Not Always on an Equal Footing: The Distribution of Voting Power as an Aspect of Fragmented Budgetary Politics

A large literature links the ‘fragmentation’ of decision making to potentially problematic budgetary outcomes. In short, when a number of actors, such as political parties, can influence the collectively adopted policy package, budgets arguably tend to expand and deficits become persistent even if those outcomes lack justification in macroeconomic or socio-economic circumstances. This is because each actor only considers a fraction of the costs associated with its spending priorities, and the mismatch between perceived benefits and costs becomes larger as the number of decision makers increases. In parliamentary democracies, the number of government parties is often argued to be especially relevant in this respect: the more parties there are in government, the stronger is the growth of public spending, taxation and debt. The role of procedural and fiscal rules in dampening these effects has been addressed in previous research. This paper approaches fragmented decision making from a different angle and focusses on the bargaining resources at parties’ disposal. In this paper, it is argued that one should consider not only the number of government parties but also the distribution of power among them. Power is here defined as a priori voting power or the ability to make or break winning coalitions, that ability depending on the decision rule and the distribution of voting weights (i.e. seat shares) in the parliament. Such voting power is crucial as it also affects the bargaining leverage that individual parties have in negotiations on the composition of budgets. The number of government parties should be relevant with respect to budgetary outcomes insofar as power is evenly distributed among them, so that each of them is able to affect policies but unable to prevent others from doing so. In contrast, when power is centred in the hands of one party or only a few parties, the number of parties should be largely irrelevant. This hypothesis is tested using data that covers the current 28 member states of the European Union. Alongside the distribution of power, a host of other factors are also taken into account, including historical experiences of communism and liberal democracy, the prevalence of corruption and other forms of partiality in the public sector, and the stringency of fiscal rules. The results suggest that the connections between the number of government parties and budgetary outcomes are indeed sensitive to the distribution of power.