Boyne posits that statistical methods used in studies cited by public choice theorists lack critical control variables and a reliable measure of competition and therefore lead to invalid conclusions. Boyne aims to reevaluate the empirical evidence on the effects of service contracting by United States local governments. Despite the deficiencies in the statistical methods of these studies, Boyne notes that public choice supporters have tremendous confidence in their results. Boynes critique begins by identifying three public choice hypotheses regarding the impact of service contracting on efficiency and cost:
Hypothesis 1: Service contracting is associated with lower spending on those services that are produced by an external agency.
The crux of this hypothesis rests on the notion that if monopolies tend to render higher costs, then the consequence of competition should be a lower level of expenditure. Boyne cites an analysis by Niskanen who implied that because monopolies produce twice the level of service required, expenditure under competition would be reduced.
Boynes rebuttal to these notions is that service contracting does not necessarily entail a decline in the level of service output, thus reduced expenditures cannot be assumed. In fact many service contracts explicitly require external agencies to provide the same level of output. Cost savings are proportional to improvements in technical efficiencies.
Hypothesis 2: Competitive tendering is associated with an improvement in the technical efficiency of service production.
Under service contracting, rival agencies submit bids for a specific service. Given a feasible proposal, the lowest bid is likely to win. The cost per unit of a service is expected to fall. In short, the bidding process itself generates alternative unit costs.
Boyne notes that although service contracting may improve technical efficiency, it has no necessary impact on allocative efficiency. Decision-making for the type of services to be provided and their distribution rest with politicians and other officials. Service contracting thus may not enhance the power of taxpayers, consumers, or the public.
Hypothesis 3: A substantial part of any efficiency gain from service contracting will be retained by local government.
Some part of the money saved by service contracting is likely to be returned to the taxpayerslower taxes or high services.
Boyne questions how much of the savings from contracting is returned to local government budgets. One hypothesis is that the retention rate for efficiency savings will be at least as great as for grants-in-aid. Another hypothesis is that if expenditure on services that are contracted out decreases, expenditure on the remaining local government responsibilities will increase, thus the locally supplied services may be oversupplied to the point where they become less efficient. It is possible the service contracting does not eliminate inefficiency; rather, it relocates the inefficiency in local service production.
Boyne then presents an examination of empirical evidence used by public choice theorists. He notes that in some studies authors draw conclusions that are a) not substantiated by their own evidence, b) weakened by the failure to control for quality, local preferences, and competition levels. He does not blame the theory but notes that some of the empirical evidence it leans on is weakened by its own findings.
