refutation of the conflict hypothesis (remember that that latter expects tolerance levels
among dominant group members to fall as the proportion of minorities increases). As
noted earlier, since the items composing the ethnic tolerance index all refer to immigrants
I believe that the index essentially captures the tolerance levels of the ethnic majority
only. To then find that ethnic majority respondents are indeed more tolerant the more
diverse their classroom is all the more supports the contact argument.
Table 3 about here
In England, however, the relation between heterogeneity and tolerance changes from a
positive to a not-significant negative one for all respondents (Model II) and for the ethnic
majority (Model III) once individual factors are included in the analysis. A possible
explanation for the different results of the white British students compared to the ethnic
German and Swedish students is the longer history of immigration in Britain. It could
well be that this longer history has made white British youth, also those in mono-ethnic
schools, become more accustomed to ethnic minorities than their counterparts in Sweden
and Germany. In other words, the longer presence of ethnic minorities in Britain may
have had the effect of leveling out attitudinal differences between students in diverse and
homogenous classes. If this is indeed the causal mechanism then we should expect to see
the diversity effect dissipate in Sweden and Germany as well with the passing of time.
Unfortunately, this hypothesis cannot be tested with the data at hand.
Another possibility is that the political tradition of liberalism and multiculturalism
in England imposes such strong norms of tolerance that the ethnic composition of
classrooms essentially has little to add. In other words, white British students in all-white
classrooms feel the same pressure to give socially desirable responses on the tolerance
items as white British students in mixed classrooms. If this conjecture is true one would
expect the overall levels of tolerance to be higher in England than in the other countries.
The descriptive statistics of Table 1, however, show that the tolerance levels of white
British students are below the international mean of ten while those of their Swedish
counterparts are above this mean. Thus this hypothesis is unlikely to have much
16