رقابت بالقوه و عملکرد بخش دولتی

This paper considers the roles of actual and potential competition in private government formation. Higher public sector costs encourage formation. Public sector rent seeking discourages formation. Although aggregate welfare always rises, nonmembers may be harmed by private government, and this effect is larger the greater is the public sector cost disadvantage. The public sector may strategically choose to deter the formation of private government as in conventional models of entry. However, it may also make strategic choices that encourage a competing private government to form. Thus, mixed markets are fundamentally different than their purely private counterparts.

This paper concerns privatization, and like most papers on this subject it must begin with a definition. In characterizing privatization, Vickers and Yarrow (1991) write that there are three kinds: “the transfer to the private sector of state-owned enterprises operating in competitive product markets…; the transfer to the private sector of state-owned enterprises with substantial market power, and [the] contracting out of publicly financed services…to the private sector.” In fact, there is another kind. It involves the creation of so-called ‘private governments,’ where a group of consumers joins together to tax itself in order to provide supplementary services or regulation. They are ‘private’ in that the policies are directed at an exclusive group. They are ‘governments’ in that their policies, and possibly their incentives as well, are like those of traditional public sector institutions. Pack (1992) estimates that there were 900 commercial private governments in various cities in the US in 1992. McKenzie (1996) estimates that there were 150 000 residential private governments in the US in 1993, containing a total of 32 million residents. These organizations provide policing, sanitation, and social services, and they collectively own and manage public facilities like parks and streets. An extreme instance of a residential private government is a gated community (Helsley and Strange, 2000). These communities provide their residents with additional security by denying access to the general public. Egan (1995) reports that as many as 4 000 000 people lived in gated communities in the US in 1995. This paper considers the roles of actual and potential competition on private government formation. Our analysis addresses three very important questions. How does the possibility of private government formation influence the policies pursued by traditional governments? Is formation or its possibility efficient, or even welfare enhancing? How does the incentive system of the public sector influence its interaction with a private government? We will answer these questions using a model that incorporates three important sources of dissatisfaction with the traditional public sector.1 One is heterogeneity in the demand for public services. Traditional governments are usually required to service the entire population, and this forces them to pursue a one-size-fits-all policy. Private government allows high demand consumers to supplement public expenditures. Another is that the public sector may have incentives that conflict with the objectives of the citizenry. Niskanen (1971) argues that bureaucracies may adopt budget maximization as their primary objective, sometimes called ‘expense preference’ behavior. There is evidence (some anecdotal) in support of this hypothesis. Wolfe (1993), pp. 73–74 reports that managers of Germany’s public television and telephone system have stated that their objective was to “raise rates and sales so as to maximize gross revenues,” and that in response to union pressures, British rail adopted an objective of “high employment per unit of service.” Romer and Rosenthal (1979) argue that the outcomes of school referenda in Oregon suggest that school boards seek to maximize their budgets. Private government may also arise in response to response to cost differences between the public and private sectors. These differences might arise because of unionization or because older organizations produce at higher costs than newer ones. Or it may be that public provision is inherently more costly. Wolfe (1993) has compiled a meta-survey of the evidence on these cost differences for a wide range of public (and private) services, drawing from Borcherding et al. (1982), Fitzgerald (1988) and Osborne and Gaebler (1992). The Wolfe survey covers 23 separate activities, ranging from airlines to weather forecasting, and in virtually every case, private provision is shown to be less costly. Wilson (1989, pp. 350) summarizes part of this literature as follows: “There are at least fourteen studies comparing public and private trash collection in the United States, Canada, Switzerland and Japan; in eleven of the cases, private collection was more efficient… The cost of water, street cleaning, ship maintenance, housing construction, school bus operation, and railroad track repair were all lower when done by private firms rather than by government agencies.” More recently, López-de-Silanes et al. (1997, pp. 447–448) write that “private contractors use fewer people than governments do to provide the same service (Savas, 1987), pay 10–20% lower wages, and offer employee benefits that are sometimes orders of magnitude lower than those in government.” Our analysis generates several interesting results. First, although private government is guaranteed to raise both member and aggregate welfare, it need not raise nonmember welfare. Paradoxically, lower nonmember welfare is more likely when the efficiency advantages of private government are greatest. When the public sector knows that some consumers are able to purchase services at a much lower cost, it may encourage them to do so by reducing its provision level. This may hurt nonmembers. This unhappy result (for nonmembers at least) is more likely to occur when the public sector places little weight on the size of its budget and instead is largely concerned with aggregate welfare. Thus, nonmembers might prefer that the public sector pursue an inefficient objective like budget maximization when the public sector must compete with a private government. Second, individual consumer activism may be inefficient. Formation of a private government is almost certain to impact nonmembers. They may be better- or worse-off. This is ignored by members when they choose to be active. Furthermore, in some specifications of the activism decision, individual members may ignore each other in their activism decisions as well. This result is related to a conclusion reached by Grossman and Hart (1980) regarding the ability of takeovers to ensure efficient corporate governance. They note that since takeovers enrich all shareholders, the takeover is a public good, and so may be underproduced. This result is related to our analysis of activism, although the microeconomics of the interactions among agents and the nature of the inefficiency are different. Third, whether or not a private government forms depends on public sector incentives, the cost difference between public and private provision, and on other parameters. When the public sector has strong expense-preferences or when the cost difference is small, there will be no room for private government, and it will not form. Thus, in a model where privatization is endogenous, private government cannot discipline a rent-seeking public sector in the situation where the public sector incentives are most distorted. This is because if a private government were to form, the public sector would continue to spend so much that it would not pay to form a private government in the first place. Conversely, when the public sector has weak expense-preferences or when the cost difference is large, formation is inevitable. In these cases, formation does not depend on strategic interactions between the public sector and the private government. The most interesting case is when the public sector must make a strategic decision regarding private government formation. In some situations, the public sector chooses a high level of output in order to deter formation. This kind of strategic deterrence is parallel to the analysis of entry in an oligopoly. This situation will arise when the public sector has strong expense preferences, but not so strong that formation never pays. But this is not the only possibility. In other situations, the public sector chooses a lower level of output in order to accommodate formation. This situation will arise when expense preferences are weak, but not so weak that formation is inevitable. This possibility is absent in the models of pre-entry behavior in industrial organization (i.e., Tirole, 1988). This is because they provide no model of the mixed public–private system that exists after private government formation, and they consider entry to be motivated only by profit, while the new private governments are likely to have different objectives. In addition to drawing from the industrial organization and privatization literatures, this paper also builds on several papers in public economics. Our model is loosely related to the literature begun by Tiebout (1956) and Buchanan (1965) that deals with group formation and fiscal competition (see Wildasin, 1986). It is more closely related to some of our own work on private government (Helsley and Strange, 1998 and Helsley and Strange, 2000). The present model is quite different in its consideration of cost differences, mixed incentives, and potential competition. The rest of the paper is organized as follows. Section 2 discusses the role of activism in the formation of private government. Section 3 considers a simultaneous game of actual competition between a private government and the public sector. Section 4 considers a sequential game with activism where potential competition matters. Section 5 concludes.

This paper has provided an equilibrium analysis of the formation of a private government and its effect on the traditional public sector. There are a number of important positive conclusions. Private government will be more likely to form the greater is the cost advantage it enjoys and also the greater is the weight that the traditional public sector places on consumer welfare. A larger cost advantage leads to a larger service level for the private government and a smaller one for the traditional public sector. Since this makes private government more attractive for its members, it also increases the amount of activism directed at the creation of private government. Our model also helps to explain why privatization occurs in some areas but not in others. As discussed in López-de-Silanes et al. (1997), the pace of privatization is notoriously uneven. Our model suggests that variations in activism costs, provision costs, and public sector incentives could each lead to different equilibrium outcomes in otherwise similar places. There are also some important normative results. Nonmembers may either gain or lose by the creation of a private government. They are more likely to gain if the public sector places a low weight on consumer welfare or if the cost difference is relatively small. Members always gain, however, and aggregate welfare is always larger. In light of the latter conclusion, it is unfortunate that private government formation is likely to be inefficient. Some of this is because activism is a public good, and therefore may sometimes be underprovided. Some of this is because the public sector commits to an output level that leaves so little room that the private government does not form. This sort of inefficient strategic deterrence is familiar from the industrial organization literature. This is not the only possibility, however. If the cost difference or the public sector’s weight on aggregate welfare is large enough, then there will be strategic accommodation of private government formation. In this case, the public sector responds to potential competition by setting an output level that leaves enough room for the entrant.