The B.E.Journal of Economic Analysis Policy,Vol.7 [2007],Iss.I (Advances),Art.62 Table 1:Selected Characteristics of Natives and Foreign-Born Variable Natives Foreign West Asia European Aust./ Born Indies N.Z. of Tot.Imm.Pop. 5.9 20.89 33.55 2.48 Years FT Education 11.24 11.91 11.19 10.89 11.34 13.48 High Education 13.12 19.67 13.81 14.43 15.60 33.87 Inter.Education 47.55 52.64 34.94 33.00 39.75 54.04 Low Education 39.32 39.12 51.26 52.58 44.64 12.09 Age 36.46 40.31 44.79 38.15 44.72 31.96 Claiming UB 7.70 8.61 14.12 12.22 7.57 4.46 a:Data:British Labour Force Survey (LFS),1983-1990.Regions definition:Asia: India,Pakistan,Bangladesh,Asian Ugandans;Europe:EU (old definition,before 1991);Australasia:New Zealand and Australia. Education:High education:university degree,higher education;Intermediate edu- cation:A-levels and O-levels and equivalent or other qualification;Low education: no qualification. the overall immigrant population in Britain.In table 1 we display some of the key characteristics of these groups,where numbers are drawn from the British Labour Force Survey (LFS).Most noteworthy is the educational composition of the immigrant population.Unlike the US and many European countries. immigration to the UK is,and has always been,relatively high skilled,with im- migrants having on average higher levels of schooling than native born whites.10 In the table,we report two measures of education:the years of full-time school- ing,and three levels of education.On average,immigrants have slightly more years of education than the native born,but there is variation across the three groups.The distribution of educational achievement is likewise similar be- tween the groups,with immigrants being 6 percentage points more likely to hold a degree or have higher education,while about equally likely to have no qualification at all.Immigrants from Australia/New Zealand have the highest level of education,with on average more than 2 years more full-time educa- tion than white natives,and with 34%obtaining university or post-secondary higher education,compared with only 13%among natives.Asians and immi- grants from the West Indies are similar to natives at the high end of the skill distribution,but have a substantially higher fraction of individuals with low educational backgrounds.Overall,Australians/New Zealanders appear to be 10Although the table considers only subgroups of immigrants,the same is true for Britain's overall immigrant population,see Dustmann and Fabbri(2006)for details. http://www.bepress.com/bejeap/vol7/iss1/art62 10
Table 1: Selected Characteristics of Natives and Foreign-Born Variable Natives Foreign West Asia European Aust./ Born Indies N.Z. % of Tot. Imm. Pop. 5.9 20.89 33.55 2.48 Years FT Education 11.24 11.91 11.19 10.89 11.34 13.48 High Education 13.12 19.67 13.81 14.43 15.60 33.87 Inter. Education 47.55 52.64 34.94 33.00 39.75 54.04 Low Education 39.32 39.12 51.26 52.58 44.64 12.09 Age 36.46 40.31 44.79 38.15 44.72 31.96 % Claiming UB 7.70 8.61 14.12 12.22 7.57 4.46 a : Data: British Labour Force Survey (LFS), 1983-1990. Regions deÖnition: Asia: India, Pakistan, Bangladesh, Asian Ugandans; Europe: EU (old deÖnition, before 1991); Australasia: New Zealand and Australia. Education: High education: university degree, higher education; Intermediate education: A- levels and O-levels and equivalent or other qualiÖcation; Low education: no qualiÖcation. the overall immigrant population in Britain. In table 1 we display some of the key characteristics of these groups, where numbers are drawn from the British Labour Force Survey (LFS). Most noteworthy is the educational composition of the immigrant population. Unlike the US and many European countries, immigration to the UK is, and has always been, relatively high skilled, with immigrants having on average higher levels of schooling than native born whites.10 In the table, we report two measures of education: the years of full-time schooling, and three levels of education. On average, immigrants have slightly more years of education than the native born, but there is variation across the three groups. The distribution of educational achievement is likewise similar between the groups, with immigrants being 6 percentage points more likely to hold a degree or have higher education, while about equally likely to have no qualiÖcation at all. Immigrants from Australia/New Zealand have the highest level of education, with on average more than 2 years more full-time education than white natives, and with 34% obtaining university or post-secondary higher education, compared with only 13% among natives. Asians and immigrants from the West Indies are similar to natives at the high end of the skill distribution, but have a substantially higher fraction of individuals with low educational backgrounds. Overall, Australians/New Zealanders appear to be 10Although the table considers only subgroups of immigrants, the same is true for Britainís overall immigrant population, see Dustmann and Fabbri (2006) for details. 10 The B.E. Journal of Economic Analysis & Policy, Vol. 7 [2007], Iss. 1 (Advances), Art. 62 http://www.bepress.com/bejeap/vol7/iss1/art62
Dustmann and Preston:Racial and Economic Factors in Attitudes to Immigration the highest educated of the four groups,followed by the Europeans.Based on these numbers,we should expect that labour market competition should be a concern for the highly skilled when assessing immigration from New Zealand and Australia,and for the low skilled for immigration from the West Indies and Asia. In the last row,we display the percentage in each of the groups that claims unemployment benefit.Here immigrants are overall more likely to be claimants than natives,but there is considerable variation across groups,with the West Indian and Asian immigrants having probabilities to claim unemployment ben- efits that are twice as high as those of white natives,while immigrants from Europe are similar,and immigrants from Australia and New Zealand substan- tially less likely.If potential new immigrants are similar in this respect to those already present,then those with the highest expectation of bearing the burden from increased welfare costs may be particularly concerned about immigration from Asia and the West Indies.The incidence of the fiscal consequences of higher welfare dependency will depend on whether the greater cost is met by raising taxes or by cutting public expenditure and on how exactly this is done. In principle,the cost could fall on the rich or poor,but if the increase is in the most progressive aspects of the tax system,then it will be those on high incomes who suffer most (Fetzer 2000,Dustmann and Preston 2005,Facchini and Mayda 2006). 3.2 The British Social Attitude Survey Our attitudinal data is drawn from seven years of the British Social Attitudes Survey (BSA)(1983,1984,1985,1986,1987,1989,1990).This gives us an eight year period over which key questions were repeatedly asked in a form which remained unchanged.Later years of the BSA ask questions on immigration less regularly,and in a form which is not consistent from year to year.We use the data for England and concentrate on white respondents only.11 There are three sets of variables which we use for our analysis.First,questions on attitudes to further immigration.Second,questions that are related to the three underlying channels by which immigration preferences may be affected. And third,questions that reflect heterogeneity across individuals in preferences due to individual and context characteristics.We describe these variables briefly below. 11Racial identity is self-assessed.Attitudes of ethnic minority individuals towards their own communities,or towards other ethnic minorities,are likely to be driven by different mechanisms.While it might be interesting to investigate their attitudes,the sample sizes within the BSA become very small when considering attitudes of minorities only. Published by The Berkeley Electronic Press,2007 11
the highest educated of the four groups, followed by the Europeans. Based on these numbers, we should expect that labour market competition should be a concern for the highly skilled when assessing immigration from New Zealand and Australia, and for the low skilled for immigration from the West Indies and Asia. In the last row, we display the percentage in each of the groups that claims unemployment beneÖt. Here immigrants are overall more likely to be claimants than natives, but there is considerable variation across groups, with the West Indian and Asian immigrants having probabilities to claim unemployment beneÖts that are twice as high as those of white natives, while immigrants from Europe are similar, and immigrants from Australia and New Zealand substantially less likely. If potential new immigrants are similar in this respect to those already present, then those with the highest expectation of bearing the burden from increased welfare costs may be particularly concerned about immigration from Asia and the West Indies. The incidence of the Öscal consequences of higher welfare dependency will depend on whether the greater cost is met by raising taxes or by cutting public expenditure and on how exactly this is done. In principle, the cost could fall on the rich or poor, but if the increase is in the most progressive aspects of the tax system, then it will be those on high incomes who su§er most (Fetzer 2000, Dustmann and Preston 2005, Facchini and Mayda 2006). 3.2 The British Social Attitude Survey Our attitudinal data is drawn from seven years of the British Social Attitudes Survey (BSA)(1983, 1984, 1985, 1986, 1987, 1989, 1990). This gives us an eight year period over which key questions were repeatedly asked in a form which remained unchanged. Later years of the BSA ask questions on immigration less regularly, and in a form which is not consistent from year to year. We use the data for England and concentrate on white respondents only.11 There are three sets of variables which we use for our analysis. First, questions on attitudes to further immigration. Second, questions that are related to the three underlying channels by which immigration preferences may be a§ected. And third, questions that reáect heterogeneity across individuals in preferences due to individual and context characteristics. We describe these variables brieáy below. 11Racial identity is self-assessed. Attitudes of ethnic minority individuals towards their own communities, or towards other ethnic minorities, are likely to be driven by di§erent mechanisms. While it might be interesting to investigate their attitudes, the sample sizes within the BSA become very small when considering attitudes of minorities only. 11 Dustmann and Preston: Racial and Economic Factors in Attitudes to Immigration Published by The Berkeley Electronic Press, 2007
The B.E.Journal of Economic Analysis Policy,Vol.7 [2007],Iss.I (Advances),Art.62 The key variables in our analysis are based on questions concerning opin- ions about immigration from different origin countries.As discussed above, distinctions are drawn between immigration from the West Indies,from India and Pakistan,from other countries in the European common market,and from New Zealand and Australia.We create binary variables for all these responses. In Appendix D,we report the full wording of the original questions and some summary statistics. We decompose these attitudes into the three factors we have discussed above,using the methodology we describe in section 2.For that purpose,we use an array of questions which are specific to each of the suggested underlying channels that affect attitudes to immigration.As we discuss earlier,questions related to labour market concerns include fear of job loss,perception of job security,perceived ease of finding a new job,and expectations of wage growth. Questions related to welfare concerns cover opinions on generosity of benefits, needs of welfare recipients,and willingness to pay higher taxes to expand welfare provision.Finally,questions related to race consist of opinions on inter-ethnic marriage,acceptability of an ethnic minority superior at work, and self rated prejudice against minorities.Again,the exact wording of the questions and summary statistics are given in Appendix D. Not all of these questions were asked in every year.The number of responses to each question in each year that can be used is summarised in Appendix B in table Bl,where usability is determined by availability of data on both regressors and dependent variables.In our estimation procedures,we make maximum use of the available data.All observations covered in table Bl are used. The survey does not allow us to distinguish between foreign born and na- tive born respondents.We do however observe ethnicity.We exclude ethnic minority individuals from our analysis.Respondents therefore still include the white foreign born.For the UK,and over the period we consider,this pop- ulation forms only a small part of the overall population(3.8 percent).We demonstrate below that this would,even for an extreme scenario of radically different attitudes of this group,hardly affect our empirical results. The survey has extensive socioeconomic information on respondents,in- cluding education,income,age,religion,and labour market status.The set of variables in the matrix X above describe the individual's own characteristics, like their income situation,labour market characteristics,education,age,sex, and religious beliefs,and variables that describe the individual's environment, like the local unemployment rate,and minority concentrations.In table 2 we report summary statistics on these variables. The two variables describing the characteristics of the local area of residence http://www.bepress.com/bejeap/vol7/iss1/art62 12
The key variables in our analysis are based on questions concerning opinions about immigration from di§erent origin countries. As discussed above, distinctions are drawn between immigration from the West Indies, from India and Pakistan, from other countries in the European common market, and from New Zealand and Australia. We create binary variables for all these responses. In Appendix D, we report the full wording of the original questions and some summary statistics. We decompose these attitudes into the three factors we have discussed above, using the methodology we describe in section 2. For that purpose, we use an array of questions which are speciÖc to each of the suggested underlying channels that a§ect attitudes to immigration. As we discuss earlier, questions related to labour market concerns include fear of job loss, perception of job security, perceived ease of Önding a new job, and expectations of wage growth. Questions related to welfare concerns cover opinions on generosity of beneÖts, needs of welfare recipients, and willingness to pay higher taxes to expand welfare provision. Finally, questions related to race consist of opinions on inter-ethnic marriage, acceptability of an ethnic minority superior at work, and self rated prejudice against minorities. Again, the exact wording of the questions and summary statistics are given in Appendix D. Not all of these questions were asked in every year. The number of responses to each question in each year that can be used is summarised in Appendix B in table B1, where usability is determined by availability of data on both regressors and dependent variables. In our estimation procedures, we make maximum use of the available data. All observations covered in table B1 are used. The survey does not allow us to distinguish between foreign born and native born respondents. We do however observe ethnicity. We exclude ethnic minority individuals from our analysis. Respondents therefore still include the white foreign born. For the UK, and over the period we consider, this population forms only a small part of the overall population (3.8 percent). We demonstrate below that this would, even for an extreme scenario of radically di§erent attitudes of this group, hardly a§ect our empirical results. The survey has extensive socioeconomic information on respondents, including education, income, age, religion, and labour market status. The set of variables in the matrix X above describe the individualís own characteristics, like their income situation, labour market characteristics, education, age, sex, and religious beliefs, and variables that describe the individualís environment, like the local unemployment rate, and minority concentrations. In table 2 we report summary statistics on these variables. The two variables describing the characteristics of the local area of residence 12 The B.E. Journal of Economic Analysis & Policy, Vol. 7 [2007], Iss. 1 (Advances), Art. 62 http://www.bepress.com/bejeap/vol7/iss1/art62
Dustmann and Preston:Racial and Economic Factors in Attitudes to Immigration Table 2:Sample Statistics Variables Mean StdD Unemployment rate,County level 0.0437 0.0203 Ethnic minority concentration,County 0.0262 0.0285 Rank in Income Distribution 0.5008 0.2877 Manual worker 0.4555 0.4980 Ever Unemployed 0.1687 0.3745 Ever Long Term Unemployed 0.0609 0.2392 Female 0.5368 0.4986 High Education Level 0.1017 0.3022 Low Education Level 0.4991 0.5000 Age 45.936 17.706 Catholic 0.1005 0.3007 No religion 0.34620.4757 are the unemployment rate and the concentration of ethnic minorities.The former is based on yearly district level unemployment statistics.The latter is interpolated from ward level census data for 1981 and 1991.We aggregate both of these variables to the county level to minimise endogeneity issues arising from location choice(see Dustmann and Preston,2001,for a discussion of endogenous location).12 The household income variable is reported in banded form in the data. Rather than calculating a continuous measure in units of income,we have computed the average percentile point of households in that band in the income distribution,for the specific year in which the individual is interviewed.When thinking about the effect of income on attitudes,we have in mind the effect of the relative position of the individual in the income distribution,rather than some absolute income measure.Our definition of household income seems therefore quite natural in this context. The individual characteristics we include in X consist of variables that are likely to reflect differences in attitude to further immigration.Age is likely to affect attitudes for several reasons.First,it is a direct measure of life ex- perience.Second,it marks the position of the individual in their economic cycle.Finally,the age variable captures cohort effects.We have added two variables on religious beliefs,reflecting whether the individual is Catholic or not religious.Attitudes may be influenced both by the high weight placed by 12County is an administrative unit,covering on average 1.27 million people,and corre- sponding plausibly to a local labour market. Published by The Berkeley Electronic Press,2007 13
Table 2: Sample Statistics Variables Mean StdD Unemployment rate, County level 0.0437 0.0203 Ethnic minority concentration, County 0.0262 0.0285 Rank in Income Distribution 0.5008 0.2877 Manual worker 0.4555 0.4980 Ever Unemployed 0.1687 0.3745 Ever Long Term Unemployed 0.0609 0.2392 Female 0.5368 0.4986 High Education Level 0.1017 0.3022 Low Education Level 0.4991 0.5000 Age 45.936 17.706 Catholic 0.1005 0.3007 No religion 0.3462 0.4757 are the unemployment rate and the concentration of ethnic minorities. The former is based on yearly district level unemployment statistics. The latter is interpolated from ward level census data for 1981 and 1991. We aggregate both of these variables to the county level to minimise endogeneity issues arising from location choice (see Dustmann and Preston, 2001, for a discussion of endogenous location).12 The household income variable is reported in banded form in the data. Rather than calculating a continuous measure in units of income, we have computed the average percentile point of households in that band in the income distribution, for the speciÖc year in which the individual is interviewed. When thinking about the e§ect of income on attitudes, we have in mind the e§ect of the relative position of the individual in the income distribution, rather than some absolute income measure. Our deÖnition of household income seems therefore quite natural in this context. The individual characteristics we include in X consist of variables that are likely to reáect di§erences in attitude to further immigration. Age is likely to a§ect attitudes for several reasons. First, it is a direct measure of life experience. Second, it marks the position of the individual in their economic cycle. Finally, the age variable captures cohort e§ects. We have added two variables on religious beliefs, reáecting whether the individual is Catholic or not religious. Attitudes may be ináuenced both by the high weight placed by 12County is an administrative unit, covering on average 1.27 million people, and corresponding plausibly to a local labour market. 13 Dustmann and Preston: Racial and Economic Factors in Attitudes to Immigration Published by The Berkeley Electronic Press, 2007