ABSTRACT
From the late 1950s to the early 1980s Niskanen, Buchanan and other New Right public-choice writers argued the state’s inexorable tendency to grow. From the mid-1970s the state apparatus stopped growing and seemed to retract. In part it did so because there was a political will at the highest levels to see the boundaries of the state recede. This political will was in part engendered by the New Right ideology-the arguments about the inevitability of state growth stopped that ostensibly relentless process. These arguments virtually self-referentially disproved themselves. But if the seemingly inexorable tendency to state growth could be altered by simple political will at the highest levels, what does this teach us about the usefulness of general models of bureaucracy? Dunleavy (1986:18) suggests that ‘There cannot be any necessary quality about bureaucratic over-supply and budget-maximization if simple change of political will at the helm of representative institutions is enough to make the state apparatus operate in a basically different way.’ This criticism goes straight to the heart of the logic of rational-choice modelling. But it reveals a very deterministic attitude towards rational-choice theory which does not have to be adopted. We could try to explain how bureaucrats who try to budget-maximize may be defeated. Whilst there is pressure upon politicians to make promises to different sections of the population, thus enabling bureaucrats to budget-maximize, once the costs of bureaucracy get too high, the gains from promising the electorate wholesale reform will outweigh the benefits of making specific election pledges to different sections of the population. Niskanen argues that bureaucrats prefer larger budgets to smaller ones and that it is difficult to control their exaggerated claims because of the nature of pressure groups’ rentseeking. But this difficulty is not an iron law of logic, just a difficulty demonstrated by his model.
