Construction, dimension reduction and uncertainty analysis of an index of trust in public institutions

Abstract

Research on social capital in general and trust in particular is markedly increased. The worldwide problem of low and decreasing levels of trust in many public institutions is greatly debated. This is a very important problem because the trust citizens have in public institutions may foster democratic practices and facilitate better provision of public services. Moreover, trust is a central element of social order and survival of any democratic regime: it affects institutional performance, wellbeing, economic development and crime reduction. Public institutions should pay very much attention to citizens’ trust in them and in particular when introducing reforms aiming at improving their functioning. Trust is difficult to be assessed due to its quite elusive meaning. An index of general trust in public institutions is proposed within the framework of composite indicators. It is shown how to reduce the dimension of the index by finding out the most and least important aspects of it. Uncertainty analysis is used to test whether dimension reduction results are robust or not with respect to plausible changes in variable values and variable transformations. An application to data from the European Social Survey is discussed. The results show the very important role of the trust in politicians and political parties in explaining the level of general trust in public institutions.