The introduction of compulsory competitive tendering in UK local government reflected conventional public choice theories of bureaucracy. These theories suggest that competition for service contracts will lead to lower expenditure and higher efficiency. Extensions and criticisms of the classical public choice model yield very different conclusions. First, the problems of bureaucratic supply are less severe than originally assumed, and secondly competition may itself generate new problems such as transaction costs, erosion of trust between principals and agents, and rent seeking. The impact of competition on spending and efficiency is therefore indeterminate, and must be established empirically. However, studies which evaluate the effects of competitive tendering in local government are few in number, cover a limited range of services, and are methodologically flawed. Therefore neither the initial imposition of competitive tendering by the Conservatives, nor its planned abolition by Labour, can be traced to a solid foundation of theoretical or empirical support. Theoretical and methodological problems that need to be resolved by further empirical studies are identified.