More Participation, Happier Society? A Comparative Study of Civil Society and the Quality of Life

These countries are: Austria (AT), Belgium (BE), Germany (DE), Denmark (DK), Spain (ES), Finland (FI), France (FR), United Kingdom (UK), Greece (EL), Hungary (HU), Israel (IL), Italy (IT), Luxembourg (LU), Netherlands (NL), Norway (NO), Poland (PL), Portugal (PT), Sweden (SE) and Slovenia (SI). Switzerland and Czech Republic are excluded because of missing data on measures of participation in voluntary associations. Ireland is excluded because of missing data on indicators on levels of satisfaction with the government.

The results of these more detailed analyses are available from the authors upon request. For the analyses of participation in single types of voluntary associations we use hierarchical logistic regression models. Though our single variables principally distinguish between weaker and stronger forms of engagement, we have re-categorised our dependent single variables. This yields binary variables in the case of single types of voluntary associations which differentiate between ‘no participation’ on the one hand and ‘any form of participation’ on the other. Generally speaking, results are robust with the notable exceptions concerning participation in some types of voluntary associations (see text).

We have also considered zero-inflated and zero-modified Poisson regression models. Whilst the former model takes into account the larger number of zeros in the dependent variable, the latter model differentiates between two dependent variables: (a) a dichotomous variable differentiating between those people who do not participate at all and those who do participate (at least to a very small extent) in civil society and (b) a count variable which examines how much participants actually participate in voluntary associations. Because the results of all three models are very similar, we do not present them here. However, we briefly comment on the most interesting differences when interpreting the results of our Poisson regression models later on.

These results are highly stable as shown in our zero-modified models. The only noteworthy additions to be mentioned are: (1) The gender gap is mainly due to the lower probability of women to participate in civil society associations. However, there are hardly any gender differences among people who participate. (2) Social class differences are slightly less pronounced concerning participation (yes/no) than the extensivity/intensivity of participation. Hence, whilst different social classes show more similar levels of general involvement on the dichotomous measure, members of different social classes are involved to varying extents. Generally speaking, these findings present further details rather than a challenge to the findings from the less complex Poisson regression models.

More detailed results are available from the authors upon request (see note 2).

The initial factor analysis at the individual level displays a KMO of 0.797 which indicates the appropriateness of factor analysing the satisfaction indicators. Based on Eigenvalues, the analyses extracts two factors, of which the first one (satisfaction with society) explains approximately 41 per cent and the second factor (subjective quality of life) explains another 12 per cent. Alternatively, we have extracted only one factor in order to assess the ‘quality of life/society’ dimension. In this model, no single factor loading is smaller than 0.5. We conclude that, for our purposes, we can stick to the one-dimensional measure because the loss of information is compensated for by a clearer interpretation of quality of society. At the country level, scores on satisfaction scales are even more strongly correlated. KMO is equal to 0.821, 75 per cent of the variation is explained by the one-factor solution where no loading is smaller than 0.77. To conclude, factor analysis shows that a ‘good society’ is established by people who are satisfied and happy with both their own lives and the state institutions and the way they operate.

More detailed results on country level data concerning quality of life/society are available from the authors upon request.