انتخاب کیفیت بیمارستان و ساختار بازار در یک انحصار دوگانه کنترل شده

This paper analyzes the optimal structure of a regulated health care industry in a model in which the regulator cannot enforce what hospitals do (unverifiable quality of health) or does not know what hospitals know (incomplete information about production costs) or both. We show that if quality is unverifiable the choice between monopoly and duopoly does not change with respect to the verifiable case but, if there are fixed costs (assumed to be quality dependent) and the monopoly is the optimal market structure, the quality level of the operative hospital decreases. Asymmetry of information introduces informational rents that can be reduced by increasing the most efficient hospital’s market share. A monopoly is chosen more often.

Financing and organization of health care markets are based on the triangular relationship between three different economic agents: consumers, providers and purchasers of health care. The separation between providers (public and private) and purchasers (health authorities, general practitioners, insurers, etc.) requires provision of health care to be governed by contracts (NHS Reform, Act 1990). These contracts lay down a complex set of behavioral incentives for providers with financial consequences. As Chalkley and Malcomson (1996) recall: “A key issue in contracting for health services is how to design contracts to induce providers to supply appropriate standards of service while keeping costs down. […] The most obvious message from this discussion is that forms of contract need to be tailored to circumstances…”. That is, the form of the optimal contract will crucially depend on market structure, on demand conditions and on information, among other things. The purpose of this paper is to analyze the role of market structure in the relationship between purchasers and providers of health care, whether purchasers can use competition as a device for increasing social welfare and how this role is affected by frequent problems that arise in the health care industry, such as unverifiability of quality and asymmetry of information about costs. The issues discussed here are significant for the health care industry. We attempt to construct a model that reflects competition between hospitals in a public system (such as the NHS system) where a fixed price per patient is paid by the state rather than by patients. Hospital services are differentiated by geographical location, and monitoring of standards or quality control can be very expensive, or indeed impossible for some aspects of service. For some specific hospital services (heart surgery, cancer treatment, etc.) everybody gets treatment and patient demand reacts only to changes in the gap between the quality of service provided by different hospitals. If hospitals care about profits, sole sourcing never achieves quality above the enforceable level, while either competition or contestability (either dual sourcing or sole sourcing with threat of entry) can help to maintain appropriate standards. We show that under complete information about hospitals’ cost and contractible quality, the benchmark, quality provision and market structure depend on patients’ transportation costs as well as on the difference in productivity between hospitals. Given a transportation cost, the more productive hospital is required to produce higher quality and obtains a higher market share than its competitor. If the difference in productivity is great enough, the gap between quality levels is such that all patients prefer to be treated in the more efficient hospital and the less efficient hospital does not provide health care (sole sourcing). The above results serve as a benchmark for assessing the effectiveness of a two-part tariff as the payment mechanism when quality is non-contractible because it is not verifiable. We show that unverifiability of quality does not affect the market structure. Moreover, under dual sourcing the absence of verifiability has no impact on quality levels. Competition between hospitals via quality can be used as a device to maintain an appropriate level of services when that level cannot be verified. The regulator uses a two-part tariff so that competition takes place and, at the same time, hospitals’ profits are limited. This result does not apply under sole sourcing. The absence of verifiability can affect quality levels depending on whether there are fixed costs. If there are no fixed costs, the regulator can indirectly control quality through contestability: verifiability does not matter. However, if fixed costs are considered, the regulator has to decide whether the benefits of contestability outweigh the fixed cost. We show that, in general, it could be better from a social welfare point of view to allow the operative hospital to reduce quality: verifiability matters. Using a principal-agent model we show that when hospitals have private information about costs they have to be paid a rent in order to induce them to reveal their private information truthfully. These rents represent an additional marginal social cost, which decreases with the hospital’s efficiency level. This implies that the more efficient hospital’s market share has to be increased through an increase in the quality gap, in which case sole sourcing is more likely. Finally, we show that unverifiability of quality under incomplete information has the same effects as under complete information. The paper confirms some of the main conclusions of the relevant literature on purchasing health services subject to unverifiable quality and asymmetric information in a model that allows for a health authority to have a choice of market structure. Recent literature on regulation focuses extensively on the analysis of unverifiable quality, asymmetric information about costs and market structure. The main results of the analyses are well known.1 Firstly, unverifiability of quality need not be a problem if demand depends on quality but not on price; a fixed price in addition to a lump sum, a two-part tariff, can achieve efficient quality of services under certain circumstances (Ma, 1994 and Chalkley and Malcomson, 1998a, in the monopoly case). However, if consumers pay directly for services, and demand therefore also depends on prices, verifiability does matter. In this context, Lewis and Sappington (1991) show that the buyer and supplier are strictly better off when quality is verifiable. In Wolinsky (1997) verifiability matters depending on the regulatory regime (either managed competition or regulated monopolies) and on information about production costs.2 Secondly, if the purchaser is less informed about the cost of treatment than the provider the presence of informational rents generates additional costs that prevent optimality from being achieved (Wolinsky, 1997; Auriol and Laffont, 1992; Auriol, 1998, etc.). Moreover, asymmetry of information favors the monopoly structure because firms have fewer incentives to inflate costs when monopoly is chosen more frequently (Auriol and Laffont, 1992; McGuire and Riordan, 1995).3 In particular this paper parallels the papers by Auriol (1998), Wolinsky (1997) and Gravelle (1999). The main difference between the former and this paper is that Auriol (1998) deals with network industries where the good is not free of charge and quality has public-good like features, whereas this paper deals with some health care industries in which the price of care is paid by the state rather than the patient and patients get different quality if they consume from different hospitals. As we do, Wolinsky (1997) considers quality provision in a duopolistic differentiated product industry. He discusses the relative performance of two different regulatory regimes: managed competition and regulated monopolies. Our work resembles the part of his paper that analyzes the managed competition regime, but it differs in the payment mechanism. In Wolinsky (1997), consumers pay the producer directly as soon as they get the good, the usual pattern in a commodity market, and demand therefore depends on prices. In our model prices play no role, as per the usual pattern in health care markets, and the purchaser has to choose a payment mechanism in order to finance health care providers’ costs taking into account that the mechanism chosen creates behavioral incentives for providers. Moreover, the decision of the regulator on market structure is not present in his model. As is the case in the present paper, Gravelle (1999) analyses the role of prospective (case-based) payments for quality provision under different market structures (monopoly, corner, competition). In contrast to this paper, he studies how market structure evolves from entry and what inefficiencies arise. He does not address provider heterogeneity in cost or informational asymmetry. Finally, a recent contribution to health economics literature which, like our model, deals with the use of contestability and its role for the provision of quality is Bös and De Fraja (2002). They analyze a model in which the health authority enters into agreements with outside providers to induce the single health care provider to increase its quality. The paper is organized as follows. In Section 2, we introduce the model. Section 3 characterizes the social optimum under complete information when the regulator is allowed to control all decision variables. This assumption is relaxed in Section 4, where we assume that the quality choice is made by hospitals. In Section 5, we obtain the optimal incentive compatible contract under incomplete information and verifiable quality. In Section 6, we assume incomplete information and unverifiable quality. We compare the solution and the market structure under complete and incomplete information in Section 7. In Section 8, we consider the existence of quality dependent fixed costs and we analyze how these costs affect the result obtained in the previous sections. Conclusions are drawn in Section 9.

We have developed a normative model of regulation in the health industry that focuses on endogenous market structure, analyzing whether one supplier of health services or two is optimal. We have shown that under complete information and verifiable quality dual sourcing can be justified by the existence of transportation costs, whereas productivity differences and fixed costs favor, in general, sole sourcing. An analysis of the impact of unverifiability of quality and of asymmetry of information on hospitals’ quality choice and on market structure leads to the following conclusions: 1. The presence or absence of verifiability of quality does not affect the market structure, moreover unverifiability need not to be a problem if the hospitals’ demand depends on the gap between hospital quality levels (duopoly) because a two part tariff can be structured to ensure optimality constrained to patients’ free choice. If patients observe quality and they have free choice of hospital, hospitals compete for them via quality when the contract has a fixed price per patient. Thus, by an appropriate choice of price, the regulator can induce hospitals to provide the same quality levels as if he directly controls quality. Competition might be seen as a means of controlling quality when it is unverifiable. 2. The above result does not necessarily hold under sole sourcing because under this market structure demand does not react to small enough decreases in quality. We show that it is possible to maintain the optimal monopoly’s quality level by contestability if there are no fixed costs. However, in the presence of fixed costs, contestability may not be given and unverifiability decreases quality. 3. In terms of social welfare, asymmetry of information is equivalent to an increase in the gap between productivity parameters. Under duopoly, the regulator either increases or induces hospitals to increase the gap between quality levels in order to reduce informational rents. The market share of the more efficient hospital increases because its marginal information cost (the increase in the social cost of information rents for a unit increase in its output) is smaller than the marginal information cost of its competitor. This leads in some cases to a change in the optimal market structure from duopoly to monopoly. 4. Asymmetry of information is harmless under monopoly. The quality level of the operative hospital under incomplete information is the same as under complete information. 5. Unverifiability of quality under incomplete information has the same consequences as under complete information. The paper confirms the salient results derived from the literature of optimal regulation of health services in a model that allows a health authority to have a choice of market structure. However, it should be noted that some of the results depend crucially on some of the hypotheses made in the model. Specifically, the assumptions of a covered market and of patients homogeneity play a determinant role in explaining result 1. As Allen and Gertler (1991) and Ellis (1998) show, when severity of illness varies across patients and hospitals are allowed to refuse patients, the presence or absence of verifiability of quality can affect quality provision even if a fixed price per patient is used to finance hospitals. Moreover, as we point out above, result 4 depends crucially on the assumption of a covered market and on the specification of the cost function, which make monopoly informational rents independent of quality.