Democratic institutions are under pressure as was also the case at the end of the sixties. But where in those days the critique was left-liberal and seeking to extend democracy, now the discomfort with democracy has concervative-populist overtones, related to the reaffirmation of exclusive, mostly national, identities. The populist critique of liberal achievements and institutions has raised questions of ethnicity and identity. The historical tension between national identiy and parliamentary democracy offers a broader frame against which the emergence of nationalist and populist movements, like the Republikaner, Front National and the Vlaams Blok can be understood. These considerations lead to some important questions: is a parlementary democracy in the European Union possible? Does democracy need a specific cultural society? If so, European unification would break the historical link between parliamentery democracy and the nation, without realizing a democracy on a higher multi-national level. Those who want to defend the liberal democracy against the challenge of populist and nationalist political movements, will need to study more closely the cultural foundations of democracy.

This article makes first a critical review of the different definitions in the scientific literature of the 'underclass' concept. It is argued that the American and British concept of underclass is not necessarily transportable to Europe. The next part adresses the question of a possible emerging urban underclass in Belgium. It is concluded that although Belgium has to fight against social problems associated with underclass formation in the US and Britain, it has until now no clearcut underclass. The main reasons for this are that economic marginality is not permanent and stable, and that economic marginality, deviant values and criminal behaviour are not systematically coinciding.

During the last years immigration has aggravated the socialproblems in many disadvantaged urban districts. High proportions of foreigners are concentrating in those areas which suffer from unemployment and bad housing conditions. The accumulation of social problems has created a climate of insecurity, social prejudices, and political dissatisfaction. Since political discontent presently is not remedied by the established political parties, it results in low voting participation and increasing proportions of right wing votes. The close connection between the intensity of social problems on the one side, low voter turnout and high success of right extremist parties on the other side, is empirically established by an ecological analysis of the recent state elections in Hamburg.

In ten years the National Front's scores rose from less than 2 percent to more than 13 percent of the valid votes and the number of its electors from a hundred thousand to almost four millions. On the basis of two surveys conducted by CEVIPOF (Centre d'étude de la vie politique française) and OIP (Observatoire interrégional du politique) (1988-1989), this paper analyses the factors that account for this electoral rise. If the National Front vote is more frequent in urban areas, it is not so much because of objective factors (more contacts with immigrant populations, exposure to crime and violence, drugs) than because of subjective factors (fear and feeling of insecurity, pessimism) in relation with sociocultural and political specificities of the National Front's electorate (poor education, atomization, limited sociability, lack of trust in institutions etc.).

Support for the extreme right in Britain has been relatively low in Britain in recent years and has not shown the surge apparent in a number of other European countries. The paper uses data from the 1979 British Election survey to examine the characteristics of care and peripheral National Front supporters at the time of their last surge in support, and then goes on to consider why support has remained low in recent years. The 1979 evidence shows that support for the National Front was strongly linked to racist attitudes but in other respects had a 'protest' character. It is suggested that the subsequent weakness of the extreme right in Britain may be due to its single-issue character and to the availability of more attractive alternatives for protest voters such as the Liberal Party.

Both in the pre-war and the post-war period right-wing extremism was not very strong in Britain. Historians, political scientist and politicians have suggested a whole range of elements to explain this failure. In the light of this limited success the victory of the British National Party in an election of the Millwall district in the London Bourough of Tower Hamlets was indeed a surprise. It raised the question whether this was the beginning of something similar to what happened earlier in France, Germany, Belgium and the Netherlands. The very specific characteristics of the London East End and ofthe Millwall district in particular make the BNP victory however quite exceptional, and do not enable a generalization of the phenomenon. This is supported by the electoral results for the London Borough and District Council of May 5 1994. Yet one can still argue that the specific danger of the BNP is not its electoral potential, but the impact of its local activities on the relations between the ethnic groups in the neighbourhoods where it is present.

This article exposes comparable social mechanisms that have generated the social construction of threatening immigrants in Europe in the thirties and in the eighties. The analysis is building on Bourdieu 's theory of the construction of social space and the genesis of social groups. This semiotic-praxiological approach is used to explain why the specific historical and socio-economical conditions in the thirties and eighties have lead to the construction of Jews and Muslims as threatening immigrants. Our discussion focuses on the exemplary caseof the 'migrant problem' in historical and actual political discourse in Flanders (Belgium). Where at the end of the thirties the notion 'immigrant' referred exclusively to Jews, in the eighties it is used for Turkish and Maroccan 'guestworkers'. In spite of the specific historical and social situation of Jewish and Muslim immigrants parallel social mechanisms and discourses emerge in the redrawing of the social space by creating 'theatening' immigrants/strangers. These mechanisms are a religious anti Judaism/anti-Islamism, rapid social economie change fueling an economical argumented antiJew/anti-muslim and (cultural) racism legitimized by an internationally disseminated ethno-nationalism.

Studies about the Flemish voter's perceptions and attitudes revealed that the ecologist (green) party "Agalev" and the radical right-wingparty "Vlaams Blok" were each other's antipodes. These two parties were perceived by the voters as extreme poles on dimensions that were mainly formed by the parties' (perceived) viewpoint about the rights of immigrants, about Flemish political autonomy, and about materialistic versus postmaterialistic policy alternatives. According to the voters' attitudes, the two electorates were polarized by their attitudes towards immigrants, (post)materialism, economic conservatism, and Flemish nationalism. The electorates of these two 'extreme' parties were very similar in age composition (more in youngest generations), urban/rural environment (more living in large urban agglomerations), and church involvement(more non-Catholics). The voters of the ecologist party were more likely to have finished higher education and to have a high occupational status. Blue collar workers with little educational formation were over-represented among the electorate of the right-wing party. Using the data collected among 2,691 Flemisch voters after the 1991 General Elections, this study analyses relevant attitudes and socio-demographic characteristics within one global logistic-regression model. The focus is on the Vlaams Blok versus Agalev with the other electorates as reference category. Controlled for social characteristics, four attitudinal variables still show significant (and opposite) net effects on the likelihood of votingfor Agalev or for the Vlaams Blok and not for the other parties: the attitude towards immigrants, Belgian/Flemish nationalism, (post)materialism, and readiness to make efforts for the preservation of the physical environment. This findings support Swyngedouw's thesis about the emergence of two new cleavages: the so called universalistic/particularistic and postmaterialistic/materialistic cleavages that are capable tocatch about 20% of the Belgian-Flemish voters in the early 90s.