The scholarly debate on the crisis of democracy in terms of its decreasing legitimacy is now over 40 years old, if the 1975 report is considered its initiation (Crozier et al. 1975). As one landmark, Putnam’s study of social capital in Italy (1992) had immense influence on decades of studies of trust, confidence in institutions and their effects on the way democracy works. Aspects of the crisis of the “disaffected” democracies were studied with increasing use of large databases (Pharr and Putnam 2000; Inglehart 1997a; Inglehart and Welzel 2005a; Newton 2001; Newton and Norris 2000), in particular World Values Surveys. But correlations of attitudes and values could be interpreted in different ways.

As trust is thus rooted in longer term distribution of scarce resources, democratization and civil society are interrelated in their development (Paxton 2002). Both benefit each other in their evolution. Democracy both involves the national institutions of competitive government and the conditions of participation (Dahl 1971b; Schumpeter 1942). Civil society and civic culture represents participation at grass-root level in associations of various kinds. To an extent, therefore, democracy and civil society conceptually overlap or come very close to one another in terms of content of social action. Generally, civil organizations endorse democracy and vice versa. Still, the determinants yet known to us cannot fully explain all variance at national level on the aggregates of trust in people, confidence in institutions and democratic values (Uslaner 2018).

The relationship between democracy and civil society has normally been studied at national level, such as Paxton’s study, or at individual level in terms of democratic values, trust and membership status, such as in the World Values Studies (Inglehart 2010). However, as some of the international educational databases include variables on democratic values and knowledge, as well as interpersonal trust, institutional trust, participation in civic organization, we may now also investigate the mechanisms of early socialization and the role of schools in the generation or formation of values related to democracy, civic culture, interpersonal and institutional trust. The critical question of how grass-root civic culture emerges and interacts with national level democracy and democratization is thereby closer to its resolution. The recently more important factor of immigration can also be included in models at student level.

Nations are of course not political-cultural blank slates as democratic values and civic and citizenship virtues are taught in schools. Instead, civic and citizenship education can be considered part of continuous nation-building projects and formation of political cultures. However, exactly what values, virtues and attitudes that contribute to what effects on democracy is a relation not yet understood in detail, and even less to at school level (see however Ainley et al. 2013; Rånge and Sandberg 2017b;Schulz et al. 2010; Skolverket 2012; Dahlin 2010; Isac et al. 2014; Lundahl et al. 2010; Quintelier and Hooghe 2013; Amnå et al. 2007; Almgren 2006). In this paper, therefore, we wish to add further to the understanding of the evolution of trust, political or institutional trust and their relation to democratic values at school level among 14-year olds, as consequences of both socialization and immigration. The purpose is to find the student, school and national factors behind support of democratic values given the unprecedented opportunity of data on all three levels. Can educational data help us determining the mechanism at individual and school levels behind the evolution of trust, institutional trust and democratic values? Here, we use the opportunity offered by the educational science dataset the International Civic and Citizenship Education Study (ICCS 2009 and 2016), combined with nation-level variables of political culture, equality and regime types.