Political clientelism in Southern Italy has shown great persistence and capacity for conditioning the entire Italian political development. Accounting for these characteristics, clientelism is better understood as the product of the incomplete capitalistic rationalization of the Southern economy. Throughout the 19th century in Sicily, the feudo remained the basis of the economic and social structure, while in the continental South the feudal system disintegrated more quickly and widely. The resulting different models of social relationships within the Mezzogiorno, make it meaningful for our purposes to distinguish between two types of clientele, which may be termed clientela mafiosa and Neopolitan clientele.

The second part of the article is concerned with the dynamics of the clientelistic system in post-1945 Southern Italy and more generally with the relationship between clientelism and political development. The transition from the clientelism of the notables to political party-directed patronage is studied both at the local level and in the context of the whole South. It is then argued that clientelism is a very poor tool for political development since it has two effects on the social structure and the political process: these effects will be called disorganic integration and exclusivism.

Despite assertions about the ubiquity of coalition behavior in politics, theories of coalition formation have not usually been tested on data pertaining to the legislative process. The paper explores advantages and shortcomings of existing coalition theory for studying law-making in a multi-party parliamentary democracy with highly cohesive parties, using Denmark as the test case. All passed government bills 1953–1970 are studied in terms of the size and composition of the winning coalitions and policy area of bills. In all cabinet periods a certain number of bills are passed by minimal majorities, but most are not, indicating that minimal majority behavior is not a principal strategy in law-making. However, the traditional left-right model of the Danish party system is very successful in accounting for the coalitions formed. Classification of bills into policy areas also shows considerable variation with respect to levels of conflict, with taxation as the most conflict-generating area. The findings suggest that if we are ultimately to understand the significance of coalition behavior in politics, it is imperative that we relax strict rationalistic behavior assumptions and concentrate efforts on developing ways for systematic study of the content and perceived consequences of decisions made by winning coalitions.

Before undertaking comparative policy analysis, one must first be clear of the justification for the inordinately large amount of work required. A variety of reasons are reviewed. In order to make comparisons, one must have a conceptual framework to identify the different parts of the policy process. Data must be placed in a process, as well as located in geographical space. It is also important to consider to what extent differences between policy areas (defence, health care, housing, economic policy, etc) influence the policy process. The issues discussed concern problems that arise in within-nation studies of policy, as well as between-nation comparisons.