رقابت کیفیت با ارائه دهندگان انگیزه و تقاضای کند
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|6869||2013||21 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Economic Dynamics and Control, Available online 9 May 2013
We study incentives for quality provision in markets where providers are motivated (semi-altruistic); prices are regulated and firms are funded by a combination of block grants and unit prices; competition is based on quality, and demand adjusts sluggishly. Health or education are sectors in which the mentioned features are the rule. We show that the presence of motivated providers makes dynamic competition tougher, resulting in higher steady-state levels of quality in the closed-loop solution than in the benchmark open-loop solution, if the price is sufficiently high. However, this result is reversed if the price is sufficiently low (and below unit costs). Sufficiently low prices also imply that a reduction in demand sluggishness will lead to lower steady-state quality. Prices below unit costs will nevertheless be welfare optimal if the providers are sufficiently motivated.
In markets for health care or education, prices are often regulated and consumer choices are mainly based on other criteria, such as travelling distance and quality. In both types of markets, competition between publicly funded providers has become an increasingly topical policy issue in recent years, as an increasing number of countries have introduced market-based reforms which give providers (hospitals or schools) incentives to compete for consumers (patients or students).1 This has, in turn, spurred a considerable body of theoretical literature studying the nature of quality competition in regulated markets.2 However, with very few exceptions, this literature has ignored two arguably important features of such markets, namely motivated providers and sluggish demand. In the literature on health care supply, it has long been recognised that providers may exhibit semi-altruistic preferences.3 For example, physicians are typically portrayed as ‘imperfect agents’ for their patients, trading off patient benefits against lower profits (see e.g., McGuire, 2000). This notion has in recent years been complemented by an emerging literature on motivated agents in the broader public sector, where the assumption of ‘mission oriented’ workers (doctors, nurses, teachers) implies that the agents (e.g., hospitals or schools) to some extent share the objectives of the principal (government, in our examples).4 Despite the emphasis given in the literature to the importance of motivated providers in the public sector in general, and in sectors like health care and education in particular, this aspect is largely absent in the existing literature on quality competition between publicly funded providers.5 A notable recent exception is Brekke et al. (2011), who analyse within a static model hospital competition with regulated prices and show that the presence of provider motivation can potentially reverse a previously established positive relationship between competition and quality.6 For both health care and education, quality is a key market variable. In health care, since consumers are insured against medical expenditures, the quality of care is usually a much more relevant variable than price for the patient's choice of provider. Similarly, in education markets tuition fees play a relatively minor role in most European countries (though they are on the rise in several countries like England or Italy), and the quality of the institution is typically much more important for the student's choice of school or university. However, since quality is much less readily observable than prices, it is also reasonable to assume that demand adjusts much more sluggishly to quality changes than to price changes. This effect may be particularly strong in the context of health care or education, due to consumer habits or trust in specific providers. If consumers have sluggish beliefs about quality, demand will adjust sluggishly to quality changes, implying that it takes some time before the potential demand increase due to an increase in quality is fully realised. The assumption that demand is sluggish and adjusts slowly to variations in providers' quality is consistent with existing empirical evidence. The assumption that providers' demand responds to quality has been tested empirically by modelling patients' choice of a hospital among a set of alternative ones using conditional logit models (Folland, 1983, Luft et al., 1990, Burns and Wholey, 1992, Hodgkin, 1996, Tay, 2003, Howard, 2005 and Sivey, 2012). These studies find that higher quality and shorter distance increase the probability of choosing a provider. Distance to hospital is systematically a key predictor of patients' choice. Critically, demand elasticities with respect to quality are positive but small for most procedures and conditions.7 Moreover, these empirical models specify provider choice as a function of past quality measures (as proxied by mortality rates, readmission rates, complication rates and waiting times) therefore introducing a dynamic element. Both the low elasticities of demand with respect to quality and the empirical specification of demand as a function of lagged quality are consistent with our dynamic modelling choice of demand as sluggish. The implications of sluggish demand for quality competition in regulated markets are analysed by Brekke et al. (2012a), using a differential-game framework where providers choose qualities in each period and demand adjusts sluggishly over time. However, that paper follows the standard assumption in the literature on quality competition in regulated markets by assuming that providers are pure profit-maximisers. In the present paper we combine the two above-mentioned key features – motivated providers and sluggish demand – in a differential-game framework where providers are funded by a combination of block grants and unit prices, and compete on quality. We consider two different solution concepts: the open-loop and the feedback closed-loop solutions. In our differential game setting, the solution rules adopted by agents (i.e., the providers) capture the intensity of competition. Under the open-loop solution, each firm commits to an optimal quality plan at the beginning of the time considered, and then sticks to it forever. This solution is plausible within institutional contexts where investments in quality are decided infrequently. Under the feedback closed-loop solution, each firm knows the quality of the competitor at each point in time, not just the initial state. The key difference lies in the degree of commitment. The feedback solution is often interpreted as a more competitive solution because firms can change their quality at each point in time in response to the quality of the competitor. The purpose of our analysis is three-fold. First, we derive the levels of quality in the different solution concepts off steady state as a function of the demand, and then compare the steady-state levels of quality in the different solution concepts to see whether more intense competition (under the closed-loop rule) yields higher quality levels. Second, we perform a welfare analysis where we derive the first-best optimal quality, both in and off steady state, and show how the optimal solution can be achieved by optimal price regulation, depending on the dynamic decision rules used by the providers. Third, we investigate the effect on steady-state quality of increasing alternative measures of the degree of competition, either through lower travelling costs (increased substitutability) or less demand sluggishness. Throughout the analysis, a main concern for us is to show how the degree of provider motivation qualitatively affects our results. When comparing the open-loop and closed-loop solutions of the game, we show that the presence of motivated providers changes the dynamic nature of quality competition and therefore makes a substantial difference. We find that steady-state quality is higher when competition is more intense (i.e., under the closed-loop solution when providers consider the interaction with their opponents at each point of time) if the price is sufficiently high. On the other hand, if price is sufficiently low, and below unit costs, this result is reversed, with steady-state quality being higher in the open-loop benchmark case. Therefore, we show that the design of the provider payment system, i.e., the combination of block grants and unit prices, plays a crucial role in determining the outcome of dynamic quality competition. This is of policy relevance since payment systems change across countries and over time. For example, hospitals in England used to be paid according to a block grant (with effectively a zero unit price) but are now paid according to an activity-based funding rule where the price varies for each type of procedure performed (known as HRG—Healthcare Resource Groups). Similarly, within Medicare in the US hospitals are paid according to a DRG – Diagnosis Related Groups – payment system. A constant unit price is paid by Medicare for each patient treated in each hospital: the price varies across different diagnoses and procedures but not with the volume of patients treated. An interesting case is Norway where for several years prices have been set at a level which ranges between 40% and 60% of the average cost: the price is set every year by the government and has been set at either 40%, 55% or 60%. Also in this case the unit price is constant but low, so that it has to be combined with a fixed budget component (otherwise the hospital would make systematically negative profits). In the past, some European countries like Italy and Spain have experimented with pricing rules where the unit price drops to 20–30% after a certain volume of activity has been reached: given high demand levels many hospitals are effectively operating at these lower unit prices. In Italy, recently, a DRG payment system with constant unit prices has been preferred in most regions (both Italy and Spain have decentralised payment rules and arrangements on payment rules may vary across different regions). We also find that, off steady state, quality (under both solution concepts) is higher for the provider with highest demand. Intuitively, if the initial demand is high, the marginal benefit from quality through the altruistic motive is high since quality affects a larger number of consumers. Thus, the provider has a stronger incentive to provide quality. Along the equilibrium dynamic path, as demand reduces over time, the provider's incentive to invest in quality reduces correspondingly, while the opposite is true for the rival provider. This result may have implications for the empirical literature mentioned above which finds that higher quality increases demand. It suggests that quality is endogenous and is a function of demand. Therefore, the positive effect of quality on demand may be biased upwards and the true responsiveness of demand to quality may be even lower than the one estimated in the current literature. In the welfare analysis, we show that the optimal price which implements first-best quality in steady state is higher under open-loop behaviour than under closed-loop behaviour, unless providers are highly motivated. All else equal, a higher price stimulates quality provision, under either solution concept. For low degrees of provider altruism, the optimal price must be relatively high in order to give the providers a sufficiently strong incentive to provide quality at the welfare-optimal level. And under either solution concept, the optimal price is in the range where steady state quality is higher in the closed-loop than in the open-loop solution, as discussed above. Therefore, the optimal price must be higher if the providers use open-loop (rather than closed-loop) decision rules. This result is reversed for a sufficiently high degree of altruism, where the optimal price is lower and in the range where steady state quality is lower in the closed-loop than in the open-loop solution. We also show that, if providers are sufficiently motivated, they are optimally funded by a combination of block grants and unit prices, where the price does not fully cover unit costs. This is important given the range of payments used in practice across a range of countries. Furthermore, the scope for the optimal price to be below unit costs is larger if providers use the closed-loop decision rule. In markets with sluggish demand, policy makers can try to reduce demand sluggishness by collecting and publishing quality indicators on a regular basis, in order to make consumers more aware of quality differences between providers. We show that if prices are above unit costs less sluggish demand increases quality, as intuitively expected. Less intuitively, the result also holds if prices are below unit costs in the presence of altruistic firms since the altruistic motive induces them to attract consumers despite the negative cost margin. However, there is a price that it is sufficiently below unit costs such that the profit motive dominates and the result is reversed: less sluggish demand will reduce steady-state quality. In such cases, as a policy measure to stimulate quality competition, publishing quality indicators to reduce demand sluggishness may be counterproductive. These results are important for policy development and design since quality indicators are increasingly used in the health care and education sectors. For example, in the health sector report cards provide mortality rates and readmission rates for specific procedures, such as coronary bypass. In the US the State of New York was among the first to introduce such cards and for this reason has been intensively investigated in the empirical literature. There is evidence that market shares may be influenced by report cards with providers with better reports having larger market shares. Cutler et al. (2004) show that hospitals with high mortality rates experience a 10% reduction in coronary bypasses but this is not the case for hospitals with low mortality rates. Mukamel et al. (2005) also finds that higher mortality rates reduce market shares. Dranove and Sfekas (2008) find that hospitals with bad reports have a smaller market share but only after accounting for the prior beliefs of the patients: i.e., the market share reduces (or increases) only if the report cards convey information which is different from the one that patients expect based on reputation and advice from family doctors. A number of hospital quality indicators in terms of mortality and re-admission rates have also been reported in England for several years (e.g., by Dr. Foster hospital guide and website). Similar analytical results are obtained when focussing on policies which reduce transportation costs, as for example in Norway where patients' travelling costs are partially reimbursed by the public funder. Lower travelling costs lead to higher quality if the price-cost margin is not too negative. In our study we assume that providers can adjust quality instantaneously. It may be argued that quality in itself is a stock variable which increases or decreases over time depending on quality investment and its depreciation. For example, within the hospital sector, providers can invest in machines, such as CT scanners or MRIs. However, many other important dimensions of quality can be adjusted instantaneously. For example, doctors can adjust quickly the effort they exert in the provision of clinical quality when treating or operating patients, the time spent while visiting the patient and the diagnostic effort, etc. We therefore focus on the latter in this paper. We prefer to model dynamically the demand by assuming that demand is ‘sluggish’ since this seems a more salient feature of the health and education sectors.8 An alternative way to model this assumption is to assume that consumers can observe only noisy signals of quality, e.g., true quality with an error term. This is the approach followed by Montefiori (2005) which assumes that perceived quality is equal to true quality with a normally distributed error term. He shows that patients' utility is decreasing in the variance of the error term and that providers' demand is also a function of the difference in the variance of the error term. However, he shows that as long as the variance of the error term is the same across the two providers, the optimal quality incentives for the providers are not affected (since variances drop out from demand functions). Compared to this and other static models (such as Brekke et al., 2011), a dynamic modelling approach offers a number of insights: (i) derivation of quality off steady state (see our discussion above about the positive association of quality and demand over time and potential implications for empirical estimation of demand functions); (ii) a comparison of open- versus closed-loop solutions, where as mentioned above, the latter can be interpreted as a more competitive environment, therefore providing an alternative definition of competition; (iii) it allows conducting comparative statics with respect to the sluggishness of demand. This parameter is only meaningful within a dynamic model and clearly important from a policy perspective. In a static one, potential demand would always coincide with actual demand and such a parameter would drop from the analysis. The focus of our model is on consumers' quality and motivated providers. In the absence of motivation (or altruism), our quality variable could alternatively be interpreted as resources spent on advertising. Like quality, advertising is costly for the provider and attracts higher demand. However, crucially, the model assumes that providers care about consumers, i.e., they are motivated or altruistic. Since advertising in itself does not improve consumers' utility, it would also not increase providers' utility through their altruistic concerns. Since the presence of provider motivation is a key assumption of the model, we believe this makes our model distinct from any model focussing on advertising. Moreover, within European countries advertising between publicly funded hospitals or schools is limited or inexistent. Advertising seems a key feature of related markets, like the pharmaceutical industry (see Brekke and Kuhn, 2006, for a related study). The remainder of the paper is organised as follows. Section 2 presents the model under open-loop and closed-loop behaviour. It also compares the steady-state outcomes of the different solution concepts. Optimal price regulation and the relationship between competition intensity and steady-state quality are explored in Section 3. Extensions of the model are presented in Section 4. Section 5 concludes the paper.
نتیجه گیری انگلیسی
In this paper we have analysed quality competition between publicly funded providers in markets with sluggish demand, the prime applications of our analysis being health care and education (hospital or school competition). We have shown that, in such markets, the presence of provider motivation makes a crucial difference for the dynamic nature of quality competition. In contrast to previous results in the literature, we have shown that steady-state quality is higher under closed-loop rules (when competition is more intense) than under open-loop rules, if the providers face sufficiently high unit prices. Any price in excess of unit costs is sufficient to produce this result. However, the result is reversed if the price is sufficiently below unit costs. In markets with sluggish demand, policy makers can try to reduce demand sluggishness by collecting and publishing quality indicators on a regular basis, in order to make consumers more aware of quality differences between providers. As a policy measure to stimulate quality competition, we have shown that this may be counterproductive if the providers face a price that is below unit costs. Therefore policies with high unit prices and policies which increase information are complements. If the price is sufficiently below unit costs, more quality-responsive demand will reduce quality in steady state, and this is more likely to happen if providers use closed-loop decision rules. Nevertheless, in our welfare analysis we have shown that the optimal design of the provider payment system implies prices below unit costs if the degree of provider motivation is sufficiently high. Our results also have implications for future empirical work. Existing empirical studies model provider choice as a function of quality and find that higher quality increases demand. Our solution under the feedback closed loop solution suggests that providers with higher demand have a stronger incentive to invest in quality due to altruistic concerns. This suggests that quality in such empirical models is endogenous and that current studies may over-estimate (the already low) demand elasticities to quality. A key assumption in our analysis is that demand increases with the quality provided. This seems reasonable when thinking about consumers choosing prospectively the provider: in the health sector patients desire to receive care in the hospital with highest quality and health outcomes; in the education sector pupils (and their parents) seek schools with the best teaching. In certain specific cases, the opposite may however hold. For example, in the health sector, poorer quality and prevention may lead to MRSA outbreaks generating a higher arrival of patients: modelling properly this scenario would require an epidemiology model of the SIR (Susceptible, Infectious, Recovered) type and is left for future research.