All pages

CHAPTER 3: Trends in anti-immigration attitudes

Previous research

Attitudes towards immigration, immigrants and ethnic minorities have been investigated quite intensively. However, research that focuses on the evolution of anti-immigration attitudes is far more scarce (for exceptions, see [Coe98], [Coe08], [Fir88], [Qui96], [Sch97], [Sem06]). That relatively little work has been done on this topic has a lot to do with the scarcity of appropriate data. After all, studying longitudinal attitude developments entails additional data requirements. We need comparable survey measurements from different time points (see Chapter 2), and such data are rare indeed. However, the availability of several rounds of the European Social Survey has created new opportunities to study over-time attitude developments in Europe.

The lion’s share of the available research has dealt with changes in whites’ attitudes to black minorities in the USA. Since the 1950s, there has been evidence of increasing support for the principle of equal treatment among white US citizens [Fir88], [Qui96], [Sch97]. At the same time, however, support for government interventions that implement equal treatment (such as affirmative action or school desegregation programmes) does not follow the same steep upward trend. This paradox between high support for principles and low support for implementation has led some scholars to conclude that, during the last few decades, traditional negative outgroup attitudes have crystallised into new forms - such as ‘symbolic racism’ [Kin81] or ‘subtle prejudice’ [Mee97].

These observed tendencies in the USA cannot be generalised in a straightforward manner to the European situation. The attitudinal changes in the USA are, at least partly, the product of the particular historical evolution of intergroup relations from slavery to a situation of legal equality.

In many European countries, the presence of sizeable ethnic minority groups is a rather recent phenomenon, as large immigration flows into Europe only arose during the second half of the 20th century [Cas03], [Hoo08] . As a result, European researchers have only recently started to ask survey questions about outgroup attitudes, and only a couple of studies focus on European trends in outgroup attitudes.

Probably the earliest European time series of measurements of unfavourable attitudes towards ethnic minorities was reported by Coenders and Scheepers [Coe98]. This study describes how support for ethnic discrimination in the Netherlands dropped sharply between 1979 and 1986. From the mid-1980s to 1993, however, the proportion of Dutch people who support ethnic discrimination again rose substantially. In a similar study on German data, Coenders and Scheepers [Coe08] conclude that opposition to the social integration of guest workers and foreigners dropped continuously between 1980 and 2000. The only exception to this downward trend was a small surge of anti-foreigner attitudes in 1996.

A more comprehensive study of 12 European countries was carried out by Semyonov et al. [Sem06]. This study showed that ethnic prejudice increased dramatically between 1988 and 1994 in all 12 European countries under study, also including the Netherlands and Germany. However, with respect to Germany these findings contradict Coenders and Scheepers’ [Coe08] study, which found that attitudes toward foreigners grew more positive during that period. These contradictory findings might be due to the fact that both studies focus on slightly different concepts. While Semyonov et al. [Sem06] used a set of indicators measuring how people perceive the consequences of immigration, Coenders and Scheepers [Coe08] study resistance to the social integration of foreigners. Alternatively, the contradictory findings could also be caused by problems with the comparability of the measurements over time. Neither Semyonov et al. [Sem06] nor Coenders and Scheepers [Coe08] tested for measurement equivalence (as we did in Chapter 2 of this module).

While ethnic prejudice increased quite uniformly between 1988 and 1994, Semyonov et al. [Sem06] report that, from 1994 onwards, trends are less pronounced and differ strongly across countries. In some countries, such as Ireland and Luxemburg, anti-foreigner attitudes grew clearly stronger between 1994 and 2000. In other countries, such as Belgium and Spain, negative attitudes toward foreigners crumbled away. Thus, this literature review makes clear that European attitudes toward outgroups - unlike those in the USA - are not homogenous. Very diverse, country-specific patterns are found instead.

Page 1

Evolving anti-immigration attitudes based on ESS data

ESS data offer excellent opportunities to deepen our current knowledge of changes in European anti-immigration attitudes, and make it possible to see how the evolution sketched by Semyonov et al. [Sem06] continued into the first decade of the third millennium.

Exercise 3.1: Calculate the evolution of anti-immigration attitudes

Our previous analyses have shown that the three anti-immigration items measure a single dimension (see exercise 1.3) and are suitable for comparisons across countries and time points (see exercise 2.1). Our scale can therefore be used to study the evolution of anti-immigration attitudes among European populations 1.

  1. Use SPSS to construct a new variable, ‘REJECT’, containing the sum of the three anti-immigration items (IMSMETN, IMDFETN and IMPCNTR). Remember that higher scores on REJECT indicate more negative attitudes towards immigration. Do not forget to save the dataset afterwards.
  2. Calculate the average score for REJECT separately for every country at every time point. Remember to use the design weight.

SPSS Syntax

  1. *Open the file you created in the first/second chapter, ESS123_immig. *Compute the variable ‘reject’ and save the file. **Please do not forget to change ‘C:\’ to the correct path.

    COMPUTE reject = imsmetn + imdfetn + impcntr.
    SAVE OUTFILE = 'C:\ESS123_immig.sav'.

  2. *Weight data and find each country’s mean value for the reject variable for each round.

    WEIGHT by dweight.
    MEANS reject by cntry by essround.

SPSS Output

Table 3.1. Mean of the variable REJECT by country and ESS round
Country - ESS round Mean N Std. Deviation
Austria - 1 7.8527 1671 2.11806
Austria - 2 7.2990 1867 2.24808
Austria - 3 7.4402 2049 2.13667
Austria - Total 7.5164 5587 2.18060
Belgium - 1 7.2628 1621 2.22334
Belgium - 2 7.4359 1553 2.38145
Belgium - 3 7.1374 1616 2.24029
Belgium - Total 7.2766 4790 2.28418
Switzerland - 1 6.5378 1576 1.77536
Switzerland - 2 6.6611 1667 1.95637
Switzerland - 3 6.7394 1377 1.94155
Switzerland - Total 6.6424 4620 1.89336
Germany - 1 6.8732 2543 2.08144
Germany - 2 7.4155 2456 2.32610
Germany - 3 7.3108 2504 2.29667
Germany - Total 7.1968 7502 2.24811
Denmark - 1 7.0469 1321 1.99110
Denmark - 2 7.1436 1351 2.03600
Denmark - 3 6.8682 1358 1.94497
Denmark - Total 7.0191 4030 1.99376
Spain - 1 7.1706 1376 2.43318
Spain - 2 7.1760 1432 2.49854
Spain - 3 7.6366 1624 2.57256
Spain - Total 7.3431 4431 2.51539
Finland - 1 7.5896 1859 2.03360
Finland - 2 7.6906 1936 2.00512
Finland - 3 7.6284 1787 1.97255
Finland - Total 7.6370 5582 2.00441
France - 1 7.4514 1234 2.10310
France - 2 7.4794 1529 2.25001
France - 3 7.4363 1725 2.15561
France - Total 7.4551 4488 2.17373
United Kingdom - 1 7.5859 1763 2.22641
United Kingdom - 2 7.4300 1615 2.27064
United Kingdom - 3 7.7632 2022 2.24488
United Kingdom - Total 7.6056 5400 2.25035
Hungary - 1 8.5644 1313 1.82187
Hungary - 2 8.5719 1231 2.27350
Hungary - 3 8.9042 1320 2.14962
Hungary - Total 8.6829 3865 2.09199
Ireland - 1 6.6475 1739 1.99185
Ireland - 2 6.6428 1975 2.25460
Ireland - 3 6.5988 1460 2.20434
Ireland - Total 6.6320 5174 2.15489
Netherlands - 1 7.2855 2111 1.99738
Netherlands - 2 7.4614 1671 2.18149
Netherlands - 3 7.5328 1660 2.24517
Netherlands - Total 7.4149 5441 2.13451
Norway - 1 6.8557 1861 1.95663
Norway - 2 6.8139 1601 1.94720
Norway - 3 6.6780 1587 2.00731
Norway - Total 6.7866 5049 1.97079
Poland - 1 7.0707 1811 2.06893
Poland - 2 6.7753 1576 2.33536
Poland - 3 6.2212 1580 2.33146
Poland - Total 6.7067 4967 2.26784
Portugal - 1 8.3959 1279 2.38004
Portugal - 2 8.6660 1706 2.30083
Portugal - 3 8.6937 1783 2.55076
Portugal - Total 8.6039 4768 2.42082
Sweden - 1 5.6455 1673 1.92080
Sweden - 2 5.6552 1679 2.09309
Sweden - 3 5.4564 1628 1.99722
Sweden - Total 5.5869 4980 2.00678
Slovenia - 1 7.2823 1240 2.09639
Slovenia - 2 7.3252 1187 2.32380
Slovenia - 3 7.3659 1208 2.26794
Slovenia - Total 7.3241 3635 2.22943
Total - 1 7.1984 27990 2.16807
Total - 2 7.2576 28031 2.31775
Total - 3 7.2771 28288 2.34036
Total - Total 7.2445 84310 2.27714

Weighted by design weight.

Questions

Solution

The means clearly show that the evolution of anti-immigration attitudes varies considerably from one country to another. Whereas anti-immigration sentiment flared up in some countries between 2002 and 2007, it decreased in others. In eight countries, the average REJECT score decreased from 2002 to 2007, indicating that attitudes have become more immigration-friendly. These countries are Austria, Belgium, Denmark, France, Ireland, Norway, Poland and Sweden. In the other nine countries studied here (Switzerland, Germany, Spain, Finland, the United Kingdom, Hungary, the Netherlands, Portugal and Slovenia), resistance to immigration has risen, as the average score for REJECT increased between 2002 and 2007. Apparently, the evolution of attitudes does not follow regional lines. Among the Eastern European countries in the study, for example, very divergent patterns of evolution can be observed: While support for immigration crumbles in Hungary and Slovenia, attitudes are becoming more open in Poland. In Scandinavia, we find a decrease in anti-immigrant feelings in Denmark, Norway and Sweden, but not in Finland.

Poland is the country with the most marked change by far in anti-immigration attitudes. Between 2002 and 2007, the average score for REJECT decreased by almost 0.85. Other countries with marked attitude changes are Austria, Germany and Spain.

Since REJECT increased in about half the countries, and decreased in the other half, it is hard to speak of a dominant trend across Europe. This is also reflected in the development of the pooled dataset shown in the previous table. For the total of all countries, the change in REJECT between 2002 and 2007 is quite small (to be precise: from 7.2771-7.1984 = 0.0786). The marked increase in anti-immigrant sentiment that has been seen in Europe between the mid-1980s and mid-1990s has thus not persisted during the first decade of the 21st century.

Page 2

Exercise 3.2: A graphical representation of attitude trends

A table showing means for 51 groups like the one presented above is not very easy to interpret. In many instances, graphs can be more insightful. Draw a line graph using a separate line to represent the evolution for every country. Put the average REJECT-score on the Y-axis, and the ESS round on the X-axis. Using the SPSS Chart Builder is the easiest way to draw the graph.

SPSS Chart Builder

Click ‘Graphs’ in the main horizontal tool bar, and then ‘Chart Builder’.

The Chart Builder wizard pops up. We start by defining the graph type. In the lower pane of the gallery, select ‘line’. Drag the second icon (with the multiple lines) into the upper pane of the chart builder. Now you can select the variables to be displayed by dragging the variable names (in the upper left corner) into the graph. Drag REJECT to the box next to the Y-axis, and ESSROUND to the X-axis. The variable COUNTRY should be dragged into the remaining box in the upper right corner (see Figure 3.2).

In the ‘Element Properties’ window, you should select 'Line1'. Make sure that the mean is calculated for the variable REJECT. Our previous analyses have shown that the averages for all country-time point combinations range between five and nine. In order to get a better view of country-specific evolution, it is a good idea to rescale the Y-axis so that only the relevant range of REJECT values is displayed. This is done by clicking ‘Element Properties’, select ‘Y-axis 1 (Line 1)’, and change the minimum and maximum values from automatic to 5 and 9. Finally, select ‘GroupColor (line1)’ in the ‘Element Properties’ window, and select ‘Show only categories present in the data’. Click 'Apply' in the 'Element Properties' window, and then OK in the Chart Builder to finalise the graph (see Figure 3.3, 3.4 and 3.5).

Figure 3.3Figure 3.4

* The SPSS syntax generated by the Chart Builder.

GGRAPH
/GRAPHDATASET NAME="graphdataset" VARIABLES=essround MEAN(reject)[name="MEAN_reject"] cntry MISSING=LISTWISE REPORTMISSING=NO
/GRAPHSPEC SOURCE=INLINE.
BEGIN GPL
SOURCE: s=userSource(id("graphdataset"))
DATA: essround=col(source(s), name("essround"))
DATA: MEAN_reject=col(source(s), name("MEAN_reject"))
DATA: cntry=col(source(s), name("cntry"), unit.category())
GUIDE: axis(dim(1), label("ESS round"))
GUIDE: axis(dim(2), label("Mean reject"))
GUIDE: legend(aesthetic(aesthetic.color.interior), label("Country"))
SCALE: linear(dim(2), min(5), max(9))
SCALE: cat(aesthetic(aesthetic.color.interior))
ELEMENT: line(position(essround*MEAN_reject), color.interior(cntry), missing.wings())
END GPL.

SPSS Output

The graph essentially contains the same information as the means table we calculated before. However, the information is far more accessible. Here, it is obvious at first sight that Hungary and Portugal are the least immigration-friendly countries in the study, and that the Swedish population has the most positive attitudes toward immigration. There has clearly also been a marked evolution in Poland.

Questions

  1. Is the evolution that countries experience uniform (i.e. changing in the same direction over the entire period studied)? Or do they tend to go up and down instead?
  2. Are attitudes toward immigration converging in Europe? Or are European populations becoming more diverse in their attitudes instead?

Solution

The graph also provides additional insights into the data that are harder to deduct from the table.

  1. Not all countries are experiencing uniform attitude trends. In Finland, Denmark and Belgium, for example, anti-immigration attitudes grew stronger between 2002 and 2004, only to decline again afterwards. The opposite pattern is found in Austria.
  2. In spite of the fact that EU policy-makers are endeavouring to harmonise immigration policies, there is no evidence that attitudes toward immigration are converging in EU member states. On the contrary, the countries located at the extremes of the rankings seem to be moving away from the European average. In Hungary and Portugal, the countries where most resistance to immigration is found, anti-immigration attitudes are becoming even stronger. And in Sweden, already the most immigration-friendly country in 2002, we see growing support for immigration.

Page 3

Exercise 3.3: The statistical significance of attitude trends

Up until now, we have considered all differences between time points as meaningful. However, the observed evolution may be caused by sampling fluctuations. To be more confident that the mean differences are not just a matter of chance, we should perform statistical significance testing. Test every country separately to check whether the means at the three time points vary significantly. Use Analysis Of Variance (ANOVA) for this purpose (also known as an F-test for mean differences between groups). HINT: use the SPLIT FILE statement to perform the analysis country-by-country in an efficient manner.

SPSS Syntax

*Weight data and test every country separately to check whether the means at the three time points vary significantly.

WEIGHT by dweight.

SORT CASES by cntry.
SPLIT FILE by cntry.
MEANS reject by essround
/statistics=anova.
SPLIT FILE off.
WEIGHT off.
SAVE outfile = 'C:\ESS123_immig.sav'

Question

For which countries are the differences between time points statistically significant?

Solution

Table 3.2. Analysis of variance: Do the means at the three time points vary significantly?
Country Groups Sum of Squares df Mean Square F Sig.
Austria Between Groups 289.064 2 144.532 30.720 .000
Within Groups 26272.060 5584 4.705
Total 26561.124 5586
Belgium Between Groups 71.057 2 35.528 6.826 .001
Within Groups 24915.424 4787 5.205
Total 24986.481 4789
Switzerland Between Groups 30.790 2 15.395 4.300 .014
Within Groups 16528.847 4617 3.580
Total 16559.637 4619
Germany Between Groups 416.271 2 208.135 41.634 .000
Within Groups 37494.135 7500 4.999
Total 37910.405 7502
Denmark Between Groups 52.891 2 26.445 6.672 .001
Within Groups 15962.638 4027 3.964
Total 16015.529 4029
Spain Between Groups 220.836 2 110.418 17.584 .000
Within Groups 27811.598 4429 6.279
Total 28032.434 4431
Finland Between Groups 9.876 2 4.938 1.229 .293
Within Groups 22412.782 5579 4.017
Total 22422.659 5581
France Between Groups 1.528 2 .764 .162 .851
Within Groups 21198.352 4485 4.726
Total 21199.880 4487
United Kingdom Between Groups 100.687 2 50.343 9.975 .000
Within Groups 27238.776 5397 5.047
Total 27339.463 5399
Hungary Between Groups 98.259 2 49.130 11.283 .000
Within Groups 16812.166 3861 4.354
Total 16910.425 3863
Ireland Between Groups 2.258 2 1.129 .243 .784
Within Groups 24017.504 5171 4.645
Total 24019.762 5173
Netherlands Between Groups 62.020 2 31.010 6.822 .001
Within Groups 24723.600 5439 4.546
Total 24785.620 5441
Norway Between Groups 28.794 2 14.397 3.710 .025
Within Groups 19579.635 5046 3.880
Total 19608.430 5048
Poland Between Groups 619.773 2 309.886 61.726 .000
Within Groups 24921.187 4964 5.020
Total 25540.959 4966
Portugal Between Groups 76.289 2 38.144 6.524 .001
Within Groups 27860.564 4765 5.847
Total 27936.853 4767
Sweden Between Groups 41.306 2 20.653 5.137 .006
Within Groups 20010.046 4977 4.021
Total 20051.352 4979
Slovenia Between Groups 4.282 2 2.141 .431 .650
Within Groups 18057.961 3632 4.972
Total 18062.244 3634

Weighted by design weight.

Analysis Of Variance (ANOVA) can be used to test whether the mean of a certain variable (here REJECT) varies across categories of a second variable (here: ESSROUND, indicating the three time points). Here, we performed a separate ANOVA for each of the 17 countries. The test basically compares the amount of variation in REJECT between time points with the amount of variation within time points. Logically, the larger the differences between time points, the more convincing is the evidence that attitude changes have taken place. The F-test can be used to test whether differences between time points are statistically significant. The null hypothesis of this test is that the means are equal at all three time points. If the p-value of the test statistic is lower than .05, this null hypothesis can be rejected. In the latter case, we can conclude with 95 per cent certainty that the observed attitude changes are not due to chance fluctuations.

The SPSS output from the ANOVA shows that, in 13 out of 17 countries, attitude evolution is statistically significant at the .05 level. Only in Finland, France, Ireland and Slovenia were the mean differences between time points too small to be conclusive.

In fact, it is not very surprising that we find many attitude changes to be statistically significant. Significance depends not only on the size of the differences, but also on the sample size. Here, we are clearly dealing with very large datasets: our analyses include information about more than 80,000 respondents. As a result, even small attitude changes become significant. We should not, therefore, rely blindly on significance tests, but also look at whether the mean differences are substantially large. Remember that the strongest attitude change was found in Poland. There, the average score on REJECT dropped by 0.85 points. Given that REJECT has a minimum value of 3 and a maximum of 12 (REJECT is the sum of three items measured on a 1 to 4 scale), this is a very substantial change indeed. We also find substantial mean differences greater than 0.30 in Austria, Germany, Spain, Hungary and Poland.

Go to next chapter >>
Page 4