Bureaucratic Theory Meets Reality: Public Choice and Service Contracting in U. S. Local Government

Public choice theory suggests that if public officials monopolize service delivery, then the result is oversupply and inefficiency (Blais and Dion, 1992; Jackson, 1982; McMaster and Sawkins, 1996; Mueller, 1989). By contrast, if services are contracted out, then the pressures of a competitive market lead to improved performance. The advocates of service contracting in local government have been influenced strongly by public choice arguments on bureaucracy that were developed in the 1960s (Downs, 1967; Niskanen, 1968, 1971; Tullock, 1965). For example, Savas (1974, 474) argues that most city agencies are monopolies, their staffs are automatically in a position to exercise that monopoly power for their own parochial advantage -- and efficiency is rarely seen as an advantage. In short, we have unwittingly built a system in which the public is at the mercy of its servants (see also Bennett and Johnson, 1979). The of public services by external agencies has grown substantially since the early 1970s. As Miranda and Andersen (1994, 28) note, although the use of contracting varies across service areas, contracts have been used for every service local governments provide. The expansion of contracting out can be viewed as a huge natural experiment that effectively tests the validity of public choice propositions concerning the behavior of public managers who possess monopoly powers. Indeed, seldom has the major practical recommendation of an abstract model of bureaucracy been so widely implemented. Public choice theory is perhaps uniquely blessed (or cursed) by policy developments that its creators may have desired, but could not have foreseen. Reviews of the evidence on local service contracting have generally concluded that it does produce the results that public choice theory predicts. For example, Stein (1990, 498) claims that research on contracting for individual goods and services has consistently found this mode of service arrangement to be more efficient than direct service provision and production (emphasis added). Similarly, Kiewiet (1991, 66) concludes that virtually every study that has ever compared the operations of public bureaus with those of private firms providing identical services has found data to support the inefficiency hypothesis (emphasis added). Finally, Bennett and Johnson (1980, 392-93) assert that exception, the empirical findings indicate that the same level of output could be provided at substantially lower costs if output were produced by the private sector rather than the public sector.... The private sector of public services offers an excellent opportunity for tax reductions without sacrifice of services; the best of all possible worlds, at least in this case, can be achieved (emphasis added; for similar arguments see Ferris, 1986; McGuire, Ohsfeldt, and Van Cott, 1987; Miranda, 1994a; Savas, 1979). Such claims on the beneficial effects of local service contracts have also been made in other countries, for example the U.K. (Boyne, 1998a). This article aims to reappraise the empirical evidence on the effects of service contracting in U.S. local government and to evaluate the implications of this evidence for public choice models of public administration. The first part of the article identifies three specific public choice hypotheses on the impact of service contracting on spending and efficiency. The second part of the article evaluates whether these hypotheses are supported by the empirical evidence on contracting with private organizations and other units of government. It is important to note at the outset that this article evaluates evidence that compares local government agencies and external contractors as producers of public services. The aim is not to compare the relative merits of wholly public and completely private provision of local services (studies that contain evidence on this issue include Hirsch, 1965 and Pier, Vernon, and Wicks, 1974, on refuse collection; Becker and Sloan, 1985, on hospitals; Nelson, 1990, Primeaux, 1977, and Yonker, 1975, on electricity; and Bruggink, 1982 and Teeples and Glyer, 1987, on water supply). …