ABSTRACT
INTRODUCTION The analysis of British governments as social welfare maximisers is not without its critics. The approach ignores the existing voting, political party and bureaucracy arrangements in a democracy. Difficulties inevitably arise in compiling a social welfare function which reflects the individual preferences of British voters and which contains no ‘dictatorial’ decisions. Who, for example, has ever seen a social welfare function for the UK? What does it look like? A simple majority voting rule does not clarify the issue since it is subject to the paradox of voting.1 Let three individuals (1, 2, 3) rank three alternative policies (A, B, C) in order of preference. The policy context could be a vote on alternative regional policies, such as more labour mobility (A), greater mobility of capital (B) or do not intervene (C). The resulting votes might be:
A majority prefers A to B and B to C. If preferences were transitive, a government would conclude that the community prefers A to C. This would be incorrect. A majority of voters prefer C to A so that the social outcome is intransitive! The simple majority-voting rule breaks down when society has to choose between more than two alternatives. Elections are also based on equality of voters so that voting is not ‘weighted’ by the strengths of individual preferences for alternative policies. Furthermore, since elections are general, an individual cannot reveal his ‘feelings’ for any single item within the ‘mix’ of issues presented to the electorate. Also, the social welfare model of British governments implies that politicians have no private motives and faithfully respond to the ‘will of the majority’. These aspects of the political system raise doubts about the explanatory value of the orthodox social welfare maximising view of governments. Since the constitutional arrangements are likely to affect the behaviour of governments, an alternative model is required which incorporates assumptions about the behaviour of voters and political parties. The choice between the two models of government behaviour and policies will depend on which yields the more accurate explanations and predictions.
