Conferences

Multilevel analyse in comparative research rests on a rich repertoire of micro level hypotheses which are mostly justified through some theory of rational action. Macro level hypotheses, however, are less well established. They can be classified in two dimensions: (1) their results which may be means of or relations between micro level variables; they are examined in intercept or slope models and will be called here level and effect hypotheses; and for each of these two (2) the arguments used for their justification.

As for level hypotheses, arguments commonly refer either to (a) social differentiation, (b) “wealth of nations”, or© decisions of corporate actors. For example, (a) social differentiation produces economic, political, and ideological cleavages which must be increasingly rep-resented in associations, which ultimately increase mean membership. The level effect is justified by a chain of social processes which progressively gain impact on decisions of persons. Or, (b) an increasing wealth of nations facilitates the establishment of infrastructures (from theatres to (data) motorways) which ultimately increases mean life satisfaction. The level effect is justified by common resources which affect individual decision in the same way as, but independently of, individual resources such as income. Or, (c) Thatcher changed the British social security system from a socialist to a liberal one which ultimately may have heightened mean mobility aspirations in Britain, as compared with other West European countries. The level effect is justified by a – in historical perspective – sudden change of the social order which is supposed to form the attitudes of each person alike.

As for effect hypotheses, nearly all of them are justified by some multiplicator mechanism which either works positively or negatively. For example, the effect of social or political trust on political participation may increase with the mean level of trust in a country. If trust is the rule in the community, trusting persons can more easily act as they like to, namely engage themselves. Thus, micro and macro effect go in the same direction, interact positively, and reinforce each other. Or, the effect of union membership on empowerment on the work place may decrease with the degree of union organization in a country. If unions are powerful, they will grant empowerment to their constituency such that personal membership is no longer needed. Thus, micro and macro effect go in the opposite direction, interact negatively, and substitute each other.

In the session, two kinds of contributions are welcome: empirical multi-level analyses which particularly stress the derivation of macro level hypotheses; and conceptual contributions which investigate different justifications for macro hypotheses with the goal of systematization.