If one wants to examine the relationship between variables in a small n sample, usual application of multiple regression analysis fail due to a small degree of freedoms and estimation problems, accordingly. We consider a sample of 25 countries of the European Union as an example for this case and try to answer the question why the Scandinavian countries in the European Union are significantly richer than Southern/Eastern European countries? In order to answer this question, we refer to a social capital perspective applying aggregated measures of social trust and corruption. In order to test this hypothesis with regard to these methodological problems, a path model is applied that is capable of showing the interrelations between corruption as manifestation of negative and trust as indication of positive social capital.
The application of the path models allows extensions for testing temporal direct and indirect effects. By this, we pick up an idea by Granger and are able to find out about the "Granger-causal" relationships between trust and corruption.

Our empirical results, referring to a sample of EU countries show that corruption might harm poor countries but is not able to affect social trust. However, corruption in itself means that resources end up in the wrong places and not in socioeconomically optimal investments. There is, therefore, a direct damaging effect of corruption on wealth. This implies that economic actors have to invest higher transaction and control costs which will bind resources to non-productive purposes and thus destroy economic wealth. Most remarkably is that the augmentation of positive social capital could work as an effective counterforce to corruption, even if it does not compensate for the economic loss done by corruption. Thus, adding the social capital perspective to human and physical capital may contribute to understand present day variation in the wealth of European nations either by the damaging effect of corrupt activities and the positive force of positive social capital.