No abstract

The study of party unity and its determinants is conceptually confusing, with terms such as 'party discipline' and 'party cohesion' used to denote both dependent and independent variables. Moreover, while the literature recognizes both anticipated sanctions and homogeneity of preferences as pathways to party unity, it ignores possibilities such as party loyalty and the division of labour within parliamentary parties. The article examines these different pathways to party unity on the basis of five waves of interviews with nearly all members of the Lower House of the Dutch parliament. The article finds least evidence for sanctions as a major determinant of party unity, with the possible exception of parties in the governing coalition. Homogeneity, loyalty and division of labour all seem to play an important role. Party unity might seem over determined, but the more likely explanation offered is that it is a case of 'different horses for different courses'.

Delegation from cabinet to ministers is complicated because the cabinet consists of the same ministers that are supposed to act as its agents. In the extreme case ministers are completely autonomous within their portfolio. This paper argues that the resulting potential for agency loss is limited, but not negated, by both hierarchy and collective decision-making in cabinet, or by establishing direct delegation relationships between legislative committees or political parties and ministers, bypassing the government. Appointments to ministerial office are the prevailing exception to ministerial autonomy. To the extent that ministerial preferences are not stable and exogenous, screening before appointments is an ineffective control, and ministerial identification with departmental interests is the most probable source of agency loss.

The mandate-independence controversy still features prominently in studies of political representation even though the problems with its theoretical foundation and empirical operationalization have long been recognized. This article proposes an alternative typology of modes of representation. By combining type of control (ex ante or ex post) with direction of the interactions (bottom-up or top-down), our study captures the most important aspects of the relationship between voters and representatives. We demonstrate how the typology can be used in a survey instrument by comparing the attitudes toward representation of Dutch members of Parliament with the attitudes held by voters, and by relating the views of the members to their behavior.

Consociationalist theory served initially as an explanation of political stability in a few deeply divided European democracies. It argued that in these countries, the destabilizing effects of subcultural segmentation are neutralized at the elite level by embracing non-majoritarian mechanisms for conflict resolution. The theory was extended as new consociational democracies were discovered, as the related but broader concept of "consensus democracy" was introduced, and as a normative component was added, recommending consociational engineering as the most promising way to achieve stable democracy in strongly segmented societies. Consociationalism has always been controversial, but rather than one great debate about its validity, there have been many small debates about the countries, the concepts, the causes, and the consequences associated with consociationalism. These debates can become more fruitful if consociational theory is formulated less inductively and at a higher level of abstraction, and if the critics of consociationalism focus more on its principles and less on the operationalizations provided by its most important theorist, Arend Lijphart. The erosion of social cleavages in many consociational democracies raises the question of whether the very logic of consociationalism should lead to a prescription of more adversarial politics in those countries. THE PUZZLE OF STABLE DEMOCRACYThe puzzle of stable democracy is this: Stability is fostered by the absence of conflict in society, and thus by social and cultural homogeneity; representative democracy, on the other hand, presupposes at least a modicum of disagreement and contestation. The same conflicting views that are the lifeblood of democracy are also threatening to its stability (Diamond 1993:24,29-32). The pluralist theory of cross pressures provides an answer to the simultaneous needs for homogeneity and heterogeneity. Social cleavages can be rendered harmless by cross-cutting each other, i.e. social groups that are homogeneous with respect to one social cleavage are heterogeneous with respect to another. Thus, in his trade union a church member interacts with secular working-class comrades, and in his church he encounters upper-and middle-class brethren. The individual is pulled in different directions; he or she is cross-pressured, experiencing cross-cutting loyalties,