A Crazy Methodology

Despite the great popularity in the social sciences of macroquantitative, comparative research during the last two decades and despite the hope laid into the approach to be able to carefully test theories about social macrofenomena, it has had only limited lasting impact on theory building. Symptomatic is the lack of robustness of findings. The cause of this problem is the difficulty of dealing with complex macrophenomena by means of statistical analysis. If between-country comparative research relates to independent and identical behavior of individuals which can be portrayed at the macro level by the idea of the representative agent, then the analysis is tricky but not impossible. For macro-level characteristics of social systems this road is closed, however, because the model cannot be based on assumptions about modal behavior. In this case, the sole solution is to accept the limits of small numbers and to improve the elaboration of implication from theoretical considerations.