UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Strangers inside our gates: public opinion towards immigration in Canada, the United States, and the… Farris, Lily 2007

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2008_spring_farris_lily.pdf [ 27.77MB ]
Metadata
JSON: 24-1.0066230.json
JSON-LD: 24-1.0066230-ld.json
RDF/XML (Pretty): 24-1.0066230-rdf.xml
RDF/JSON: 24-1.0066230-rdf.json
Turtle: 24-1.0066230-turtle.txt
N-Triples: 24-1.0066230-rdf-ntriples.txt
Original Record: 24-1.0066230-source.json
Full Text
24-1.0066230-fulltext.txt
Citation
24-1.0066230.ris

Full Text

STRANGERS INSIDE OUR GATES: PUBLIC OPINION TOWARDS IMMIGRATION IN CANADA, THE UNITED STATES, AND THE UNITED KINGDOM by LILY FARRIS B.A., The University of British Columbia, 2005 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (Sociology) THE UNIVERSITY OF BRITISH COLUMBIA December 2007 © Lily Farris, 2007 ABSTRACT Using 2005 data from Gallup public opinion surveys on attitudes toward immigration policy in Canada, the United States, and the United Kingdom this study explores the factors that impact attitudes. Additional analysis is conducted on the United States exploring how economic, political and associative measures impact attitudes. 11 Table of Contents Abstract^ Table of Contents^  .iii List of Tables ^ .iv Acknowledgements .v Dedication^  .vi CHAPTER I Introduction and Literature Review ^ 1 1.1 Introduction^  1 1.2 Objectives ^2 1.3 Measuring Public Opinion ^ 3 1.4 Comparing Canada, the United States, and the United Kingdom ^ 5 1.5 Forces in Canada^ 6 1.6 Forces in the United States^ .16 1.7 Forces in the United Kingdom 25 1.8 Conclusion^  .32 1.9 Bibliography 34 CHAPTER II Shouting across the Divide: Cross-National Variation in Attitudes toward Immigration Policy in Canada, the United States and the United Kingdom, 2005 2.1 Introduction^ 35 2.2 Dimensions of Attitudes toward Immigration^ 44 2.3 Objectives  .48 2.4 Data and Measures^ .49 2.5 Strategy of Analysis .53 2.6 Results ^ 54 2.7 Discussion/Conclusion ^ 57 2.8 Bibliography  .61 CHAPTER III Dimensions of Hostility: Analyzing the Factors that Shape American Public Opinion toward Immigration Policy, 2005 3.1 Introduction^ 66 3.2 Dimensions of Attitudes toward Immigration ^ 69 3.3 Objectives 76 3.4 Data and Measures^ .79 3.5 Dependant Variable 79 3.6 Individual-Level Independent Variables^ 80 3.7 Strategy of Analysis ^ .84 3.8 Results ^  84 3.9 Discussion/Conclusion  . 90 3.10 Bibliography 94 CHAPTER IV Conclusion 4.1 Conclusion^ 104 4.2 Conclusions from Chapter 2^ 104 4.3 Conclusions from Chapter 3 .105 4.4 Limitations/Further avenues of research^ 106 4.5 Overall Conclusions^  110 4.6 Bibliography .112 LIST OF TABLES Table 2.1 Percentage of each group and their levels of support for immigration in Canada, the United States and the United Kingdom^ 64 Table 2.2 Multinomial Logistic Regression of Immigration Levels and Socio- Demographic Characteristics (versus Stay the Same) ^65 Table 3.1 Percentage of each group who want given levels of immigration in the United States, 2005^ 100 Table 3.2 Percentage of each group who want given levels of immigration in the United States, 2005 101 Table 3.3 Percentage of each group who want given levels of immigration in the United States, 2005^ 102 Table 3.4 Percentage of each group who want given levels of immigration in the United States, 2005 102 Table 3.5 Percentage Logit model Estimates of Immigration Attitudes (versus 'stay the same')^ 103 iv ACKNOWLEDGEMENTS I offer my enduring gratitude to the faculty, staff and my fellow students at UBC, who continually inspire me. I owe particular thanks to Dr. Rima Wilkes and Dr. Neil Guppy who have both asked me the hard questions and provided me with the opportunity to find out the answers for myself It has been an honour to work with these two amazing researchers. Thanks to Jill Lambert, Sandra Enns, and Kieran O'Doherty for comments on the earlier drafts of this paper. Thank you to my family Marta, Steve and Sean Farris for their continual support and enthusiasm. A special thanks to my grandmother Ruth Oligvie. Thank you to Danielle Mashon, Amanda Hessler, Martin Arnesen, Devon Bouchard, Alia Husain, and Antoaneta Nimoh. Thank you to RIIM and Metropolis for funding, Eric Neilson of Gallup International Polling Organization and the UBC Department of Sociology for research facilities and equipment. v DEDICATION To Sean Farris vi Chapter 1: Introduction and Literature Review 'Strangers in my Homeland': Comparing the deep social and structural forces that impact attitudes toward immigration in Canada, the United States and the United Kingdom 1.1 Introduction Americans tighten illegal immigration regulations. British citizens with immigrant parents are charged with acts of terrorism. Canadians debate whether to accept HIV positive immigrants. These examples illustrate the increasing complexity involved in understanding how views on immigration policy are formed. Resistance to change in immigration policy is on the rise internationally, and policy formation remains a politically contentious topic in many Western nations (Simon and Lynch, 1999; Crawley, 2005). In this climate of growing attention to immigration, mapping public attitudes toward immigration policy may help to illuminate some of the characteristics that shape attitudes. To further understand the construction of attitudes towards immigration this thesis analyze, the ways in which both individual and contextual factors impact perceptions of immigration policy. In chapter 2, I replicate and furthers cross-national literature on attitudes toward immigration by comparing three structurally similar countries: Canada, the United States, and the United Kingdom. In chapter 3, I analyze attitudes toward immigration in the United States, testing the impact that individual-level dimensions of demographic, economic, political, and personal contact have on attitudes. The work in this thesis uses 2005 Gallup International Polling Organization to analyze, in multinomial logit models, the individual-level characteristics that impact support and hostility toward immigration. The goal of this work is to retest measures that have previously been found 1 to significantly impact attitudes toward immigration, while also providing new insight into the complexities of public opinion formation. Through the comparison of positive and negative views toward immigration further insight is gained into how attitudes are fostered both cross-nationally and cross-sectionally. I begin by analyzing public opinion as a tool for analysis, and I also compare the historical contexts in each country. By providing background on the development of immigration in each country the ways in which public opinion differs between these three countries is highlighted. I argue that public opinion cannot be disentangled from the context in which it develops. 1.2 Objectives The two papers in this thesis use secondary survey data to determine the impact of individual-level characteristics on attitudes toward immigration policy. The first paper compares how age, gender, race, religion, income and employment status impact attitudes in Canada, the United States, and the United Kingdom. The second paper focuses on how interracial interaction might impact attitudes towards immigration policy. The specific research questions addressed in each paper are outlined below. Chapter 2 Research Question 1: Are attitudes towards immigration similar or different in Canada, the United States and the United Kingdom? Research Question 2: How, if at all, do individual-level characteristics affect attitudes toward immigration policy differently in Canada, the United States and the United Kingdom? Chapter 3 Research Question 1: Do levels of support for immigration in the United States vary according to individual-level characteristics? If so what impact do demographic, economic, political and personal contact dimensions have on views toward immigration? 2 1.3 Measuring Public Opinion This thesis uses public opinion poll data collected by Gallup International Polling Organization to analyze attitudes toward immigration. Public opinion provides a way of determining meaningful measures of individual perceptions as well as possible predictions about their future behavior (Kelman, 1961). Theories about the processes through which individuals determine and represent their opinions can be tested using public opinion poll data. Through sampling a representative proportion of the population, in each country via telephone or in-person interviews, public opinion poll data can provide a barometer of current community views on particular topics. Often only the overall trends of poll results and their descriptive statistics are reported. This thesis addresses this gap by providing a more in-depth analysis, in comparison to contrasting frequencies cross-nationally (Lynch and Simon, 2003), of public opinion formation toward immigration policy, a politically contentious issue. One of the challenges of conducting a secondary data analysis of attitudes toward immigration policy is the availability and structure of previously collected data. In the case of this thesis, the opportunities for analysis are constrained by the measures available in the data and the uniformity of the measures included in the surveys by country. Public opinion data, as a one time survey, provides a one time snapshot on a contentious topic such as immigration. The analysis of multiple surveys over time may provide more insight into the consistency of attitudes as well as their development overtime. The analysis in this thesis will focus on understanding national public opinion toward legal immigration in each of the three research countries. The previously collected data sets selected for analysis were chosen for their inclusion of a uniform measure which 3 asks respondents for their individual views on immigration policy. The specific question asks respondents 'In your view, should immigration be kept at its present level, increased or decreased?' There are several benefits to the use of this particular research question, beyond its common presence in the literature providing opportunities for comparison (Espenshade and Hempstead, 1996; Mayda, 2004; Tucci, 2005; Citrin et al., 1997). The question format personalizes the issue of immigration policy asking respondents to determine what they themselves would do if they had control of immigration policy decisions, instead of asking their level of approval for the decisions made by policy- makers. In addition this measure allows respondents who are dissatisfied with current immigration policy to determine what they would prefer the immigration level to be. The question also does not specify the type of immigration (refugees, entrepreneurs family sponsorship etc.), legal status (illegal immigrants etc.), or outline the exact number of immigrants entering the country. This undefined question structure makes the cross- national comparison of this measure more robust capturing respondents overall views toward immigration not relating them to any specific topic. This format is beneficial when comparing attitudes between countries which may frame the topic of immigration in very different ways. There are limitations to the use of a single dependent variable to determine attitudes since past research suggests that attitudes sometimes vary by question format (Crawley, 2005). The analysis in this thesis focuses on reporting attitudes toward immigration policy saving comparisons between measures of immigration for later work. When analyzing this dependent variable in multinomial logistic regression models one of the three response categories must be excluded for comparison purposes. The 4 analysis conducted here uses 'remain the same' as the excluded category based on the opportunity this provides to measure the impact of the factors that support changing immigration policy whether to increase or decrease it in comparison to individuals who want policy to stay the same. Notice that the response options for this variable were 'remain the same', 'increase', 'decrease', 'other' or 'don't know'. If respondents selected the last two options they were excluded from the analysis since they did not indicate a preference for a particular immigration policy level. These three response options or 'remain the same', 'increase' or 'decrease' can also be viewed as judgments by respondents on the overall levels of support for current immigration policy. This interpretation might associate the responses of 'increase', 'remain the same' and 'decrease' with 'too little', 'the right amount' or 'too many' which could potentially reframe these responses as valued proportions which could be placed in ascending order. By giving values and placing them in proportional order these response categories could be using Ordinal Regression (OLS) models. This method, however, does not allow for the level of comparison available in multinomial logistic regression which determines both the positive and negative factors that impact immigration. 1.4 Comparing Canada, the United States, and the United Kingdom There is a need on the part of both policy-makers and academics to better understand the connections between attitudes towards immigration and issues of integration, legality, discrimination and national identity. Connections are often drawn between openness towards immigration and increased receptivity towards diversity generally (Buck et. al., 2003). While public opinion toward immigration has long been a 5 divisive issue for the Canadian, American and British publics this topic raises different related issues in each country. Canada and the United States have constructed their national identities as 'immigrant' nations, while the United Kingdom based its identity in the European tradition of a homogenous cultural construction (Castles, 2000). As Mackey notes "within modernity a nation must be seen to 'have' a distinct culture in order to be recognized" emphasizing the challenges faced by Canada and the United States in finding their distinct identity within their multiculturally diverse reality (Mackey, 1999:11). Part of creating citizenship is highlighting factors that separate citizens from non-citizens. In the case of the United Kingdom citizenship is associated with ethnicity. While in Canada and the United States it is more complex, since their immigration policies select in large part based on their ability to contribute to the economy, often producing a more ethnically and linguistically diverse group of residents. Unlike the United Kingdom, the United States and Canada have formed their national history in response to their experiences as colonies of the United Kingdom. For this reason, the two North American countries struggle to form their independent identities in the face of their increasing diverse realities (Castles, 2000). 1.5 Forces in Canada In this section the development of Canadian immigration policy will be outlined, highlighting how these changes in policy might inform current views on immigration. Canada, was built on the principles of "peace, order and good government" (paix, ordre et bon government see section 91 of the Constitution Act of 1867). While there is no evidence that the founders of Canada were significantly different to the founders of 6 the United States in terms of education, wealth, economic involvement or class differences in founding principles highlight the ways in which these similar countries were already on differing value trajectories (Grabb and Curtis, 2005). In contrast to Canada's motto which encourages the acceptance of the federal control of law the Americans "life liberty and the pursuit of happiness" and French ("liberte, 6galite, fraternite liberty, equality, fraternity) focus on the rights of the individual. The phrase of "peace, order and good government" was replicated in other British Acts of Parliament including the New Zealand Constitution Act of 1852, the Colonial Laws Validity Act of 1865, the British Settlements Act 1887, the Commonwealth of Australia Constitution Act 1900, and the West Indies Act of 1962. Canada received its independence from the United Kingdom in 1867 in part because Britain was providing colonies with their independence as a way to cut costs (Grabb and Curtis, 2005). While founded at different times and for different reasons Canada and the Untied States in the 1880s had a great deal of similarities. Both had resource based economies dominating industry and the need for settlers to populate their physically expansive national territories. Daily activities in both countries were also relatively similar with parallel proportions of the population in different industries and similar class distributions in which the majority of the population were at either extreme of rich or poor (Grabb and Curtis, 2005). However, over time the Canadian colonies fell behind the American republic economically, as the American advantage of climate, geography, and proportion of high quality land accentuated the opportunities available to Americans. This economic gap between Canada and the United States continued widening into the late 1880s, expanding even further after the American Civil War when 7 the United States matched and surpassed Britain as the world's most powerful economy. This increased American wealth accentuated class and capital differences within the United States, which raised the overall standard of living, while creating greater extremes in poverty rates. Canada's similarities and differences to the United States provide on important point of comparison both to understand historical similarities and to hierarchical differences. One of the distinctive aspects of Canadian society is the presence of the French in Canada. After the Seven Years War (1756-1763) between France and England, French territory, which already had permanent French colonies established, was transferred to Britain in the Treaty of Paris 1763. Thus British colonial leaders established rule over a French majority whose population would continue to expand significantly even though French-speaking immigration was restricted significantly after the English gained control. This isolation both from France and English-speaking Canada created a separate identity for French-Canadians governed by alternate social structures and significantly influenced by the Roman Catholic Church. Parallels developed between the French Canadian entrepreneurial, business, and state practices and procedure in the United States. However, the influence of Roman Catholic Church also insured that aspects of French society remained. The differences in language use between French and English Canada led to distinct levels of idea exchange, through which English-Canadian and American publics were exposed to more information on liberty fostering less support for authoritarianism in contrast to the French acceptance of the influence of the political elites. Not only did language divide Canadians ethnically, the regional focus of the French language geographically isolated these communities. In 1775 American military 8 forces invited Quebec to become the 14th colony in the rebellion against Britain. However, lack of support lead to rejection of this idea because local French leaders wanted to retain the autonomy that they had under the British, while the public was unsupportive of the idea due to the prospect of potential uncertainty associated with a revolution and finally the discouragement of the catholic church. These French and English charter groups have been characterized by some as the founding groups of Canada giving them a specific place within the social hierarchy of Canada (Porter, 1979). This prevailing position has fostered what Porter has called a 'vertical mosaic' structure in Canadian society through which the founding groups occupy the top social tier and First Nations, Metis and Inuit groups occupy the bottom tier (Porter, 1979). Immigrants, based on their visible, social and cultural distance to these founding groups occupy the middle positions in this social hierarchy with the assumption that these immigrants will have limited opportunities to take over positions of power within Canadian society. Taylor argues that it is in the best interests of Canadians to present a multicultural national identity through which all Canadians can see themselves reflected in the presentation of history (1992). Beyond the presence of the French in Canada, the United States and Canada were ethnically very similar in their early years. Early migrants to Canada and the United States were predominately English speaking from the British Isles (although Canada received relatively more Scottish than English) and Protestant. The prevalence of Protestantism in both countries fostered a perspective on life that each individual should work within their given field, impacting the social perspective on the role of individuals within society. The time of 1840-1900 saw over a million 9 Canadians move to the United States, due to increased opportunities and value for 'American' ideals accentuating the value different between the two countries (Grabb and Curtis, 2005). Canadian religion has been characterized by the monolithic, authoritarian and state-supported religion, and the connections between the Roman Catholic Church and the Quebec government provides a distinct way of understanding religion. In the late 1800s, labour shortages and settlement planning practices created the need for more immigration than desired source countries (the United States, the United Kingdom, and France) had available (Knowles, 1997). In order to meet labour market needs, immigration opportunities were extended to citizens from some of Canada's less preferred origin countries including Southern and Eastern European nations and Asia (although migration from Asia was highly regulated). Flows of migrants from China, South Asia, Africa and the Caribbean were regulated through the Head Tax, the Chinese Immigration Act of 1885, the Exclusion Act of 1923, the 1908 continuous passage requirement, and the 1910 Immigration Act (which used 'race' as a category for acceptance) (Thobani, 2000; Green, 1976). Introduced as an entry tax in 1903 the Chinese Head tax charged an additional immigration fee to Chinese immigrants upon entering Canada (Li, 1998). Designed first to limit, then to exclude Chinese immigration, the Head Tax was raised from $50 in 1888 to $500 in 1905 (Li, 1988). This change in entrance costs has been linked to the needs of the Canadian National Railroad as a means to regulate the inflow of Chinese labour. The tax peaked, once the railroad was complete, at $500 stopping Chinese immigration (Green, 2000). Those Chinese that moved to Canada during to this period faced systematic discrimination both in the labour force and in the social spaces of Canadian society (Li, 10 1998). While the Chinese were a valuable labour contribution to the development of Canada, immigration policy was designed to encourage immigrants not to settle permanently because of fears that their permanent presence would be shift the ethnic construction of Canadian society. The Opium and Narcotic Drug Act of 1922 allowed the deportation of 'domiciled aliens' (or immigrants who had been in Canada for five years of more) who head drug related convictions providing a way for the government to sent home immigrants who had been in the country too long (especially Chinese immigrants who were more likely to consume opium). An order council issued January 1923 excluded the entry of immigrants of 'Asiatic' race (which includes Middle Eastern countries such as Turkey and Syria) with the exception of farm workers, female domestic servants or the wives and children of individuals who were legally allowed to enter Canada. In 1923, Chinese immigrants were banned from entering the country altogether with the exception of diplomats, students, children of Canadians and investors, by the Chinese Immigration Act which was enacted on July 1 st known in the Chinese Canadian community at 'Humiliation Day' (Li, 1998). In September 1930 an Order in Council stopped the entry of all 'Asiatic' origin immigrants with the exception of wives and children of citizens. Following this in 1931 an Order in Council required Chinese and Japanese born individuals to renounce their former citizenship as part of Canadian naturalization process. At this point in time Japan had no policy through which to renounce citizenship thus restricting Japanese access to Canadian citizenship. The Railway Agreement put together by the Canadian Pacific and Canadian National Railway was designed to allow for the active recruitment of 11 immigrants to meet railway development needs. Between 1925-1929 Central and Eastern European immigrants entered Canada through changes in these regulations. In 1938 the Anglican Church, the United Church, the YMCA, the Co-Operative Commonwealth Federation (CCF), Jewish community groups and regional organizations called for the Canadian government to accept Jewish refugees from Europe. This proposition was opposed by the Native Sons of Canada, Leadership League, Canadian Corps, and groups in Quebec which were especially unsupportive of the refugees. When the St. Louis sailed from Germany with 930 Jewish refugees on board no country in the Americas would take them including Canada. The boat was forced to return to Europe where the majority of the passengers were killed by the Nazis. In February 1942 Japanese Canadians were required to move at least 100 miles from the Pacific due to Canada's was status with Japan. Many of these individuals who ended up in detention centers in British Columbia's interior. This situation continued until the end or the war while the Canadian government encouraged Japanese-Canadians to move back to Japan. The years after WWII saw the acceptance of large numbers of Polish veterans, the establishment of the Canadian Citizenship Act 1946 (making Canada the first Commonwealth country to create citizenship separate from Britain), the repeal of the Chinese Immigration Act 1947 (1930 regulations were re-implemented which allowed only the sponsorship of wives and children), and the creation of the Department of Citizenship and Immigration 1950. A June 1950 Order in Council changed immigration policy giving a great deal of autonomy to immigration officers while maintaining English, Irish, French and American citizens as preferred immigrants. Agreements were made between Canada and the governments of India, Pakistan and Ceylon (the name of 12 Sri Lanka before 1972) to accept a limited number of immigrants beyond those accepted under the Asiatic class. A new Immigration Act was passed in 1952 which did not change policy a great deal but did officially sanction the refusal of immigrant's based on their nationality, ethnicity, geographical area of origin, cultural customs, lack of suitability with the Canadian climate, and potential lack of ability to assimilate to Canada, while explicitly excluding Homosexuals, and prostitutes. This pattern of systematic marginalization affected Canadian conceptions of the place of immigrant participation in society especially when contrasting recent waves of immigration against previous situations. The Canadian immigrant selection process was in part of Canadian society who are racially or ethnically similar to the founding groups are preferred since their preference would contribute to the Canadian population while allowing it to remain ethnically homogenous (Lynch and Simon, 2003). As Prime Minister Mackenzie King noted, in May 1946 "Immigration is for economic development ... [and]... Immigration must not distort the present character of the Canadian population" (Lynch and Simon, 2003). This policy of selecting immigrants based on their ability to contribute to the Canadian population still holds in the present immigration discourses in which Canadians prefer immigrants that will have a positive impact on the economy (Mackey, 1999; Li, 2003). With the rise of the Canadian welfare state in the 1940s and 1950s sustained levels of population and economic growth were required, which were only possible at the time through increasing birth rates (Thobani, 2000). Canada's racially discriminatory immigration policy (selecting immigrants based on country of origin) was dissolved on July 19 th , 1962 by Ellen Fairclough (Freeman, 1995). The policy shift altered the 13 requirements for un-sponsored immigrants, rating applicants based on their education and skill-levels instead of their place of birth. Implemented on February 1st, this change in policy shifted regulation away from the use of race as selection criteria making Canada the first major immigrant receiving society to change its racially discriminatory immigration policy. By contrast, the White Australia policy in Australia would not change until 1973 and the American Immigration Act was implemented until 1976 while J.F. Kennedy was President (Freeman, 1995). Building on the 1962 policy, in 1978 Canadian immigrant selection criteria was expanded to include demographic, economic, family and humanitarian factors (Griego, 1994). Setting maximums on the total amount of immigrants that could enter the country each year, entrance levels were tied to the fertility rates of the domestic Canadian population (Grabb and Curtis, 2005). This act also established the three separate classes of immigrants: (1) family class, which includes immediate family and dependant children (2) refugees and asylum seekers and (3) economic immigrants. During this some time period in the United States was drafting soldiers for the war in Vietnam. Canada was a popular destination for individuals who wanted to avoid the draft, although many went underground in Canada if they couldn't meet the immigration requirements. Between 1966-1972 roughly 16,000 Americans aged 16-25 immigrated to Canada. Migration levels were so high that in 1972 a special one year program was implemented to give immigrant status to 80,000 individuals, many of them deserters. Current immigration policy in Canada developed from the 1967 Immigration Act (revised in 1974, 1978, 1985, and 1996) which has maintained the point system as well as the separate entrance categories (Lynch and Simon, 2003; Knowles, 1997). In addition, to 14 leading the way toward immigration policy Canada's multicultural policy sets it apart internationally. Canada is unique among these three countries because of its official Multicultural Policy. This policy was first developed in 1971 by Liberal Prime Minister Pierre Elliot Trudeau, and enacted by Conservative Prime Minister Brian Mulroney in 1988 (Knowles, 1997). The policy has challenged Canadians to learn about the possible Canadian identities. Taylor argues that this policy is an important step forward for Canadian society since it encourages acknowledgement of the diverse lived experiences in Canada society and limits the cultural privilege of such groups (1992). Article 27 of the Canadian Charter of Rights and Freedoms also includes an outline of rights "in a manner consistent with the preservation and enhancement of the multicultural heritage of Canadians" encouraging the recognition of the diverse group of individuals who consider themselves Canadians. Today Canada has one of the most diverse groups of foreign-born residents in the world. With around 18% of the population foreign-born. Canada is third in the world for accepting the most immigrants after New Zealand and Australia (Statistics Canada 2001). Over 50% of immigrants admitted to Canada are economic immigrants selected based on points attached to their occupational skills determined by the Designated and Open Occupations List maintained by the Ministry of Citizenship and Immigration outlining the needs of the Canadian domestic labour market (Meissner et al., 1993). Within the economic class there is also a subset of migrants who are rated not just on their ability to be employed but on their commitment to invest domestically in Canadian business. 15 Thus, Canada's position as a receiving country, open to immigration has been fostered in some ways through the development of immigration in Canada. The first of the major immigrant receiving countries to change its policy on immigration encouraging more egalitarian immigration policy. 1.6 Forces in the United States Formed out of a rebellion against British Colonial control, the United States has sought to foster individual freedom from its inception (Adams, 2003). In this section, some of the key factors that have shaped attitudes toward immigration in the United States are explored. America's long history as a social, political and economic immigrant receiving country has significantly impacted the construction of its own national identity. This construction of immigrants as part of the American identity has had a significant impact on American public views toward their own relationship with immigration and immigrants' participation in the process of state construction. Founded as a colony of the United Kingdom, the United States asserted its independence in 1776 citing the British government's restrictive immigration policies to the colonies as one of the reasons for seeking independence (Lynch and Simon, 2003). The writers of the Declaration of Independence charged that King George III tried to keep the population in the colonies small and refused to recognize naturalization acts passed by the colonies. In addition the constitution also stipulated which positions in the federal government immigrants could hold, opening all positions up to immigrants with the exception of the office of president (Meissner et. al. 1993). Congress established the first policy to allow foreign-born individuals to enter the country in 1790 and the first limitations on the quantity of immigrants entering the country in 1875 which included the 16 restrictions on the entrance of criminals and prostitutes. The boundaries between Mexico and the United States have changed over the years and with them the definition of who is not included in the United States has shifted. Up to the Mexican War of 1846 large parts of the West Coast and the Southwest of the United States belonged to Mexico. When the modern Mexican-American border was created it divided relatives and immediate family members from each other. This began a long tradition of cross-border migration for Mexican nationals between kinship groups in both countries. In 1891 Congress established the Immigration Service which gave control of the process of accepting new immigrants to the federal government. The early American reliance on the southern plantation system made it one of the largest and most lucrative agricultural systems in the world. Plantation society hinged on the presence of a large cheap labouring force of slaves taken from Africa. Americans faced challenges justifying the use of slavery both locally and internationally as citizens of a country founded on the principals of universal liberty. Slavery was justified by characterizing slaves as 'heathens' thus making them inferior and separating them from real people who could qualify for Christian ideals of mercy and charity. The practice of slavery legally enforced differences in social status and access to the rights of citizenship based on race. Some argue that the very concept of race (distinguishing individuals based on skin color) emerged with development of slavery as a justification for the exploitation of black people by white people. Previous to the plantation system white indentured servants had been employed by landowners who would pay passage for workers in exchange for services. These workers were free to go after they had paid off their debt, while the use of 17 slaves provided a sustained cheap labour force to support the agricultural needs of the south. In the first 150 years after independence, the United States accepted around forty million immigrants (Lynch and Simon, 2003). Most came from Northern and Western Europe seeking political, economic and social freedom. The founding principles of the United States encouraged religious freedom welcoming the religiously persecuted which included a range of religious groups. Religiously, the Anglican Church had relatively little control over the colonies in comparison to Britain. In addition many individuals to the colonies came from diverse origins including Methodists, Baptists, Presbyterians, Quakers, reform Protestants (Grabb and Curtis, 2005). These religious groups were often regionally located and sometimes relatively intolerant toward different religious groups. This occurred more frequently in the United Sates than in Canada where the emphasis on personal freedom of religion was stronger. Other immigrants moved to the United States to escape the professional restrictions of trade unions and guilds, while others came seeking exemption from the rural traditions of land-claim and agricultural practices (Meissner et al., 1993). At the same time that immigration was encouraged from Europe there was an extensive set of restrictions on immigration from China. The Chinese Exclusion Act of 1882 restricted the entrance of all Chinese immigrants and excluded all foreign-born Chinese from receiving citizenship (Lee, 2003; Brown, 2003). This policy change did not halt Chinese immigration but instead pushed it underground (Lee, 2003). Due to the high demand in China for the opportunities available through immigration in the United States a system of illegal immigration was established which created false identities and 'paper' 18 families in order to allow fabricated family sponsorship networks to allow immigration. No other nationality in the history of American immigration policy received a specific act to restrict their entrance into the country. In addition to the Chinese Exclusion Act, the Foran Act restricted recruitment and passage-payment for foreign-born unskilled labourers effectively limiting all immigration from Asia, where the majority of the population could not afford to pay the cost of passage before arrival (Lynch and Simon, 2003). Between 1896 and 1917, the American federal government debated and implemented a series of literacy restrictions as a means to restrict immigration. Beginning with the 1917 Immigration Act immigrants over sixteen years of age were required to provide proof that they could read and write in English or another language. The Act also completely restricted the immigration of "Asiatics" regardless of their language abilities (including Indians, Indochinese, Afghanis, Arabians, and East Indians). All Asians were restricted from immigrating to the United States with the exception of Filipinos who received American national status when the Philippines became an American territory after the Spanish-American War (Brown, 2003). At the same time that the United States was restricting immigration from Asia, shifts were occurring in the traditional European source countries. In the 1880s migration patterns changed making Eastern and Southern European countries the major source countries for immigration up till WWI. These large waves of immigrants in the early 1900s fostered public anxiety which legislators responded to by passing a series of policies which decreased the quantity of immigrants that were allowed to enter the United States (Brown, 2003). 19 Immigrants were excluded from entering the United States not only based on their country of origin, education, and language skills but also on their potentially morally and socially undesirable qualities. In the late 1800s criminals, prostitutes, paupers, lunatics, idiots (1882), polygamists (1891), and anarchists (1903) were barred from entering the United States (Lee, 2003). Similar policies were also enacted in the Canadian context, restricting individuals who might be considered undesirable from entering Canada (Knowles, 1997). This pattern of restrictive policies for perceived unproductive or undesirable members of society was part of an immigration system which selected immigrants based on perceived contribution to the American state, and culture. Policies which explicitly exclude certain groups indicate how these groups were perceived by the society was well as highlighting what was determined to be desirable qualities for future generations of Americans to have (Lynch and Simon 2003; Knowles, 1997). The quota system, which placed new limits on immigration based on country of origin, were implemented in the 1921 Johnson Act and revised in 1924. U.S. national census was used to determine the immigration rate that was limited to 3% of the total number of foreign born individuals in the United States from a particular country. Regions of the world were given total quotas limiting to the amount of immigrants from South America and Asia. While the quota system placed caps on the number of immigrants from specific countries, some policies also completely excluded the entrance of specific groups, the act placed restrictions on the number of immigrants based on geographic regions of the world (Asia, South America, and Africa). A second aspect of the 1921 Quota Law was the focus on family reunification which allowed immediate relatives of U.S. citizens and some extended family members to immigrate regardless of 20 the quota numbers from their particular country or allowing them to be favoured for immigration within the per country number restrictions. In times of labour shortages the Brocetro Program (instigated by Roosevelt, 1942-1965) created a system of short-term work visas issued to Mexican labourers allowing than to migrate to the U.S. for short periods of time in order to meet labour market needs (Brown, 2003). Many of these brocetros outstayed their visas and remained in the United States illegally. A system of illegal immigration from Mexico was fostered in which kinship networks would provide immigrants with refuge in the U.S. The United States-Mexico became a gateway for illegal migration from other parts Central and South America. The War Brides Act of 1945 and 1947 allowed members of the military to bring their foreign-born spouses and children to the United States as American citizens after WWII (Lynch and Simon, 2003). This major influx of foreign-born women and 'mix-lineage' children marked the first major wave of migration from (Meissner et al., 1993). After World War II and the Cold War, America emerged as an economic and political world superpower (Brown, 2003). All immigrant nationalities, but especially Asians, benefited from the 1952 McCarran-Walter Act which allowed relatives and spouses of American residents to enter the country irrespective of the national quota totals by country (Brown, 2003). This fostered what continues to be an important avenue of migration to the United States which is through the ties of family and marriage. The Immigration and Nationality Act Amendments of 1965 abolished the national-origin quota system. Instead the system focused on family reunification (spouses, children under 21 and parents of citizens over 21 were not subject to the same restrictions on immigration) and for immigrants with needed job skills. Numerical restrictions were not eliminated and 21 countries in the Eastern Hemisphere (Europe, Asia, Africa, and Australia) had per- country total immigration limits as well as restrictions based on immigration categories. Countries by contrast in the Western hemisphere did not have total caps by preference category or country. The Hart-Celler Act of 1965 set an annual limit on the total level of immigration and a general per country quota of 20,000 (Brown, 2003). This policy shift equalized the quota system for immigrants from the Northern and Southern hemispheres. New categories were created separating immigrants into groups based on their type of immigration claim such as family reunification or professionally skilled allowing different immigration requirements to apply to each (Lee, 2003). When restrictions on immigration were reduced after 1965, the proportion of legal migrants from Mexico increased. With the reduced restrictions in post-1965 immigration policy, the proportion of legal immigrants arriving from Mexico also increased significantly. This influx of Hispanic migrants placed economic strains on the welfare and education systems in California and Texas into the 1990s. Not until 1976 was the immigration policy changed so that the categorical preference system was applied to Western Hemisphere applicants as it had been to Eastern Hemisphere applicants (Lynch and Simon, 2003). Finally, in 1978 a single total was established for the absolute number of immigrants that could enter the country from both the Eastern and Western Hemispheres. Since World War lithe United States has been involved in accepting specific groups of refugees. One example is the 1975 Indochinese Refugee Resettlement Program began bringing 200,000 Indo-Chinese refugees into the United States after the Vietnam War. This large influx of Vietnamese was met with hostility by some Americans based on fears that the welfare system would be strained by this large influx of immigrants. The 22 1980 Refugee Act gave Congress and the president the ability to decide refugee policy on a comprehensive and yearly basis. The policy shift placed American refugee policy accepted the expanded definition of refugee from the 1967 Protocol to the 1951 United Nationals Refugee Convention. In 1986 the Immigration Reform and Control Act attempted to regulate the arrival of unauthorized immigrants, placing sanctions on employers who consciously hired illegal workers, and developed amnesty programs for seasonal agricultural worker. Individuals would have worked in some agricultural jobs for at least 90 days could apply for the Seasonal Agricultural Workers amnesty program. The Illegal Immigration Reform and Immigration Responsibility Act of 1996 increased the number of border patrol agents as well as reducing government benefits to immigrants and allowing employment and social service agencies to determine citizenship status via telephone or internet. The Homeland Security Act of 2002 created the Department of Homeland Security and in so doing repositioned the Immigration and Naturalization Services (INS) since this new agency took over control of immigration services, border enforcement and inspections. This restructuring divided the INS responsibilities of citizenship and naturalization into the Bureau of Citizenship and Immigration Services, and border enforcement into the Bureau of Customs and Border Protection and the Bureau of Immigration and Customs Enforcement. Economic recessions in Mexico in the 1980s and 1990s propelled many Mexicans to immigrate to the United States whether legally or illegally in order to seek financial opportunities. A pattern of seasonal migration developed whereby Mexican agricultural labourers would work for American farmers illegally for part of the year, then return to Mexico with their earnings. American immigration debates in the 1990s centered on 23 rising discomfort of residents with the demographic construction of post-1965 immigration. War terminology such as "invasion", "conquest", and "save our state" was increasingly used to describe issues of immigration. Several high-profile cases of boats carrying illegal Chinese migrants attempting to arrive in American parts created a great deal of debate on how to limit illegal immigration in the 1990s (Brown, 2003). Regulations such as California's Proposition 147 were designed to exclude illegal immigrants from accessing health and welfare services. Since the majority of illegal immigrants were of central and South American origin, United States authorities began raiding neighbourhoods with high proportions of foreign-born individuals seeking out illegal immigrants who may have found support networks within the immigrant communities. These raids constructed Hispanic-American communities as outsiders who harboured illegal residents, and fostered hostilities within these communities to the way immigrants are threatened. The 1990 Immigration Act added a category of admission focusing on diversity which increased the total number of immigrants who could enter the country and reestablished the cap as a 'flexible' limitation. This flexibility allows the United States to accept more immigrants in specific categories in years when not all the visas from the family-sponsored and employment-based categories were used the year before. Today, American immigration offers two avenues for non-citizens to gain legal entry into the country, either through the application for permanent residency (lawful permanent residents, LPN or green card holders) or through temporary access. As in keeping with previous policies the U.S. has an immigration objective of family reunification for spouses and immediate family members. U.S. citizens can also sponsor 24 other relatives under the family sponsorship program which is subject to federal limitations. Employment-based preferences are still present in U.S. immigration policy selecting immigrants based on their professional skills. For the majority of these applicants, an employer must submit a labour certification request to the Department of Labour, establishing that there are not enough individuals available locally to fill the necessary employment needs and that employing this foreign-born individual would not impact wages or working conditions within the United States. The employer must then file a petition with the USCIS on behalf of the immigrant. Current U.S. immigration policy also continues to accept refugees and asylum-seekers who because of race, religion, nationality, membership in a particular social group, or political opinion do not want to return to their home country. Refugee numbers are set by congress and the president on an annual basis, the policy in practice has been to admit about half the number of refugees identified by the U.N. High Commissioner for Refugees, leaving part of the refugee proportion unapplied allowing for unexpected resettlement needs. 1.7 Forces in the United Kingdom The British public government has increased restrictionist policies, while the British public has growing hostilities toward immigration (Freeman, 1995; Crawley, 2005). This section outlines some of the pertinent historical developments which have helped to shape current attitudes toward immigration. Beginning with the development of the British Empire, this historical contextualization presents information on the development of immigration. As David Brent a British actor from the television series The Office said to the American Golden Globe audience after winning his acting award "As you can guess, I don't come from around these parts. I come from a little place called 25 England. We used to run the world before you did," summarizing the residual ties of colonization which still informs British identity today and the perspective of former British colonies. Founded by Celtic and Pict tribes, England was later settled by the Romans, Jutes, Angles, Saxons, and Vikings who brought the early-French language with them shaping the development of British government and law. It was the age of exploration which established the British ties with the rest of the world through the establishment of international trade routes. This system supplied American plantations with the needed slave labour necessary for agricultural production and shuttled the raw materials and profits of the colonies back to England. Black slaves were sometimes brought back to England to accompany the children of plantation workers at boarding school or as domestic workers. The status of these early immigrants was unclear although their status identification was tied to their owners and their status as Christians. This proportion of African-origin residents grew with the increased prosperity of the American colonies leading to the emergence of an absolutist movement. In 1772, although the courts did not want to rule on immigration, they recognized that slaves should not be forcibly transported to the United Kingdom, although results of this ruling were rarely enforced. The end of slavery in Britain came in two parts first in 1833 parliament banned the trading or slaves, but not slavery, and in 1833 Parliament banned slavery in the whole British Empire. This stopped the arrival of African-origin individuals while immigrants from the rest of Europe continued to arrive. As Britain continued to grow as an international trading power African and Asian sailors established small communities in British port cities such as London in part because they were often abandoned by 26 employers upon arrival in England. The years 1830-1850 saw massive waves of immigration from Ireland as migrants attempted to escape poverty. Beyond the United Kingdom's relationship with its colonies the United Kingdom historical patterns of migration has helped to shape current views on immigration. Beginning in the 19 th and early 20 th century, waves of British citizens moved to North America seeking economic opportunity and religious freedom (Meissner et al., 1993). Between 1850 and 1920 roughly forty percent of the population emigrated, the highest of any European country including Italy and Scandinavia (both with 30%) (Meissner et al., 1993). It was this emigration of the surplus population that created the redistribution of the British population necessary for the industrial revolution transforming Britain from an agricultural to an industrial society. With the onset of the World Wars, the need for soldiers drew recruitment from across the British Empire including India. Many of these soldiers chose to stay in England between the wars establishing the first major Indian communities in England. The presence of these non-white British citizens was not welcomed as evident by the tensions that erupted in the 1919 race riots. British labour shortages after World War II led to the first major recruitment and acceptance of immigrants, with the arrival of large groups of migrants. Many former soldiers from the West Indies also arrived in England and while they were met with public hostility these arrivals fulfilled the needs of the British economy. During this same time period limited finances prompted the United Kingdom to provide independence to many of its colonies. Immigration during this time period was characterized by unskilled migrants from Poland, Ukraine, the Caribbean and India filling gaps in the British labour 27 force. In 1947 India gained independence prompting waves of immigration to the United Kingdom. The quantity of Indian-origin immigrants continued to increase into the 1970s when Idi Amin expelled 90,000 Gujarti Indians from Uganda. The June 22 nd 1948 arrival of the Empire Windrush carrying West Indian men planning to rejoin the RAF or look for work marked the beginning of the era of mass migration to the United Kingdom. The 1948 British Nationality Act identified two classes of individuals trying to immigrate, foreign-born commonwealth citizens and non-commonwealth citizens. The 1949 Ireland Act recognized the Republic of Ireland but classified its citizens as not citizens of a foreign country, restricting Ireland from being a member of the Commonwealth. Many British passports from the Empire and Commonwealth entered the country without a problem although the government began to feel more and more pressure to regulate non-white migration. The 1956 Hungarian revolution brought waves of Hungarian refugees. Previous to the 1962, Commonwealth Immigration Act, Commonwealth citizens could arrive and settle in the United Kingdom without limitation. This act separated citizens of the United Kingdom and colonies (CUKCs) by whether their visas were directly issued by the United Kingdom government from individuals whose passports issued by the governor of their colony. The 1962 Commonwealth Immigration Act restricted the immigration of British citizens living abroad. Entry become conditional on access to job vouchers which required pre-arranged jobs, special skills and were determined based on labour market needs. Immigration regulations in the 1970s focused on tightening the access of Black and Asian migrants to the United Kingdom. In 1971, citizenship was limited to "partials". Included in this definition were citizens of the United Kingdom and its colonies. This 28 policy favoured Commonwealth citizens over former colonies providing an avenue for restricting the access to citizenship of non-white individuals. The policy changed again in 1972 further restricting entry to work permit holders or individuals with parents or grandparents born in the United Kingdom. This effectively curtailed migration from the Commonwealth countries. In 1972 the Uganda Asian Crisis brought more Asian immigrants (Dustmann and Preston, 2002). In 1981, three classes of citizens were created. A first category consisted of people born in the United Kingdom or whose parents or grandparents have been born in the United Kingdom. A second class dependant territory citizens or individuals whose partners were born in one of these territories or people whose parents were born in one of these territories. Finally, the last category of citizens were children of British citizens born outside of the United Kingdom; these foreign dwelling parents were the only ones who could not pass. British immigration policy in the 1980s focused on two sometimes contradictory characteristics: restricting entry of new residents and protecting the rights of ethnic and racial minorities. Work permit preference was given to highly skilled individuals or those with professional trades. The majority of immigrants in the 1980s were from the United States, Australia, New Zealand, South Africa or South Asia (this was mostly medically trained individuals). In April llth, 1981 Afro-Caribbean youth rioted in Brixton, claiming that the police targeted them for street crime. Enacted in 1983 the British Nationality Act 1981, drew distinction between British Citizens, whose citizenship was not 'by decent' allowing them to pass on their citizenship, and British oversees territory citizens whose citizenship was by decent and who could not pass on their citizenship to their children 29 (although some were able to register their children). Joining the European Union, the United Kingdom agreed to the 'Four Freedoms' one of which is free movement of people. May 1St, 2004 marked the entry of Central and Eastern Europe into the EU increasing the push to immigrate to the United Kingdom. The Worker Registration scheme allows member states to place temporary limitations on immigration although these restrictions will end by 2011. When Romania and Bulgaria enter the EU in 2007 they will not face the same restrictions but instead the Home Office indicated that it will select for students, the self-employed, highly skilled workers, as well as agricultural workers when it decide who is allowed in the country. Today, there are several ways for non-citizens to work in the United Kingdom. The United Kingdom Ancestry Entry Clearance provides work permits to individuals whose grandparents were born in the United Kingdom, Channel Islands or Isle of Man at any time or the Republic of Ireland on or before March 31st, 1922. Once in the country, these non-citizens can apply for indefinite leave after five years of employment. The second method of entry for non-citizens is through the Highly Skilled Migrant Programme (HSMP) which gives point values to skill levels of an offer of employment or an employer sponsorship visa. Points are given for education, work experience, professional recognition, partners achievements, age (younger than 28) and doctors working in the United Kingdom. Even though immigration has been severely restricted in British society, issues of immigration are still contentious today, especially in relation to asylum seekers. The number of asylum seekers rose to the 1990s and the major increase in the number of 30 applicants places train on the Home Office, which was ill equipped to handle this volume of applicants (Lynch and Simon, 2003). The state supports claimants while they are waiting for their applications to be processed. This raises debate in the public about the cost of immigrants and the economic strain of immigration. In order to reduce the strain of major influxes of asylum seekers in immigrant receiving areas, the government created a program that dispersed immigrants throughout the country, placing some in remote villages with little or no foreign-born residing in the community. These implanted residents with undetermined citizenship sometimes faced animosity from local residents in these small towns that were often struggling economically. Since the foreign arrivals were supported by the British government, they were often viewed as draining resources that might otherwise be available for local residents (Lynch and Simon, 2003). Overall, rates of asylum seeker applications have been increasing in recent years despite a consistent debate on the subject in political circles (UK Office of National Statistics, 2001). Non-citizens can only enter the country through the following classes: family reunification, work permits, and student visas. Family members need to get an entry certificate and show proof of relationship in order to sponsor a family member. This bureaucratic mechanism created a systematic selection process which favoured applicants from the global north because it was harder for individuals from remote regions to provide proof of relationship through birth and marriage certificates (Fleras and Elliot, 2002). In many rural areas and in specific countries detailed records were not well kept, reproduction or widely available, making the process of migration much more challenging. 31 Work permits were only issued to applicants with recognized qualifications or a high degree of skill/experience. Immigrants were selected based on their age (23-54 years old preferred) and occupation (depending on what type of work was needed most in the labour market). Work permits were issued for 12 months and extensions could be granted by the Home Office for up to 5 years (Lynch and Simon, 2003). Immigrants must remain at the same job for the whole five years of the work permit and may not switch positions (Lynch and Simon, 2003). Because of immigration restrictions, most work permit holders don't apply for settlement (Meissner et al. 1993). This creates a pattern of short-term immigrants who arrive and boosting the economy but who are restricted from accessing British social services. Special visas are issued to holiday workers, seasonal and agricultural workers, investors, entrepreneurs and students. 1.8 Conclusion A great deal of research has been conducted on public opinion toward immigration policy and the work in this thesis can provide a useful contribution to further knowledge on this topic. The aim of this introduction was to provide an overview of immigration policy development in each of the three countries included in this thesis. Through contextualizing attitudes towards immigration, the way Canadian, American, and British respondents understand immigration can be better understood. Specifically, this thesis focuses on attitudes towards immigration policy in 2005 public opinion poll data. The factors that shape attitudes are first analyzed in three countries separately as well as comparing them cross-nationally to determine the ways in which open or intolerant attitudes are formed. Following this attitudes in the United States are analyzed in more detail with the inclusion of additional measures to further 32 understand what factors impact American attitudes toward immigration. To answer these questions, it is valuable to understand the contribution historical trajectory makes the development of perceptions toward immigration. Thus this instruction provides the background for a better understanding of the ways in which favorable and unfavorable attitudes toward immigration developed. The historical patterns of restrictionist immigration policies in the United Kingdom as well as an ethnically centered national there, suggests that residents in the United Kingdom will be more likely to hold negative attitudes towards immigration then their compatriots in Canada or the United States (Freeman, 1995). While hostilities toward immigration grow in the United States and support for immigration is not increasing a great deal in Canada, the United Kingdom remains relatively consistent on its limitations toward immigration. Thus, understanding attitudes toward immigration can be furthered through knowledge of the background in which perceptions develop. As such, analyzing the way attitudes are shaped cross-nationally and cross-sectionally is an important exercise in the process of examining the development of tolerance. 33 1.9 Bibliography Adams, Michael. 2003. Fire and Ice: The United States, Canada, and the Myth of Converging Values. Toronto: Penguin Books. Brown, D.Clayton. 2003. Globalization and America since 1945. Wilmington, Delaware: Scharlorly Resources. Castles, Stephen. 2000. Ethnicity and Globalization. London: SAGE Publications. Crawley, Heaven. 2005. "Evidence on Attitudes to Asylum and Immigration: What We Know, Don't Know and Need to Know. Centre for Migration, Policy and Society." Work Paper No. 23. University of Oxford. WP-05-23. Fleras, A. and J.L. Elliot. 2003. Unequal Relations: An Introduction to Race and Ethnic Dynamics in Canada. 4th Ed. Toronto: Prentice Hall. Freeman, Gary. 1995. Modes of immigration politics in liberal democratic states. International Migration Review 19.4: 881-902. Macmillian. Knowles, Valerie. 1997. Strangers at our gates: Canadian immigration and immigration policy, 1540-199 7. Toronto: Dundurn Press. Lee, Erika. 2003. At America's Gates: Chinese Immigration During The Exclusion Era, 1882-1943. Chapel Hill and London: the University of North Carolina Press. Li, Peter. 1998. Chinese in Canada. 2nd Edition. Oxford University Press. Lipset, S. M. 1963. The first new nation. New York: Basic Books. Lynch, James P. and Rita Simon. 2003. Immigration the World Over: Statutes, Policies, and Practices. Lanham: Rowan & Littlefield. Mackey, Eva. 1999. The House of Difference: Cultural Politics and National Identity in Canada. London and New York: Routledge. Meissner, Doris M., Hormats, Robert D., Garrigues Walker, Antonio, Ogata, Shijuro. 1993. International Migration Challenges in a New Era. New York: Trilateral Commission. Taylor, C. 1992. Multiculturalism and The Politics of Recognition'. Princeton: Princeton University Press. Thobani, Sunera. 2000. "Nationalizing Canadian: Bordering Immigrant Women in the Late Twentieth Century." Canadian Journal of Women and Law 12: 279-315. Porter, J. 1979. The Measure of Canadian Society: Education, Equality and Opportunity. _Canada: Gage publishing. 34 Chapter 2 Shouting across the Divide: Cross-National Variation in attitudes toward immigration policy in Canada, the United States and the United Kingdom, 2005 2.1 Introduction Internationally, anti-immigration sentiment is on the rise (Dustmann and Preston, 2001; Semyonov, Raijman, and Gorodzeisky, 2006; O'Rourke and Sinnott, 2004) and in many nations it is one of the most politically contentious issues (Crawley, 2005). Despite these anti-immigration sentiments, migratory flows into advanced industrial countries continue, and in some countries have increased recently (e.g. Canada). This conflict between public sentiment and immigration policy challenges receiving countries and the process of integrating immigration. Stepping beyond the media catch phrases, the messages of elected officials and the agendas of special interest groups, this paper addresses the growing need to understand public views toward immigration policy. While levels of public support for immigration policy are often reported engaging at a more in- depth level with the dynamics of these attitudes provides important insight for the policy development process. In order to further understand the rise of anti-immigration sentiment in parts of Europe (Semyonov, Raij man, Gorodzeisky, 2006) and North America (Lapanski et al., 1997), I compare three socially and historically similar countries: Canada, the United States and the United Kingdom. The strong ties between these three countries make them a useful case study through which to expand our understanding of the impact that individual-level characteristics and national context have on attitudes. 35 Building on the extensive body of literature on determinants of attitudes toward immigration, this cross-national study focuses on the analysis of negative and positive attitudes in three similar nations (Freeman, 1995; Grabb and Curtis, 2005). In this paper, Canada, one of the few receiving countries with consistent levels of support for immigration contrast against the United States and the United Kingdoms increasing hostility towards immigration (Simon and Lynch, 1999; Crawley, 2005; Lapinski et al., 1997). Previous research found that individual characteristics impacted attitudes toward immigration (Palmer, 1996; Fortrin and Loewen, 2004; Bauer, Loftstrom, Zimmerman, 2001; Chandler and Tsai, 2001; Citrin et al., 1997; Espenshade and Hempstead, 1996; Scheve and Slaugher, 2001; Dustmann and Preston, 2001; Dustmann and Preston, 2004b; Tucci, 2005; Hernes and Knudsen, 1992; Gang, Rivera-Batiz and Yun, 2001; Lynch and Simon, 2003) and that attitudes toward immigration vary cross-nationally (Quillian, 1995; Brenner and Fertig, 2006; Bauer, Loftsrom, Zimmerman, 2001; Mayda, 2004; O'Rourke and Sinnott, 2004; Card, Dustmann, and Preston, 2005; Dustmann and Preston, 2004b; Dustmann and Preston, 2001; European Commission, 2003). This paper makes three key contributions to the literature on attitudes toward immigration. First, I replicate previous cross-national public opinion research using 2005 data, focus on three countries with similar national characteristics, and control for the impact of country. First, through the replication of pervious research methods, conclusions about attitudes towards immigration in Canada, the United States and the United Kingdom can be retested analyzing attitudes towards immigration remain consistent in their levels of hostility. Building on this previous work, attitudes are analyzed in each country individually, as well as cross-nationally, measuring the impact 36 of demographic and economic variables on attitudes toward immigration, retesting. By retesting previous results I contribute to understanding of the consistency of attitudes towards immigration. The second contribution to the literature made by this paper is the focus on three countries with strong historical and social ties (Grabb and Curtis, 2005; Paul 1997; Freeman, 1995). Canada and the United States are both liberal democratic receiving states constructed with support from British migration (Feeman, 1995; Lynch and Simon, 2003). This historical power relation, between colonizer (the United Kingdom) and colonies (Canada and the United States) speaks to their common ties of "law, language, and literature" (Churchill, 1998: 973). These countries represent an interesting opportunity for comparison between three countries with similar cultural, linguistic, political and legal traditions but differing recent social trajectories (Adams, 2003; Grabb and Curtis, 2005). Currently, these countries share populations where the majority of the population identifies as Anglo-origin, English-speaking, Judeo-Christian and the population growth levels are near or below replacement levels (Adams, 2003; Freeman, 1995; US Census Bureau, 2000; UK Office of National Statistics, 2001; Statistics Canada, 2001). Instead of comparing attitudes between countries with extremely different immigration policies and a range of historical migration patterns (Lynch and Simon, 2003), this paper analyzes three countries with strong historical ties and very clear cross- national relationships. This provides insight into how attitude differences develop in similar natural settings (Grabb and Curtis, 2005). The purpose of this focus is to further explore some of the subtler ways in which the relationship between individual-level 37 characteristics and national context shape attitude construction (Grabb and Curtis, 2005; O'Rourke and Sinnott 2001; Lynch and Simon 2003). The third contribution to the literature made here is the inclusion of country as a variable in the regression model. This test for the impact of country on attitudes, determining if there is a statistically significant distinction between attitudes Freeman argues that the United Kingdom should have distinctly different attitudes toward immigration than Canada and the United States based on immigration policy differences (1995), although Lapinski has found that the level of hostility in the United States is growing (1997), paralleling patterns in the United Kingdom (Crawley, 2005). The United Kingdom shares immigration characteristics with larger groups of European nations who only started receiving mass in-migration after World War II when they accepted temporary workers and short term migration to ease labour market demands (Paul, 1997; Dunlevy et al., 2006). This paper tests the impact that these differences in historical development and national context have on attitudes and in so doing furthers our understanding of the relationship between national setting and views on immigration. Building on the work of other researchers who include country as a variable in the regression models this paper replicates this procedure in the context of these three particular countries (Semyonov, Raijman, Gorodzeisky, 2006). Comparing Canada, the United States and the United Kingdom Canada, the United States, and the United Kingdom share many similarities as well as experiencing some significant differences as nations. This section will explore three key points of comparison between these three countries: state formation, patterns of migration and construction of immigrant's participation in society. These structural 38 factors foster the development of "everyday life" reinforcing common national ideas, values, and beliefs in the population which shape the core values of individuals in a particular national setting (Braudel, 1979; Grabb and Curtis, 2005). The shared language and social structural bases fostered by shared ties to the British state have encouraged a higher levels of idea exchange between these three countries, further emphasizing the development of common political processes (Lipset, 1963). Braudel argues that differences in the core values upon which each country was founded impact the experience of its public and how attitudes toward immigration develop. When there is a lack of difference in the way states formed, public perspectives are expected to be similar (Braudel 1979; Lipset, 1963; Grabb and Curtis, 2005). Canada, the United States, and the United Kingdom are all democracies, although the way their states were formed is important to understand when comparing views of citizenship and immigration (Dunleavy et al., 2006; Grabb and Curtis, 2005). The United States and Canada were formed as colonies of the United Kingdom, gaining independence in different ways, and points in time (Freeman, 1995). The United States, formed by revolutionaries, asserted its independence from Britain in 1776. Canada, a refuge for loyalists from the United States, was given independence by the British in 1867 during a period at a time when the United Kingdom was providing colonies with their independence in order to cut costs (Grabb and Curtis, 2005; Dunleavy et al., 2006). This difference in the way these two states attained independence impacted their core values of independence and collectivism shaping their national identities. For example, American democracy arose out of a revolution against Great Britain. Canadian democracy developed out of acceptance of ties to colonial Britain and Pre-Revolution 39 France. Canada and the United States are characterized by governments with strong support for individual rights, competitive party systems and regular elections, intended to be open to public opinion, involvement and debate (Freeman, 1995). State formation has shaped each states construction of citizenship, while patterns of migration have impacted how immigration policy is perceived and developed. As a colonizer, Britain helped to form the modern nations of Canada and the United States, through supplying these countries with a large part of their first major waves of immigrants (Paul, 1997; Dunleavy et al., 1997). This established Britain's economic and social ties to the colonies and allowed Canada and the United States to recognize economic and demographic contribution of immigration. Early policy in the Canada and the United States gave preference to British immigrants, implementing a citizenship selection process which favoured immigrants with ethnic and racial similarities to the founding groups. This also fostered an early pattern of encouraging assimilation whereby recent immigrants would be viewed as outsiders, but overtime they would come to be included as full citizens. By contrast, the United Kingdom's historical relationship with immigration differs, based on the nations pre-WWII experience as a sending society which only started receiving large numbers of immigrants in the last half of this century (Freeman, 1995; Paul, 1997). The post-war labour shortages created a system which cycled temporary workers from Eastern Europe and other parts of the world through the United Kingdom, while discouraging permanent settlement. Over time some of these short-term migrants began to take up residence in the United Kingdom and sponsor their relatives. The slow permanent settlement of these short term migrants fostered hostilities in the British public, encouraging fears that the British government did not have full 40 control over the regulation of migrants (Crawley, 2005; Paul, 1997). This concern furthered a range of increasingly restrictionist immigration policies allowing only a select few access to citizenship based on their skills and British heritage. In summary, while Britain has a historical pattern of increasing restrictions on immigration, the United States and Canada have both moved toward a more transparent immigration system that accepts significant quantities of immigrants based on their ability to participate in society (Freeman, 1995; Crawley, 2005; Paul, 1997). In addition, to the historical development of immigration, attitudes toward immigration are often related to immigrants perceived potential integration into receiving societies (Crawley, 2005). For example, if immigrants are not perceived to be integrating into society, they are often viewed as challenging the general social identity (Esses et al. 2001; Crawley, 2005). These definitions of in-group and out-group perceptions of local and national identity, as well as the guidelines for integration, are country specific. Attitudes vary based on the perceived participation of immigrants in their national and local communities which is connected to their resource and welfare service use, crime participation, perceived engagement in deviant behavior such as use of illegal substances, health status, language skills, educational attainment and community involvement (Taylor, 1996; Espenshade and Hempstead, 1996; Day, 1990; Lapinski et. al., 1997; Hjerm, 2000). Thus, through lived reality and political and media assertions, the ways in which immigration is framed differs. For example, Canadian immigration debates are often related to discussions of the contribution of immigrants to the Canadian economy (Palmer, 1996), while American immigration debates focus on the legal status of new residents (Lapenski et al., 1997). Attacks on British citizens by immigrants and the 41 children or immigrants such as the July 7 th , 2005 bombings in London or the foiled July, 2007 bombings further isolate the immigrant communities and intensify hostilities toward these small groups of immigrants. Examples such as these highlight the differing contexts used to frame debates around immigration shaping views on policy development. Building on their historical and social similarities, these three countries share several demographic parallels important to this paper these are language, race and religion. First, the United States and Canada's inheritance of English from the United Kingdom fosters strong communication ties between these three countries (Churchill, 1998). This common language has accentuated the process of exchanging ideas between these three countries strengthening their social and political ties (Lipset, 1963; Grabb and Curtis, 2005). The most common first language spoken in all three countries is English (Grabb and Curtis, 2005; Dunleavy et al., 2006), although the distribution of the other languages spoken differs cross-nationally (Statistics Canada, 2001; US Census Bureau, 2000; UK Office of National Statistics, 2001). Canada has the lowest rate of residents whose mother tongue is English, sharing official language status with French, which is spoken by a significant portion of its population, in addition to a large and diverse group of residents whose first language is neither English nor French (Statistics Canada, 2001). In the United States, English is spoken by the majority, but is not the official language, while the other common languages are Spanish followed by Chinese (US Census Bureau, 2000). The United Kingdom has by far the largest portion of English speakers of the three countries with 95% of the country reporting English as their first language (UK Office of National Statistics, 2001). Even though the minority languages in each country are 42 different and the proportion that speaks English varies, the English language main mode of communication in all three countries. In addition to language, the racial and ethnic structure in each of these countries impacts how minority groups of all types are viewed. As with language, the majority populations in each country share a common ethnicity, White, although differences exist cross-nationally between the other minority groups. In Canada, the majority of the population identifies themselves as Canadian or English, although second to these choices are a wide range of ethnicities including French, Scottish, Irish, German, Italian, Chinese, Ukrainian and First Nations (Statistics Canada, 2001). The United States has the largest proportion between these three countries of visible minorities, where most are African American or Hispanic (US Census Bureau, 2000). In contrast, the United Kingdom has the largest White proportion of these three countries with roughly 90% identifying as White, while individuals who identify as Indian are the largest non-white group in the United Kingdom (UK Office of National Statistics, 2001). A third demographic point of cross-national comparison in this paper is religion in the shared common Judeo-Christian roots. This common religious heritage resulting from the United Kingdom as a found force in Canada and the United States (Grabb and Curtis, 2005). Canada and the United States differ in their relationship to religion Canada was founded with religion as a symbolic part of the state, while the United States was established with the principals or religious freedom clearly outlined the constitution (Adams, 2003). In Canada, the majority of the religious population is Catholic, with Protestantism as a close second (Statistics Canada, 2001). While the largest proportion of the religious population in the United States is also Catholic, followed by Baptist and a 43 wide range religious sects many with affiliation to Protestantism (US Census Bureau, 2000). Finally, the majority of the British population also identifies themselves as Christian (UK Office for National Statistics, 2001). Overall, these demographic points of comparison provide further insight into how similar these three states are, as well as were differences might emerge. While the majority in each country is English-speaking, White and Christian, the construction of the minority groups in each of these three countries is distinct and may impact how minority groups in general are viewed and reacted to. No doubt, the construction of the relationship between minority and majority groups in each society will impact attitudes towards immigration (Taylor and Moghaddom, 1994). 2.2 Determinants of Attitudes a) Age and Gender The first two variables included in the model are age and gender, which have previously been found to have a significant relationship with attitudes toward immigration (Espenshade and Calhoun, 1993; Tucci, 2005; Scheve and Slaughter, 2001; Espenshade and Hempstead 1996; Citrin et al., 1997). Age has been found to sometimes have a significant impact on attitudes with some studies finding that older respondents are less supportive of immigration (Espenshade and Calhoun, 1993; Tucci, 2005), although other researchers did not find age to have a significant impact on attitudes (Scheve and Slaughter, 2001). Gender, in some studies, had a weak impact on attitudes (Espenshade and Hempstead 1996; Scheve and Slaughter, 2001; Citrin et al., 1997; Hiebert, 2003) with some researchers finding women to be less supportive of increasing immigration than men (Palmer 1996). These negative attitudes among women has been 44 linked to greater labour market insecurity of women. Other researchers suggest that men, as the main providers in many households, would be more likely to be hostile toward immigration due to the pressure placed on them by dependants to maintain a secure place in the labour market (Crawley, 2005). b) Race As noted previously, Canada, the United States and the United Kingdom each have predominately white populations (Statistics Canada, 2001; US Census Bureau, 2000; UK Office of National Statistics, 2001). Differences in the proportions of Whites will be important to note since racial and ethnic demographics can impact community formation and relations between ethnic groups (Grabb and Curtis, 2005). For the purposes of this paper distinctions will be drawn between the white majority and the non- white minority (Taylor, 1996). The types of minority group vary by country however overall, support appears to vary between white and non-white respondents (Taylor, 1996). Charter or founding groups, in the case of these three countries with white heritage, are expected to be more likely to harbor hostilities toward visible-minority immigrants whose presence might threaten their demographic majority and cultural dominance in society while fostering greater support for White migrants who may not be framed as immigrants (Esses et al., 2003; Schmid, 2003). Non-White residents in each of these countries are expected to be more supportive of immigration due to their potential ability to relate to the feelings of isolation associated with the marginalized experience of being a minority whether racially or through immigrant status (Flexter, 2003). In Canada, there is an association between region and race as attitudes differ between French and English Canadians (Taylor, 1996). In the United States, African Americans and in some 45 cases Asians (Espenshade and Hempstead, 1996) were more supportive of immigration than white respondents (Schmid, 2002; Diamond, 1998; Espenshade and Hempstead, 1996; Citrin et al. 1997; Dion and Kawakami, 1996). Some researchers found that Hispanic respondents had similar levels of support for immigration to white respondents (Schmid, 2003; de la Garza et al., 1992), while others reported higher levels of support among Hispanics due to the ethnic solidarity they might feel associate with recent immigrants from Central and South America (Citrin et al., 1997). In the United Kingdom race was found to have a stronger impact on attitudes than economic measures (Dustmann and Preston, 2002). While the proportion and demographic construction of minority groups is distinct in each of these countries, the non-white minority in each state is expected to share common feelings of isolation and solidarity with the marginalized experience of immigrants (Fletzer, 2000). c) Religion Previous research reports a significant relationship between religion and views on immigration in which Protestant respondents are less supportive of immigration than other religiously affiliated individuals (Daniels and von der Runr, 2005). Research using the European Social Survey for a cross-national study of attitude determinants found that Christians were less supportive of immigration than other respondents (Card, Dustmann, Preston, 2005). Based on this research, religious individuals, especially those that identify as Protestant or Catholic will be more supportive of immigration than non-religious respondents. d) Employment Status/Income 46 Based on the economic labour market theory, individuals will act in their own self-interest, selecting their policy level choice based on what is in their best interests, thus an individual with a more insecure place in the labour market will be less supportive of immigration due to the potential threat to their personal economic situation (Palmer, 1996; Citrin and Green, 1990; Espenshade and Calhoun, 1993). Based on this theory, individuals who are not in the labour force full-time will be less likely to support immigration although previous research has not found a significant association between attitudes and unemployment status. Occupational labour market status is expected to have a significant impact on attitudes, as measured through employment status, and income (Dustmann and Preston, 2002). Although some researchers did not find a significant association between attitudes and unemployment status or occupation (Fertig and Schmidt, 2001; Dustmann, and Preston, 2004b; Card, Dustmann, Preston, 2005; Hernes and Knudsen, 1992; Bauer, Loftstrom, Zimmerman, 2001). The impact of employment status appears to vary with some studies finding an association with position in the labour force (Brenner and Fertig, 2006), while other researchers did not find this affect (Citrin et. al., 1997). Some research found lower income levels to be associated with less support for immigration (Dustmann and Preston, 2002), while other studies found no correlation between attitudes and income (Hernes and Knudsen, 1992; Sanossui et al., 1998; Bauer, Loftstrom, Zimmermann, 2001). Overall, individuals in more economically disadvantaged situations, have been found to be less supportive of immigration (Citrin, 1990; Espenshade and Calhoun, 1993). In Italy and Germany those respondents experiencing more economic strain were more supportive of negative views on immigration (Gang, Rivera-Batiz, and Yun, 2002). 47 2.3 Objectives 1. Research Question: Are attitudes towards immigration similar or different in Canada, the United States and the United Kingdom? Attitudes are expected to differ between Canada, the United States and the United Kingdom (Freeman, 1995; Grabb and Curtis, 2005; Adams, 2003; Simon and Lynch, 1999) despite their historical and social similarities. Research suggests that attitudes toward immigration will be more supportive in Canada than in the Untied States or the United Kingdom where hostilities will be greater (Simon and Lynch, 1999; Heibert, 2003; Crawley, 2005; Palmer, 1996). Attitudes are expected to be especially anti- immigration in the United Kingdom in part as a result of the history of restrictionist policies, the strong white majority, and pattern of increasing hostility toward immigration in previous research (Freeman, 1995; Paul, 1997; Crawley, 2005). 2. Research Question: How are attitudes expected to differ based on individual-level characteristics between Canada, the United States and the United Kingdom? a) Age and gender are expected to have a weak impact on attitudes towards immigration based on previous research in which younger respondents and males were more supportive of immigration (Espenshade and Calhoun, 1993; Tucci, 2005; Quillian, 1995; Card, Dustmann, and Preston, 2005; Chandler and Tsai, 2001). b) Race is expected to have a stronger impact on attitudes toward immigration. Visible minority respondents are expected to be more supportive of immigration than the White respondent's majority based on the potential of these groups to be able to relate to the marginalized experience of immigrants (Schmid, 2003; de la Garza et al. 1992; 48 Espenshade and Hempstead, 1996). In addition, experience as a visible minority is expected to increase tolerance for other minority groups (Flexter, 2002). In addition white respondents have been previously been found to have more support for decreasing immigration than other groups and are more likely to view immigrants as a threat to their social position based on the association between visible minority and immigrant status (Esses et al., 2001; Li, 2003). c) Religion is expected to have a stronger impact on attitudes toward immigration with attitudes differing significantly between religious and non-religious respondents (Dunlevey et al., 2006), as well as differences between Protestant and Catholic respondents (Daniels and von der Runr, 2005). Overall, non-religious individuals are expected to be the most supportive increasing immigration while Protestant and Catholic respondents are anticipated to be the least supportive. d) Building on the hypotheses presented in the literature more economically insecure individuals will be more likely to harbour negative attitudes toward immigration (Fertig and Schmidt, 2001; Dustmann, and Preston, 2004b; Card, Dustmann, Preston, 2005; Hernes and Knudsen, 1992; Sanossui et al., 1998; Bauer, Loftstrom, Zimmerman, 2001; Citrin et. al., 1997; Dustmann and Preston, 2002). Hostilities toward immigration are expected to be higher among respondents with lower income levels and who are not participating in the labour force, based on their potentially insecure place in the labour market. 2.4 Data and Measures To test the hypotheses outlined above, data from national telephone surveys fielded by the Gallup Polling Organization in 2005 were analyzed. These surveys were 49 selected based on the inclusion of the identical question on attitudes toward immigration policy and their availability of independent variables with uniform question formats. Gallup's second wave of data for 2005 was completed in Canada in August 22-31 (1,005), in the United Sates in June 6-25 (2, 264), and in the United Kingdom in August 26-Sep 8 (1,010). All results are based on telephone interviews collected from a nationally representative sample of adults aged 18 and above. These three datasets were pooled into a single dataset with a final sample size of 4,279 (after cases were omitted for missing data). When appropriately weighted, the survey participants constitute a reasonably representative cross-section of persons 18 years of age or older in households with phones for 2005 (Statistics Canada, 2001; US Census Bureau, 2000; UK Office of National Statistics, 2001). Dependant Variable This paper focuses on one key dependant variable, which measures views toward legal federal immigration levels (Espenshade and Hempstead, 1996). The question is worded as follows: ' Would you like to see the level of immigration in this country increased, decreased, or remain about the same?' Response categories were 'increase,' 'decrease, "remain the same, "other,' and 'no opinion.' Individuals who selected 'other' or 'no opinion' were excluded from the sample since their responses could not be compared to categories which indicated a specific level of support. The response categories in this variable were recoded in descending order as follows 'increase' (3), 'remain the same' (2), 'decrease' (1). In the multinomial logistic regression models 'remain the same' was selected as the excluded category, which allows comparison between the factors that impact positive and negative views on immigration. The goal in 50 examining this variable was to determine how much respondents supported current immigration policies and if they do not support current policies in what direction would they like to see policy change. There are limitations to the use of one measure to determine attitudes toward immigration however this particular question is commonly used in the literature on immigration levels and due to the use of previously collected survey data, this paper is limited to the questions uniformly asked in the survey. Independent Variables Several independent variables were included in the models reported in this paper. The frequencies for the six independent variables included are presented in Table 2.1. Age (18-82) and sex (male, female) were included in the model as they were asked in the survey with age entered as a continuous variable and male entered as the comparison category due to the anticipated higher support levels. Race, however, was recoded into white and non-white (i.e., all individuals who identified their association with any other racial category other than white were recoded as non-white). As noted previously while the majority of the population and the survey sample were white, the other minority groups varied by country which made the division of responses into these two categories preferable (US Census Bureau, 2000; UK Office for National Statistics, 2001; Statistics Canada, 2001). While grouping all non-white minorities together for analysis has its limitations, this will test whether there is an overall heightened level of hostility toward immigration among the dominant racial group, who may view immigrants as a threat to their place in the social hierarchy (Esses et al., 2001) and the minority groups who may be more likely to empathize with the marginalized experience of immigrants and thus support higher levels of immigration (Fletzer, 2003). Religion (Protestant, Catholic, 51 Other, None) is included in the model as an additional measure which has been found to previously impact attitudes and has been recoded merging all religious individuals who were not Catholic or Protestant together (Daniels and von Ron, 2005). This grouping was intended to separate religious from non-religious individuals while still keeping the response categories large enough to be useful in a regression model. The second type of variable included in the model is economic measures to test the economic labour market theory (Dustmann and Preston, 2002). Through measures of employment status and income the economic insecurity of respondents is measured determining how secure their labour market position is. Employment status was recoded into respondents participating in the labour market full-time, and respondents in all other categories including students and retired respondents. This format allows for comparison of those individuals who are participating in the labour market full-time, which may provide them with more access to economic capital and heightened feelings of economic security and respondents who for a variety of reasons are not in that situation. Income has been included in this model in three collapsed categories of high, medium and low. The three categories were created by combining the original response options in each country into three categories in which the attempt was made by the researcher to evenly distribute the responses between the three categories. Thus, respondents in each country were divided into the top, middle, and lower of the income brackets allowing for a crude, but comparable, measure of income since it effectively reduces the information detail we have for individual respondents. However, this format allows for the testing of the overall differences in attitudes between those at the top tier of these countries income levels in comparison to respondents with less annual income. In 52 summary, age, sex, race, religion, employment status and income are included to measure the impact that demographic and economic variables have on attitudes. 2.5 Strategy of Analysis First, the descriptive statistics, are presented showing the distribution of the data as well as the significant bivariate relationships (Table 2.1-2.3). The bivariate relationships between attitudes toward immigration policy and age, gender, race, religion, employment status, and income were analyzed separately in each country so the differences in these variables cross-nationally could be compared. The independent variables in each country were analyzed in individually in additional models not included here to test the consistency of the factors that impact attitudes. The survey data from each country was merged and the final multinomial logistic regression model was run with country included as a variable testing the significance of the impact that country had on attitudes. In this paper Chi-Squared values greater than 5 are considered large, while Eta2 values greater than .05 are considered substantial. P-values are included distinguishing between results that have p<.05, p<.01, p<.001. For the purposes of this paper, multinomial logistic regression was selected as the method of analysis. Since this paper models the probability that someone will hold a particular view toward immigration policy based on their individual-level characteristics, logistic regression was the best procedure available. The dependant variable in this data set is categorical with three response options which are presented in size order. Multinomial (or polytomous) logistic regression allows for the analysis of each combination of values (or covariate pattern) with the independent variables (Norusis, 2005; Espenshade and Hempstead 1996; Mayda 2004; Tucci 2005). 53 2.6 Results Table 2.1 presents the distribution of attitudes towards immigration in Canada, the United States, and the United Kingdom. The differences in levels of support for increasing and decreasing immigration highlight the distinct patterns of attitudes in each of the research countries. In Canada, the majority of respondents (57.6%) supported keeping immigration at its current level, while the majority of American (58.7%) and British (61.5%) respondents supported decreasing immigration. Canada is also unique because of its higher level of support for increasing immigration (17%), in comparison to the United States (5.7%) and the United Kingdom (4.1%). Table 2.3 reports on the cross-tabulations of immigration attitudes and each of the independent variables by country, as well as measures of significance. In the cross- tabulation tables of attitudes toward immigration, female respondents were consistently less supportive of immigration than male respondents. The relationship between attitudes toward immigration and gender appears to vary by country. In Canada, fewer males (21.3%) than females (31.0%) wanted immigration to decrease, while there was no significant relationship between attitudes toward immigration and gender in the United States. In the United Kingdom were also a smaller proportion of males (56.6%) than females (63.4%) that supported decreasing immigration. Race also had a significant and varying impact on attitudes in all three countries. In Canada, 26.8% of non-whites wanted to increase immigration in comparison to 15% of white respondents, while 13.4% of non-white respondents wanted to decrease immigration in comparison to 28.6% of white respondents. This same pattern of support for immigration among non-white respondents was repeated in the United States and the United Kingdom. In the United 54 States, 13.6% of non-white respondents wanted to increase immigration in contrast to 4.2% of white respondents. While 61.6% of white American respondents wanted immigration levels to decrease, 42.7% of non-white respondents wanted immigration levels to decrease. British white respondents were far more supportive of decreasing immigration (62%) than non-white respondents (34%). Overall, race consistently impacted attitudes cross-nationally with non-white respondents supporting increasing immigration and not supporting decreasing immigration. Religion had a stronger impact on attitudes toward immigration in the United States and the United Kingdom than in Canada. In Canada, the greatest distinction in attitudes was in the levels of support for increasing immigration between non-religious (25%) and religious respondents (12.6% Protestant, 15.8% Catholic, and 15.5% other). In the United States, non-religious individuals were the lest supportive of decreasing immigration (46.7%), in comparison to Protestants (64.5%) who were the most supportive. This was also the case in the United Kingdom where 50% of non-religious respondents wanted to decrease immigration which was low compared to 64.8% of Protestant respondents. Protestants were overall the least supportive of immigration in comparison with the other religious and non-religious groups in all three countries. Next we turn to the issue of whether or not these bivariate relations are retained once other control variables are introduced. Considering first, the regression model for Canada in Table 2.4, there is evidence that the factors that impact positive attitudes follow different patterns than those that shape negative attitudes. In the case of Canada, factors that impact negative attitudes include age and gender. While, white respondents and religious individuals especially 55 Protestants, were also less likely to support increasing immigration. Model 2 for the United States, yielded a distinctly different pattern of support. White and Protestant respondents, as in the case of the United Kingdom, were both more likely to harbour hostilities toward immigration. Unlike in Canada, in the United States attitudes appear to be impacted by two key factors: race and religion. The third model in Table 2.4 reports on the results of the regression model run in the United Kingdom. Turning to the factors that impact hostilities toward immigration in Britain, white and Protestant respondents were more likely to hold hostile views toward immigration. In addition, respondents with low or medium income levels were more likely to want immigration to decrease than respondents in the higher income brackets. The United Kingdom is the only country where income significantly impacted attitudes. Interestingly age and employment status impacted openness toward immigration. For this survey sample, older respondents were more likely to want immigration to increase while respondents in the labour force were less likely to support more openness toward immigration. Finally, the last model in Table 2.4 includes data from merging the surveys of all three countries and includes the country dummy variables. In this model, there appears to be significant differences between attitudes in the United Kingdom and Canada. Canadians are more likely to be open toward immigration and less likely to be hostile toward immigration than British respondents. Overall, there are differences in support and hostility between Canada, the United States and the United Kingdom. While the types of factors that impact hostilities and openness in the United States and Canada are similar namely race and religion. The way these factors impact attitudes appear to be different based on the significantly greater 56 levels of hostilities in the United States. Differing historical experiences with immigration (Paul, 1997) policy development and construction of immigrant's participation (Freeman, 1995) may have helped to foster these differences in openness toward immigration. Interestingly, how attitudes are shaped in the United States and the United Kingdom are significantly different, while the levels of hostility toward immigration are quit similar. Thus, this paper provides further evidence that even when levels of support for immigration are similar, the factors that shape these attitudes develop quit differently. 2.7 Discussion/Conclusion This paper explores the relationship between attitudes toward immigration, country and individual-level characteristics of respondents. The strength of the mechanisms that impact attitudes were tested and the differences by country explored. The results suggest that individual-level variables have different impacts on attitudes toward immigration cross-nationally. This paper makes several contributions to the immigration policy literature, through replicating previous research on attitudes toward immigration, comparing three countries with similar historical characteristics and controlling for the impact of country on attitudes. From this analysis, the ways in which attitudes follow similar or different value trajectories are highlighted. To conclude, the two hypothesizes presented in the paper are reviewed and general conclusions drawn. Beginning with the hypothesis of cross-national differences, Canadians were significantly more open to immigration than British respondents supporting Freeman's assertion that differences in immigration policy would shape attitudes. There is not, however, support for the assertion that attitudes would be 57 significantly more open in the United States than the United Kingdom (Freeman, 1995). The increasing hostilities toward immigration in the United States, as mapped by other researchers (Citrin et al., 1997), may illustrate a possible trend in the American public toward more hostile views that are characteristic of the restrictionist views present in European public opinion (Semyonov, Raijman, Gorodzeisky, 2006). From this analysis, which highlights the different ways in which a small set of factors impact attitudes we can start to understand how two countries, such as Canada and the United States, with similar immigration policies, and parallel historical trajectories can have such different views on immigration (Adams, 2003; Lipset, 1963). The second hypotheses raised in this paper were that differences would exist in the way demographic and economic characteristics might impact attitudes cross- nationally. Through the comparison of regression models in each country these differences were highlighted, as well as the ways in which these factors impacted support and hostility toward immigration differently. Gender and age had a minimal impact on attitudes with older Canadian respondents reporting higher levels of support for decreasing immigration, while older British respondents were more supportive of increasing immigration. Only in Canada were males less likely to want immigration to decrease immigration than females. Overall, the ethnic and racial status of respondents consistently impacted attitudes although the strength of this impact varied. In all three countries, demographic characteristics had a more significant impact on attitudes than measures of economic insecurity. Economic insecurity only impacted attitudes in the United Kingdom where those with lower income levels were less supportive of immigration. One of the interesting observations from this paper that the way in which 58 attitudes differ between countries, such as the differences in the ways the some factors shape attitudes in contrasting ways cross-nationally. For example, the United States and the United Kingdom have similar levels of support and hostility for immigration but very different factors shaping those attitudes. In sum, the results from this paper provide some interesting evidence for both the demographic and economic hypotheses as well as providing opportunities for further investigation. This paper has mapped some of the social, historical and demographic similarities between Canada, the United States and the United Kingdom (Lipset, 1963; Grabb and Curtis, 2005; Simon and Lynch, 1999) highlighting the growing distance in public opinion trends between Canada and the United States as well as the emerging similarities between the United States and the United Kingdom. Further research is needed to determine the impact of socio-demographic characteristics on positive and negative attitudes toward immigration cross-nationally. Additional measures of demographic characteristics such as different variables might add further detail and insight into how economic and racial characteristics impact attitudes. Additionally, research might include other liberal democratic countries such as Australia and New Zealand with similar characteristics to the United States and Canada (Freeman, 1995), as well as additional European countries with similar historical patterns of migration to the United Kingdom. The inclusion of more variables may also provide further insight into the interaction between multiple factors in shaping attitudes toward immigration. Part of what these results reveal is that the situation in which public perception toward immigration develops is complex vary cross-nationally and are challenging to unpack in relation to specific dynamics. In analyzing attitudes cross- 59 nationally this paper highlights the reality that while issues of immigration are becoming more globalized, the profile of differences in attitudes varies. 60 2.8 Bibliography Adams, Michael. 2003. Fire and Ice: The United States, Canada, and the Myth of Converging Values. Toronto: Penguin Books. Bauer, Thomas K., Magnus Lofstrom, and Klaus F. Zimmermann. 2001. "Immigration Policy, Assimilation of Immigrations, and Native's Sentiments Towards Immigrants: Evidence from 12 OECD Countries." Working Paper 33.San Diego: University of California. www.ccis-ucsdorg/PUBLICATIONS/wrkg33.pdf Braudel, Fernand. 1979. The Structures of Everyday Life: the Limits of the Possible, Vol. 1: Civilization and Capitalism, 15' -1 8th Century. New York: Harper and Row. Brenner, Jan and Michael Fertig. 2006. "Identifying the Determinants of Attitudes towards Immigrants: A Structural Cross-Country Analysis." IZA Discussion Papers. 2306. Card, David, Christian Dustmann and Ian Preston. 2005. "Understanding attitudes to immigration: the migration and minority module of the first European Social Survey." CReAM Discussion Paper. Castles, Stephen. 2000. Ethnicity and Globalization. London: SAGE Publications. Chandler, Charles R. and Hung-mai Tsai. 2001. "Social Factors influencing immigration attitudes: An analysis of data from the General Social Survey." The Social Science Journal 38: 277-288. Citrin, Jack, Donald P. Green, Christopher Muste, Cara Wong. 1997. "Public Opinion Immigration Reform: The Role of Economic Motivations." Journal of Politics 59.3: 858-881. Citrin, Jack and Donald P. Green. 1990. "The Self-Interest of Motive in American Public Opinion." In Research in Micropoltics. 3 (eds.) Samual Long. Greenwich, CT: JAI Press. Crawley, Heaven. 2005. "Evidence on Attitudes to Asylum and Immigration: What We Know, Don't Know and Need to Know." Centre for Migration, Policy and Society. Work Paper No. 23. University of Oxford. WP-05-23. Daniels Joseph P., and Marc von der Ruhr. 2005. "God and the Global Economy: Attitudes toward Trade and Immigration in the United States." Socio-Economic Review 3:467-489. Day, C. I. 1990. "Ethnocentrism in the Era of Limits: Nativits Reactions to the 'New' Immigration." Working Papers in US.-Mexican Studies 3. San Diego: Center for U.S.-Mexican Studies, University of California. de la Garza, R., DeSipio, L., Garcia, F., Garcia, J. and Falcon, A. 1992. Latino Voices: Mexican, Puerto Rican, and Cuban Perspectives on American Politics. Westview Press, Boulder, CO. Dion, Kenneth L., and Kerry Kawakami. 1996. "Ethnicity and perceived discrimination in Toronto: Another look at the personal/group discrimination discrepancy." Canadian Journal of Behavioral Science 28.3: 203-213. Dunleavy, Patrick, Richard Heffernan, Philip Cowley, and Colin Hay. 2006. Developments in British Politics: Bk. 8. Palgrave Macmillan. Dustmann, Christian and Ian Preston. 2001. "Attitudes to Ethnic Minorities, Ethnic Context and Location Decisions." The Economic Journal 111:353-373. 61 Dustmann, Christian and Preston, Ian. 2002. "Racial and Economic Factors in Attitudes to Immigration." University College London and Institute for Fiscal Studies. Dustmann, Christian. and Ian Preston. 2005. "Is Immigration Good or Bad for the Economy? Analysis of Attitudinal Responses. forthcoming." Research in Labour Economics. CReAM Discussion Paper No.06/04. Dustmann, Christian and Ian Preston. 2004b. "Is Immigration Good or Bad for the Economy? Analysis of Attitudinal Responses." CReAM Discussion Espenshade, Thomas J. and Katherine Hempstead. 1996. "Contemporary American attitudes toward U.S. immigration." International Migration Review 30.2: 535— 570. Espenshade, T.J. and Calhoun, C.A.1993. "An analysis of public opinion toward undocumented immigration." Population Research and Policy Review 12:189— 224. Esses, Victoria, John Dovidio, Lynne Jackson, and Tamara Armstrong. 2001. The Immigration Dilemma: The role of perceived group competition, ethnic prejudice, and national identity. Journal of Social Issues 57:389-412. European Commission. 2003. European Citizens and Freedom, Security and Justice: Qualitative Survey of Citizens of the 15 Member States and the 13 Applicant Countries. europa.eu.int/comm/justice_home/doc_centre/intro/docs/eurobaro_qual itatif en.pdf Fertig, M., and C.M. Schmidt. 2001. "First- and Second-Generation Migrants - What Do We Know and What Do People Think." in Migration Policy and the Economy: International Experiences. (ed.^) by R. Rotte and P. Stein, Neuried (ars et unitas). Fertig, Michael and Christoph M. Schmidt. 2002. "The Perception of Foreigners and Jews in Germany — a structural Analysis of a Large Opinion Survey." IZA discussion papers series 431. Fetzer, Joel S. 2000. Public Attitudes Towards Immigration in the United States, France, and Germany. New York: Cambridge University Press. Freeman, Gary. 1995. "Modes of immigration politics in liberal democratic states." International Migration Review. 19.4: 881-902. Gang, Ira, Francisco Rivera-Batiz, and Myeong-Su Yun. 2002. "Economic strain, ethnic concentration and attitudes towards foreigners in the European Union." Rutgers University Department of Economic Working Papers. 14. Grabb, Edward and James Curtis. 2005. Regions Apart: The Four Societies of Canada and the United States. Oxford: Oxford University Press. Hernes, Gudmund and Knud Knudsen. 1992. "Norwegians' Attitudes Toward New Immigrants." Acta Sociologica. 35: 123-139. Hiebert, Dan. 2003. Are Immigrants Welcome? Introducing the Vancouver Community Studies Survey. Vancouver: Research on Immigration and Integration in the Metropolis, Working Paper Series. 03-06. Hjerm, Mikael. 2000. "Multiculturalism Reassessed." Citizenship Studies 4.3:357-381. Lapinski, John S., Pia Peltola, Greg Shaw, and Alan Yang. 1997. "The 62 Polls-Trends: Immigrants and Immigration." Public Opinion Quarterly 61:356- 383 Li, Peter. 2003. Destination Canada: Immigration Debates and Issues. Toronto: Oxford University Press. Lipset, S. M. 1963. The first new nation. New York: Basic Books. Lynch, James P. and Rita Simon. 2003. Immigration the World Over: Statutes, Policies, and Practices. Lanham: Rowan & Littlefield. Mayda, Anne Maria. 2004. Who Is Against Immigration? A Cross-Country Investigation of Individual Attitudes towards Immigration. IZA Discussion Paper No. 1115. O'Rourke, Kevin H. and Richard Sinnott. 2001. "The Determinants of Individual Trade Policy Preferences: International Survey Evidence." Economic Papers. 20010. Trinity College Dublin: Department of Economics. Palmer, Douglas. 1996. "Determinants of Canadian attitudes toward immigration: More than just racism?" Canadian Journal of Behavioral Science 28: 180-192. Paul, Kathleen. 1997. Whitewashing Britain: Race and Citizenship in the Postwar era. Cornell University Press. Scheve, Kenneth F. and Matthew J. Slaughter. 2001. "Labor Market Competition and Individual Preferences Over Immigration Policy." The Review of Economics and Statistics 83: 133-145. Schuster, L and Solomos, J. 2004. Race, immigration and asylum: New Labour's agenda and its consequences. Ethnicities 4.2: 267-300. Schmid, Carol. 2003. "Immigration and Asian and Hispanic Minorities in the New South: An Exploration of History, Attitudes, and Demographic Trends." Sociological Spectrum 23: 129-157. Semyonov, Moshe, Rebecca Raij man, and Anastasia Gorodzeisky. 2006. "The Rise of Anti-Foreigner Sentiment in European Societies, 1988-2000." American Sociological Review 71: 426-449. Simon, Rita J. and James P. Lynch. 1999. "A Comparative Assessment of Public Opinion Towards Immigrants and Immigration Policies." International Migration Review 33: 455-467. Statistics Canada. 2001. Census of Canada. http://www12.statcan.ca/english/census/Index.cfm Taylor, Donald. 1996. Diversity with Justice and Harmony: A Social Psychological Analysis. Prepared for Strategic Policy, Planning a research and Metropolis Project. Citizenship and Immigration Canada. June. Taylor, D.M. and Moghaddam, I.M. 1994. Themes of intergroup relations — international Social psychological perspectives. (second edition). Preager: Westport, Connecticut. Tucci, Ingrid. 2005. "Explaining Attitudes Towards Immigration: New Pieces to the Puzzle. Berlin." German Institute for Economic Research Discussion Paper 484. US Census Bureau. 2000. United States Census 2000. http://www.census.gov/ UK Office of National Statistics. 2001. National Statistics. www.statistics.gov.uk . Crown Copyright. 63 Appendices 1 Note that the racial categories as they are identified in this paper are not the classification of this researcher but are the labels as they are presented in the survey. The terms have been kept in their original form in order to better represent the options available for selection by respondents. There are other methods of racial and ethnic categorization as well as other terms for classifying these various groups. This paper does not intend to promote the use of a particular racial term but instead attempts to best represent the data as it was presented to the respondent. 2 Income categories in each countries Canada The United States The United Kingdom Table 2.1: Precentage of each group and their levels of support for immigration in Canada, the United States, and the United Kingdom, 2005 Canada^United States  United Kin dom Less Same More N Less Same More N Less Same More N 25.3 57.7 17.1 987 58.7 35.5 5.7 911 61.5 34.4 4.1 985 Gender Male 22.9 59.5 18.7 491 57.0 36.8 6.2 443 56.6 37.5 5.9 470 Female 31.0 54.8 14.1 514 61.0 34.3 4.8 476 63.4 34.3 2.3 540 Race White 28.6 56.4 15.0 909 61.6 34.2 4.2 795 62.0 34.9 3.1 947 Non-White 13.4 59.8 26.8 86 42.7 43.6 13.6 111 34.0 49.1 17.0 57 Religion Protestant 26.4 61.0 12.6 325 64.5 31.8 3.7 480 64.8 32.7 2.5 532 Catholic 30.2 54.0 15.8 417 60.5 36.0 3.5 183 61.4 34.2 4.4 115 Other 24.0 60.5 15.5 134 47.9 43.0 9.1 147 54.0 40.5 5.5 179 None 24.1 50.9 25.0 120 46.7 40.0 13.3 96 50.0 43.1 6.9 170 Employment Status In the labour force 26.6 57.5 15.9 614 58.1 36.3 5.6 564 59.5 36.7 3.7 616 Not in the labour force 28.1 56.0 16.0 389 60.4 34.2 5.4 353 61.6 34.5 3.8 393 Income Low 29.0 57.5 13.5 334 59.7 34.1 6.2 211 65.4 31.3 3.3 243 Medium 27.8 54.6 17.6 313 59.3 35.8 4.9 204 62.2 34.4 3.4 323 High 22.6 59.8 17.7 266 58.9 35.9 5.2 440 49.8 43.9 6.3 239 64 Table 2.2: Multinomial Logistic Regression of Immigration Levels and Socio-Demographic Charactersitics (versus Stay the Sam Canada^U.S.^U.K.^All Countries Decrease^Increase^Decrease^Increase^Decrease^Increase^Decrease^Increase Intercept -2.114 *** -0.552 -0.438 -0.931 -1.460 * -2.898 * -0.841 ** -1.474 ** Age 0.015 * 0.012 0.003 0.012 0.009 0.038 ** 0.006 * 0.012 * Demographic Male -0.388 * 0.156 -0.080 0.114 -0.193 0.508 -0.208 * 0.233 White 0.631 -0.691 * 0.607 ** -0.987 ** 0.753 * -0.916 0.677 *** -0.853 *** Protestant -0.203 -1.012 *** 0.538 * -1.067 * 0.458 * -0.364 0.307 * -0.891 *** Catholic 0.056 -0.611 * 0.330 -1.399 * 0.481 0.033 0.317 * -0.621 ** Other -0.176 -0.724 * 0.011 -0.596 0.125 -0.659 -0.019 -0.661 * Economic In Labour Force 0.170 -0.076 -0.093 -0.075 0.062 -0.864 * 0.037 -0.154 Low Income 0.198 -0.306 0.004 -0.023 0.746 *** -0.356 0.306 * -0.249 Medium Income 0.270 0.144 -0.001 -0.180 0.511 ** -0.232 0.262 * 0.031 Counhy Canada -1.246 *** 1.016 *** U.S. 0.179 0.232 R-Squared 0.06 0.07 0.08 0.17 N 1,005 921 1,010 2,936 65 Chapter 3 Dimensions of Hostility: Analyzing the Factors that Shape American Public Opinion toward Immigration Policy, 2005. 3.1 Introduction American public hostility toward immigration is on the rise, challenging the process of incorporating new residents into society (Buck et al., 2003). The major source countries for American immigration are shifting and recent waves of immigrants are more racially, ethnically, and linguistically distinct from the majority of Americans (Bobo, 2000; Oliver and Wong, 2003). This multi-ethnic reality affects how the public views immigration policy since future policy decisions are determined based on current perceptions toward immigration policy. Policy-makers need to understand how ideas about the regulation of immigration and the policing of out-group members (a group considered significantly different to an individual's group) access American society. To do so furthers the process of discouraging exclusion and intolerance, while defusing possible interracial violence. To understand how views toward out-groups such as immigrants develop, it is useful to explore how interaction with out-groups can affect perceptions toward out- groups reflected in immigration public opinion (Quillian, 1995; Oliver and Wong, 2003). Recent research suggests that views on immigration policy are influenced no only by racial prejudice, but also by the racial construction within respondents area of residence (Quillian, 1995). To build on this work, I analyze the impact of perceived interaction with diverse groups on attitudes toward immigration. I use respondent's estimates of the 66 proportions of minorities in the neighbourhood instead of the government's estimates of minority presence measured through census data to test the differences in the resulting estimates. Specifically, I test the significance of Fezter's distinction in the ways that (strong social ties) romantic relationships and (weak social ties) community interaction impact attitudes differently (2000). The demographic make-up of American immigration is changing, and with these shifts come new challenges in the process of easing society through the growing pains of integration. The common American "melting-pot" metaphor encourages the non- dominant group to culturally and socially assimilate to the expectations of the dominant group. However, such situations can sometimes result in more of an inter-group pressure cooker effect than a melting process, increasing the possibility of hostility and violence (Berry, 2001). In contrast, the multicultural model of integration encourages mutual accommodation of social practices by the dominant and non-dominant groups (Berry, 1984). The behaviour pattern shift for both groups can result in a shared change in perspective, promoting value and respect between groups. Integration and inter-group tolerance requires low levels of public prejudice and exclusionist views. Here, I measure tolerance through public views on immigration policy or the process through which out- groups such as immigrants are excluded or included in American society. Previous research, the Realistic Group Threat theory that the significant presence of particular out- group, such as immigrants, will lead to greater hostility toward that out-group based on the perceived threat to the social and economic position of in-group members (Esses et al., 2001). The proportion of black residents in white respondent's neighbourhoods had a significant impact on how they view immigration (Quillian, 1995). In this paper I add to 67 the immigration research literature by determining the impact of perceived interracial experience on attitudes toward immigration in contrast to using survey data. Previous research has highlighted the reality that the public sometimes inaccurately estimates the presence of particular ethnic and racial groups (Crawley, 2005); thus the way in which individuals interpret their interracial interactions may affect attitudes differently than census measured race demographics race demographics developed from census data. Using Gallup 2005 data, I test the impact that demographic, economic, political, and contact factors have on attitudes toward immigration. In multinomial logistic regression (MNL) models I am able to compare how individual-level factors affect positive and negative views toward immigration. The variables I include in the model make two contributions to the literature on attitudes toward immigration. First, building on previous research, I test the extent to which individual-level characteristics shape attitudes. In particular, I evaluate the impact of demographic, economic and political measures on attitudes toward immigration. Since these factors have previously been found to have a significant impact on attitudes I include them in this paper to retest their significance with 2005 data and measure how the impact on attitudes changes depending on how attitudes are framed, allowing me to test the extent to which the impact of these factors varies by the way the topic is framed. Second, I argue that in order to understand attitudes toward immigration, we need to look at how individuals perceive their interaction with diverse groups perceptions of interaction with diverse groups. I analyze perceived interaction, instead of testing inter-group involvement using census data (Oliver and Wong, 2003) or through interactions at work (Fezter, 2000), in order to present a new way of understanding the impact of contact with out-group members on 68 attitudes towards immigration. I hypothesize that less interaction with visibly different individuals will increase hostility toward immigration. 3.2 Trends in Immigration Attitudes a) Demographic Understanding individual attitudes toward immigration begins by acknowledging how individual-level demographic characteristics shape how the public forms views on a range of topics, including immigration. Demographic characteristics vary substantially in their impact on immigration policy attitudes, producing a range of interpretations. Across multiple articles, the following demographic variables have been found to have a significant impact on attitudes: age (Quillian, 1995; Card, Dustmann, and Preston, 2005; Chandler and Tsai, 2001; Bauer, Loftstrom, Zimmermann, 2001; Stonewall, 2003), gender (Chandler and Tsai, 2001), marital status (Brenner and Fertig, 2006), religion (Daniels and von der Runr, 2005; Card, Dustmann, Preston, 2005), and immigration status (Card, Dustmann, Preston, 2005). In some studies, there was an overall increase in negative attitudes as respondents age (Quillian, 1995). Some studies reported males to be more supportive of immigration than females, a trend they associated with women's increased participation in the short term labour market (Hiebert, 2003; Palmer, 1996; Quillian, 1995; Card, Dustmann, and Preston, 2005). Men are more likely to be supporting financial dependents than women, which might increase their hostility toward potential labour market competition (Bauer, Loftstrom, Zimmermann, 2001; Stonewall, 2003). The impact of religious affiliation on attitudes toward immigration policy has been associated with the differences in social capital and belief systems between religious groups which influence how members view out-groups such as immigrants (Daniels and 69 von der Runr, 2005). Daniels and von der Runr found religiously affiliated respondents to be more supportive of open immigration than non-religious respondents (2005) and with the religious respondents fundamentalist Protestants and some Christians were supportive of immigration. The race of respondents has been found to have a significant impact of attitudes toward immigration policy although the strength of the association varied by the context. Some researchers have connected attitudes toward out-group members such as immigrants in relation to how similar the new group members are to the dominant group. From this the views of immigrants may also be tied to the perceived threat of new immigrants to the racial position (Card, Dustmann, Preston, 2005; Citrin et al., 1997; Diamond, 1998; Espenshade and Hempstead, 1996; Schmid, 2003; Dion and Kawakami, 1996). Immigrants who are ethnically similar to the majority in society receive more support than visibly different immigrants who can be easily classified as out-group members (Dustmann and Preston, 2002; Dustmann, Preston, 2005; Simon and Lynch, 1999; Card, Dustmann, and Preston, 2005). Immigrants with the same racial and ethnic background as the majority of residents receive more public support than immigrants who were visibly different from residents (Card, Dustmann, Preston, 2005). Of the American racial groups White and Hispanics were the less supportive of immigration than Blacks (Citrin et al., 1997). Although no racial group, including Blacks, had a supportive majority toward immigration (Dimond, 1998; Espenshade and Hempstead, 1996; Schmid, 2003; Dion and Kawakami, 1996). In fact, evidence suggests that immigration has heightened the economic threat experienced by blacks in the United States, which has not been in increased negative attitudes among African Americans reflected in increasing 70 negative attitudes toward immigration. Overall, race has a significant impact on attitudes, in some studies having a stronger impact on policy decisions than economic position (Dustmann and Preston, 2005). Respondents area of residence also significantly impact their attitudes toward immigration views toward the proportion of outsiders accepted into the country is tied to the area of the country in which respondents live and the urban or rural nature of that region. At the regional level in the U.S., Californians were more supportive of decreasing immigration. This hostility has been associated with concentrated levels of recent immigrants in this part of the country, encouraging public perceptions that the country is receiving too many immigrants (Hood and Morris, 1997; Valentine and McDonald, 2004; Olzak, 1992). The settlement patterns of recent immigration waves have concentrated the presence of immigrants into specific regions of the country furthering the perception in some areas that immigration levels are high (Passel and Fix, 1994). In addition to regional differences, respondents vary in their views toward immigration depending on the urban, suburban, or rural context of their residence. Urban dwellers have been found to be more supportive of ethnic diversity generally and immigration specifically. This perspective within the urban community has been associated with increased opportunities to interact with diverse groups encouraging a more cosmopolitan view and discouraging ethnocentric perspectives (Haubert and Fussell, 2006). b) Economic An extensive body of research has tracked the association between public views toward immigration and economic factors. Immigration policy attitudes have been associated with individual economic context as well as views toward the national 71 economy. Some argue that the link between views on immigration policy and results from the individual self-interest tying views toward immigrants to their perceived economic impact on the context of individuals. Individuals who perceive immigrants to be an economic or social threat will be more likely to want immigration levels reduced to limit this potential challenge to their personal position. Economic insecurity has also been associated with variations in attitudes toward immigration. Specifically the following economic factors have been found to have a significant impact on attitudes toward immigration, these include education (Bauer, Lofstrom, and Zimmermann, 2001; Dustmann, and Preston, 2004b; Mayda, 2004; Scheve and Slaughter, 2001; O'Rourke and Sinnott, 2004; Citrin et al., 1997; Gang, Rivera-Batiz, Yun, 2002; Fertig and Schmidt, 2002; Daniels and von der Ruhr, 2005; Espenshade and Hempstead, 1996; Hainmueller and Hiscox, 2007; Haubert and Fussel, 2006; Chandler and Tsai, 2001), income (Hernes and Knudsen, 1992; Sanoussi et al., 1998; Bauer, Loftstrom, Zimmerman, 2001), skill-level (Borjas, 1991; Borjas 1997; Scheve and Slaughter, 2001; Palmer, 1996), and occupation (Facchini and Mayda, 2006; Dustmann and Preston, 2004). Research in the U.K. suggests that changes in immigration rates have different impacts on individuals based on the part of the labour market they occupy. Respondents with high education and skill levels are more supportive of education and skill levels are more supportive of immigration since they are less likely to have their economic position threatened by immigrants and more likely to benefit from an increase in immigration (Dustmann and Preston, 2005). Highly skilled or educated individuals have reported greater support for immigration based on their greater security in the labour force due to the lack of individuals with whom to compete (Dustmann and Preston, 2005). Beyond the 72 effect that education has on labour force position, some argue that education fosters greater overall tolerance and understanding of individuals different to ones self (Dustmann and Preston, 2005). Highly educated individuals may also be more likely to socialize in settings with white-collar workers who discourage the overt use of ethnocentric statements (Haubert and Fussell, 2006). The education level of respondents parents also had a significant impact on attitudes may result from the secure economic position of the parents which encouraged tolerance and an economically secure perspective among their children. The relationship between skill-level and attitudes toward immigration varies by the skill-level of native workers and the type of contribution made by immigration to the labour force (Mayda, 2004, Scheve and Slaughter, 2001; O'Rourke and Sinnott, 2004). A pattern of migration employment has emerged where recent immigrants are more likely to contribute to the unskilled or the short term work components of the labour market. Based on this pattern native workers in the parts of the labour market immigrants are more likely to join will hold more negative views toward immigration policy. Anti- immigration sentiment is stronger among native workers who have skill levels that are equal or lower skill levels to respondents, due to the increase in competition and possible drop in wages that immigrants present (Facchini and Mayda, 2006). Some researchers found that participation in the labour market, as measured by employment status, was not significantly associated with views toward immigration (Fertig and Schmidt, 2002; Dustmann, and Preston, 2004b; Card, Dustmann, Preston, 2005; Citrin et al., 1997; Hernes and Knudsen, 1992; Sanoussui et al., 1998; Bauer, Loftstrom, Zimmerman, 2001) with some exceptions (Brenner and Fertig, 2006). Researchers also found a lack of 73 association between attitudes and income level (Sanoussi et al., 1998; Bauer, Loftrom, Zimmermann, 2001; Hernes and Knudsen, 1992), while some connected lower income levels with greater hostility toward immigration. Beyond measures of individual economic context associations have also been found between policy preferences and views on the national economic context. National level variables which have been associated with immigration include views on the national economy (Citrin et. al., 1997; Zaller, 1992; Borjas, 1994; 1999b; Friedberg and Hunt, 1995; Olzak, 1992; Espenshade and Hempstead, 1996; Chandler and Tsai, 2001; Magda, 2005; Scheve and Slaughter, 2001; O'Rouke and Sinnott, 2003; Borjas, Freeman, Katz, 1996; Borjas, 1999), international trade (Espenshade and Hempstead, 1996), taxes (Citrin et al., 1997) and the potential impact of immigrants on the welfare system (Citrin et al., 1997; Passel and Frix, 1994; Dustmann, and Preston, 2004b; Borjas, 1999a; Studlars, 1977; Simon, 1989). Associations between attitudes toward immigration policy and views on the state of the nation have been perceived to result from public perceptions about the potential impacts on current cultural and economic trajectories. Negative views toward immigration policy have been tied to perspectives that immigrants will depress wages and shift domestic markets (Kiewiet and Rivers, 1985; Scheve and Slaughter, 2000). c) Political The third hypothesis tested in this paper is that ideological perspectives such as political views will significantly affect toward immigration. Previous research has found an association between political orientation and attitudes toward immigration (Semyonov, Raijman, and Gorodzeisky, 2006; Citrin et al., 1997; Dustmann, and Preston, 2002; 74 Espenshade and Hempstead, 1996; Scheve and Slaughter, 2001). In U.S. studies Conservative respondents and Republicans (Scheve and Slaughter, 2001) were more supportive of decreasing immigration than liberals, or Democrats. In a cross-national study of attitudes in European countries, anti-immigration sentiment was associated with support for right-wing parties (Semyonov, Raijman, and Gorodzeisky, 2006). Support was particularly high for parties whose political agenda focused on free market development and ethnocentrism (O'Rourke and Sinnott, 2004). Isolationist, patriotic, chauvinistic and racist views have also been associated with anti-immigration attitudes (Espenshade and Hempstead, 1996; Esses et al., 2001; Burns and Gimpel, 2000; Citrin et al., 1997; Dustmann, and Preston, 2004b; Borjas, 1997). Overall, political views that frame immigrants as outsiders or fame immigrants as different from residents are more likely to be associated with negative attitudes toward immigration. d) Contact This paper makes a unique contribution to the literature on attitudes toward immigration by testing the impact of interracial contact on attitudes. Contact theory suggests that interacting with visibly or ethnically different individuals will change individuals attitudes toward outsiders such as immigrants. Fetzer builds on the contact theory by disguising between the type of contact and individual engages in (2000). Fetzer's description of the contact theory draws distinctions between "acquaintance" and "casual" levels of interaction. At the "acquaintance" level tolerance is expected to the "casual" level which is expected to foster hostility due to the indirect nature of community interaction (Fezter, 2000). In a study using MORI 2005 data in the United Kingdom, respondents who knew someone of a particular minority group had greater 75 tolerance for that particular group as well as other minority groups generally (Stonewall, 2003). Contact without engagement did not foster tolerance and some argue that this may even create inter-group hostility due to the ways in which superficial interactions reinforce stereotypes (Valentine and McDonald, 2004). 3.3 Objectives How do demographic, economic, political, and contact factors impact attitudes differently? a) Demographic Based on previous research, significant associations are expected between attitudes toward immigration and socio-demographic characteristics including age, gender, marital status, religion, region, and urban area. Past results suggest that younger respondents, males, single respondents, non-whites, non-religious respondents, respondents from Eastern states and urban residents will be more supportive of immigration policy (Quillian, 1995; Card, Dustmann, and Preston, 2005; Chandler and Tsai, 2001; Bauer, Loftstrom, Zimmermann, 2001; Stonewall, 2003; Brenner and Fertig, 2006; Daniels and von der Runr, 2005; Card, Dustmann, Preston, 2005). b) Economic In the model reported in this paper I test the impact of individual economic context and perceptions of the national economy on attitudes toward immigration. By including these two types of measures I am able to assess the extent to which these framing of economic context impacts attitudes toward immigration policy (Hernes and Knudsen, 1992). The economic labour market insecurity hypothesis suggests that individuals in insecure personal and financial positions will be less supportive of 76 increasing immigration and more supportive of decreasing immigration than respondents who are financially secure. Based on this hypothesis I expect support for immigration at the individual level to be higher among people with higher levels of education income and greater labour force participation (Dustmann, and Preston, 2004b; Dustmann and Preston, 2004a; Mayda, 2004; Scheve and Slaughter, 2001; O'Rourke and Sinnott, 2004; Dustmann and Preston, 2001). At the contextual level, hostility toward immigration will be higher amoung respondents who believe that the economy is doing poorly (Citrin et. al., 1997; Zaller, 1992; Borjas, 1994; Espenshade and Hempstead, 1996; Chandler and Tsai, 2001; Mayda, 2004; Scheve and Slaughter, 2001; O'Rouke and Sinnott, 2004; Borjas, Freeman, Katz, 1997; Borjas, 1992). Negative views toward immigration have been linked to perceptions of the impact that immigrants have on receiving societies (Dustmann and Preston, 2004a; Dustmann, Fabbri, and Preston, 2004); thus attitudes will be expected to differ based on one's economic position as well as one's perspective on that economic position. c) Political Existing literature suggests that the political ideology of respondents and their views on the state of the nation will impact how attitudes toward immigration policy develop (Semyonov, Raijman, and Gorodzeisky, 2006; Citrin et al., 1997; Dustmann and Preston, 2002; Espenshade and Hempstead, 1996; Scheve and Slaughter, 2001). The framing of immigration policy debates are informed by the national agendas of political parties, and ideological values encouraged about immigration. Based on the existing literature I anticipate fiscally conservative and ethnocentric respondents to be less supportive of immigration. The self-interest hypothesis may also apply encouraging 77 respondents who do not feel secure in the state of the nation or do not want support for minority groups to be less supportive of immigration. d) Contact The contact hypothesis asserts that support for immigration will be higher among individuals who have intimate or "acquaintance" relationships with someone of a different racial or ethnic background, while "casual" interracial encounters will foster less tolerance for out-groups (Flezter, 2000; Allport, 1954). In the case of this data set, individuals who are visibly different to the majority of the population will be expected to be more supportive of immigration. Since the majority of the United States population is white (US Census Bureau, 2001), living in an area with a high proportion of residents who are black, Hispanic, or recent immigrants will decrease tolerance based on Fezter's hypothesis due to the superficial nature of interaction at the community level. These low tolerance levels are anticipated based on the superficial nature of interracial community interactions, since living among particular ethnic group may not lead to direct interpersonal interaction further reinforcing stereotypes (Crawley, 2005). Living in a community with a particular ethnic group may not breed direct interaction or questions about the further reinforcing proportions or particular ethnic minorities the respondents perceives stereotypes to be present in their community. I will test the extent to which "acquaintance" level interracial interaction has a different impact on views toward out- groups such as immigrants. Measured in this paper through interracial dating "acquaintance" interaction may foster strong social ties to a visibly or ethnically different group fostering greater understanding at the interpersonal level may encourage the practice or mutual accommodation at the interpersonal level necessary for the fostering of 78 tolerance. Thus, the process of accommodating an ethnically different partner may increase acceptance of other individuals who are different to ones self (Flezter, 2000; Valentine and McDonald, 2004). 3.4 Data and Measures To test these hypotheses, I analyzed previously collected national telephone survey data. The survey was fielded in the United States by Gallup Polling Organization in 2005. Selected for analysis based on the large representative sample of the American public and the inclusion of a variety of measures, the questions asked in this survey make it an excellent research tool for understanding the range of dimensions which can impact the formation attitudes. The Gallup Polling Organization completed this survey in June 6- 25 (2, 264) collecting a representative sample of American adults aged 18 and over. The sample was weighted based on age and sex to provide a representative sample of American population for analysis (after cases with missing data were omitted). 3.5 Dependant Variable Replicating and building on pervious research, this paper uses a measure of immigration policy preference levels to assess attitudes toward immigration (Espenshade and Hempstead, 1996; Mayda, 2004; Tucci, 2005; Citrin et al., 1997). The question to respondents was "Would you like to see the level of immigration in this country increase, decrease, or remain the same?" with the response categories of "decrease", "remain the same" and "increase." Respondents who selected "other" or "don't know" were excluded from the sample. By analyzing responses to this question I measure approval for current immigration policy as well as how respondents would prefer to see immigration change. 79 Note that this question does not distinguish between legal and illegal immigration or include immigrant's country of origin. 3.6 Individual-Level Independent Variables The dataset selected includes a range of demographic, economic, political and contact measures allowing me to test the extent to which these factors impact attitudes. The variables included in the multinomial logistic regression model are shown in tables 3.1-3.4. These variables have been divided into separate tables based on the dimension of attitudes they measure. Basic demographic measures such as age, gender, race, marital status, religion, country of birth, region and area have been associated with immigration attitudes. 2 Age, gender, and area are included in the model as they appear in the original survey, while the other demographic measures were recoded, merging categories with small response rates. The foreign-born status of individuals and the foreign-born status of their parents were combined into a single variable which divides respondents into three categories: U.S. born respondents with U.S. born parents, U.S. born respondents with a parent who was born outside in the U.S., and foreign-born individuals. This measure intends to capture personal and immediate family ties to individuals born-aboard many who may have moved to the U.S. as immigrants. Respondents are not asked for their residency status or citizenship creating no distinctions between illegal immigrants and American citizens born abroad, providing a rough measure of individuals born abroad in a variety of contexts. By including this variable I test the impact that having at least one foreign-born parent might have on attitudes toward immigration in contrast by being born abroad your self. Race was also recoded merging respondents into white, black, Hispanic and "other" categories. Only respondents who identified themselves as exclusively 80 Hispanic were coded as such. If a second racial category was mentioned respondents were coded into the "other" racial category. Marital status, religion, and area were also included based on previously significant results (Quillian, 1995; Daniels and von der Runr, 2005; Fertig and Schmidt, 2002). I recoded marital status combining "married" and "living as married" into a single category representing all individuals who live with their partners. "Divorced," "widowed," and "separated" respondents were also merged into a single category based on their low response rates and shared potential interpersonal instability. Finally, I kept "single" respondents as a separate category based on the impact that interpersonal independence might have encouraging self reliance and financial independence increasing the possibility of hostilities toward labour market threats. Religion was also recoded combining all religious afflictions that were not Protestant, Catholic or non- religious into a single response category. Catholics and Protestants are the largest religious groups in the U.S. so separating out how these two major American Christian denominations view immigration provide further detail into the development of attitudes (US Census Bureau, 2000; Daniels and von der Ruhr, 2005). Finally, region was recoded into four US census area groups: North, South, East, and West (2001). To asses the influence of economic factors on attitudes, I include variables which measure economic position, respondent's perceptions toward their personal finances and opinions on the state of the national economy. By including a range of question types I compare how attitudes are impacted by respondent's personal economic situation and their views on the national economic context. To determine individual economic context I include measures in the model of education-level, income, and employment status.3 81 Education and income were recoded, and I combined the categories with low response rates. Employment status was also recoded, merging the response categories into respondents who participated in the labour force full-time and respondents who did not. I also included questions about respondent's assessments of their financial situation. Respondents were asked about their financial situation in general as well as their ability to pay for food, clothing, and medical bills. By including questions about respondent's perceptions toward their finances I am able to discern an additional way of understanding financial insecurity. Previous research suggests that attitudes are also influenced by confidence in the national economy (Citrin and Green, 1990). I measure this factor here through variables asking for respondents views on the economy as well as how they think the economy is changing. I consider the influence of political perspective on attitudes by including measures of political philosophy, satisfaction with the United States, and the role of government in advocating for minorities. Political philosophy has previously been found to affect attitudes, resulting in greater support for immigration among left leaning respondents (Semyonov, Raijman, and Gorodzeisky, 2006; Citrin et al., 1997; Dustmann and Preston, 2002; Espenshade and Hempstead, 1996; Scheve and Slaughter, 2001). I entered a collapsed from of the political philosophy variable into the model which combined "very conservative" with "conservative" and "very liberal" with "liberal". The resulting variable has three categories of conservative, moderate and liberal respondents providing a general barometer of political orientation. Respondents satisfaction with the U.S. overall was also included in the model testing the significance of the impact that respondents general dissatisfaction general dissatisfaction with the state of the nation on 82 attitudes. Previously ethnocentric individuals were less supportive of immigration than respondents who supported the rights of diverse groups. The importance of views toward minorities in shaping attitudes is measured here including individual views on minority advocacy programs and affirmative action for minorities. Since previous researchers found that ethnocentric individuals were less supportive of immigration (Semyonov, Raijman, and Gorodzeisky, 2006). I included measures of respondent's views toward minority advocacy policy and affirmative action. These measures allow me to determine how much individuals support and recognize the issues facing marginalized groups. Finally, to determine how interracial contact affects attitudes toward immigration, I evaluate how perceived interaction with out-groups influences attitudes. 5 By analyzing the estimated presence of racial groups in respondents' communities, I can determine how an individual's perceptions toward heterogeneous communities affects their attitudes. Note that the question format does not ask respondents to provide an accurate estimate of racial groups in their area it only requests they estimate the overall presence of out-groups. The definition of "area" is also unspecified, allowing respondents to define the parameters of their community based on what they consider appropriate. Beyond testing the impact of interracial interaction, I include interracial dating patterns as a measure of "acquaintance" level contact with out-group members. Respondents were asked if they ever dated someone of a different racial or ethnic background. The undefined nature of this question format captures respondent's engagement in interracial dating without making distinctions based on sexuality, level of commitment, length of relationship or type of relationship. I am trying to measure perceived out-group interaction at the community ("casual") and interpersonal ("acquaintance") level in order 83 to evaluate how contact impacts attitudes and how attitudes might differ depending on the type of contact one has experienced. 3.7 Strategy for Analysis To test these hypotheses, multinomial logistic models were regressed on individual-level variables of demographic, economic, political, and contact factors. The technique of assigning values to response categories and analyzing the outcome with a single regression model has been previously used in the literature. However, ordinary least squares (OLS) estimates, a common choice of modeling technique, inefficiently models correlates since differences in logits between categories are not reflected, losing valuable detail. Instead, I estimate the model using multinomial logistic regression, which can assess the logits between response categories. Dividing the variance into separate categories allows for further assessment of the exact differences in attitudes. I exclude "remain the same" as the comparison response category in this analysis, allowing comparison between the factors that affect positive and negative views toward immigration. The response categories of "don't know" or "no answer" were coded as missing data. 3.8 Results Table 3.1 presents the bivariate trends in immigration attitudes using 2005 data. Attitudes toward immigration are shown to be associated with demographic, political, economic and contact measures. Evidence in Table 3.1 indicates that gender, race, religion, marital status, region and area all had significant associations with attitudes towards immigration. The following economic variables also had a significant association with attitudes toward immigration: education, employment status, income, respondents' 84 concerns about personal finances, assessment of the economic context in the U.S. and perceived economic changes. All of the measures of respondent's political perception were significantly associated with views on immigration; this included overall satisfaction with the United States, views toward affirmative action and the role of the government in changing the situation of minorities. Finally, of the contact variables, the measures of association were only significant between attitudes and respondents perception that there was a high quantity of Hispanic, Asian and recent immigrant residents living in their area. In addition, never having dated someone of a different racial or ethnic background fostered support for decreasing immigration. The multinomial logistic regression analysis which uses immigration policy attitudes as the dependant variable is shown in table 3.5. Models 1-3 show the results of previously significant dimensions of attitudes while the theoretically unique contact variables are added in model 4. Model 1 provided estimates of the total variance in immigration attitudes by age, gender, country of birth and country of parent's birth, religion, race, marital status, region and area. In model 1, younger respondents were less supportive both of increasing immigration and decreasing immigration, suggesting that overall, younger respondents are less supportive of change to immigration policy. In model 3 and 4 when political orientation and contact variables were added, only younger respondents resistance to decreasing immigration remained significant. Respondents who were born in the U.S. (in contrast to foreign-born respondents) whose parents were also born in the U.S. as well as respondents who had a parent born outside the U.S. were significantly more supportive of decreasing immigration than foreign-born respondents. Foreign-born individual's views on immigration policy were 85 significantly different than individuals born outside of the U.S. Having one or more parents born outside of the United States did not appear to foster significantly higher levels of support for immigration. Thus, the experience of having a foreign-born parent did not have a significant impact on the perceptions of their children toward current immigration policy. Supporting previous research, white respondents were significantly more supportive of decreasing immigration than respondents classified as "other" (Dimond, 1998; Espenshade and Hempstead, 1996; Schmid, 2003). Whites, as the racial majority in the U.S., may be more likely to view the continuous inflow of immigrants, especially non-white immigrants, as a threat to their social position within U.S. society (US Census Bureau, 2000; Esses et al., 2001). In models 2 and 3, blacks were also significantly more likely to support decreasing immigration than respondents in the "other" racial category. Religion had a varying impact on attitudes toward immigration, the impact of which changed between models. In models 1-3, Protestant and Catholic respondents were significantly more supportive of decreasing immigration than respondents who identified themselves with "other" religions or respondents who were non-religious. However, religion was not significant in model 4 when measures of contact were added. Relationship status had a significant impact on attitudes towards immigration. Single respondents were less supportive of decreasing immigration and more supportive of increasing immigration than "Divorced/Separated/Widowed" respondents, pointing to greater support for policy change among single respondents. Support for decreasing immigration among "Divorced/Separated/Widowed" respondents could be associated 86 with self-interest, as divorced, widowed, and separated individuals may experience greater personal economic insecurity. Respondents from western states were more likely to support reductions in immigration levels than eastern states. The western states were chosen as the excluded category based on their high levels of immigration and concentrated immigration settlement patterns, especially in states such as California (Hood and Morris, 1997). The regional concentration of immigrants in the western U.S. may affect perceptions toward the proportion of immigrants present regionally, leading to greater support for decreasing immigration Area also had a significant impact on attitudes toward immigration, with urban and suburban respondents indicating more support than rural respondents for increasing immigration and less support for decreasing immigration. This result supports the cosmopolitan hypotheses that even when controlling for factors such as education and income, the area in which respondents live will have an important impact on their attitudes. Model 2 adds economic variables, testing how they affect attitudes and how significant this dimension is in relation to other factors. Three types of economic variables are included, measuring personal economic insecurity, perspectives on personal finances and views toward the economy. An interesting pattern emerges through which attitudes are significantly affected by all three types of factors. At the personal level, respondents who completed some high school or graduated from high school were more supportive of change to immigration than university graduates, supporting both increases and decreases to immigration. University graduates were less supportive of decreasing immigration than individuals who had not yet graduated from university. Overall, 87 individuals with higher levels of education were more supportive of keeping immigration at its current level while individuals with less education were more supportive of change in immigration. Support for increasing immigration was much higher among respondents in the labour force than those not in the labour force. Income also had a significant impact on attitudes towards immigration, as individuals with low and medium income levels were more supportive of keeping immigration the same than high income respondents. Model 2 also include questions about respondents' views toward the national economic situation. Respondents who worried about whether they would have enough money to meet the needs of their family "most of the time" were more supportive of decreasing immigration than individuals with less financial worries. To determine how the level of financial worry impacts attitudes I created a scale of payment ability which gave a higher score to individuals who expressed concerns about payments in more than one of the arenas of food, clothing and medical care. Since this scale did not have a significant affect on attitudes I can assume the first measure which asks respondents for their frequency of financial concern has a stronger impact on attitudes than the range of financial worries that they have. Measures of economic perspective were also added in model 2. Views on the national economic situation did not have a significant impact on attitudes, while opinions on economic change did have an impact. Respondents who thought the economy was getting worse (in comparison to those who thought the economy had remained the same) were significantly more supportive of decreasing immigration. Views toward change in 88 the economy more than rating of the economy had a more significant impact on attitudes than rating of the overall economy. Model 3 added political perspective variables to the model, including self- identified political orientation, satisfaction with the situation in the United States, views on affirmative action, and attitudes toward government involvement in minority advocacy. At the personal level, political views had a strong impact on attitudes; conservative and moderate respondents were more likely to support decreasing immigration than liberal respondents. At the contextual level, respondents who were satisfied with the way the United States was going overall were less likely to support decreasing immigration. Respondents who thought that the government should play a minor role in improving the position of blacks and other minority groups were less supportive of increasing immigration. In summary, political orientation had a significant impact on attitudes, with the strongest predictors being political orientation, views toward the U.S. overall and opinions on government involvement in improving the position of minorities. In model 4, measures of interracial contact were added to the model. Having controlled for race as a demographic variable, the results from this model show that the perceived presence of whites, Hispanics, and recent immigrants living in a respondents' community significantly affect their immigration attitudes. Respondents who perceived the level of white residents to be high were more supportive of decreasing immigration. In contrast, respondents living in areas with "many" Hispanic residents were more likely to support increasing immigration. In addition, respondents who lived in areas that they perceived to have "many" or "some" recent immigrants were more likely to support 89 increasing immigration. To summarize, if respondents felt there was a high proportion of whites in their area, they had a lower probability of supporting immigration while respondents who perceived their area to have a high concentration of Hispanics or recent immigrants were more likely to support immigration. Interracial dating was also added to model 4 as a measure of "acquaintance" out-group contact. Thus, the area in which an individual lived has a stronger impact on their attitudes toward immigration than their dating patterns. Engagement in homogenous behavior or the experience of not dating someone of a different racial or ethnic background fostered less support for immigration. The results of table 3.5 showed the consistent impact of the factors I have reported as well as the increased predictive power of the model with the addition of attitudes dimensions. 3.9 Discussion/Conclusion Hostility toward immigration in the United States remains high and most Americans do not want increases in immigration (Lapinski et al., 1997). Traditional measures for determining attitudes toward immigration remain significant in this paper, particularly economic variables. The addition of interracial contact has further illuminated the ways in which the multiethnic American reality affects perceptions toward policy for new respondents. This paper provides additional components necessary for a complete picture of the factors that shape immigration attitudes. By extending the types of measures included in immigration policy attitude models to measure perceived interaction with interracial groups, I highlight the importance of out-group contact on explaining immigration attitudes. Beyond, providing an analysis of individual-level perceptions of immigration 90 policy in the United States, this paper makes two key contributions to the literature. I retest significant attitude dimensions using measures of respondent's social position as well as their perceptions toward their social position and I test the impact that perceived community and interpersonal interracial contact has on attitudes. Within the demographic variables, age, gender, country of birth, parental country of birth, race, religion, marital status, region and area all significantly affected perceptions toward immigration. Overall, demographic characteristics were stronger predictors of negative attitudes toward immigration than positive attitudes. Support for decreasing immigration was strong among older, white, Protestant, male, single respondents as well as individuals born in the United States with parents who were born in the United States. These results reinforce the conclusions from other empirical studies that found the demographic characteristics of individuals significantly affect their attitudes toward immigration (Quillian, 1995; Card, Dustmann, and Preston, 2005; Chandler and Tsai, 2001; Bauer, Loftstrom, and Zimmermann, 2001; Stonewell, 2003). Turning to the economic dimension of attitudes toward immigration, measures of an individual economic context, individual economic security and perspective on the national economy all affected attitudes. The economic situation of individuals, measured here by income, education, and employment status, had a consistently strong impact on attitudes. Highly educated respondents did not want immigration policy to change, while respondents with higher income levels tended to support policy change. Respondents who worried about having enough money for their family were more likely to support decreasing immigration. Unlike previous research, there was no significant relationship between attitudes toward immigration and views toward the national economy, although 91 respondents who believed the economic situation was getting worse were more supportive of decreasing immigration. Thus, while income was a good predictor of positive and negative attitudes toward immigration, other economic measures included were better predictors of negative attitudes toward immigration. The political dimension, as measured by political philosophy and views on minority advocacy, also significantly affected attitudes. Politically conservative respondents were more likely to support decreasing immigration as found in previous research. Satisfaction with the situation in the United States led to less support for decreasing immigration. Supporters of minority advocacy programs were also more supportive of immigration, tying views on immigration to larger issues of minority rights. Previous research found a significant association between views immigrants' use of the welfare system and attitudes towards immigration (Dustmann and Preston, 2002; Chandler and Tsai, 2001), while views on the role of the government in advocating for minorities provide an alternative way of understanding perceptions toward minorities. In this paper, the assumption held that respondents who supported programs for minority groups would be more likely to support immigration. Finally, contact with diverse racial and ethnic groups also had an interesting impact on attitudes. The area in which respondents lived affected both support and hostility toward immigration policy. Respondents who lived in predominately white areas were less receptive toward immigration, while respondents in areas with large Hispanic or recent immigrant populations were more likely to support increasing immigration. The ethnic and racial structure of the area in which respondents lived significantly affected their views, more so than interpersonal relationships. Attitudes are shaped differently, 92 depending on the perceived demographic construction of the area in which respondents lived. This challenges the contact hypotheses that "casual" interaction such as the presence of a particular racial or ethnic group in the community might foster hostility toward other diverse groups (Fezter, 2000) and supports previous research that found distinct attitude trends toward out-groups by race (Oliver and Wong, 2003). At the "acquaintance" level, dating someone of a similar ethnic racial or background led to less support for immigration. In summary, community interracial interactions affect attitudes differently, depending on the race of the dominant group in the community, while no interracial interaction at the interpersonal level leads to greater hostility toward immigration. 93 3.10 Bibliography Allport, G. 1954. The Nature of Prejudice. Reading, MA: Addison & Wesley. Bauer, Thomas K., Magnus Lofstrom, and Klaus F. Zimmermann. 2001. "Immigration Policy, Assimilation of Immigrations, and Native's Sentiments Towards Immigrants: Evidence from 12 OECD Countries." Working Paper 33. Berry, J.W. 1984. "Cultural relations in plural societies: Alternative to segregation and their socio-psychological implications." In N. Miller & M. Brewer (Eds.), Groups in contact. New York: Academic Press. Berry, J.W. 2001. "A Psychology of Immigration." Canadian Journal Behavioral Science 28: 171-179. Brimelow, P. 1995. Alien Nation. New York: Harper Perennial. Brenner, Jan and Michael Fertig. 2006. "Identifying the Determinants of Attitudes towards Immigrants: A Structural Cross-Country Analysis." IZA Discussion Papers. 2306. Bobo, Lawrence. 2000. "Race and Beliefs about Affirmative Action: Assessing the Effects of Interests, Group Threat, Ideology, and Racism." In David 0. Sears, Jim Sidanius and Lawrence Bobo, eds., Racialized Politics: The Debate About Racism in America. Chicago: University of Chicago Press. Borjas, G. J. 1999. "The Economic Analysis of Immigration." Handbook of Labor Economics. Harvard. Borjas, G. J. 1997. "The economic progress of immigrants." unpublished paper. Harvard University. December. Borjas, G. J. 1994. "The economics of immigration." Journal of Economic Literature. 32.4:1667-1717. Borjas, G. J. 1992. "Ethnic capital and intergenerational mobility. Quarterly Journal of Economics. 107.1:123-150. Borjas, G. J. 1991. "Immigration and self-selection." in: John M. Abowd and Richard B. Freeman, eds. Immigration, Trade, and the Labor Market. Chicago: University of Chicago Press: 29-76. Borjas, G. J., R. B. Freeman, and L. F. Katz.1997. "How much do immigration and trade affect labor market outcomes." Brookings Papers on Economic Activity1:1-67. Burns, P., and J.G. Gimpel. 2000. "Economic Insecurity, Prejudicial Stereotypes, and Public Opinion on Immigration Policy. " Political Science Quarterly 00323195 Summer 115. 2. Buck, Beverly, Melanie Hobbs, Anne Kaiser, Susan Lang, Darrel Montero, Kathy Romines, and Tonia Scott. 2003. "Immigration and Immigrants: Trends in American Public Opinion, 1964-1999." Journal of Ethnic & Cultural Diversity in Social Work 12.3: 73-90. Card, David, Christian Dustmann and Ian Preston. 2005. "Understanding attitudes to immigration: the migration and minority module of the first European Social Survey." CReAM Discussion Paper. 03.05: 1-45. Chandler, Charles R. and Yung-mei Tsai. 2001. "Social Factors Influencing Immigration Attitudes: An Analysis of Data from the General Social Survey." Social Science Journal 28: 177-88. Citrin, Jack , Donald P. Green, Christopher Muste, and Cara Wong. 1997. "Public 94 Opinion Toward Immigration Reform: The Role of Economic Motivations." The Journal of Politics 59.3: 858-881. Citrin, Jack and Donald P. Green. 1990. "The Self-Interest of Motive in American Public Opinion." In Research in Micropoltics. 3 (eds.) Samual Long. Greenwich, CT: JAI Press. Crawley, Heaven. 2005. "Evidence on Attitudes to Asylum and Immigration: What We Know, Don't Know and Need to Know." Centre for Migration, Policy and Society. Work Paper No. 23. University of Oxford. WP-05-23. Day, C. L. 1990. "Ethnocentrism, economic competition, and attitudes toward U.S. Immigration Policy." Paper presented at the Annual Meeting of the Midwest Political Science Association. Chicago, Illinois. April 5-7. Daniels, Joseph Patrick, and Marc von der Ruhr. 2005. "God and the global economy: attitudes towards trade and immigration in the United States." Socio-Economic Review 3:467-489. Dimond, Jeff. 1998. "African-American Attitudes towards United States Immigration Policy." International Migration Review 32.2: 451-62. Dion, Kenneth L., and Kerry Kawakami. 1996. "Ethnicity and perceived discrimination in Toronto: Another look at the personal/group discrimination discrepancy." Canadian Journal of Behavioral Science 28.3: 203-213. Dustmann, Christian and Ian Preston. 2002. "Racial and Economic Factors in Attitudes to Immigration." University College London and Institute for Fiscal Studies. Dustmann, Christian and Ian Preston. 2004a. "Racial and Economic Factors in Attitudes to Immigration." CReAM Discussion Paper No. 01/04. Dustmann, Christian, F. Fabbri and Ian Preston. 2004. "Racial Harassment, Ethnic Concentration and Economic Conditions." CReAM Discussion Paper. No. 05/04. Dustmarm, Christian and Ian Preston. 2004b. "Is Immigration Good or Bad for the Economy? Analysis of Attitudinal Responses." CReAM Discussion Paper No. 06/04. Dustmann, Christian. and Ian Preston. 2005. Is Immigration Good or Bad for the Economy? Analysis of Attitudinal Responses. forthcoming. Research in Labour Economics. CReAM Discussion Paper No.06/04. Espenshade, Thomas J. and Katherine Hempstead. 1996. "Contemporary American attitudes toward U.S. immigration." International Migration Review 30.2: 535- 570. Esses, Victoria M., John Dovidio, Lynne M. Jackson, and Tamara Armstrong. 2001. "The Immigration Dilemma: The Role of Perceived Group Competition, Ethnic Prejudice, and National Identity." Journal of Social Issues 57.3: 389-412. Facchini, Giovanni and Anna Maria Mayda. 2006. "Individual Attitudes toward Immigrants: Welfare-State Determinants Across Countries." CReAM Discussion Paper. 04/06. Fertig, Michael and Christoph M. Schmidt. 2002. "The Perceptions of Foreigners and Jews in Germany - A Structural Analysis of a Large Opinion Survey." IZA Discussion Papers. 431. 95 Fertig, Michael, and Jan Brenner. 2006. "Identifying the Determinants of Attitudes toward Immigrants - A Sturctural Cross-Country Analysis." RWI Discussion Papers 0047. Rheinisch-Westfalisches Institut fUr Wirtschaftsforschung. Fetzer, Joel S. 2000. Public Attitudes Towards Immigration in the United States, France, and Germany. New York: Cambridge University Press. Gang, Ira, Francisco Rivera-Batiz, and Myeong-Su Yun. 2002. "Economic strain, ethnic concentration and attitudes towards foreigners in the European Union." Rutgers University Department of Economic Working Papers. 14. Hainmueller, Jens and Michael J. Hiscox. 2007. "Educated Preferences: Explaining Attitudes Toward Immigration In Europe." International Organization Winter 61.2. Haubert, Jeannie and Elizabeth Fussell. 2006. "Explaining Pro-Immigrant Sentiment in the U.S.: Social Class, Cosmopolitanism, and Perceptions of Immigrants." International Migration Review Fall. 40.3: 489-507. Hernes, Gudmund and Knud Knudsen. 1992. "Norwegians' Attitudes Toward New Immigrants." Acta Sociologica. 35: 123-139. Hood III, M.V. and Irwin L. Morris. 1997. "Amigo o Enemigo?: Context, Attitudes, and Anglo Public Opinon toward Immigration. Social Science Quarterly June 78.2: 309-323. Lapinski, J. S., P. Peltola, G. Shaw, and A. Yang. 1997. "The Polls-Trends: Immigrants and Immigration." Public Opinion Quarterly 61:356-383. Mayda, Anna Maria. 2004. "Who Is Against Immigration? A Cross-Country Investigation of Individual Attitudes towards Immigration." IZA Discussion Paper No. 1115. Olzak, S. 1992. The Dynamic of Ethnic Competition and Conflict. Stanford, CA: Stanford University Press. O'Rourke, K. H. and R. Sinnott. 2004. "The Determinants of Individual Attitudes Towards Immigration." Economic Papers. Trinity College Dublin: Department of Economics. Palmer, Douglas L. 1996. "Determinants of Canadian attitudes towards immigration: more than just racism." Canadian Journal of Behavioral Science 28:180-192. Quillian, Lincoln 1995. "Prejudice as a Response to Perceived Group Threat: Population Composition and Anti-Immigrant and Racial Prejudice in Europe." American Sociological Review 60: 586-611. Sanoussi, B. et al. 1998. "Determinants of attitudes towards immigration: a trade theoretic approach." Centre for Economic Policy Research Discussion Paper No. 1877, Centre for Economic Policy Research. Scheve, Kenneth F. and Matthew J. Slaughter. 2001. "Labor market competition and individual preferences over immigration policy." The Review of Economics and Statistics February 83.1: 133-145. Scheve, Kenneth F., and Matthew J. Slaughter. 2001. "What Determines Individual Trade-Policy Preferences." Journal of International Economics. Schmid, Carl. 2003. "Immigration and Asian and Hispanic Minorities in the New South: An Exploration of History, Attitudes, and Demographic Trends." Sociological Spectrum 23: 129-157. Semyonov, M., R. Raijman, A. Gorodzeisky. 2006. "The Rise of Anti- 96 foreigner Sentiment in European Societies, 1988-2000. "American Sociological Review Albany. June. 71. 3: 426-460. Simon, Rita J. and James P. Lynch. 1999. "A Comparative Assessment of Public Opinion Towards Immigrants and Immigration Policies." International Migration Review 33.2: 455-467. Stonewall. 2003. Profiles of Prejudice — The Nature of Prejudice in England: In-depth Findings, London: Stonewall. Tucci, Ingrid. 2005. "Explaining Attitudes Towards Immigration: New Pieces to the Puzzle "Berlin. German Institute for Economic Research Discussion Paper 484. US Census Bureau. 2000. United States Census 2000. http://vvww.census.gov/ Valentine, Gill, and Ian MacDonald. 2004. Understanding prejudice. London: Stonewall. Zaller, J. 1992. The Nature and Origin of Mass Opinion. New York: Cambridge. 97 Appendices 1^Note that the racial categories as they are identified in this paper are not the classification of this researcher but are the labels as they are presented in the survey. The terms have been kept in their original form in order to better represent the options available for selection by respondents. There are other methods of racial and ethnic categorization as well as other terms for classifying these various groups. This paper does not intend to promote the use of a particular racial term but instead attempts to best represent the data as close as possible to the original terminology. 2^a) Demographic Sex (Female, Male) Race (White, Black, Other) Religion (Protestant, Catholic, Other/None) Marital Status (Married/Living, Single, Divorced/Separated/Widowed) Region (East, Midwest, South, West) Area (Urban, Suburban, Rural) 3^b) Economic Education (some high school, high school graduate, some college, university graduate), Employment Status (in the labour force, not in the labour force) Income (low, medium, high) Economic conditions are getting better or worse (getting better, getting worse, same), How often do you worry that your family income will not be enough? (all of the time, some of the time, almost never) Last year did not have enough money for food (yes, no), Last year did not have enough money for clothing (yes, no), Last year did not have enough money for medical bills (yes, no). 4^c) Political Political philosophy (conservative, moderate, liberal) Satisfaction with the way things are going in the United States (excellent, good, only fair, poor) Favour Affirmative Action Programs for Racial Minorities (Favour/Don't Favour), Government role in improving the position of blacks and minorities (Major role/Minor role/No role) The measure included in this paper is a crude description of political orientation (dividing respondents into conservative, moderate, and liberal); more detailed measures have been used in other contexts, and have found results to vary based on the structure of the political orientation variable, which might lead to different results (Golder, 2003; Kitschelt, 1997; Carter, 2005). 5^d) Contact 'How many Whites live in your area?' (many, some, few) 98 'How many Blacks live in your area?' (many, some, few) 'How many Hispanics live in your area?' (many, some, few) 'How many recent immigrants live in your area?' (many, some, few) Dated someone of a different racial/ethnic background (yes, no) Dated someone of the same racial/ethnic background (yes, no) 99 Table 3.1: Percentage of each group who want given levels of immigration in the United States, 2005 Less Same More N 48.1 35.3 16.6 2184 Gender Male^ 51.2 30.7 18.1 1064 Female 45.4 39.3 15.3 1200 Country of Birth Born in the United States and both parents born in the United States^52.1 33.8 14.1 1724 Born in the United States and one parent born outside the United States 37.5 37.2 25.3 280 Born outside the United States^ 30.3 44.9 24.8 247 Race White^ 50.4 34.6 15.0 1797 Black 46.4 33.5 20.2 247 Other 32.6 39.5 27.9 133 Religion Protestant^ 53.4 34.8 11.8 1074 Catholic 46.3 33.5 20.2 574 Other/None 40.3 38.2 21.5 585 Martial Status Married/Living As Married^ 46.2 36.5 17.3 1461 Divorced/Seperated/Widowed 56.5 29.0 14.5 397 Never Married^ 45.5 38.0 16.5 390 Region East^ 40.0 42.5 17.5 471 Midwest 51.0 30.4 18.5 535 South 51.4 36.1 12.6 722 West^ 47.7 32.7 19.6 536 Area Urban 44.9 36.9 18.2 746 Suburban^ 45.2 34.6 20.2 1088 Rural 61.3 33.9 4.8 430 100 Table 3.2: Percentage of each group who want given levels of immigration in the United States, 200 Less Same More Education Less than High School 53.3 28.3 18.4 899 High School Graduate 52.3 35.2 12.5 740 Some College 48.7 32.5 18.9 268 College Graduate 25.5 55.1 19.4 255 Employment Status Full time 47.6 33.5 18.9 1169 Part time/Not in the labour 48.8 36.9 14.3 1084 Income Less than $20k 45.9 40.1 14.0 298 $20K-Less than $30K 50.2 34.4 15.4 296 $30K-Less than $50K 50.8 31.9 17.4 619 $50K-Less than $75K 43.6 41.9 14.4 297 $75K and above 49.3 29.3 21.4 587 How often do you worry that you will not have enough money for your family All of the time 50.1 28.7 21.2 406 Most of the time 54.9 29.4 15.7 254 Some of the time 48.6 35.5 15.9 974 Almost never 43.8 41.2 15.0 625 Last year did you not have enough money for food? Yes 43.2 43.8 22.0 370 No 49.1 35.3 15.5 1893 Last year did you not have enough money for medical bills? Yes 45.1 34.1 20.8 647 No 49.0 35.9 15.1 1603 Last year did you not have enough money for clothes? Yes 46.8 33.4 19.8 530 No 48.4 35.9 15.7 1,725 Economic Conditions in the United States Excellent/Good 44.1 39.8 16.0 805 Only Fair 48.6 34.1 17.3 1012 Poor 54.4 29.3 16.3 437 Change in the Economic Conditions Getting Worse 53.9 30.1 16.0 1,267 The same/Getting better 39.9 42.4 17.7 952 101 Table 3.3: Percentage of each group who want given levels of immigration in the United States, 2005 Less Same More Political Orientation Conservative 57.0 31.0 11.7 859 Moderate 45.0 39.2 15.8 858 Liberal 37.8 38.0 24.2 476 Satisfied with the way things are going in the United States Satisfied 40.4 43.5 16.1 823 Dissatisfied 52.4 30.1 17.5 1375 Favor affirmative action for racial minorities Favor 39.3 39.5 21.2 1133 Oppose 56.7 30.8 12.5 957 Government role in imporving the situation of minorities Major role 35.6 40.8 23.7 828 Minor role 54.5 35.2 10.4 1000 No role at all 56.3 25.8 17.9 395 Table 3.4: Percentage of each group who want given levels of immigration in the United States, 2005 Less^Same^More How many Whites live in your area? Many 47.1 35.7 17.2 1718 Some 48.6 34.8 16.7 287 Few/None 53.6 34.5 11.8 242 How many Blacks live in your area? Many 47.7 35.2 17.1 745 Some 49.3 36.5 14.1 704 Few/None 47.7 34.1 18.2 811 How many Hispanics live in your area? Many 42.3 37.9 19.8 779 Some 51.5 29.4 19.1 595 Few/None 49.6 37.7 12.7 846 How many Asians live in your area? Many 44.2 31.5 24.2 268 Some 46.7 34.9 18.4 580 Few/None 49.8 36.0 14.1 1368 How many recent Immigrants live in your area? Many 41.6 31.8 26.5 370 Some 47.0 32.3 20.7 588 Few/None 51.9 35.8 12.3 1178 Dated Someone of Different Racial/Ethnic Background Yes, have 46.2 35.5 18.3 1093 No, have not 50.0 35.0 15.0 1171 Dated Someone of the Same Racial/Ethnic Background Yes, have 48.5 35.6 15.9 2082 No, have not 43.8 31.3 25.0 182 102 Table 3.5: Multinomial Logit Model Estimates of Immigration Attitudes (versus 'stay the same') Model 1 Decrease Increase Model 2 Decrease Increase Model 3 Decrease Increase Model 4 Decrease Intercept -0.750 * -1.967 *** -1.471^**" -1.849 *** -0.946 -2.274 ** -1.543 Age -0.010 ** -0.014 ** -0.010 ** -0.007 -0.013 ** -0.001 -0.010 Last year did you have enough money for (food, clothing, medical bills) -0.086 0.012 -0.020 0.071 -0.040 Sex (Female) Male 0.538 *** 0.421 ** 0.657 **" 0.357 * 0.502 *** 0.312 0.482 *** Country of Birth/Parents country of birth (Foreign-born) U.S. born, U.S. born parents 0.874 *** 0.048 0.734 **" 0.197 0.734 ** 0.252 0.892 *** U.S. born, one parent foreign-born 0.705 ** 0,460 0.710 ** 0.710 * 0.525 0.822 * 0.621 Race (Other) White 0.369 -0.171 0.800 ** -0.331 0.696 * -0.495 1.079 ** Black 0.364 0.136 0.628 -0.129 0.616 -0.328 0.630 Religion (Other/None) Protestant 0.293 * -0.329 * 0.437 ** -0.312 0.520 *** -0.098 0.433 * Catholic 0.320 * 0.067 0.525 ** 0.254 0.414 * 0.585 ** 0.320 Marital Status (Divorced/Seperated/Widowed) Married/Living as Married 0.117 0.433 * -0.048 0.128 -0.048 0.442 -0.092 Single 0.849 *** 0.687 * 0.891^**" 0.638 * 0.864 *** 0.687 0.836 ** Area (rural) Urban -0.415 ** 1.462 *** -0.564 **" 1.321 *** -0.537 ** 1.244 ** -0.649 *** Suburban -0.380 ** 1.601 *** -0.298 * 1.707 *** -0.295 1.642 ** -0.301 Education (University graduate) Some high school 1.614^**" 1.456 *** 1.389 *** 1.969 **' 1.905 *** High school graduate 0,990 **" 0.566 ** 0.628 *** 0.710 ** 0.702 *** Some university 0.835 **" 0.005 0.770 *** 0.350 0.707 **" Employment Status (Not in Labour Force) In Labour Force -0.217 0.146 -0.235 0.389 * 0.058 Income ($75K and above) Less than $20k -1.461^**" -1.375 *** -1.592 *** -1.884 **' -1.622^*** $20K-Less than $30K -1.175^**" -1.149 *** -1.085 *** -1.558 **' -1.089 *** $30K-Less than $50K -0.648 **" -0.578 ** -0.768 *** -0.713 ** -0,633 a* $50K-Less than $75K -0.629 **" -0.598 * -0.764 *** -1.283 **' -0.674 ** Worry about family income (Almost never) All of the time 0.234 0.427 0.216 0.484 0.349 (Almost never) Most of the time 0.687 ** 0.435 0.862 *** 0,409 0.864 *** Some of the time 0.151 -0.059 0.183 -0.104 0.310 Rating of Economic Conditions (Poor) Excellent/Good -0.023 -0.241 0.027 -0.092 0.117 Only Fair -0.115 -0.187 -0.111 -0.120 -0.140 Economic conditions (Getting better/the same) Getting worse 0.730 *"" 0.166 0.661^*** 0.208 0.760 *** Political Philosophy (Liberal) Conservative 1.151^*** -0.329 1.221^*** Moderate 0.266 -0.205 0.282 Satisfaction with the US (Not Satisfied) Satisfied -0.942 *** -0.105 -0.941^*** Favour Affirmative Action Programs for Minorities (Favour) Don't Favour -0.196 0.407 * -0.212 Govemment role in minority situation (Na Role) Major role -0.441^* -0.152 -0.412 Minor role -0.078 -0.676 ** -0.011 How many Whites live in your area (Few) Many -0.186 Some 0.230 How many Blacks live in your area (Few) Many 0.403 • Some -0.185 How many Hispanics live in your area (Few) Many -0.022 Some 0.446 * How many Asians live in your area (Few) Many 0.180 Some M.247 How many recent immigrants live in your area {Few) Many 0.034 Some 0.)75 Dated someone of the same racial/ethnic background (Na) Yes -0.370 NagelkerkeV2 0.12 0.22 0.33 0.39 Chi-Square 217.50 *** 397.96 *** 566.45 *** 662.23 *** 103 4.1 Conclusion Many recent public opinion reports have described attitudes toward immigration though few have provided detailed evidence on factors that impact positive and negative views toward immigration cross-sectionally and cross-nationally. The aim of this thesis is to provide further insight into the complexity involved in understanding attitudes toward immigration policy by analyzing the determinants of attitudes cross-nationally in Canada, the United States, and the United Kingdom and a cross-sectionally in the United States. Using existing Gallup 2005 survey data, this thesis furthers knowledge on public understanding of this politically contentious topic. Beyond reporting the descriptive results, these two papers provides specifics on how attitudes toward immigration are impacted by different types of individual and contextual factors. To conclude, the results from each paper are summarized, followed by the overall concluding points and research opportunities which emerge from this work. 4.2 Conclusions from Chapter 2 This paper focused on determining the impact of demographic and economic individual-level characteristics on attitudes toward immigration in Canada, the United States, and the United Kingdom. Support and hostility toward immigration was compared, controlling for country as a contextual variable. Age, gender, race, income and country were found to have a significant impact on attitudes toward immigration. Particular variables included in the model did not have a consistent impact on attitudes. When country was added as a variable in model 4 it highlighted the fact that country matters a great deal in understanding attitudes and the influences are attitudes are distinct 104 between countries. The large pseudo-R2 in model 4 provides additional support for the impact of country on attitudes as the values are significantly lower in models 1-3. Building on previous research comparing attitudes cross-nationally, this paper provides additional detail into the impact of demographic and economic characteristics on attitudes and how distinct this impact can be cross-nationally. These findings prompt a number of important observations. First, Canadians are overall more supportive of immigration policy than American or British respondents. Economic measures such as income highlight distinct ways attitudes are influenced within the three countries included in the paper. The results of this paper suggest that the trends of increasing anti- immigration sentiment in the United States and the United Kingdom (Adams, 2003; Grabb and Curtis, 2005; Crawley, 2005), while hostilities in Canada are moderate by comparison. 4.3 Conclusion from Chapter 3 The second paper included in this thesis focused on analyzing multiple dimensions of attitudes toward immigration in the United States. This work tested the impact of demographic, economic, political and contact variables in multinomial logistic regression models on positive and negative attitudes toward immigration. Of the variables included in the analysis age, gender, foreign-born status, parents foreign-born status, region, area, education, income, economic insecurity, political ideology, satisfaction with the way the United States is going, interaction with diverse groups and interracial dating significantly impact of attitudes toward immigration. Overall, measures in each of these dimensions had a significant impact on attitudes toward immigration, emphasizing the range of facets involved in the process of shaping public opinion. The results of this 105 paper suggest that perceptions toward immigration are shaped both by multiple types of factors which impact positive and negative attitudes in different ways. This paper also includes measures of economic and political dimensions of attitudes at the personal and contextual level tracking the variation in attitudes between these distinct ways of framing questions of public perception (Hernes and Knudsen, 1992). Building on previous research this paper replicates and furthers hypotheses on the forces that influence American public opinion (Chandler and Tsai 2001; Citrin et al., 1997; Espenshade and Hempstead 1996; Scheve and Slaugher 2001). 4.4 Limitations/Further avenues of Research While these papers have made significant contributions to expanding our understanding of attitudes toward immigration, this research project faced limitations as well as uncovering opportunities for further investigation. The use of existing survey data limited analysis opportunities to the data and measures available in the datasets selected the major limitations of this particular research project centre around nature of the data analyzed, which created different challenges in the process of answering the research questions of each paper. In the case of Chapter 2, the data constrained both what countries could be included in the analysis and what measures could be compared. To conduct this analysis cross-nationally uniform questions and response categories were required in each century. For example education was not included in the data as a measure of all three courses a restriction which meant that either the country that did not have education as a variable had to be excluded or the variable of education had to be excluded or the variable of education had to be excluded from the cross-national analysis. Challenges also arose in the process of comparing measures such as race and income 106 cross-nationally. The distribution of race and income in each country was different, limiting the level at which differences could be contrasted (Statistics Canada, 2001; US Census Bureau, 2000; UK Office of National Statistics, 2001). By merging the response categories for income and race into broader easier to compare options, some of the nuances imbedded in these response options were lost. Although even while limited by the lack of uniform data important and interesting results were still produced. The analysis in Chapter 3 was also constrained by the data available in the single survey analyzed, although there was more measure included in this data set than in the combined data found in Chapter 2. The range of measures included in the dataset used in Chapter 3 required that the data be analyzed so that only the most useful data was included. Additional measures of occupation, professional security and industry sector might have provided further insight into how more nuanced distinctions in labour market position could have shaped attitudes (Dustmann and Preston, 2002). Further Research While the analysis reported here raises a range of research questions there are three key avenues for further investigation that build directly from this work. Additional research could include expanding the number of countries compared, including more measures to understand the ways in which different individual-level dimensions impact attitudes, and compare the results reported here to similar models available in other data sets. The first way to expand the work done in this thesis would be to add more countries to the cross-national comparison. The purposed distinction between liberal democratic and European sending societies could be further explored through the addition of more countries from these two groups (Freeman, 1995). Australia and New Zealand share 107 common characteristics with the United States and Canada as liberal democratic receiving societies and former British colonies while other Western European states such as Germany, France, Switzerland, the Netherlands, Sweden, and Belgium might add depth to the profile of European sender societies in addition to Britain. Contrast could also be made between ex-colonial European states with strong temporary worker policies and Southern European states with historical moments of labour surpluses and mass emigration (Paul, 1997). Beyond these attitude comparisons between European sender societies and Western receiving society's further work could compare attitudes toward immigration between countries that have experienced major out-migration at different points in their history, such as Mexico and China exploring how public perceptions might impact the rate at which individuals leave the country. Continental differences may also exist between 'sender' countries in Asia and Europe which may require further exploration as well as countries that have only recently received immigration. Beyond including more countries in the analysis of attitudes the inclusion of additional measures could also provide more insight into exactly how attitudes toward immigration are shaped. Further analysis might find more measure of views on immigration as well as additional variables that might impact attitudes. Views on immigration can be measured in a variety of ways beyond questions of policy levels. Attitudes may differ if respondents are asked their views on refugees and asylum seekers, immigrants from different countries of origin, estimates of immigration levels, measures of knowledge of immigration, perceptions toward immigrant's participation in society as well as their potential engagement in illegal activity (Crawley, 2005). While attitudes may vary based on the type of question asked, they may also vary by the type of 108 immigration dimensions measured. Questions that focus on different ways of framing contextual and individual factors may also shape attitudes (Hernes and Knudsen, 1992). Additional measures can provide insight into the impact of socio-demographic, factors and political views. Finally, further research could replicate the methodology used in this thesis with different data sources. Research institutions and polling organizations have collected data on views toward immigration in each country independently as well as other available cross-national datasets. In Canada, a number of organizations have conducted social surveys with questions on concern for immigration. The General Social Survey has been (GSS) conducted annually in Canada since 1985 by Statistics Canada and is publicly available. While questions of immigration have not always been directly measured, the data provides a barometer of change in social values. The World Values Survey collected data on attitudes toward immigration in a variety of countries in several waves of data beginning in the 1980s. The World Values Surveys includes a range of measures related to immigration capturing views on a variety of social issues measures views on immigration as well as a variety of social values. Included in these data sets are measures of respondent's views on immigrant participation in their neighbourhoods and participation in the labour market. Versions of these surveys were fielded in the United States and the United Kingdom as well, providing opportunities for confirmation of the results found in this work. Public opinion polls such as the ones analyzed in this thesis have been fielded by such organizations as Gallup, Environics and Angus Reid and if available could provide further avenues for analysis. Other public opinion research includes the Listening to Canadians surveys conducted nationwide in Canada three times 109 a year, measuring views on a variety of public policy issues. In the United States there are also a range of organizations conducting social surveys on public policy and immigration. The National Opinion Research Centre (NORC), based out of the University of Chicago has administered 25 waves of the General Social Survey since 1972 providing measures of trends in overall social values. As well a variety of polling agencies including Gallup research public opinion on immigration. 4.5 Overall Conclusions This thesis has worked to further rearticulate and unpack public views toward immigration policy in order to underscore the need for understanding into what shapes attitudes. Through mapping how these attitudes are shaped, insight can be gained into the process of fostering tolerance. Researchers, policy makers and the public should continue to recognize and discourage the ill treatment of immigrants and all outsiders. This thesis provides further tools for understanding how intolerance develops and how these patterns discouraged. It is encouraging that all three of these nations are striving to foster tolerance and support for diverse communities such as immigrants through their legal policies, educational practices, and on-going research (Li, 2003; Crawley, 2005; Meissner et al., 1997). The reality of superstition, lack of information and violent crime between visibly and ethnically different emphasis the need for expert insight into policy development. Clearly the construction of a positive national immigration policy experience requires a policy agenda which includes addressing the disadvantaged experiences of immigrants as well as the interactions of native-born residents with immigrant groups. 110 The results from this thesis are relevant to questions regarding the extent to which distinct patterns of support for immigration exist in each of these three countries. In addition, this work highlights how we should challenge ourselves to understand the dynamics that affect attitudes cross-nationally. More work is needed to understand how these results compare to other major immigrant receiving countries especially to understand how the impact of country varies. While the future cannot be predicted, the current levels of diversity experienced in each of these three countries could not necessarily have been anticipated. This lived diversity challenges receiving societies to work toward better understanding of the ways in which safe and stable communities can be fostered. As the range of countries from which Canada and the United States accept immigrants expands constructing a clear national identity becomes more challenging (Castles, 2000). Although, the United Kingdom and other countries that have long focused their national identity as sender societies with uniform ethnic and racial constructions must find ways to shape the view of the nation which embraces the diverse reality. Since the future of immigration is only expected to become more present and diverse, it is important for researchers to understand how to ease societies through the growing pains of diversification. 111 4.6 Bibliography Adams, Michael. 2003. Fire and Ice: The United States, Canada, and the Myth of Converging Values. Toronto: Penguin Books. Castles, Stephen. 2000. Ethnicity and Globalization. London: SAGE Publications. Chandler, Charles R. and Hung-mai Tsai. 2001. "Social Factors influencing immigration attitudes: An analysis of data from the General Social Survey." The Social Science Journal 38: 277-288. Citrin, Green, Muste and Wong. 1997. "Public Opinion Toward Immigration Reform: The Role of Economic Motivations." The Journal of Politics 59: 858- 881. Crawley, Heaven. 2005. "Evidence on Attitudes to Asylum and Immigration: What We Know, Don't Know and Need to Know." Centre for Migration, Policy and Society. Work Paper No. 23. University of Oxford. WP-05-23. Dustmann, Christian and Preston, Ian. 2002. "Racial and Economic Factors in Attitudes to Immigration." University College London and Institute for Fiscal Studies. Espenshade, Thomas J. and Katherine Hempstead. 1996. "Contemporary American attitudes toward U.S. immigration." International Migration Review 30.2: 535- 570. Freeman, Gary. 1995. Modes of immigration politics in liberal democratic states. International Migration Review. 19.4: 881-902. Grabb, Edward and James Curtis. 2005. Regions Apart: The Four Societies of Canada and the United States. Oxford University Press. Hernes, Gudmund and Knud Knudsen. 1992. Norwegians' Attitudes Toward Li, Peter. 2003. Destination Canada: Immigration Debates and Issues Toronto: Oxford University Press. Lynch, James P. and Rita Simon. 2003. Immigration the World Over: Statutes, Policies, and Practices. Lanham: Rowan & Littlefield. Meissner, Doris M., Robert D Hormats, Antonio Garrigues Walker, and Shijuro Ogata. 1993. International Migration Challenges in a New Era. New York: Trilateral Paul, Kathleen. 1997. Whitewashing Britain: Race and Citizenship in the Postwar era. Cornell University Press. Scheve, Kenneth F. and Matthew J. Slaughter. 2001. "Labor market competition and individual preferences over immigration policy." The Review of Economics and Statistics February 83.1: 133-145. Scheve, Kenneth F., and Matthew J. Slaughter. 2001. "What Determines Individual Trade-Policy Preferences." Journal of International Economics. Statistics Canada. 2001. Census of Canada. http://www12.statcan.ca/english/census/Index.cfm US Census Bureau. 2000. United States Census 2000. http://www.census.gov/ UK Office of National Statistics. 2001. National Statistics. www.statistics.gov.uk . Crown Copyright. 112

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0066230/manifest

Comment

Related Items