted in penalties being imposed on undocumented migrants and legal non-immigrants while increasing the relative advantages for naturalized citizens and legal immigrants.3 At the
2 For example, Blau (1984); Simon (1984); Borjas & trejo (1991); Borjas & hilton (1996); Borjas (1999).3 the effectiveness of these sanctions however has been questioned.
Ather h. AkBArI
130
same time, changing economic and political conditions around the world have resulted in an increase in the relative share of undocumented migrants and legal non-immigrants in the total foreign-born population of the United States. It should be noted that separate data on the socioeconomic charac-teristics of these two status groups are not available from either the popu-lation census nor from the Current Population Survey micro data, both of which have been used extensively to analyze the economic performance of immigrants. Aggregating across all groups, however, (regardless of immi-grant status), much of the recent research on the economic performance of this overall immigrant stock has documented a declining economic perfor-mance of immigrants.4
Given the focus of this book and the noted data limitations, in the present chapter I analyze the economic performance of only one group of immigrants in the United States, i.e., naturalized citizens. Of course, this is the most legally-advantaged group of United States immigrants. I will also provide comparative results for non-citizens, and the native-born citizen population. I argue in particular that if the results of the present study, which concentrates on citizenship acquisition, differ from the conventional immigration research as cited above, then future research on the economic performance and impact of immigration must pay attention to the different legal status of this segment of the United States population.
I next discuss the possible factors that motivate immigrants to acquire citizenship in the United States. I follow this discussion with an analysis of the impact of some well-known demographic and economic characteristics on the decision to acquire United States citizenship and note that citizen-ship status may affect the economic performance of naturalized immi-grants. In a fourth section I analyze the impact of naturalization on the labour market earnings of foreign-born individuals by their citizenship status. this analysis is motivated by Borjas (2002a and 2002b) who obser-ves that welfare participation rates of naturalized citizens are greater than those of non-citizens and so, in consequence, this analysis of post-citizen-ship income can provide a measure of a representative naturalized immigrant’s tax payments as an offset for the presumed increased receipt of public money. I then present my results followed by some policy conclu-sions derived from this study.
4 For example, Borjas (2002a).
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
131
Motivation for citizenship acquisition for a United States immigrantUnder the Immigration and Nationality Act (INA, 1965), the United States Congress has established certain requirements that must be fulfilled before ascension to citizenship. In sum, these requirements include a period of residency and physical presence, good moral character, attachment to the constitution, language, an adequate knowledge of United States govern-ment and history, pledging an Oath of Allegiance and the renouncing of citizenship of the country of origin.5
Just as in other countries, United States citizenship has its attendant privileges. A citizen can vote, be elected to public office, serve on a jury, enjoy government protection while traveling abroad, and -- most importantly for most immigrants -- gain expedited entry for family members to the United States. In addition, certain government benefits that are not available to non-citizens are available to citizens who can also meet potentially preferential tax requirements that are different from those of non-citizens.6
Despite the above-noted privileges that are available to citizens, the economic value of citizenship acquisition in the United States has diminis-hed over time due to the increasing rights of permanent residents and more relaxed naturalization rules. For example, in 1986 the required residence time in the United States necessary for a United States citizen married to an alien to be able to automatically transmit United States citizenship to a child born abroad was reduced from ten years to five years. Also, the period of time after naturalization before a naturalized citizen can reside abroad was also reduced from five years to one year in 1986. Until 1994, a natura-lized citizen could lose citizenship by setting up a permanent residence abroad within one year of obtaining United States naturalization. this pro-vision was repealed by Congress in 1994. Furthermore, as Yang (1994) has noted, occupations including practicing as an accountant, architect, attor-ney, dentist, physician, private detective, funeral director, liquor dealer, etc. that were available to only United States citizens in the past, can now be practiced by all permanent residents, including citizens and non-citizens.7
to offset some of the dilution in the economic value of citizenship ascension, the United States Congress introduced the Personal responsibi-
5 For a review of current rules of naturalization in the United States, see Bloemraad (2006).6 http://uscis.gov/gtraphics/citizenship/becoming/htm 7 two other examples of relaxing citizenship rules include 1) rights granted to an unmar-ried mother to transmit her United States citizenship to a child born abroad, and 2) faster eligibility to gain citizenship for immigrant military personnel who enlisted after Septem-ber 11, 2001 (the $320 application fee was also waived and soldiers were permitted to attend naturalization ceremonies at military bases and consulates abroad).
Ather h. AkBArI
132
lity and Work Opportunity reconciliation Act (PrWOrA) in 1996. that Act excluded non-citizens, even if permanent residents, from welfare pro-grams and allowed the states to introduce further restrictions.8 In the same year, a federally-funded Illegal Immigration reform and Immigrant responsibility Act (IIrIrA) was introduced which made it more difficult for legal non-naturalized immigrants to sponsor their relatives for entry to the United States, which, of course, placed a premium on citizenship acqui-sition to accelerate family reunification.9
the results of the above two reforms created a flux of applications for naturalization toward the end of the 1990s. At the same time, a new program for Green Card renewal after ten years and the eligibility for citi-zenship of the 3 million people legalized under the Immigration reform and Control Act of 1986 were additional motivating factors for ascension to United States citizenship. this outcome generated a new debate in the United States’ political and academic circles over the motivation to natura-lize, as discussed below.
Borjas (2002a) holds that eligibility for participation in welfare pro-grams is a strong motive for naturalization. his study shows that welfare participation rates among households headed by naturalized citizens rose after 1996 and at the same time a decline was observed among households headed by non-citizens.10 this result led Borjas to conclude that “many immigrants will become citizens not because they want to fully participate in the United States’ political and social systems, but because naturaliza-tion is required to receive welfare benefits.”11 In particular, Balistreri and hook (2004) examined whether or not the 1996 welfare reforms changed the social and economic determinants of naturalization among Mexican immigrants. Based on the Current Population Survey data from 1994/95 to 2000/01, their econometric study showed that the social and economic determinants of naturalization did not change in the case of Mexican immi-grants. these authors also analyzed the impact of IIrIrA and concluded that policies that restricted an immigrant’s ability to sponsor relatives for legal migration encouraged the propensity to naturalize.12
8 the PrWOrA had the broader objective of reducing the potential work disincentive effects of welfare schemes and applied to all United States residents.9 For further details of how new naturalization rules compare with those in the past, see Jasso and rosenzweig (1990) for rules that applied in the 1980s and to Bloemraad (2006) for more recent rules. 10 Welfare payments included cash benefits, Medicaid, and food stamps.11 Before this study, espenshade et al. (1997) had drawn similar implications derived from the welfare reforms.12 Prior to the above two studies, Yang (1994) had also noted that accelerated proces-sing of family reunification immigrants to the United States was a very strong motivation for immigrants in general to acquire United States citizenship. An immigrant permanent resident who is a non-citizen can only sponsor his or her spouse and unmarried children as immigrants within a numerically limited category. Sponsorship under this category
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
133
A third and a more recent motivating factor for naturalization in the United States may be the anxiety arising from the attack of September 11, 2001. this event resulted in strict anti-terrorist measures that may have prompted many foreign-born United States residents to become citizens due to the perception that doing so might help avoid unpleasant integra-tion experiences. however, no evidence of this exists in the literature to date.
In summary, the recent literature portrays a complex set of motives (economic, social and political) that motivate recent United States immi-grants to ascend to citizenship. I will, however, argue that many of these forces can be characterized as either increasing the costs on non-citizenship or raising the benefits of citizenship acquisition, and I will develop a broad economic model to incorporate these features.
econometric analysis of Citizenship ascension Since it is the thesis of this chapter that a cost-benefit analysis best informs citizenship acquisition, I will employ a modified human capital model to further the analysis. this model argues that an individual’s demographic, social and economic characteristics, as well as the economic conditions in the immigrant’s country of origin, determine citizenship acquisition in the United States. At least two reasons can be advanced to highlight the importance of analyzing the citizenship acquisition decision in the United States. First, in the presence of post-naturalization benefits, it could be argued that naturalization policies partially determine the long-term bene-fits of immigration for the individual as well as for United States society as a whole. Second, since United States citizens are allowed to sponsor their foreign-born parents, siblings, spouses and children, a greater number of immigrants ascending to United State citizenship could result in an increase in immigrant applications from this section of its foreign-born population. If this is the case, then greater participation in the political process and in the country’s welfare programs can be expected. For this reason, naturali-zation rates are of crucial interest to politicians, who may like to see greater numbers of voters in their constituencies.
Demographic characteristics that may affect an individual immigrant’s naturalization decision are reflected in this study by four variables, namely: age, and a set of three dummy variables representing marital status, gender,
accounts for only 26 percent of all numerically limited categories. On the other hand, a citizen can sponsor not only his or her immediate relatives with no numerical restriction, but also their unmarried or married adult children and their siblings. Such sponsorships make up about 54 percent of all numerically limited categories. In sum, the earlier the immigrant becomes a citizen, the shorter the waiting time for his / her relatives before immigrating to the United States.
Ather h. AkBArI
134
and the presence of children under the age of 15. An older immigrant – up to a point – is more likely to naturalize as the resulting benefits increase and the costs of forgoing their previous citizenship decrease with age. Yang (1994) has further argued that a married immigrant is more likely to natu-ralize than a single person because such a person exists in a more stable environment and also because families have more extensive contacts with United States institutions, thereby making it easier to integrate.
Some research has focused on the role of gender in the naturalization process, with ambiguous results. Some studies report that the rate of natu-ralization is significantly higher among women than among men (Alvarez 1987; Yang 1994). Portes and Curtis (1987) have also shown that the pro-pensity to naturalize increases with the presence of children in the house-hold. In contrast, Yang argues that men would be more likely to naturalize as they may receive greater economic benefits from ascension. the effect of this variable, therefore, is theoretically ambiguous.
economic and labour market characteristics that may influence an immigrant’s ascension decision are reflected in this study by the immigrant’s earned income, weeks worked during the year, and a set of three dummy variables representing education, occupation, and home ownership status. these variables in total reflect the immigrant’s degree of economic adapta-tion to the United States and consequently raise the costs of returning home.
Citizenship acquisition hinders an immigrant’s ability to return home in the absence of dual citizenship, and Portes and rumbaut (1990) have argued that less favorable conditions in an immigrant’s country of origin also reduce the reversibility of a migration flow and therefore increase the probability of naturalization. So, in order to incorporate the effect of eco-nomic conditions of the country of origin, I have included the ratio of GDP per capita in country of origin to that in the United States.13 In sum, I argue that the cost of naturalization is inversely related to this ratio.
the probability of naturalization may also be a function of time spent in the United States as greater social and economic integration is expected with extended length of stay and hence the number of years since migra-tion variable is also added to my model, implying that as more time passes it is more expensive to return home.
Finally, many authors have suggested that an immigrant’s propensity to naturalize in the host country is also affected by dual citizenship laws in his/her country of origin as well as in the host country. If an immigrant must sacrifice their citizenship in their country of origin to ascend to United
13 A more precise statement of sending country conditions would also incorporate the effect of the political conditions in the country of origin. One may expect this to be vaguely captured by the country’s GDP per capita.
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
135
States citizenship, this may impose two costs on the individual: first, the individual could lose access to labour market opportunities, wealth holding and social benefits in the country of origin; and second, he or she could incur a psychological cost by being viewed as a foreigner among friends and family members in the country of origin. It is therefore reasonable that the possibility of losing one’s citizenship status in the country of origin could be a deterrent for naturalization in the United States.
Acquisition of United States citizenship requires renouncing any foreign allegiance which has an implied effect on holding dual citizenship.14 this requirement is generally interpreted to mean that naturalized citizens cannot hold dual citizenship. Some literature (Jasso & rosenzweig 1990; Bloemraad 2006) suggests that this is not a true requirement. Immigrants are not required to provide any documents proving official rejection of their foreign citizenship. As a result, one may expect some confusion by the individual immigrant over the interpretation of the United States government’s stand on dual citizenship. however, one may unambiguously assess the impact of dual citizenship laws by considering if the country of origin allows dual citizenship. For this reason, I include a dummy variable in the ascension model to represent if the country of origin allows dual citizenship to reflect the increased cost of becoming a United States citizen.
Data the study is based on micro data obtained from a 5 per cent sample of the United States population census conducted in 2000. this sample is inclu-ded in the Integrated Public Use Microdata Series (IPUMS) produced by the Minnesota Population Center, University of Minnesota.15 the census sample is a weighted national sample of approximately 6,184,438 house-holds and 14,081,466 personal records. the analysis was limited to 578,248 immigrants 25 to 65 years of age, who by the census date had lived in the United States a sufficiently long period of time to become eli-gible for naturalization.16 the unit of analysis is the household head who had worked for at least one week during 1999 and for whom wages and salaries earned during that year were positive. Mexican immigrants were excluded from this citizenship study for two reasons: first, since Mexico is
14 the Oath of Allegiance begins as “I hereby declare, on oath, that I absolutely and entirely renounce and abjure all allegiance and fidelity to any foreign prince, potentate, state or sovereignty of whom or which I have heretofore been a subject or a citizen…” (Jasso & rosenzweig 1990).15 the IPUMS consists of 38 high-precision samples of the American population drawn from 15 federal censuses and from the American Community Surveys of 2000-2004. For detailed information about this database, the reader is referred to http://www.ipums.org/usa/.16 to be eligible to apply for United States citizenship, a person must have lived there for at least five years and be a legal resident.
Ather h. AkBArI
136
a part of the North American Free trade Agreement (NAFtA), Mexicans may view the benefits of United States citizenship differently from other immigrants;17 and second, a large number of Mexican immigrants in the United States are residing there illegally and no separate information on their legal status is available in the population census or in the population survey, hence their exclusion from our analysis reduces potential bias in our estimates.
Sample characteristics
table 1 provides the mean values for the outlined selected demographic and economic characteristics for naturalized citizens and non-citizens in the United States, which broadly reflect the human capital factors cited above that I feel will influence the costs and benefits of citizenship acquisi-tion. these persons have been broadly grouped as originating from develo-ped and developing countries. We find some noticeable differences between the characteristics of naturalized citizens originating in a developing versus a developed country.
the average naturalized citizen in either development group is about 4 years older than a non-citizen and is more likely to be married and with fewer children aged under 15 than a non-citizen. An immigrant coming from a developing country tends to acquire United States citizenship sooner. this finding is consistent with DeVoretz and Pivnenko’s (2006) finding for Canada and confirms that immigrants from developing countries are more likely to benefit from citizenship acquisition than their developed country counterparts. this result could also be partly explained by the fact that immigrants from poor countries tend to have larger families so they tend to benefit more from acquiring the right to sponsor relatives which, as noted, is derived from United States citizenship acquisition. We also plot this result in Figure 1 for selected developed and developing country immi-grants.
17 this may be especially so because of the greater mobility of workers across US-Mexico borders allowed under NAFtA.
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
137
table 1. Descriptive statistics for immigrants to the US, age 25-65, over 5 years since immigration (Mexican immigrants excluded)
Developed countries Developing countries
Naturalized citizen
Not a citizen
Naturalized citizen
Not a citizen
Sample size 83,943 57,463 250,931 175,898
Age, mean 48.16 43.86 44.29 40.34
Years since immigration, %
6-10 3.92% 25.34% 11.43% 39.75%
11-15 7.35% 19.09% 17.32% 26.66%
16-20 9.72% 14.59% 22.54% 17.51%
21-25 12.60% 11.30% 18.17% 7.55%
26-30 13.98% 9.13% 13.07% 4.45%
over 30 52.43% 20.55% 17.47% 4.08%
Marital status
Married, spouse present, % 71.08% 68.58% 68.54% 59.61%
Married, spouse absent, % 1.71% 2.18% 3.48% 6.21%
Separated, % 1.70% 2.17% 3.04% 4.76%
Divorced, % 11.08% 10.79% 9.26% 7.90%
Never married, % 10.86% 14.00% 13.11% 19.32%
Females, % 55.61% 55.18% 52.57% 49.12%
Presence of eldest child under 15, %
19.43% 27.14% 28.15% 31.47%
Occupation
Professional, % 23.27% 24.33% 24.81% 16.28%
Managerial, % 16.21% 16.82% 12.22% 8.49%
Education
Above high school or diploma, % 27.32% 26.64% 26.22% 18.58%
Bachelor degree, % 19.80% 20.38% 21.48% 13.00%
Master’s or professional degree
11.42% 11.27% 12.23% 8.93%
Doctorate, % 2.21% 3.75% 2.12% 2.33%
Home ownership, % 78.63% 66.35% 68.01% 43.28%
Weeks worked, mean 36.71 36.35 38.50 35.28
Wage and salary income, mean $32,597 $32,219 $30,614 $21,372
Total personal income, mean $41,375 $38,945 $36,192 $24,398
Notes: Source countries are divided into two groups based on $14000 GDP per capita threshold. Frequencies were weighted using the weights of the sample.
Source: United States population census (2000) microdata as included in IPUMS.
Ather h. AkbAri
138
Figure 1. Cumulative percentage of naturalizations among permanent immigrants from high income countries (UK, Canada, Germany, Italy, Netherlands) and low income countries (China and India)
Source: United States population census (2000) microdata as included in IPUMS.
Once naturalized, a developing country immigrant is more likely to work in a professional or managerial occupation. in contrast, the occupational status does not vary by citizenship status for an immigrant who comes from a developed country, as is also the case with respect to his/her educa-tional attainment. On the other hand, a naturalized citizen who arrives from a developing country is more likely to have acquired a university degree than his/her developed country counterpart. in addition, a naturali-zed citizen is more likely to own a house in the United States. Moreover, the effect of citizenship acquisition on home ownership is stronger for citi-zens who originate from a developing country.
the labour market variables indicate that a naturalized citizen who comes to the United States from a developing country works more weeks than his/her non-citizen counterpart and also more than a person from a developed country. Such a citizen also earns about 50 percent more income than a non-citizen, while citizenship status has no effect on the income of the person who comes from a developed country.
in summary, an immigrant from a developing country who eventually naturalizes may accumulate more human capital, own more housing and work more weeks in anticipation of exploiting the prospect of post-citizen-ship earnings gain. Moreover, these cited characteristics of naturalized United States citizens are consistent with those found for naturalized citi-zens in Canada (DeVoretz & Pivnenko 2006). We now turn to a formal analysis of the impact of these characteristics on the propensity for natura-lization.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 5 10 15 20 25 30 35 40 45 50 55 60
Years since immigration
Cu
mu
lati
ve %
UK, Canada, Germany, Italy, Netherlands China and India
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
139
econometric model for citizenship acquisitionFollowing the existing econometric literature, we estimate a logistic regres-sion model to analyze the effects of the explanatory variables on the proba-bility of citizenship acquisition in the United States.18 the model takes the form:
Ln [ pi / (1 - pi ) ] = a + Σ bi Xi + e
where Ln [ pi / (1 - pi ) ] is the logged odds of becoming a naturalized citizen in the United States, a is the intercept, Xi’s are the independent vari-ables, bi’s are the coefficients and e is the error term. the coefficients are interpreted as the change in the logged odds of becoming a naturalized United States citizen when there is a unit change in the independent vari-able, keeping the other independent variables constant. In cases where the independent variable is a dummy variable, the effect is measured when the corresponding variable changes its value from 0 to 1.
to facilitate the interpretation, we consider the odds ratios, which are the anti logs of the logged odds. Odds ratios are interpreted as the change in the odds of becoming a United States citizen associated with a one-unit change in the independent variable, net of other variables in the model. In cases where the independent variable is a dummy variable, the coefficient represents the change in odds between categories. An odds ratio above 1 indicates increased odds of becoming a naturalized citizen for a unit change in the corresponding independent variable and, in cases where the indepen-dent variable is a dummy variable, it represents the increased odds between the categories. An odds ratio of 1 indicates no difference and a ratio below 1 indicates decreased odds. Since the sample size is very large, tests of signi-ficance (t-tests) are not very useful, therefore our interpretation of results will largely focus on the patterns and the effective size of the odds ratios.
As I argued earlier, the costs and benefits of citizenship acquisition may vary by the immigrant’s country of origin, and therefore separate equations are estimated for developed and developing country immigrants.19
18 Aldrich and Nelson (1984) as well as hanushek and Jackson (1977) suggest the esti-mation of logistic regression if the dependent variable is a dummy variable, as in this case where a dummy variable representing citizenship status is used (the value of this variable equals 1 if the individual is a naturalized citizen and 0 otherwise). Afifi and Clark (1990) further suggest that logistic regressions do not require any assumptions regarding the distributions of the predictive variables. hence, logistics regressions are appropriate where the predictive variables are a combination of discrete and continuous variables. 19 Bueker’s (2005) analysis also shows that country of origin is a statistically significant determinant of citizenship acquisition in the United States.
Ather h. AkBArI
140
results of citizenship acquisition model the results for the logistic regression are reported in table 2, which also includes the results for the odds ratios.
Dev
elop
ed c
ount
ry im
mig
rant
sD
evel
opin
g co
untr
y im
mig
rant
s
Coe
ffici
ent
t-va
lue
Odd
s ra
tio
Coe
ffici
ent
t-va
lue
Odd
s ra
tio
Con
stan
t0.
448
6.41
11.
565
-2.6
50-6
5.14
00.
071
Age
-0.0
256
-15.
100
0.97
40.
003
2.44
61.
003
Age
squ
ared
0.00
014
8.50
01.
000
4.98
11.
000
Yea
rs s
ince
imm
igra
tion
0.08
714
8.61
01.
091
0.09
419
4.53
1.09
8
Fem
ale
-0.0
15-1
.230
.098
50.
220
30.4
251.
246
Mar
ried
0.10
57.
754
1.11
10.
159
19.1
001.
172
Pres
ence
Chi
ldre
n U
nder
15
-0.0
60-3
..687
0.94
2-0
.025
-2.8
700.
976
Uni
vers
ity
Deg
ree
0.16
111
.174
1.17
50.
411
43.7
751.
508
Prof
essi
onal
Occ
upat
ion
0.06
83.
.650
1.05
90.
152
14.5
431.
164
hom
e ow
ners
hip
0.29
021
.754
1.33
70.
580
80.2
091.
786
Nat
ural
log.
of
tota
l inc
ome
0.04
07.
693
1.03
70.
063
21.9
111.
065
rat
io o
f th
e so
urce
cou
ntry
GD
P pe
r ca
pita
to
that
of
the
US
-2.9
20-6
4.00
30.
054
-0.8
67-1
7.51
30.
420
Dua
l cit
izen
ship
-0.3
50-2
9.22
60.
704
-.05
2-6
.375
0.94
9
Log
Lik
elih
ood
func
tion
-186
750.
2-4
7944
7.2
Cox
& S
nell
r-s
quar
e)0.
2316
30.
216
Chi
-squ
ared
4801
9.8
1052
93.8
Not
es: S
ee t
able
1 f
or r
espe
ctiv
e sa
mpl
e si
zes.
Odd
s ra
tio
is t
he a
nti-
log
of e
ach
coef
ficie
nt a
nd is
inte
rpre
ted
as t
he c
hang
e in
the
odd
s of
bec
omin
g a
natu
raliz
ed U
S ci
tize
n as
soci
ated
wit
h a
one-
unit
cha
nge
in t
he in
depe
nden
t va
riab
le (
in c
ase
of
dum
my
inde
pend
ent
vari
able
, the
cha
nge
is f
rom
0 t
o 1)
. S
ourc
e: A
utho
r’s
esti
mat
ions
bas
ed o
n th
e U
nite
d St
ates
pop
ulat
ion
cens
us (
2000
) m
icro
data
as
incl
uded
in I
PU
MS.
tabl
e 2.
Log
istic
reg
ress
ion
resu
lts (d
epen
dent
var
iabl
e: C
tZN
[1=
US
citiz
en, 0
=oth
erw
ise]
)
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
141
the goodness of fit statistics associated with the two equations are repor-ted at the bottom of table 2. the explanatory power of each model is given by the chi-squared statistic, with degrees of freedom equal to the number of independent variables (excluding the intercept term) in the model. It is observed that each chi-squared value is statistically significant at 0.01 or 0.05 level of significance, which gives strong confidence to the explanatory power of each model. the predictive value of the developing country immigrant equation is greater than the estimated equation for the developed country immigrant.
It is seen from table 2 that the developing country immigrants’ logistic equations all contain variables which obtain t-values greater than 2, i.e., the critical t-value at 0.05 level of significance. In the case of the developed country immigrant equation, only the “female” variable has a lower t-value and hence does not pass the t-test of statistical significance. however, as we mentioned before, the large sample sizes used for estimation may be causing the t-values to be high, thereby reducing the credibility of the t-test of signi-ficance; hence, we will only interpret the odds ratios.
Most of the reported odds ratios in both models are greater than 1, indicating that the included demographic and adaptation characteristics of immigrants adequately predict the likelihood of citizenship acquisition. the odds of acquiring United States citizenship are similar for individuals who have identical post-schooling experience, regardless of whether they come from a developing or a developed country. Immigrant women coming from developed countries are less likely to naturalize than their male coun-terparts, while the opposite is true for immigrant women who come from developing countries. the presence of children under 15 years of age reduces the odds of acquiring United States citizenship in both immigrant groups, which is a surprising result. If the immigrant has a university edu-cation, professional qualifications, owns a home and earns a high income, then the odds of acquiring citizenship increase for both groups. Moreover, the impacts of these variables are greater for developing country immi-grants. As revealed by the coefficient of the GDP ratio, the odds of acqui-ring citizenship decrease as one comes from a relatively prosperous country in each group. the presence of dual citizenship in the country of origin reduces the odds of acquiring United States citizenship, a finding which is also contrary to our expectations.20
the greatest differential impacts derived from any variables across the two income groups are observed in the home ownership and education variables. When I compare the impact of home ownership, the odds of
20 It should be noted that dual citizenship is more likely in developed countries and this could be the reason that the dual citizenship status as defined by the eligibility in the sending country leads to this odd result.
Ather h. AkBArI
142
acquiring citizenship increase for a homeowner who came from a develo-ping country by 1.34 times more than for a developed country immigrant. In addition, persons with a university degree in the developing country sample are 1.29 times more likely to ascend to United States citizenship. We now turn to an estimation of how the labour market rewards citizen-ship acquisition in the United States.
Wage equations for citizens and non-citizensWhile citizenship acquisition increases the entitlements of individuals in welfare schemes, it may also enhance their labour market opportunities. hence, any analysis of the potential public treasury effect of naturalization should not only consider the effect on welfare receipts but also the impact on the income of individuals, which in turn determines their public tre-asury contribution and their risk of consuming welfare. In the present section, we analyze the impact of naturalization on one indicator of the labour market performance of immigrants, i.e. the labour market earnings, and compare the same with the earnings of non-citizens.
Separate earnings equations are estimated for naturalized citizens and non-citizens within each group. I have also performed separate estimations for developed and developing country immigrants. Furthermore, to isolate the compounding effect of gender discrimination on annual wages, I also estimate separate equations for males and females. the specification of each wage equation is based on human capital theory, which views educa-tion as the single most important determinant of earnings.
to capture the effect of citizenship acquisition on annual wages, an expanded wage equation is specified in which the natural logarithm of annual wages is considered as the dependent variable. the independent variables include age (as a proxy for post-schooling experience); age-squa-red (to account for a possible concavity of age-earnings profile); years since migration (to isolate the effect of citizenship status from the United States labour market experience effect on earnings); the logarithm of weeks worked during the year (to control for any labour supply effect on ear-nings); and a set of dummy variables for the citizenship status, different education levels, occupations, and full-time weeks worked. results of our estimation are provided above in table 3.
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
143
table 3. OLS estimation of log-linear earnings model: citizenship effect on immigrant earnings (dependent variable is log of earnings)
Developed countries(GDP per capita >
$13,999)
Developing countries(GDP per capita
< $14,000)
(1)Females
(2)Males
(3)Females
(4)Males
(Constant) 4.660(76.519)
4.456(66.070)
4.935(158.331)
5.175(170.437)
Age .026(10.024)
.071(25.059)
.026(18.581)
.041(31.028)
Age squared -0.0003(-9.804)
-.001(-22.137)
-0.0003(-19.353)
.0004(-31.164)
Years since immigration .003(9.676)
.001(4.021)
.008(35.463)
.010(49.376)
LN(weeks) .969(153.274)
.825(90.744)
.879(265.610)
.766(201.780)
Naturalized citizen (reference group is non-citizens)
.021(2.973)
-.010(-1.341)
.092(23.809)
.107(29.317)
Education (Reference group is high school graduate or less)
Above high school or diploma
.195(24.289)
.121(13.035)
.201(45.477)
.158(36.387)
Bachelor degree .388(39.647)
.361(35.569)
.384(75.107)
.331(64.877)
Master’s or professional degree
.544(43.706)
.555(46.102)
.539(81.168)
.620(102.866)
Doctorate .611(25.234)
.539(30.377)
.607(41.427)
.668(69.150)
Occupation ( Reference group is skilled and low-skill occupations)
Professional occupation .263(30.899)
.234(23.967)
.371(78.835)
.344(69.938)
Managerial occupation .322(33.402)
.399(44.984)
.310(50.316)
.320(64.559)
Status of weeks worked (Reference group is of those who worked part-time weeks, e.g., 34 hours or less)Full time weeks worked .847
(111.557).927
(67.927).750
(166.812).816
(127.351)
Adjusted r Square .544 .363 .542 .442
F-statistics 4857.743 2367.483 15061.919 11692.251
Note: t-statistics are given in brackets. Source: Author’s estimations based on the United States population census (2000 microdata as included in IPUMS.
Ather h. AkBArI
144
All OLS regression coefficients yield the predicted signs owing to human capital theory, as shown in table 3. Of primary interest for this study is the coefficient of the dummy variable representing citizenship acquisition. In the case of immigrants arriving from developed countries, the magnitude of this coefficient is small and the corresponding t-value is below its critical value of two.21 hence, citizenship acquisition alone does not have a statistically-significant effect on the log of annual wages of deve-loped country immigrants. however, a larger and statistically significant effect of citizenship acquisition on the log of annual wages is found in case of immigrants arriving from developing countries. In both samples, the direction of results do not vary by gender.
In order to investigate any possible occupational effect on the impact of post-citizenship earnings, I further augment my model by adding an interaction dummy variable which is obtained by multiplying the dummy variable for naturalized citizenship with that for professional occupation status. this interactive variable represents an individual who is a naturali-zed citizen and has worked in a professional occupation. thus the coeffi-cient of this variable should indicate the earnings effect of working in a professional occupation given that citizenship acquisition has occurred. the results of the augmented model are presented in table 4.
21 All tests of significance are performed at a 0.05 level of significance.
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
145
table 4. OLS estimation of log-linear earnings model: citizenship effect on immigrant earnings (dependent variable is logarithm of annual wages)
Developed countries(GDP per capita > $13,999)
Developing countries(GDP per capita < $14,000)
(1)Females
(2)Males
(3)Females
(4)Males
(Constant) 4.672(76.629)
4.471(66.201)
4.928(158.032)
5.171(170.243)
Age .026(9.997)
.071(25.019)
.026(18.569)
.041(30.997)
Age squared .0003(-9.766)
-.001(-22.096)
.0003(-19.339)
-0.0005(-31.125)
Years since immigration
.003(9.415)
.001(3.797)
.008(35.681)
.011(49.594)
Naturalized citizen .005(.642)
-.027(-3.062)
.104(24.028)
.116(28.769)
Naturalized citizen in professional occupation
.055(3.892)
.061(4.012)
-.051(-6.232)
-.043(-5.375)
Above high school or diploma
.195(24.300)
.121(13.093)
.199(45.079)
.156(35.984)
Bachelor degree .387(39.591)
.361(35.542)
.383(74.906)
.330(64.628)
Master’s or professional degree
.543(43.597)
.554(46.050)
.538(80.876)
.618(102.337)
Doctorate .616(25.413)
.544(30.592)
.601(40.893)
.663(68.349)
Professional occupation
.230(19.342)
.197(14.851)
.406(55.229)
.372(52.007)
Managerial occupation
.323(33.432)
.398(44.896)
.309(50.115)
.319(64.362)
LN(weeks) .969(153.299)
.825(90.749)
.879(265.602)
.766(201.758)
Full time weeks worked
.846(111.535)
.926(67.906)
.750(166.836)
.816(127.406)
total degrees of freedom
48,877 49,852 152,917 176,948
Adjusted r Square .544 .363 .542 .442
F-statistics 4486.535 2187.268 13909.738 10796.771
Note: t-statistics are given in brackets.Source: Author’s estimations based on the United States population census (2000) micro-data as included in IPUMS.
In both samples, the coefficient of the interaction variable is statistically significant, indicating the presence of occupational effects on the returns to citizenship acquisition. In the developing country case, this effect is nega-tive, indicating that the returns to citizenship are lower for those develo-ping country immigrants who work in professional occupations. this result
Ather h. AkBArI
146
probably reflects the lack of foreign credential recognition, which will be discussed in the concluding section. however, as was indicated by the coef-ficient of the professional occupation dummy variable in table 4, the overall returns to professional occupations are greater for developing country immigrants (once the effect of citizenship has been isolated).
the wage equation results showed that naturalization has a statisti-cally significant and positive effect on the earnings of immigrant men and women who arrive from developing countries. For developed country immigrants, the effect is statistically significant only in case of women. human capital theory attributes earnings differentials between individuals mostly to human capital endowments (education and labour market expe-rience). therefore, I analyze the effect of human capital variables on the wage differential observed between naturalized citizens and non-citizens. to achieve this task, I decompose the wage differential observed between citizens (C) and non-citizens (N) in each group into the component responsible for differences in human capital endowments and the compo-nent responsible for differences in market rewards for human capital endowments. this decomposition allows one to assess if citizenship acqui-sition results in greater returns to human capital investment. the technique used for this decomposition of wage differential follows the technique first suggested by Oxaca (1973)22. Using his technique, the difference in loga-rithm of wage earnings between naturalized citizens and non-citizens in each group can be written as follows:
€
lnWC − lnWN = (X C − X N )T ˆ β C + X NT ( ˆ β C − ˆ β N )
the first part of the above equation, i.e.,
€
(X C − X N )T ˆ β C , is the component of wage differential that is attributed to differences in human capital endowments between citizens and non-citizens in each group. the second part of the above equation, i.e.,
€
X NT ( ˆ β C − ˆ β N ), is the component of wage dif-
ferential that is attributed to the differences in market rewards for human capital endowments of citizens and non-citizens. In other words, the second part measures the difference in logarithmic wage earnings of those individuals in each group who possess the same capital endowments. the anti-log of each part of the above equation is used to compute the per-centage difference. results are reported in tables 5 and 6 using the equa-tion results reported in tables 3 and 4, respectively.
22 Oaxaca (1973) used this technique to decompose the earnings differential between men and women in the United States into the component attributed to differences in productivity-related characteristics and the component attributed to differences in labour market reward for each productivity-related characteristic. this technique has now become a standard technique in labour market literature to analyze earnings differences between demographic groups such as citizens and non-citizens.
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
147
table 5. Decomposition of wage differentials between naturalized citizens and non-citizens in the US: population of foreign-born employees 18-65 years old who immigrated over 5 years ago
Source country status
human capital endowment effect
“Discrimination” component
Wage differential
Females
Developed 8.48% 2.04% 10.53%
Developing 26.83% 7.05% 33.87%
Males
Developed 9.07% -2.12% 6.95%
Developing 28.55% 8.12% 36.67%
Source: Author’s calculations based on Oaxaca decomposition using results of Table 2 and averages of education and age variables.
table 6. Decomposition of wage differentials between naturalized citizens and non-citizens in the US: population of foreign-born professional employ-ees 18-65 years old who immigrated over 5 years ago
Source country status
human capital endowment effect
“Discrimination” component
Wage differential
Females
Developed 8.73% 7.12% 15.85%
Developing 15.52% 3.51% 19.03%
Males
Developed 8.92% 3.64% 12.56%
Developing 13.82% 5.22% 19.04%
Source: Author’s calculations based on Oaxaca decomposition using results of Table 3 and averages of education and age variables.
Ather h. AkBArI
148
All of these reported decompositions of wage differentials show that human capital variables cause the earnings of citizens to be greater than those of non-citizens. the effect on the wage differential is greater for developing country immigrants than in the case of developed country immigrants. the effect is also greater for men. however, the acquisition of citizenship raises the returns from human capital more for developing country immigrants than for developed country immigrants. Furthermore, most of the wage differential attributed to human capital variables arises due to differences in human capital endowments and is not due to differential market rewards for human capital endowments.
effects of naturalization upon public coffersAs mentioned earlier, Borjas (2002b) has shown that foreign-born citizens of the United States tend to participate more in welfare programs than do native-born citizens. In fact, he also uses this evidence to argue that eligibi-lity to participate in welfare programs is a strong motivation for naturali-zation. however, in order to review the overall impact on public coffers, one should compare the receipts of transfer payments and use of public services by naturalized citizens with their tax payments. We provide this comparison in the present section. For this purpose, we use micro-data based on the latest United States population census conducted in 2000. Our unit of analysis is a household. Following previous literature (Simon 1984), we define an immigrant household to be one in which either only the male spouse or both male and female spouses are immigrants.23
the first part of our analysis concentrates on the use of public transfers and services by immigrant and non-immigrant households. the census asked questions about the receipts of important transfer payments. It also collected information about participation in Medicare and Medicaid pro-grams, which I was able to combine with published data on per capita costs of these programs to obtain dollar estimates for immigrant and non-immigrant households. the census did not ask questions about the partici-pation of children in each household in elementary/secondary schools. however, we were able to use the number of school age children, i.e., child-ren aged 6-18, with the published data on per pupil cost to estimate the taxpayers’ cost incurred by each household for the provision of public edu-cation. We obtained all estimates of transfer payments, health care costs and educational costs (at elementary/secondary levels) separately for each immigrant entry cohort defined by citizenship status and for non-immi-grants.
23 households in which only the wife is an immigrant are excluded from the analysis as their inclusion may lead to double counting.
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
149
For the sake of brevity, however, we have decided not to present the dollar estimates of public transfers and services but to present only the relative estimates in table 7. these data show the receipt of each important component of public transfer payments by a household headed by a natu-ralized citizen relative to those headed by a non-citizen. the most recent entry cohort in these data entered during 1990-94 since the census asked questions about household incomes pertaining to 1999 and citizenship acquisition in the United States requires a minimum stay of five years by an immigrant. As seen, the total receipts of public transfers and services by households headed by naturalized citizens remains lower than that of non-citizens in most entry cohorts, although this gap narrows with length of stay. the most noticeable are the use of public welfare, unemployment compensation/workers’ compensation, veterans’ benefits and Medicaid, which are significantly lower for most entry cohorts of naturalized citizens at least until they have stayed for 40 years in the country. On the other hand, the receipts of social security and child support and Medicare are greater for most entry cohorts of households headed by naturalized citi-zens.
Ather h. AkBArI
150
tabl
e 7.
rec
eipt
s of
tran
sfer
pay
men
ts b
y ho
useh
olds
hea
ded
by n
atur
aliz
ed c
itize
ns r
elat
ive
to th
ose
head
ed b
y no
n-ci
tizen
s, U
nite
d St
ates
, 199
9*Y
ears
of
entr
yU
nem
ploy
men
t/W
orke
r’s
Com
pens
atio
nV
eter
an’s
ben
efits
(1)
Wel
fare
(Pub
licA
ssis
tanc
e)(2
)
Supp
lem
enta
lSe
curi
ty
inco
me
(3)
Chi
ldsu
ppor
t(4
)
Food
stam
ps(5
)
Soci
alse
curi
ty(6
)
Med
icar
e*(7
)M
edic
aid*
(8)
tota
lh
ealt
hca
re(9
)
Scho
olin
g co
sts
for
elem
enta
ry/
seco
ndar
yst
uden
ts(1
0)
tota
l pu
blic
tr
ansf
ers/
serv
ices
(11)
1990
-19
941.
060.
232.
721.
460.
711.
262.
720.
760.
940.
840.
93
1986
-19
890.
680.
023.
060.
440.
450.
451.
330.
650.
730.
900.
85
1980
-19
850.
610.
000.
871.
640.
151.
081.
890.
450.
610.
780.
74
1975
-19
790.
781.
261.
231.
452.
141.
891.
670.
760.
950.
971.
06
1970
-19
740.
860.
260.
751.
730.
412.
022.
140.
620.
910.
810.
94
1965
-19
690.
710.
061.
061.
580.
432.
092.
420.
571.
170.
630.
99
1960
-19
640.
660.
660.
760.
290.
332.
032.
250.
551.
170.
571.
01
Bef
ore
1960
1.34
1.00
0.44
0.55
0.40
1.46
1.33
0.44
1.05
0.50
1.16
*A v
alue
gre
ater
tha
n 1
mea
ns t
hat
dolla
r va
lue
rece
ived
by
the
hous
ehol
d he
aded
by
a na
tura
lized
cit
izen
exc
eede
d th
at r
ecei
ved
by a
hou
seho
ld h
eade
d by
non
-na
tura
lized
cit
izen
.i.
Col
umns
(1)
thr
ough
(6)
, com
puta
tion
s w
ere
base
d on
dol
lar
valu
es o
btai
ned
from
the
200
0 U
.S. p
opul
atio
n ce
nsus
mic
ro-d
ata
ii.
For
colu
mns
(7)
and
(8)
, nu
mbe
r of
per
sons
cov
ered
per
hou
seho
ld (
2000
U.S
. po
pula
tion
cen
sus
mic
ro d
ata-
data
) w
as m
ulti
plie
d by
the
res
pect
ive
per
ca
pita
val
ues,
$77
5 an
d $7
30, w
hich
in t
urn
was
bas
ed o
n da
ta p
rovi
ded
by C
ente
rs f
or M
edic
are
and
Med
icai
d (2
005a
, 200
5b)
acco
rdin
g to
whi
ch n
atio
-
nal p
er c
apit
a he
alth
car
e ex
pend
i-tu
res
in 2
000
wer
e $4
,560
, div
ided
into
Med
icar
e an
d M
edic
aid
(inc
ludi
ng C
hild
ren’
s h
ealt
h C
are
Prog
ram
) as
17 a
nd 1
6
perc
ent,
res
pect
ivel
y.iii
. Fo
r co
lum
n (1
0),
num
ber
of c
hild
ren
aged
6-1
8 ho
useh
old
(200
0 U
.S.
popu
lati
on c
ensu
s m
icro
-dat
a) w
ere
mul
tipl
ied
by t
he p
er p
upil
cost
of
$6
,911
(N
atio
nal C
ente
r fo
r e
duca
tion
Sta
tist
ics,
200
1).
Sour
ce: A
utho
r’s
calc
ulat
ions
as
disc
usse
d ab
ove.
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
151
In sum, one may conclude from the above results that, overall, households headed by naturalized citizens tend to be the recipients of lower transfer payments and public services at least until 20 years after arrival in the United States when compared with a household headed by a non-citizen. After 35 years of residence naturalized citizens receive benefits similar to households headed by non-citizens. however, receipts of some components of the public transfers emulate non-citizens faster.
We now turn to the overall balance of public transfers from an immi-grant household to a non-immigrant household. For this purpose, we con-sider the dollar values of the receipts of public transfers and services as well as the payment of taxes. estimation of tax payments was conducted by applying a constant tax rate of 29.6 percent to average household incomes in each entry cohort as obtained from the census micro-data. the rationale for the 29.6 percent rate is discussed in the notes provided at the end of table 8. the data are reported separately for households headed by natura-lized citizens and non-citizens.
Ather h. AkBArI
152
tabl
e 8.
Bal
ance
of p
ublic
tran
sfer
s an
d se
rvic
es fr
om a
n im
mig
rant
hou
seho
ld (b
y ci
tizen
sta
tus)
to a
non
-imm
igra
nt h
ouse
hold
, U
nite
d St
ates
, 199
9*Y
ears
of
ent
ryD
olla
rs d
iffe
renc
e in
tax
es p
aid
by
imm
igra
ntan
d no
n-im
mig
rant
hou
seho
lds
(1)
Dol
lar
diff
eren
ce in
pub
lic s
ervi
ces
and
tran
sfer
s re
ceiv
ed b
y im
mig
rant
s an
d no
n- im
mig
rant
hou
seho
lds
(2)
Net
tra
nsfe
r fr
om im
mig
rant
to
non-
imm
igra
nt h
ouse
hold
s ($
)**
(3)
hea
d is
na
tura
lized
citi
zen
hea
d is
non
-ci
tize
nh
ead
is
natu
raliz
edci
tize
n
hea
d is
non
-ci
tize
nh
ead
is
natu
raliz
edci
tize
n***
hea
d is
non
-ci
tize
n***
1990
-94
-139
2-3
735
727
95-6
65-3
640
1986
-89
324
-369
228
2-1
251
606
-494
3
1980
-85
4162
-412
287
7-1
872
5039
-599
4
1975
-79
1078
-105
2-1
328
-71-
21-2
50-1
773
1970
-74
2612
-204
3-7
45-1
315
1867
-335
8
1965
-69
1889
-902
-116
9-1
277
720
-217
9
1960
-64
1825
-483
-733
-640
1092
-112
3
Bef
ore
1960
-307
4-2
674
-464
7-2
823
-772
1-5
497
*In
1999
, an
aver
age
non-
imm
igra
nt h
ouse
hold
rec
eive
d $8
769
in p
ublic
tra
nsfe
rs a
nd s
ervi
ces
and
paid
$16
,603
in a
ll ta
xes.
**C
ontr
ibut
ion
tow
ards
pu
blic
goo
ds is
igno
red.
***A
pos
itiv
e nu
mbe
r m
eans
a t
rans
fer
of f
unds
fro
m a
n im
mig
rant
to
a no
n-im
mig
rant
hou
seho
ld, w
hile
a n
egat
ive
num
ber
mea
ns t
he r
ever
se is
tru
e.
Cal
cula
tion
s:
i.
For
tax
calc
ulat
ion:
acc
ordi
ng t
o O
eC
D (
2005
) go
vern
men
t ta
x re
venu
e as
a p
erce
ntag
e of
GD
P w
as 2
9.9
perc
ent
in 2
000
whi
ch w
e as
sum
ed t
o al
so
ho
ld
for
1999
. t
his
perc
enta
ge
was
ap
plie
d to
av
erag
e ho
useh
old
inco
me
in
each
co
hort
, ob
tain
ed
from
th
e 20
00
U.S
. po
pu-
la
tion
cen
sus
mic
ro-d
ata.
In
the
Oe
CD
cla
ssifi
cati
on,
taxe
s ar
e cl
assi
-fied
by
the
base
of
the
tax:
inc
ome
and
profi
ts,
payr
oll,
prop
erty
, co
n-
su
mpt
ion
and
othe
r ta
xes.
Soc
ial s
ecur
ity
cont
ribu
tion
s pa
id t
o ge
nera
l gov
ernm
ent
reve
nue
are
also
cla
ssifi
ed a
s ta
xes.
ii.
For
calc
ulat
ion
of d
olla
rs v
alue
s of
pub
lic t
rans
fers
and
ser
vice
s re
cei-
ved,
see
not
es a
t th
e en
d of
tab
le 7
.So
urce
: Aut
hor’
s ca
lcul
atio
ns a
s di
scus
sed
abov
e.
IMMIGrANt NAtUrALIZAtION AND ItS IMPACtS ON IMMIGrANt LABOUr MArket PerFOrMANCe AND treASUrY
153
In the case of households headed by naturalized citizens, tax payments are below those reported by non-immigrants only for the most recent and oldest cohorts. however, none of the entry cohorts of households headed by non-citizens paid more taxes than households headed by non-immi-grants. this result is consistent with the result reported earlier that natura-lized citizens tend to have an income advantage over non-citizen immi-grants and so will pay greater taxes.
On the other hand, receipts of public transfers and services by immi-grant households headed by naturalized citizens begin to exceed those of non-immigrant households after they have remained in the United States for at least 20 years. Immigrant households whose head is a non-citizen are seen to consume transfer payments and public service in excess of non-immigrants after 10 years of residence in the United States.
In sum, the data on usage of public funds indicate that immigrant households headed by naturalized citizens tend to consume fewer public services than do non-immigrants until after 20 years of the head’s arrival in the United States, while in the case of households headed by non-citizens, this catch-up occurs only within 10 years of arrival.
Finally, to obtain the net public fund transfers from an average immi-grant to a non-immigrant household, we add up the above results in column 3 of table 8. these estimates may be viewed as conservative esti-mates since no account of the possible increased provision of public goods has been made. 24
It is observed in table 8 that most entry cohorts of immigrant house-holds headed by naturalized citizens remain sources of public fund trans-fers for non-immigrants until at least after their 40th year of arrival in the United States. Conversely, a household headed by a non-naturalized immi-grant remains a net recipient of public fund transfers from non-immigrants, starting from the initial entry date.
In sum, the above results indicate that there is a benefit to American taxpayers with the acquisition of citizenship by immigrants. the importance of distinguishing between the categories of immigrant classes in any research evaluating the economic performance of immigrants is also high-lighted in these results.
24 By virtue of their presence, immigrants increase the provision of such public goods as national defence, research and development, policing, etc., without affecting the level of their use. hence, an immigrant household reduces the tax payment of a non-immigrant household, and this tax saving should also be included in net transfer calculations.
Ather h. AkBArI
154
Discussion and ConclusionsNaturalized citizens form the most legally-advantaged group of immigrants in the United States. Yet, as noted by Bloemraad (2006), only about 40 per cent of immigrants reported in the 2000 U.S. census were naturalized citi-zens. results of the present study indicate that socio-economic characteris-tics play an important role in determining the probability to naturalize. Immigrants from less-developed countries are more likely than those from developed countries to acquire United States citizenship and also reap the most economic gain. homeownership, which is an indicator of wealth accumulation, increases the odds of acquiring citizenship status. however, the result that dual citizenship in country of origin reduces the odds of acquiring United States citizenship is puzzling. We attribute this result to a probable confusion in the understanding of dual citizenship laws of the United States.
All immigrants who obtain United States citizenship earn higher labour market incomes because 1) they tend to possess greater human capital endowments, and 2) they also enjoy greater rewards for their human capital investment than do non-citizens. these rewards are greater for immigrants originating in developing countries, probably because prior to naturaliza-tion, their earnings were much lower than those of developed country immigrants. A possible explanation for this result may be that acquisition of citizenship by an immigrant is viewed by an employer as an indicator of a greater knowledge of local customs and traditions that is essential for a firm’s success.
Acquisition of citizenship status by immigrants increases their contri-bution to the public coffers of the United States. While the previous litera-ture has shown that naturalization in the United States increases welfare participation, the present study has shown that for a household headed by a naturalized citizen, the dollar values of public transfers remain below those received by a household headed by a non-citizen or a non-immigrant for a long period after the head has arrived in the country. Citizenship acquisition also results in greater tax payments by an immigrant house-hold. thus, when we add up the transfer receipts and tax payments, a household headed by a naturalized citizen makes a positive treasury trans-fer between the10th to the 40th year in the United States.
In summary, the results of this study confirm the suggestion of Massey and Bartley (2005) that any analysis of immigrants’ economic performance must pay attention to the legal status composition of immigrant population.
155
Chapter 6
the eCONOMICS OF CItIZeNShIp: a SYNtheSIS
Pieter Bevelander & Don J. DeVoretz
IntroductionVarious disciplines within the social sciences have a long and detailed history of analyzing the relationship between immigrant citizenship acqui-sition and outcomes. Political scientists focus on analyzing recent naturali-zed citizens’ voting patterns and cross border political activities. Sociolo-gists emphasize the various concepts of citizenship and challenge the notion of a nation-based citizenship in favor of more exotic views of world or neighborhood citizenships. Economists have occasionally entered the field in an ad hoc manner with speculations about the possible earnings or employment effects that may derive from immigrant citizenship ascension.
Based on these disparate strains in the social science literature an inter-disciplinary workshop was held under the aegis of the Willy Brandt Profes-sorship at Malmö University in 2004. This exploratory workshop brought together European and North American economists, sociologists, political scientists and an anthropologist, and its aim was to draw up a common set of research questions and an acceptable interdisciplinary methodology to explore the motivation for immigrant citizenship acquisition. After ten paper presentations, and animated discussions, the group concluded that there exists no common methodology shared across all disciplines to inves-tigate both the causes and effects of immigrant citizenship acquisition. However, two broad questions emerged which piqued the interest of all participants. Namely, why do immigrants ascend to citizenship at differen-tial rates, and what are the political, social and economic consequences of this ascension?
This book attempts to go beyond these two prosaic questions and poses the broader question of what constitutes the economics of citizen-ship. In short, what can economics contribute to the understanding of both citizenship acquisition and the labour market and public finance impacts
PIETEr BEVElANdEr & doN J. dEVorETz
156
derived from naturalization? The underlying theory outlined in this book argues that the decision to ascend to citizenship is based on the prospects that the discounted long-run economic benefits of acquiring citizenship outweigh the more immediate economic costs. Moreover the presumed economic benefits are partially derived from the immigrant’s acquisition of general and specific human capital acquired while resident in the host country during the normal waiting period prior to citizenship acquisition. The most obvious forms of general human capital accumulated by poten-tial naturalized citizens are host country language and job skills while spe-cific skill acquisition would include host country professional and trade certificates. In other words, a selective three stage dynamic process is envi-sioned in this economic theory of citizenship acquisition.
In the first stage, self-selected immigrants arrive in their host countries and, given their existing stock of human capital, they may or may not accu-mulate more host-country-specific and general human capital while await-ing naturalization. Human capital investment may be accelerated by the host country providing subsidies for general human capital, or even speci-fic human capital, while the immigrant awaits naturalization. Given this investment, it is argued that, if the costs of citizenship ascension are lower than the benefits derived from ascension, an immigrant will adopt the host country’s citizenship. After naturalization the newly naturalized citizen may realize increased economic rewards for this human capital acquisition as heretofore labour market restrictions on non-citizens are relaxed in the public sector and private entrepreneurs offer a wage premium to citizens.1 This increase in earning power should in turn lead to naturalized citizens contributing more to the host country’s treasury than non-citizens do.
This economic theory now allows us to analyze the forces behind dif-ferential citizenship acquisition across immigrant host countries for similar ethnic groups and differential naturalization patterns by immigrant cohorts within a country. The central thesis of this book is that comparing the “old” and the “New” settled regions would provide a laboratory to test this economic theory of citizenship acquisition. Figure 1 depicts the diffe-rential patterns of citizenship acquisition across old and New settled regions.
1 See deVoretz (2008).
THE ECoNoMICS oF CITIzENSHIP: A SYNTHESIS
157
Figure 1. Naturalization rates across selected countries, 2001-2003
Source: OECD 2005.
Figure 1 illustrates that two newly settled countries (Canada and Australia) have ascension rates of approximately 70% of their resident foreign born populations circa 2001-2003. Most “old World” countries have naturali-zation rates of approximately 40% for their foreign-born stock. However, there are interesting exceptions to this “New World/old World” dicho-tomy. For example, Sweden and the Netherlands have higher naturaliza-tion rates (63%) than the United States (45%).
To some extend the “New World/old World” dichotomy in naturali-zation rates could arise due the presence of differential naturalization hurdles. In other words, the individual country rules and regulations for citizenship may affect rates of ascension. In Figure 2 we picture three of the most important rules and regulations. The first rule depicted is the minimum number of required years of residency in the host country before an immi-grant can apply for citizenship. The second rule illustrated in Figure 2 refers to the legality of dual citizenship in host countries. Finally, Figure 2 reports the method of citizenship acquisition either in terms of birthplace of an immigrant’s parents (jus sanguinis) or by birthplace of the immigrant (jus soli) or via a mixed set of principles.2
2 None of the countries reported above have a pure jus soli citizenship policy.
Pieter Bevelander & don J. devoretz
158
Figure 2. Minimum number of years of residence required for citizenship, dual citizenship and citizenship at birth (jus sanguinis) and mixed (jus soli and jus sanguinis)
notes:Grey = single citizenship toleratedBlack = dual citizenship toleratedStriped = Mixed (jus soli and jus sanguinis)
Sources: Bauböck 2006; Bloemraad 2006; DeVoretz & Pivnenko 2006; Department of Internal Affairs New Zealand.
For the countries in Figure 2 the range of residence requirements prior to citizenship acquisition varies from two to ten years. in the new World countries of australia, new-zealand and Canada, the residency period is minimal. Belgium and ireland are “old World” countries that have short minimum residence demands. the remaining reported countries with long minimum years of required residence before naturalization are all southern european.
Figure 2 also reports that countries that require fewer years of residence also tolerate dual citizenship and utilize mixed birthplace of parents and immigrants policies to grant citizenship. in fact, many countries with long minimum residency requirements prohibit dual citizenship and apply the so-called restrictive jus sanguinis citizenship principle for citizenship ascension.
australia, Canada, new zealand and, in europe, Belgium and ireland, have the most liberal citizenship/naturalization policies. restrictive policies towards naturalization/citizenship ascension, on the other hand, are found in austria and in the southern european countries: Spain, Portugal, italy and Greece. if we combine the findings reported in Figures 1 and 2, we
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Chile Ethiopia
Iran Iraq
Vietnam Norway
Figure 2.4 chapter 4
0
2
4
6
8
10
12
Australi
a
Belgium
Canad
a
Irelan
d
New Zea
land
France
Netherl
ands
Sweden
United Kingdom
USA
Finland
German
y
Norway
Denmark
Austria
Spain
Greece Ita
ly
Portugal
Figure 2 I kapitel 6
THE ECoNoMICS oF CITIzENSHIP: A SYNTHESIS
159
note that Portugal, Canada and Australia have high naturalization rates, while New zealand, Spain and Belgium have relatively low levels of natu-ralization. This leads us to conclude that liberal or restrictive policies in themselves do not explain the cross country naturalization levels observed in Figure 1.
Given this background we now move to the task of explaining these macro-level ascension trends for our selected countries and ask;
How do the differing socio-economic characteristics of the resident foreign-born stock across these countries affect citizenship acquisition rates?
Furthermore,
Given heterogeneous foreign-born populations and varying labour market conditions across “New” and “Old” World countries, how do the econo-mic impacts of citizenship acquisition vary across the new and old regions?
In particular, we speculate that a generous combined jus soli and jus sanguinis principle for citizenship acquisition policies in the New World, plus dual citizenship provisions, should enhance the prospects of naturali-zation in the New World (deVoretz, 2008).
In addition, differential immigrant selection policies across these countries will affect both their rates of citizenship acquisition and the eco-nomic outcomes derived from naturalization. In particular, the fact that many European and United States immigrants select themselves while the majority of Canadian immigrants are doubly selected implies different levels of acquired human capital stock across these countries’ resident immigrant populations. Indeed in countries where immigrants are doubly selected, first via self-selection and then selected by the host country, we should observe immigrants whose human capital stocks exceed those found in immigrants of singly selected countries in Europe.3 The greater human capital stock derived from double selection in turn implies both higher naturalization rates and a greater economic impact derived from naturali-zation in the “New” versus “old” World where single selection reigns.
The act of citizenship ascension in turn implies a third (second) self-selection process in the “New” (“old”) World. However, this selection process could result in positive or adverse selection depending on the incenti-ves build into the citizenship entitlements of each country (see Appendix). For example, if socially financed entitlements (health, social security, edu-cation, welfare) are withheld from foreign-born non-citizens, then immi-
3 Europe is dominated by refugees who are not screened nor selected by the host countries for their economic attributes. Hence only a limited second selection process exists in Europe for some immigrant entrants.
PIETEr BEVElANdEr & doN J. dEVorETz
160
grants with a smaller human capital are more likely to apply for citizenship to gain these benefits. This adverse selection into citizenship could be exa-cerbated if the new naturalized citizens are also given accelerated sponsor-ship rights for family members who are singly selected. In fact, these are the conditions which prevail in the United States where adverse selection should occur at the third, or citizenship selection, stage.
In Europe where the majority of foreign admissions are similarly based on self-selection, adverse selection into citizenship is likely: since immigrant citi-zens will have a smaller human capital stock than either the native-born population or resident EU foreigners. This adverse selection will be exacer-bated in the EU context since many of the foreign-born are either refugees or reunified family members who have little or no economic incentive to return home.
Where Canada is concerned, the possibility of triple selection implies an optimistic case of positive self-selection into citizenship. The triple selec-tion arises in this context since the near majority of Canadian immigrants are screened prior to entry, thus insuring a large amount of high human capital embedded in the potential citizens.4 Next, given that the acquisition Canadian citizenship is costly, we argue that only those immigrants who have the human capital to recoup these costs in the host labour markets will ascend into citizenship. There is one caveat to this possible case of positive self-selection. If there exists a large and free set of public entitle-ments emanating from citizenship acquisition, then adverse selection into citizenship could also arise in Canada. one free public good readily avail-able after ascension to citizenship is a Canadian passport which embodies all the “good will” of previous passport holders.5 Thus, foreign-born resi-dents in Canada who still hold their original passports with weak mobility rights could self-select into citizenship regardless of their human capital endowments to secure a more recognized passport.6
In sum, the triple selection process inherent in citizenship ascension in Canada should insure positive self-selection for its naturalized citizens unless large and important externalities exist for an important immigrant arrival group. We will now turn to the empirical evidence reported in this book on our two “New World” cases, and then we review the “old World” cases to confirm the existence of positive or adverse selection into their citi-zenship status.
4 In fact an experiment conducted by the authors with 2001 Canadian Census data demonstrated that most Canadian-born citizens aged 25 to 50 could not obtain the points necessary to enter Canada due to a lack of human capital. 5 See deVoretz and zhang (2004).6 This is clearly what happened in the case of the Hong Kong Chinese who immigrated to Canada and became Canadian citizens prior to 1997. Many became citizens to obtain a Canadian passport as an insurance scheme in case they could never return to Hong Kong. See deVoretz and Guo (2007).
THE ECoNoMICS oF CITIzENSHIP: A SYNTHESIS
161
Canada
deVoretz and Pivnenko clearly established that the relevant socio-econo-mic arguments outlined above condition the ascension to citizenship in Canada. They confirmed the economic model of citizenship ascension in the Canadian context when they demonstrated that the naturalization deci-sion is conditioned by the expected wage gain, immigrant educational attainment, marital status, age and presence of children in the immigrant household. At the macro level, their study focused on the implications of Canadian citizenship on the earnings of naturalized immigrants and their lifetime contributions to public finance contributions. They reported that all immigrants, regardless of their source country group and citizenship status, made a positive contribution to Canada’s treasury circa 1996 over their life cycle. Contributions to the Canadian treasury of naturalized citi-zens from oECd countries exceeded those of the Canadian-born by more than $14,000 over their lifetime. In addition, naturalized citizens made larger net federal treasury contributions than their non-citizen counterparts regardless of source country.
Taken collectively, these results supported the existence of a triple selection process for Canadian immigrants who naturalize. Thus, a positive selection bias surrounds the decision to ascend to Canadian citizenship. Furthermore, deVoretz and Pivnenko reported oaxaca decomposition results that demonstrated the importance of differential human capital endowments across citizenship status to explain differential citizenship/non-citizenship earnings differences. In sum, our initial speculations about the positive self-selection inherent in the Canadian immigration and citi-zenship experiences were borne out by the deVoretz and Pivnenko study.
United States
The United States is an intermediate case of triple immigrant selection. Alt-hough the United States yearly receives the greatest absolute number of immigrants and has a long history as an immigrant-receiving country, their selection and integration policies are ambivalent. In fact, the United States does not doubly select most permanent immigrants since the majority of the yearly intake of immigrants is either illegal (singly self-selected in the extreme), members of the family reunification class (again, single selection), or temporary migrants who may become permanent and eventually be eli-gible for citizenship. Thus, on balance the immigrant stock in the United States has been singly selected for admission. Hence, predictions derived from our triple selection economic model may not apply to the United States.
PIETEr BEVElANdEr & doN J. dEVorETz
162
A further deterrent to citizenship ascension in the United States is the ambiguous nature of its dual citizenship policy. Although a dual citizenship policy is observed, there exist several caveats about its actual existence. Nonetheless, the United States precludes the receipt of federally funded benefits by immigrants and, in turn, this prohibition offers a strong econo-mic incentive to naturalize. First, non-citizens can not enjoy welfare bene-fits and sponsorship of relatives is near impossible. Hence, naturalization which allows new citizens to access welfare and accelerates family reunifi-cation may encourage adverse selection into citizenship ascension in the United States. on the other hand, some research suggests that labour market incentives also enhance the prospects of naturalization and promote positive selection into citizenship.
In short, there is no a priori case for positive or adverse selection into United States citizenship, and only empirical evidence can confirm the pre-sence or absence of adverse selection.
In the absence of double selection in the United States context Akbari still reports results which support our economic model of citizenship ascen-sion. The typical arguments underlying the human capital model of citizen-ship ascension apply to the United States, along with a positive effect derived from home ownership. The impact of naturalization on the United States labour market is positive for most immigrants, with males from the developed world being the exception. Akbari further notes that naturalized United States immigrants earn more because they have greater human capital endowments and enjoy greater rewards from their human capital than non-citizens. In this sense positive discrimination prevails on the United States labour market.
Citizenship acquisition in the United States also increases both immi-grant contributions to the treasury and welfare participation. However, the net effect of these two phenomena is that tax payments exceed transfer payments for naturalized citizens between their 10th and 40th year of resi-dence.
In sum, we can conclude that, even in the absence of a dominant triple selection path to citizenship, naturalized citizens in the United States largely appear to behave like their counterparts in Canada.
For its part, the “old World” seems populated by singly selected immi-grants who adversely select into citizenship. Under these conditions we would expect weaker support for our economic model of immigrant citi-zenship ascension and even less support for positive economic outcomes derived from “old World” citizenship acquisition. We now turn to the studies of “old World” countries to see if these predictions are confirmed.
THE ECoNoMICS oF CITIzENSHIP: A SYNTHESIS
163
Netherlands
The results reported for the Netherlands are based on two surveys conduc-ted in 2002 and 2003 with two immigrant cohorts. The older cohort of immigrants consists of labour migrants, family reunion migrants and mar-riage migrants, whereas the newer cohort of immigrants consists of refu-gees and family reunion migrants. Clearly the majority of arrivals were singly selected.
Citizenship in the Netherlands has been fairly generously granted at times within the framework of a mixed jus soli and jus sanguinis with peri-odic dual citizenship privileges. Between 1992 and 1997, when dual citi-zenship was permitted, a remarkable rise in the number of new citizens was registered. After 1997, citizenship acquisition was conditioned by the immigrant’s schooling and the location of the educational institution. Par-ticipation in the Netherlands ‘integration programs’ did not correlate with a higher naturalization rate, but being a women and living longer in the Netherlands positively conditioned citizenship acquisition.
Moreover, citizenship acquisition in the Netherlands had a positive and significant effect on the immigrants’ employment opportunities. However, while naturalization had no significant effect on the income prospects of immigrants, it increased refugees’ earnings. This is a peculiar outcome in terms of our theory since refugees were singly selected and with little human capital.
In sum, the Netherlands case provides evidence to support our position that single selection leads to limited economic returns in the context of citi-zenship acquisition.
Norway
The study of Norway highlights a unique case of ‘single selection’ for admission. Moreover, ascension to citizenship in Norway is particularly difficult because of the eight-year residency requirement. In addition, citi-zenship is granted on the basis of birth (jus sanguinis) and, in principle, dual citizenship is not allowed. Nonetheless, Hayfron’s results show that economic factors were important determinants of citizenship ascension in the 1990s study period. In fact, length of stay in Norway, age, predicted wage, the prospect of employment in jobs reserved for Norwegian natio-nals, and immigrant admission status were important determinants of citi-zenship ascension rates in Norway. In sum, Norway does not conform to our expectations that single selection blunts the impact of economic forces on citizenship acquisition, but rather strongly supports our economic inter-pretation.
PIETEr BEVElANdEr & doN J. dEVorETz
164
Sweden
Admission into Sweden was initially a ‘double selection’ process of labour migrants. This policy was followed by a period of ‘single selection’ for more recent refugee admissions. Swedish naturalization legislation during the study period required short residency, and dual citizenship has been permitted since July 1st, 2001. Scott’s estimates suggest three patters of Swedish citizenship acquisition. First, immigrants from industrialized and non-labour-surplus countries acquire Swedish citizenship at initially low, but gradually increasing, rates. Second, immigrants from labour-surplus countries show generally high ascension rates which slow overtime. Third, refugees immediately ascend to citizenship at high rates.
The estimated economic effects derived from Swedish naturalization are small, and only a small income premium occurs after ascension. Finally, Scott’s results show that the so-called ‘naturalization effect’ found in other studies may largely be a selection effect and not a function of citizenship itself. Thus, Sweden conforms to our single selection prediction model.
Conclusions and ImplicationsIn sum, we can conclude from the cited studies that, even in the absence of ‘triple selection’ into citizenship status, the naturalized United States citi-zens appear to follow the path of Canadian immigrants. In fact, irrespec-tive of admission status and economic development of country of origin, these “New World” immigrants seem to experience a ‘citizenship premium’. results derived from the studies on “old World” countries support only a weak positive relationship between economic integration and citizenship ascension. In addition, singly selected immigrants from developing countries to the “old World” are largely refugees and family reunion migrants who derive small or no economic benefits from citizenship ascension.
Before delving into the policy implications of our case studies, we need to remind ourselves that the studies contained in this book were based on a variety of data sources. The advantage of data diversity is that a broader variable set could be used to identify naturalization propensities and their economic consequences. on the other hand, we must exert extreme caution when making comparisons.
overall three types of data were used in the preceding chapters: (1) administrative data or population registers for Sweden and Norway7 con-nected to other individual data sets, (2) census data for Canada and the United States, and (3) survey data for the Netherlands with a focus on immigrants and a variety of integration indicators. The advantage of popu-lation registers, or longitudinal data in general, is that the individual natu-
7 For Norway only the year 1999 is analysed.
THE ECoNoMICS oF CITIzENSHIP: A SYNTHESIS
165
ralization decision as well as the economic integration can be studied as a sequence of events. This type of sequential analysis is important because it can disentangle effects of naturalization from effects of self-selection, i.e., factors that make upwardly mobile immigrants more likely to naturalize. Census data, on the other hand, is among one of the most reliable data sources and often contain many other types of indicators that can be con-nected to the citizenship decision of immigrants. However, census data is retrospective and only reports citizenship status, not what preceded it. Survey data, while often limited in size, include topics not covered by offi-cial statistics, such as the intention to naturalize, and expectations tied to the acquisition of citizenship and dual nationality. Thus, we argue that a clear lesson from this book is that we need strong longitudinal data to study immigrant ascension to citizenship in a comparative framework.
Taking into account concerns surrounding the data used in our studies, what policy lessons have we learned on the economics of citizenship? Tra-ditionally, the transition of immigrants to full citizens is analyzed by politi-cal scientists and sociologists using a legal or normative perspective which we still consider important. In this book we aimed to complement previous studies by offering an economic viewpoint. our results indicate that the immigrant selection processes influence naturalization patterns and affect the size of the economic premium derived from citizenship acquisition. Thus, the design of countries’ immigration and citizenship policies enhan-ces or impedes the economic integration of their potential citizens.
It is important to note that both “New” and “old” Worlds current citizenship policies were designed to meet the individual countries’ self-concepts as ethnic or civic nations, and that they were not used to produce any ‘economic premiums’ to contribute to economic growth. In fact, the results of this book point to a policy trade-off between immigrant and citi-zenship acquisition policies. If a country applies a rigorous screening device for immigrant entry, then an economic citizenship premium can be achie-ved under a liberal citizenship regime. on the other hand, if a country only singly selects its immigrants, then only a rigorous citizenship screening policy will yield an economic premium from naturalization.
Moreover, each country of either the “New” or the “old” World must recognize that the passive selection of immigrants and of citizenship candi-dates leads to poor economic integration prospects, as adverse selection into citizenship will result. In other words, over time citizenship ascension must be seen as a part of the ongoing immigrant integration process instead of as its capstone.
appendixCitizenship provisions and benefits in Canada, the United States, Sweden, Norway, and the Netherlands
Can
ada
Uni
ted
Stat
esSw
eden
Nor
way
Net
herl
ands
Cit
izen
ship
rul
e(j
us s
oli,
jus
sang
uini
s)M
ixed
Mix
edJu
s sa
ngui
nis
Jus
sang
uini
s
J
us s
angu
inis
Nat
ural
izat
ion
age
18 a
nd o
ver
18 a
nd o
ver
18 a
nd o
ver
18 a
nd o
ver
18 a
nd o
ver
or m
arri
ed/
been
mar
ried
dua
l cit
izen
ship
Sinc
e 19
77Si
nce
2001
Not
in g
ener
al, b
ut p
ossi
ble
in c
erta
in c
ases
Bet
wee
n 19
92-1
997,
sti
ll po
ssib
le in
cer
tain
cas
esE
ligib
ility
req
uire
men
ts
r
esid
ence
Thr
ee o
f th
e la
st f
our
year
s be
fore
app
lyin
gFo
ur y
ears
and
nin
e m
onth
s be
fore
app
lyin
gl
ive
lega
lly (
perm
anen
t re
side
nt p
erm
it)
in t
he
coun
try
and
five
year
s be
fore
app
lyin
g.N
ordi
c ci
tize
ns o
nly
two
year
s.r
efug
ees
or s
tate
less
in
divi
dual
s on
ly f
our
year
s
liv
e le
gally
in N
orw
ay f
or
seve
n ye
ars.
Nor
dic
nati
onal
s tw
o co
nsec
utiv
e ye
ars
Nat
ural
izat
ion
proc
edur
e:
liv
e (r
esid
ent
perm
it)
lega
lly in
the
Net
herl
ands
, T
he d
utch
Ant
illea
ns o
r A
ruba
five
yea
rs b
efor
e ap
plyi
ng.
onl
y th
ree
year
s w
hen
spou
se/p
artn
er is
dut
ch
l
angu
age
Eng
lish
or F
renc
h su
ffici
ent
for
com
mun
icat
ion
Spea
k, r
ead,
wri
te, a
nd
unde
rsta
nd b
asic
Eng
lish
Bas
ic d
utch
, abi
lity
to
unde
rsta
nd, s
peak
, rea
d an
d w
rite
; nat
ural
izat
ion
test
oth
erN
o co
nvic
tion
s of
an
indi
ctab
le c
rim
inal
off
ence
in
the
pas
t th
ree
year
s
and
know
ledg
e ab
out
Can
ada
and
know
ledg
e ab
out
the
righ
ts a
nd r
espo
nsib
iliti
es
of c
itiz
ensh
ip
”Goo
d m
oral
cha
ract
er”:
la
st fi
ve (
thre
e) y
ears
of
tax
es p
rope
rly
paid
,
mal
es b
etw
een
18-2
6 ye
ars
old
have
reg
iste
red
for
the
Sele
ctiv
e Se
rvic
e,
no
cri
min
al r
ecor
d, e
tc.,
an
d
mus
t be
abl
e to
de
mon
stra
te k
now
ledg
e an
d un
ders
tand
ing
of
the
fund
amen
tals
of
U.S
. hi
stor
y an
d pr
inci
ples
of
gove
rnm
ent
Con
vict
ions
of
an
indi
ctab
le c
rim
inal
off
ence
le
ads
to p
roba
tion
of
obta
inin
g ci
tize
nshi
p
rec
ord
of g
ood
cond
uct
No
high
deb
ts
No
conv
icti
ons
nor
high
fin
es d
urin
g th
e la
st f
our
year
s
Res
pons
ibili
ties
Ren
unci
atio
n of
form
er
citiz
ensh
ip if
allo
wed
by
coun
try o
f citi
zens
hip
Ren
unci
atio
n of
form
er
citiz
ensh
ip if
allo
wed
by
coun
try o
f citi
zens
hip
THE ECoNoMICS oF CITIzENSHIP: A SYNTHESIS
167
oth
er b
enefi
tsC
itiz
ens
can
peti
tion
for
m
ore
fam
ily m
embe
rs
to c
ome
to t
he U
.S. w
ith
shor
ter
wai
ting
list
s fo
r gr
een
card
s.
Cit
izen
s w
ho r
etir
e ab
road
re
ceiv
e fu
ll So
cial
Sec
urit
y be
nefit
s (w
here
as la
wfu
l pe
rman
ent
resi
dent
s re
ceiv
e on
ly h
alf
thei
r be
nefit
s) a
nd
citi
zens
may
be
subj
ect
to
few
er r
estr
icti
ons
on e
stat
e ta
xes.
Cit
izen
s ar
e el
igib
le f
or
mor
e pu
blic
ben
efits
, suc
h as
Sup
plem
enta
l Sec
urit
y In
com
e an
d Fo
od S
tam
ps,
and
cert
ain
type
s of
sc
hola
rshi
ps a
nd fi
nanc
ial
aid.
Cer
tain
occ
upat
ions
- in
clud
ing
polic
e,
prof
essi
onal
sol
dier
s an
d so
me
safe
ty s
ervi
ce o
ffice
rs-
rest
rict
ed t
o ci
tize
ns
Cer
tain
occ
upat
ions
- in
clud
ing
polic
e,
prof
essi
onal
sol
dier
s an
d hi
gher
ser
vice
offi
cers
- re
stri
cted
to
citi
zens
Cer
tain
occ
upat
ions
- in
clud
ing
polic
e,
prof
essi
onal
sol
dier
s, a
nd
high
er s
ervi
ce o
ffice
rs-
re
stri
cted
to
citi
zens
res
pons
ibili
ties
ren
unci
atio
n of
for
mer
ci
tize
nshi
p if
allo
wed
by
coun
try
of c
itiz
ensh
ip
ren
unci
atio
n of
for
mer
ci
tize
nshi
p if
allo
wed
by
coun
try
of c
itiz
ensh
ipPr
ivile
ges
run
or
vote
in e
lect
ions
run
or
vote
in e
lect
ions
Unc
ondi
tion
al r
ight
to
live
and
wor
k in
Sw
eden
.r
un o
r vo
te in
nat
iona
l pa
rlia
men
tary
ele
ctio
ns.
Prot
ecti
on b
y au
thor
itie
s
run
or
vote
in n
atio
nal
parl
iam
enta
ry e
lect
ions
. Pr
otec
tion
by
auth
orit
ies
run
or
vote
in n
atio
nal
an
d pr
ovin
cial
ele
ctio
ns.
Prot
ecti
on b
y au
thor
itie
s
169
RefeRences
Afifi, A.A. and V. Clark. 1990. Computer-Aided Multivariate Analysis. 2nd ed. New York: Van Nostrand Reinhold.
Aftenposten. 2005. http://www.aftenposten.no/english/local/article1260069.ece.
Aldrich, J.H. and F.D. Nelson. 1984. Linear Probability, Logit, and Probit Models. Newbury Park, CA: Sage.
Alvarez, R. 1987. A profile of the citizenship process among Hispanics in the United States. International Migration Review, 21: 327-51.
Balistreri, K.S. and J.V. Hook. 2004. The more things change the more they stay the same: Mexican naturalization before and after welfare reforms. International Migration Review, 38: 113-30.
Barkan E.R. and N. Khokhlov. 1980. Socioeconomic data as indices of naturalization patterns in the United States: A theory revisited. Ethni-city, 7: 159-90.
Barth, E., B. Bratsberg and O. Raaum. 2004. Identifying earnings assimi-lation of immigrants under hanging macroeconomic conditions. The Scandinavian Journal of Economics, 106: 1-22.
Bauböck, R. 1994. Transnational Citizenship: Membership and Rights in International Migration. Aldershot: Edward Elgar.
Bauböck, R. (ed.) 2006. Migration and Citizenship, Legal Status, Rights and Political Participation. IMSCOE Reports: Amsterdam University Press.
Bengtsson, T. and K. Scott. 2006. Immigrant consumption of sickness benefits in Sweden, 1982–1991. Journal of Socio-Economics, 35: 440-57.
Bernard, W.S. 1936. Cultural Determinants of Naturalization. American Sociological Review, 1: 943-53.
170
Bevelander, P. 2000. Immigrant Employment Integration and Structural Change in Sweden, 1970-1995. Almqvist & Wiksell International, Sweden.
Bevelander, P. and H.S. Nielsen. 2001. Declining employment success of immigrant males in Sweden: Observed or unobserved characteristics? Journal of Population Economics, 14: 455-71.
Bevelander, P. and J. Veenman. 2006. Naturalisation and Employment Integration of Turkish and Moroccan Immigrants in the Netherlands. Journal of International Migration and Integration, 7: 327-50.
Blau, F.D. 1984. The use of transfer payments by immigrants. Industrial and Labour Relation Review, 37: 222-39.
Blinder, A. 1974. Wage Discrimination: Reduced Form and Structural Estimates. The Journal of Human Resources, 8: 436-55.
Bloemraad, I. 2006. Becoming a Citizen: Incorporation of Immigrants and Refugees in the United States and Canada. Berkeley and Los Angeles: University of California Press.
Bloemraad, I. 2002. The North American Naturalization Gap: An Institu-tional Approach to Citizenship Acquisition in the United States and Canada. International Migration Review, 36: 193-228.
Borjas, G.J. 2002a. Welfare reform and immigrant participation. Interna-tional Migration Review, 36: 1093-1123.
Borjas, G.J. 2002b. The impact of welfare reform on immigrant welfare use. Center for Immigration Studies (March).
Borjas, G.J. 1999. Immigration and the Food Stamp Program. Cambridge, MA: Harvard University.
Borjas, G.J. and L. Hilton. 1996. Immigration and the welfare state: Immigrant participation in means-tested entitlement programs. Quar-terly Journal of Economics, 111: 575-604.
Borjas, G.J. and S.J. Trejo. 1991. Immigrant participation in the welfare system. Industrial and Labour Relations Review, 44: 195-211.
Bratsberg B., J.F. Ragan and Z.M. Nasir. 2002. The effect of naturaliza-tion on wage growth: A panel study of young male immigrants. Journal of Labor Economics, 20: 568-97.
171
Bueker, C.S. 2005. Political incorporation among immigrants from ten areas of origin: The persistence of source country effects. International Migration Review, 39: 103-40.
Centres for Medicare and Medicaid Services. 2005a. National Health Expenditures and Selected Economic Indicators, Levels and Annual Percent Change: Selected Calendar Years 1998-2014. U.S.
Centres for Medicare and Medicaid Services. 2005b. The Nation’s Health Dollars, 2003. U.S.
Chiswick, B. 1978. The effect of Americanization on the earnings of foreign-born men. Journal of Political Economy, 86: 897-921.
De Volkskrant, September 3, 2004. Kabinet blijft fel tegen dubbele natio-naliteit. Retrieved October 4, 2004: http://www.volkskrant.nl/denhaag/printversie/1094102585211.html
Department of Internal Affairs of New Zealand web site: http://www.dia.govt.nz/diawebsite.nsf/wpg_URL/Services-Citizenship-Requirements-for-New-Zealand-Citizenship?OpenDocument
DeSipio, L. 1987. Social Science Literature and the Naturalization Process. International Migration Review, 21: 390-405.
DeVoretz, D.J. 2008. The economics of citizenship: a common intellectual ground for social scientists. Journal of Ethnic and Migration Studies, 34: 679-93.
DeVoretz, D.J. and S. Guo. 2007. Chinese immigrants in Vancouver: Quo vadis? Journal of Immigration and Integration, 7: 425-47.
DeVoretz, D.J. and Z. Ma. 2002. Triangular Human Capital Flows between Sending, Entrepot and the Rest of the World Regions. Cana-dian Population Studies, 29: 53-69.
DeVoretz, D.J. and S. Pivnenko. 2006. The economic causes and conse-quences of Canadian citizenship. Journal of International Migration and Integration, 6: 435-68.
DeVoretz, D.J. and S. Pivnenko. 2004. Immigrant public policy transfers: A Comparative Analysis by City. Canadian Journal of Urban Research, 13: 155-69.
DeVoretz, D.J. and K. Zhang. 2004. Citizenship, passports and the brain exchange triangle. Journal of Comparative Policy Analysis, 6: 199-212.
172
Ekberg, J. and B. Gustafsson. 1995. Invandrare på arbetsmarknaden. Stockholm: SNS.
Espenshade, T., J. Baraka and G. Huber. 1997. Implications of the 1996 welfare and immigration reform acts for U.S. immigration. Population and Development Review, 23: 769-801.
Esping-Andersen, G. 1990. The Three Worlds of Welfare Capitalism. Oxford: Polity Press.
Groeneveld, S. and Y. Weyers-Martens. 2003. Minderheden in beeld, SPVA-02. Instituut voor Sociologisch-Econmisch Onderzoek (ISEO), Rotterdam.
Guide to Naturalization pamphlet 04/99:400 (revised April 1999): www.udi.no/ upload/English/EngPublications/usortert/Norwegian%citizenship. pdf
Gustafsson, P. 2005. International Migration and National Belonging in the Swedish Debate on Dual Citizenship. Acta Sociologica, 48: 5-19.
Hanushek, E.A. and J.E. Jackson. 1977. Statistics Methods for Social Sci-ences. New York: Academic Press.
Hayfron, J.E. 2001. Language training, language proficiency and earnings of immigrants in Norway. Applied Economics, 33: 1971-79.
Hayfron, J.E. 1998. The performance of immigrants in the Norwegian labor market. Journal of Population Economics, 11: 293-303.
Heckman, J. 1976. The common structure of statistical models of trunca-tion, sample selection and limited dependent variables and a simple estimator for such models. Annals of Economic and Social Measure-ment, 5: 475-92.
Hsiao, C. 1992. Analysis of panel data. Econometric Society Monographs No. 11. Cambridge University Press.
Jasso, G. and M.R. Rosenzweig. 1990. The New Chosen People: Immi-grants in the United States. New York: Russell Sage Foundation (The Population of the United States in the 1980s: A Census Monograph Series).
Jasso, G. and M.R. Rosenzweig. 1986. Family reunification and the immi-gration multiplier: U.S. immigration law, origin-country conditions, and the reproduction of immigrants. Demography, 23: 291-311.
173
Kelley, J. and I. McAllister. 1982. The decision to become an Australian citizen. Australian and New Zealand Journal of Sociology, 18: 428-39.
Kogan, I. 2003. Ex-Yugoslavs in the Austrian and Swedish labor markets: The significance of period of migration and the effect of citizenship acquisition. Journal of Ethnic and Migration Studies, 29: 595-622.
Kommunal- og regionaldepartementet. 2003. Report to the Storting #17: Asylum and Refugee Policies in Norway. April 5.
Lee, L.F. 1978. Unionism and wage rates: a simultaneous equations model with qualitative and limited dependant variables. International Econo-mic Review, 19: 415-33.
Levitt, S and S. Dubner. 2005. Freakonomics. London: Penguin books.
Liang, Z. 1994. Social contact, social capital, and the naturalization process: evidence from six immigrant groups. Social Science Research, 23: 407-37.
Longva, P. and O. Raaum. 2003. Earnings assimilation of immigrants in Norway – A reappraisal. Journal of Population Economics, 16: 177-93.
Maddala, G.S. 1986. Limited dependent variable models using panel data. Journal of Human Resources, 22: 307-38.
Massey, D. and K. Bartley. 2005. Changing legal status distribution of immigrants: a caution. International Migration Review, 39: 469-84.
Mata, F. 1999. Patterns of acquiring citizenship. In Immigrant Canada: Demographic, Economic and Social Challenges edited by S. Halli and L. Driedger. Toronto: University of Toronto Press.
National Centre for Education Statistics. 2001. Early Estimates of Public Elementary and Secondary Education Statistics: School Year 2000-01. U.S. Department of Education 9NCES 2001-331.
Neuman, S. and R. Oaxaca. 2003. Estimating labor market discrimination with selectivity corrected wage equations: Methodological considera-tions and an illustration from Israel. The Pinhas Sapir Center for Deve-lopment Paper No. 2-2003.
Oaxaca, R. 1973. Male-female wage differentials in urban labor markets. International Economic Review, 14: 693-709.
174
Pivnenko, S. and D.J. DeVoretz. 2003. The recent economic performance of Ukrainian immigrants in Canada and the U.S. IZA Discussion Paper No. 913. Bonn: IZA.
Portes, A. and J.W. Curtis. 1987. Changing flags: naturalization and its determinants among Mexican immigrants. International Migration Review, 21: 352-71.
Portes, A. and R. Mozo. 1985. The political adaptation process of Cubans and other ethnic minorities in the United States: A preliminary analy-sis. International Migration Review, 16: 35-63.
Portes, A. and R. Rumbaut. 1990. Immigrant America. Berkley and Los Angeles: University of California Press.
Proposition 1968. Kungl. Maj:ts proposition 1968: 142. Riktlinjer för utlänngspolitiken m.m. [His Royal Majesty’s Proposition 1968: 142. Guidelines for Immigrant Policy, etcetera]. Stockholm: Swedish Parlia-ment. P. 142.
Robinson, C. and N. Tomes. 1982. Self-Selection and Interprovincial Mig-ration in Canada. Canadian Journal of Economics, 15: 474-502.
Rooth, D.-O. 1999. Refugee Immigrants in Sweden: Educational Invest-ments and Labour Market Integration. Doctoral Dissertation, Lund University, Sweden.
Rosholm, M., K. Scott and L. Husted. 2006. The Times They Are A-Changin’: Declining Immigrant Employment Opportunities in Scan-dinavia. International Migration Review, 40: 318-47.
Schuck, P.H. 1998. Citizens, Strangers, and In-Betweens: Essays on Immi-gration and Citizenship. Boulder, CO: Westview.
Scott, K. 1999. The Immigrant Experience: Changing Employment and Income Patterns in Sweden, 1970-1993. Lund: Lund University Press.
Shamsuddin, A.F.M. and D.J. DeVoretz. 1999. Wealth accumulation of Canadian and foreign-born households in Canada. Review of Income and Wealth, 44: 515-53.
Simon, J. 1984. Immigrants, taxes, and welfare in the United States. Popu-lation and Development Review, 10: 55-69.
SOU. 1999. Svenskt medborgarskap [Swedish Citizenship]. Stockholm: Fakta Info Direkt. P. 34.
175
Soysal, Y.N. 1994. Limits of Citizenship: Migrants and Postnational Membership in Europe. Chicago: University of Chicago Press.
Van den Maagdenberg, V. (ed.) 2004. Jaarrapport Integratie 2004. ISEO, Rotterdam.
Van der Erf, R. 2000. Naturalisatie in de EU, Nederlanderschap blijkt populair. Demos, 16: 1-7.
Watsula, M. 2005. Naturalization in the United States: who, why, and how? Multicultural Center Prague. April.
Yang, P.Q. 1994. Explaning immigrant naturalization. International Migration Review, 28: 449-77.
Zhang, K. 2006. Recognizing the Canadian Diaspora. Canada Asia Com-mentary No. 41. Asia Pacific Foundation of Canada, Vancouver.