Ro INNO TWD HEALTH PLANNING An Issues Paper Judith R. Gelman Division of Industry Analysis Bureau of Economics Federal Trade Commission April 1982 Disclaimer: The opinions expressed in this paper do not necessarily reflect those of the Federal Trade Commission, individual Commissioners or other members of the FTC staff. ~~ Sh. 1 COMPET TION AND HEALTH - PLANNING. ACKNOWLEDGMENTS I owe many thanks to Sherrie Williams, who served as head secretary on this project in the now-defunct Office of Policy Planning. Her organizational skills, as well as her typing, are much appreciated. Thanks are also due to the Word Processing Center in the Bureau of Economics for help in typing and editing the final product. The advice, assistance, and support of my two knowledgeable colleagues Mike Pollard and Carol Scott helped shape this report. Keith Anderson, Howard Beales, Richard Craswell, Steven Salop, Terry Winslow, and many other members of the FTC staff provided useful comments on various drafts. Lincoln Williams and David Helms of the Alpha Center for Health Planning also assisted me by providing examples and comments. Responsibility for remaining errors and omissions, of course, resides with the author. II. | | TABLE OF CONTENTS Introduction . . . . . . . . A. The Statutory Role of Competition in Health Planning + « +. + + Competition and Planning ee eee ee se eee ee. 1. Special Characteristics of Health . 2. A Note on Antitrust Understanding the Competitive Process . . . A. B. The Competitive Ideal . Immobility of Resources: Entry and Exit Barriers . . . . . . 1. Entry Barriers . . . 2. Exit Barriers . . . Natural Monopoly . . . . 1. Identifying a Natural ee ese eae Monopoly . . . 2. Entry in Markets Characterized as Natural Monopolies eee eee Costly or Inaccessible Information . . . 1. Costs and Methods of Gathering 2. Effects of Costly and Information . . . 3. Solutions to Problems Caused by Costly Incomplete ee se ee se and Incomplete Information . . . wii Information 11 13 14 18 19 25 29 30 34 38 39 40 43 TABLE OF CONTENTS (continued) HY : E. Heterogeneity of Products . . . . . . . . . . . . oe Varieby + « + « « « + «+ vn % vv » w vw + 4 + 2. QUAlILY + « « » © + 3» » + ow ow ow ow WEE x ow a. The Cost/Quality Tradeoff . . . . . . . . b. Quality Differences With Imperfect Information . : + + + + + + « + + + F. Pricing Distortions and Externalities . . . . . . 1. Health Plans: The Effect on Consumers’ Cholo@s + « + © + + + « » @ = % wm = » + « =» 2. Reimbursement Schemes: The Effect on Providers' Decisions . . . . . . . . . . . III. Regulations and Their Interaction With Competition . A. Regulation of Costs and Prices . . . . . . . . . 1. The Purpose of Cost and Price Regulations . . 2. Types of Cost and Price Regulations . . . . . a. Reimbursement Based on the Individual Provider's Costs . . . . . . . . . . . b. Reimbursement Based on Costs of Prototypical Provider . . . . . . . . 3. Effects of Cost and Price Regulation . . . . 4. Alternatives to Price and Cost Regulation for Institutional Providers . . . . . . . 5. Interaction of Price and Cost Regulation With Competition . . . . . . . . . . . . . 54 56 61 64 7% 80 83 84 86 86 88 90 95 97 TABLE OF CONTENTS (continued) B. Regulation of Capital Investment . . . 1. The Purpose of Current Regulatory Schemes. . +. . + + + ¢ ¢ ¢ ¢ 4 os + oo» a. Rate-of-Return Regulation and Increased Investment . b. Reasonable Cost Reimbursement and Increased Investment . . . . . . . c. "Technology" Competition and Overinvestment . . . . . . . . . . d. Roemer's Law and Overinvestment . . 2. Programs to Control Capital Investment 3. Effects of Capital-Investment Regulation a. The Effect of a Programs . . . “eo . . b. The Effect of Section 1122 Review Programs . . . . . . “eee ee c. The Effect of Private Programs . . . d. General Effect of Capital-Investment Regulations . . . . . . . . . . 4. Capital-Investment Regulations and the Competitive Approach . . . . . . . . C. Regulation of Institutional Quality . . . . . 1. The Purpose of Institutional-Quality Regulations . . . . . . . . . . . . . 2. Schemes Regulating Institutional Quality a. Licensure . . « « w 3 » 2 & + © % + « DD. Cortification » «+ « i» w 5 # # + » + + Ce Accreditation « «+ « + « + + « + ig 100 101 103 103 107 109 113 114 124 126 126 126 127 TABLE OF CONTENTS (continued) 3. Empirical Evidence on Regulation of Institutional Quality . . . . . . . . 4. Alternatives to the Current System . . . 5. Interaction of Competition With Institutional-Quality Regulations. . . Regulation of Personnel . . . . + « + « « «+ 1. The Purpose of Regulating Health-Care Personnel . . . . + + + + + ee ee 2. Regulations Addressing the Qualifications of Health-Care Personnel . . . . . . . Bs LICENSULE « »s + » ¢ + w v w « = o + + PD. Certification : « « + + + 5 ss +» + + + 3. Evidence on the Effects of Regulation of Personnel . . « + «+ « + « oo + eo oe 4. Alternatives to Regulating Personnel . . 5. The Interaction of Regulation of Personnel With Competition . . . . . . Regulation of the Provision of Health Care . 1. Types of Regulations Governing Provision OE CATE « uw » » o # %» 5 + + % & & ® » a. Medical Audit . . . +. + + + + + o . b. Utilization Review . . . . . . . . . c. Claims Review . . «. +. « + « « « «+ « 2. The Effect of Regulations Governing Provision of Care . . . . +. « «+ «+ + 3. Alternatives to These Regulations . . . . 4. Interaction of Provision-of-Care Regulations and Competition . . . . . —-—v- 134 135 137 142 144 145 146 146 148 148 149 152 153 TABLE OF CONTENTS (continued) Page Questions About the Market: A Summary . . . . . . . . 155 A. Introduction . . . . . . . . . ..........155 B. Reimbursement Schemes . . . . . . . . . . . . . . 159 C. The Consumer: How Price Sensitive Is Demand for the Service? . . . . . . . . . ...... 166 D. Competitiveness of Providers . . . . . . . . . . . 176 I. INTRODUCTION This paper discusses the potential for increasing the use of competition in local health-care markets. The goal is to provide a guide for analyzing health-care markets and for identifying situations where competitive forces can improve market perform- ance. The paper is aimed principally at planners who work for health systems agencies and State health planning and development agencies. The economic analysis contained in this paper may also assist health benefits officers, health plan administrators, and others concerned with how market forces can be used in regulated health-care markets. For example, understanding the competitive process in health care can help corporate decisionmakers in designing employee benefit packages and reimbursement schemes with incentives for cost containment. Health planners have a special interest in the use of com- petition to improve the allocation of health-care resources because of the admonition to consider competition contained in the 1979 Amendments to the National Health Planning and Resources Development Act. The paper starts with a brief discussion of the statutory basis for the competition approach as it affects planners operating under the 1974 Health Planning Act. The use of competition in the health planning context as an alternative to centralized regulatory schemes is discussed. The main body of the paper consists of three parts. The first section discusses the advantages and disadvantages of competition. It examines the ideal conditions for using competition to allocate resources and the resulting benefits. These ideal conditions exist in few markets. In the health sector, deviations from this ideal commonly involve entry and exit barriers, large economies of scale, costly and inaccessible information, variations in product quality, and third-party-reimbursement schemes. These deviations are discussed at length and their effects on the desirability of using competition-based approaches to particular problems in the health sector are examined. The second section discusses the major alternative to competition--regulation. Five classes of regulations are dis- cussed: (1) regulation of costs and prices, (2) regulation of capital investment, (3) regulation of institutional quality, (4) regulation of personnel, and (5) regulations governing provision of care. Although these regulations commonly coexist, for clarity they are examined separately. The primary purpose of each type of regulation is examined. The regulatory schemes used to achieve these goals and empirical evidence on their effectiveness are dis- cussed. Alternative regulatory schemes and the interaction of each regulatory scheme with competition are also reviewed. The third section contains a list of questions to ask in conducting an analysis of a specific market. Each question is followed by a discussion of its importance in assessing the potential for taking a competitive approach to the market. These questions summarize the material contained in the two preceding sections and, where possible, highlight techniques for improving the market's competitive performance. A. The Statutory Role of Competition in Health Planning Dissatisfaction with the cost, quality, and accessibility of health care has been fairly constant and widespread in the past two decades. In the late 1960's and early 1970's, numerous State and Federal governments responded to this dissatisfaction by adopting regulatory mechanisms. Although regulation of health care was not new, regulatory controls on the health sector grew tremendously during this period. Health planning emerged as one means for coordinating the impact of these regulatory activities and for resolving conflicts among the participants in the process. The Health Planning Act of 1974 created new local entities known as "Health System Agencies" (SA's) .l These private non- profit agencies were intended to serve as advisors on federally mandated State regulatory programs and to develop local plans for health services. At the time HSA's were created, Congress appears to have assumed that competition could not contain costs in the health-care industry.? However, despite increased regulatory and planning activities following passage of the Act, dissatisfaction with the performance of the health-care sector grew. The ability 1 pyblic Law No. 93-641, 42 U.S.C., § 300k-1 to 300k-1-5 (1981). 2 conference Report, H. R. Rep. No. 1640, 93d Cong., 2d sess. (1974), p. 3. -3- of regulatory schemes to allocate health-care resources effi- ciently and to control costs was questioned by some individuals who believed that market mechanisms should be encouraged in cer- tain health-care markets. The 1979 Health Planning and Resource Development Amendments adopted this view, in part, by encouraging planners to promote "competitive forces". The 1979 Senate Report states: Despite the fact that the health care industry has not to date responded to classic market- place forces, the committee believes that the planning process--at the federal, state and local level--should encourage competitive forces in the health services industry where- ever competition and consumer choice can con- structively serve to advance the purpose of quality assurance and cost effectiveness. . « . With these provisions the Committee introduces into the health planning law the recognition that planning agencies at all levels have an important responsibility to promote competition among health providers as well as the obligation to encourage cost containment, facility closure or shared services. The experience with other public regulation authorities have been that of creating industry cartels which emphasize market stability rather than innovation and consumer preference. The Committee intends to protect against such tendencies in health sector regulation. [Emphasis added.]3 Section 1502(b) of the Act outlines Congressional findings concerning the role of competition in health planning and is the key provision requiring planners to promote competition. In this provision, Congress set forth its finding that prevailing methods 3 s. Rep. No. 96, 96th Cong., 1st sess. (1979); reprinted in U.S. C.C.A.N. 1979, No. 2, at 1306. of paying for health services diminish the effect of competition on the supply of health services and facilities, resulting in “duplication and excess supply of certain health services and facilities, particularly in the case of inpatient health services. "4 For inpatient institutional services, Congress concluded that the ability of competition to achieve a desirable allocation of resources may be limited by prevailing payment mechanisms and large economies of scale. The Amendments direct planners to continue their previous functions in allocating the supply of these services.® For noninstitutional and outpatient services--including such services as home health care, out-of- hospital laboratory tests, and ambulatory care--Congress found that a competitive market might already exist or might be created. In such situations, planning agencies are instructed to "take action which would strengthen the effect of competition on the supply of services."?’ The Act further directs HSA's to assess the ability of competition to allocate resources efficiently in 4 public Law No. 96-79, § 1502(b)(1). 5 gee discussions in sections II.C., II.F., and IV.A. 6 public Law No. 96-79, § 1502(b)(2)(1979). 7 public Law No. 96-79, § 1502(b)(3)(1979). particular markets when developing plans, reviewing certificate- of-need (CON) applications, and conducting Appropriateness Review for these services.8 Because the various classes of regulations interact, it is difficult to quickly institute a competitive approach in an industry as heavily regulated as health care. As discussed in section III of the paper, reforming these regulatory schemes is often the most direct way to institute a competitive approach to the market. However, the administration and reform of most regulations are outside the scope of the authority of most planners. If Federal and State health-care regulations change, planners may be increasingly able to incorporate the competitive approach in their decisionmaking. For example, prevailing financ- ing mechanisms in this sector diminish the sensitivity of demand and supply to market forces. If payment mechanisms are restruc- tured to communicate the correct incentives to consumers and providers, the benefits of using competition in the market for inpatient services may increase.? The 1979 Amendments give planners little specific guidance on how to incorporate competition into actual planning decisions. 8 see Public Law No. 96-79, § 1513(a)(11), (12), and (15) for planners' specific responsibilities in preserving and improving competition in the health services area. These sections also instruct Federal regulators to consider improvements and innova- tions that foster competition and advance the purposes of quality assurance and cost effectiveness in promulgating new regulations. 9 see H. R. Rep. No. 190, 96th Cong., lst sess. (1979), pp. 5-7. -6- The Senate Report on the 1979 Amendments outlines the thinking underlying the Amendments and provides some examples of actions planners might take; but neither the Act nor the Report tells planners how to assess the appropriateness of using competition to allocate specific resources in specific markets. The purpose of this paper is to provide general guidance for conducting such assessments. 10 The remainder of the paper outlines the factors to consider in assessing the competitive potential of particular services in particular geographic areas. Some HSA's have already begun this evaluation, and (as they are well aware) there is no single correct set of questions to be asked. In addition, the answers differ substantially from one geographic area to the next and from one service market to another. It is hoped that this paper can guide planners and others concerned with the health sector as they develop their own procedures for encouraging competition, while ensuring the availability of accessible, cost-effective care in their planning areas. 10 Lawyers, planners, economists, and health-policy experts disagree about the extent to which the 1979 Amendments obligate planners to use competition as a tool. At the center of this dispute is a legal issue of what the law requires and a policy issue of how the goals of health planning can best be achieved. This paper takes no position on legal obligations to use competition. On the policy issue, the author supports the view that whenever possible, regulations should be reformed to create market incentives and to allow the use of competitive approaches. wi B. Competition and Planning Planning and competition are both approaches to allocating resources. The results expected from the competitive process are often desired under a planned system, but different mechanisms are ordinarily used to achieve them. Planning typically uses centralized decisionmaking to direct the allocation of resources and ensure the accessibility of services. This approach relies largely on a regulatory system rather than on competitive forces in the market to achieve its goals. Under a basically regulatory system, it is typically difficult to encourage consumer choice. In contrast to regulation, competition relies on the decen- tralized decisions of individual buyers and sellers to determine who gets what product at what price. Competitive systems typic- ally achieve efficient resource allocation and consumer choice without administrative intervention in the market. However, to achieve other social goals, such as universal access to health services, income redistribution schemes may be necessary in a competitive market. In many cases, a combination of centralized and decentralized decisionmaking techniques may be the most appropriate approach. In the planning context, using competition means identifying those functions best performed centrally and those best left to the market. In determining the best approach to a particular market, factors to consider include: the amount of information available to consumers; the degree to which insurance blunts price sensitivity; the ability of consumers to assess their own needs and to search for appropriate care; and the ability of regulation to create appropriate incentives. Each of these factors is dis- cussed in detail in this paper. In every case, the question is whether it is possible to design regulations that result in outcomes superior to those resulting from unfettered market interactions. 1. Special characteristics of health. Opposition to the introduction of competition into health-care markets is often based on the belief that "health is different" and that it is too important to be left to the decentralized decisionmaking process of the market.ll Health care does differ from other service industries in a number of ways, but it still offers many opportunities for competition. Demand for most health services results from illness, which is often unpredictable. Some illness has the potential to inflict tremendous harm, even death. In addition, consumers' desires to avoid unexpected financial liabilities make insurance an important aspect of the demand for health care. Urgency in receiving treatment, combined with the prevalence of insurance, reduces price sensitivity in the demand for many health services. 11 For an excellent discussion of the elements that make health care different from other goods or services, see K. J. Arrow, "Uncertainity and the Welfare Economics of Medical Care," Am. Econ. Rev. 53 (1963):941-69. Receiving medical attention does not necessarily assure restoration of health. Not only does the efficacy of even the best treatment vary from case to case, but medical personnel may differ in their competence in diagnosing and administering treatment. Patients may be unable to verify the quality of care they receive. As a result, faith and trust are often integral parts of the physician-patient relationship. This means that medical ethics and quality regulations can play important roles in assuring that patients neither undergo unnecessary treatment nor suffer harm from low-quality care. While the juxtaposition and gravity of these factors are probably unique to health care, similar problems occur in other industries, too. Concerns about quality and professional ethics are important in many service industries, from legal services to auto repairs. Unpredictable demand and potentially large consumer injury also characterize the market for repairs of various sorts. Demand in many other markets also is affected by insurance. In no other service industry has regulation supplanted competition to the degree it has in health care. Yet most markets sharing characteristics found in the health-care sector operate reasonably well with less governmental intervention and with fewer regulatory restraints. While few would contend that competition is appropriate for all segments of the health industry, there is no reason to think that regulation is appropriate in all health- care markets either. The most suitable approach must be determined on a service-by-service basis. Often the most suitable wm Qe approach involves a mix of competition and regulation. To decide whether competition or regulation or some combination of tech- niques provides the best allocation of resources in a particular market, planners and others concerned with the allocation of resources in the health sector should strive to understand both the effect of certain market characteristics on the efficacy of competition and the impact and efficacy of regulations used to remedy perceived problems in the market. 2. A note on antitrust. The decisions planners make about the most desirable mechanism for allocating resources are not made in a vacuum. The antitrust laws affect the relationship between competition and planning in two fundamental ways. 12 First, anti- trust seeks to prevent market participants from engaging in anticompetitive behavior. As such, antitrust is an important enforcement instrument for keeping markets operating competi- tively. Second, antitrust law establishes limits on the tech- niques available for developing and implementing health plans which may be anticompetitive. For example, under the antitrust laws, competing providers cannot decide among themselves how to reduce or allocate health services. Instead, if anticompetitive actions are desired, the action must usually be an explicit part 12 For a brief but clear discussion of antitrust in the health planning context, see J. Sims and K. Grimm, "A Health Care Planner's Guide to Antitrust," Hospital Practice Management, forthcoming. Also see M. Pollard, "The Essential Role of Antitrust in a Competitive Market for Health Services," Milbank Memorial Fund Quarterly/Health and Society 59 (Spring 1981):305-63. -11- 374-115 0 - 82 - 2 : QL 3 of a plan adopted and supervised by the State, with the resulting displacement of competition specifically contemplated, before actions pursuant to the plan are exempt from antitrust scrutiny.13 Antitrust thus affects the techniques that can be used to gain compliance with the health plan. By making anticompetitive actions illegal unless they are specifically authorized and supervised by a State, antitrust may make a planning approach based on informal negotiations among competing providers less attractive than an approach based on competition. 13 In National Gerimedical Hospital and Gerontology Center v. Blue Cross of Kansas City et al., 49 U.S.L.W. 4672 (June 16, 1981), the Supreme Court held that the federal health planning laws do not create a total, blanket exemption from the antitrust laws. It is generally agreed, however, that the planning laws create an implied antitrust exemption in some areas. In addition, the "state action" doctrine limits the antitrust laws applicability when conduct is regulated by the states. wi] Jo II. UNDERSTANDING THE COMPETITIVE PROCESS The antitrust laws and the 1979 Amendments to the Health Planning Act reflect the belief that resources are generally allocated most effectively by markets unhindered by undue concen- trations of market power or by collusive agreements among the actors. To be sure, the economic assumptions of the perfect- competition model, without modification, fit only a few indus- tries. However, the model is presented here in detail because it is a useful tool for gaining a better understanding of the impor- tant factors in the competitive process. In addition, it provides a norm against which to judge the severity of market imperfections. When an industry departs from this norm, the policy question that arises is whether a better allocation of resources can be achieved by using another mechanism in place of or to supplement the market. The answer depends on two factors. First, the degree to which the market in question differs from the ideal is important because minor imperfections are likely to result in correspondingly small inefficiencies. Second, the availability of an alternative resource-allocating mechanism that can be expected to generate more desirable results is crucial to a decision to reject the market. Because regulatory alternatives cannot achieve perfect allocative efficiency, the relevant question becomes, "Which system achieves the 'second-best' result?" In general, the further the institutions of the market are from the competitive ideal, the more likely it is that some regulatory intervention technique may improve the ww) Bue allocation of resources. In many cases, however, the same conditions that create distortions in the operation of the market also thwart the effective administration of regulatory schemes. The six sections in this part of the guide discuss the com- petitive ideal and various market imperfections commonly found in the health market. These include entry and exit barriers, natural monopolies, costly and inaccessible information, problems associ- ated with quality, and pricing distortions created by third-party payment schemes. For the sake of clarity, the discussion below lays out each market imperfection alone and in its most extreme form. The reader should keep in mind that such extreme distor- tions rarely occur. In actual markets, the degree to which the market solution departs from the optimum (or perfectly competi- tive) allocation depends crucially on the degree of the imperfec- tion. When a market has more than one imperfection, the distor- tion is typically compounded; but in many cases, this compounding makes regulatory schemes even more difficult to administer. A. The Competitive Ideal The allocation of resources in the ideal or perfectly compet- itive market has two extremely desirable characteristics. First, all resources used in such a market are put to their most produc- tive uses. That is, there is no alternative allocation of resources which would result in a greater production of all goods and services. Second, the perfectly competitive market allocates resources so that no market participant can be made better off wo Ll without making another participant worse off. While other alloca- tions of resources may be more equitable, the perfectly competi- tive market results in the most efficient use of resources, given the current income distribution. Because there is no resource allocation that all participants would agree is superior, the allocation achieved by the perfectly competitive market is referred to as optimal and it provides a standard for judging the allocations achieved under all other systems.l4 Achieving this optimal allocation depends on the market having a number of characteristics. The importance of a few of these characteristics is explained briefly in the next few para- graphs. The effect of the absence of these characteristics is then investigated in detail in the rest of this part of the paper. The perfectly competitive market is characterized by a large number of well-informed buyers and sellers exchanging identical products. Because products within each market are identical in this theoretical model, participants in the market use price alone in deciding with whom to trade. Because each actor is small relative to the market, he ignores the impact of his own actions on the price charged or received. Under these ideal conditions, no producer would want to sell a product at less than the market price, because each producer is small enough to sell as much as 14 For the reader interested in examining the workings of the perfect-competition model further, see George J. Stigler, The Theory of Price, 3d ed. (New York: Macmillan Publishing Co., Inc., 1966), especially Ch. 10: "The General Theory of Competitive Prices." -15- desired at the market price. Charging more than the market price results in no sales at all. Thus, perfect information about prices and qualities, together with a large number of market participants, maintains a single price in the market. Easy entry and exit are important characteristics of the perfectly competitive market. Easy entry insures that any excess profits are quickly competed away, because temporary profits attract additional resources to the market. Entry stops when resources in the industry earn the same rate of return available in their next-best use. Easy exit means that losses are only temporary as well, because individuals can shift their resources to other, more profitable uses. Industries with production processes that rely on physically mobile resources and resources with multiple uses generally have these characteristics.l5 In such industries, potential entry alone can keep prices, quality, and profits at the competitive level. Finally, actors in perfectly competitive markets pay the full cost of the resources they use. In such a market, consumers choose which goods and services to purchase by balancing their decisions in light of their limited budgets. Whether this budget is determined by their own earnings or by a transfer payment from 15 por example, airlines can quickly shift planes from one market to another, according to changes in demand. While not literally mobile, fertile land often has several uses, and the farmer's choice of what crop to plant may largely depend on crop price forecasts. In hospitals, beds easily shifted from the medical service to the psychiatric service are an example of a mobile resource. -16= the Government or an insurance company, the cost of their choices must be reflected dollar for dollar in the prices they pay. For prices to communicate the correct incentives, the consumer must not receive a larger income or transfer payment when he chooses one alternative rather than another.l® The perfectly competitive market yields the most efficient resource allocation because each participant makes the choices best suited to his own tastes in reaction to incentives created by the market. In short, in the ideal market, prices play an important role in resource alloca- tion. They reflect the full cost of each market participant's decisions--including the effect of his actions on other market participants. The description of the perfectly competitive market contains many assumptions that do not always apply to conditions in most actual markets. Characteristics found in the health sector that violate these assumptions include immobility of resources, natural monopoly, heterogeneity of products, costly or inaccessible information, and pricing distortions and externalities. The discussion that follows explains how each of these characteristics affects the resource allocation in markets in which it occurs. It also explains the conditions under which specific forms of regulatory intervention by planners might be appropriate. 16 This condition is not met when, for example, consumers pay a higher copayment when a procedure is performed on an outpatient basis than when it is performed on an inpatient basis. -17- B. Immobility of Resources: Entry and Exit Barriers When a perfectly competitive market is in equilibrium, sellers earn the same rate of return they would earn were their resources invested in their next-best use. This condition is often called the "zero" profit condition because sellers earn no profits above the "normal" rate of return--the rate just necessary to keep them in the market. Profits above this rate act as a signal to sellers already in the market to expand capacity and as a signal to additional potential sellers to enter the market. When expansion and/or entry can occur freely, equilibrium is quickly reestablished after any increase in demand or decline in costs, with sellers once again earning the normal rate of return. Losses--profits below the normal rate of return--induce sellers to leave the market. When sellers can freely leave one market and enter another, losses are short-lived, because a sufficient number of sellers quickly exit, restoring the equilibrium profit level. Because deviations from the "normal" profit rate act as a signal for entry or exit, the profit rate serves an important allocative function in competitive markets. Conversely, the ability of the competitive market to allocate resources effi- ciently depends upon the ability of profits to induce entry and exit. When the mobility of resources is limited, the speed with which deviations in the profit rate induce entry and exit is reduced, and the market adjusts more slowly to the new resource allocation dictated by changes in supply or demand conditions. LB "Entry and exit barriers" refers to legal, institutional, technological, financial, and marketing factors that impede the free flow of resources. Such barriers--many of which are created by regulations--occur throughout the economy. Their impact on the ability of a particular market to achieve the optimal resource allocation depends on the degree to which the barriers prevent buyers and sellers from responding to deviations from the normal rate of return. This section discusses the effect of entry and exit barriers on resource allocation, the types of barriers frequently found in health-care markets, the benefits of competition in markets with barriers, and the role for planners in lowering or eliminating barriers. 1. Entry barriers. One of the key purposes of the 1974 Health Planning Act was to create entry barriers in the form of certificate-of-need (CON) requirements.l” Because a major function of Health Systems Agency planners is to make recommenda- tions on CON applications, and because loosening the interpreta- tion of "need" is a major tool planners have for adopting a competitive approach to particular markets, the subject of entry barriers is particularly important to planners. Other entry barriers found in the health sector include large capital requirements, limited access to specialized resources, and such regulatory restrictions as licensure. 17 For discussion of CON and other regulations limiting capital investment, see section III.B. 1 Ge Entry barriers restrict the ability of additional providers to respond to high rates of return because they thwart access by new suppliers to the market. When entry cannot occur, providers already in the market can respond to an increase in demand by raising prices. Providers protected from competition from entrants have less incentive to pass along cost reductions to consumers, to maintain quality levels, or to produce the most desired mix of services. When the market already has a large number of sellers capable of freely expanding capacity, restrictions on entry may not create serious distortions, because competition among existing providers may serve to keep price in line with costs and to maintain quality levels.l8 However, the threat of entry by new sellers is the most effective force for insuring competitive results because it reduces the rewards to market participants who cooperate to maintain high profits.l9 Access to large amounts of capital is a prerequisite for entering many markets throughout the economy. This requirement 18 For a full discussion of quality issues, see section II.E.2. 19 A tacit agreement among existing providers to raise rates might increase profits in the short run. In a market with free entry, however, additional providers would quickly respond to the high profit rate by entering the market. To gain customers, the new entrants might compete by lowering rates or raising quality. The incumbents participating in an agreement not to compete would be handicapped in their ability to respond. Knowing that entry is possible whenever profits rise, incumbents might compete harder now for their share of profits rather than cooperating as they might in situations where there were longrun profits to be shared. Where high profits can quickly attract entrants, the gains to collusion are small and short-lived. -20- may successfully block the entry of othewise competent potential entrants into health-related industries, such as manufacturing diagnostic equipment or developing new drugs. Outside the health product and equipment markets, large capital requirements rarely act as a barrier in the health sector. Many firms have access to sufficient amounts of capital to fund undertakings of the size commonly found in hospitals, nursing homes, and other health services. For example, Morton Foods and Del Monte are both major providers of meals for the elderly. Upjohn Pharmaceuticals and ARA Services both have subsidiaries providing home health services and nursing-home management . 20 The entry of these firms into the provision of health-care services suggests that the pool of corporations with access to the resources needed to enter these markets is large. As in other industries, potential entrants may be discouraged from investing in health-care markets when interest rates are high. This situation should not be confused with one in which large capital requirements act as a barrier to entry. Rather it may reflect the fact that the rate of return in the industry is lower than the cost of capital. Although profits may exist in the industry, entrants borrowing at high rates have higher fixed costs than incumbents and therefore may be at a competitive dis- advantage. A similar disadvantage is created in health markets 20 Health Advocate, no. 122 (July 1981):3. Published by Health Research Group, Public Citizen, Inc., Washington, D.C. mh where one class of providers, such as hospitals, has access to financing at below-market rates through municipal revenue bonds or through cross-subsidies and these sources of funding are not available to another class of providers, such as surgicenters. In some cases, such factors put entrants at a cost disadvantage unrelated to their efficiency in providing the service. Limited access to specialized resources can create entry barriers in some health-care markets. For example, a surgeon's ability to compete in the treatment of certain conditions may depend on access to a computerized tomographic scanner. In markets where a limited number of facilities have this specialized equipment, having privileges at one of those facilities is necessary for performing some services. A policy of limiting access by a facility with unique equipment may create barriers to entry in related markets, such as the market for professional services. In HSA's where planners review certificate-of-need proposals for such equipment, taking the access policy of facili- ties into account may alleviate the severity of such entry barrier. Most of the entry barriers in the health-care sector, includ- ing the example of the scanner (above), are imposed by regulations and other forms of governmental involvement. As mentioned earlier, certificate-of-need laws were specifically designed to create entry barriers for those facilities and services to which De they apply .21 State occupational licensure poses a serious entry barrier for individuals wishing to compete in the supply of health-care services in many markets. Federal programs and some State programs limit the number of health professionals entering the market by placing limits on the size of medical and nursing schools and on the number of students entering medical specialties, as a condition for receiving funding. 22 While some of these barriers may serve legitimate functions--such as assuring that new entrants meet certain minimum levels of competence--less restrictive alternatives which allow competition to better perform its allocative function are sometimes available.23 One major impediment to entry for health-care professionals has been eroded. Until recently, many States had laws prohibiting advertising by professionals. 24 These laws reduced the ability of professional providers to attract new customers, thereby 21 gee section IV.B. for further discussion. 22 Because medical education has been heavily subsidized, there may be too many students relative to the social costs of their education. Some controversy exists over whether the combination of entry restrictions and subsidies raises or lowers the number of students, relative to what it would be without governmental inter- vention. Because more students can afford a subsidized education, schools can be more selective. Thus, it is argued that the combination of subsidies and restrictions may raise the quality of medical students. 23 gpecific alternatives to licensure--such as certification-- are discussed in the section on regulatory mechanisms. See section IV.D. 24 advertising and its role in information provision are also discussed in section II.D. -23- restricting entry. Most of these laws have been repealed or modified in the past 5 years in the wake of the Bates decision, which held such laws to be unconstitutional restrictions of First Amendment rights of free speech. 25 Professional codes of ethics have also prohibited advertising and other commercial practices. In a case recently upheld by an equally divided Supreme Court, the Federal Trade Commission ruled that private agreements to restrict advertising and prohibit certain forms of commercial activities by professionals violate the antitrust laws. 26 Various additional governmental and private restrictions on commercial practices by professionals exist. Many of these restrictions ban innovative forms of practice, such as locating in retail establishments or practicing under a trade name. Some of these restrictions, such as limits on setting up satellite offices, make it more difficult to market professional services. These restrictions often hinder new entrants more strongly than they hinder incumbents because entrants lack established reputa- tions and thus have more difficulty competing using only practices allowed by the ethical codes. Others, such as bans or severe limits on the use of paraprofessionals to perform routine func- tions, may raise the cost to all professionals of providing services but also may protect their hold on the market. 25 Bates v. State Bar of Arizona, 433 U.S. 350 (1977). 26 American Medical Association, 94 F.T.C. 701 (1979), aff'd, 638 F.2d. 443 (2nd Cir., 1980), aff'd by an equally divided Court, 50 U.S.L.W. 4313 (U.S., March 23, 1982). -24- Restrictions on commercial aspects of professional practice may limit the scope of innovation in some service markets.27 2. Exit barriers. In some States, certificates of need are required when health facilities reduce the services they provide as well as when they expand them. In these States, CON consti- tutes an exit barrier as well as an entry barrier. Exit barriers are factors imposing costs on firms that would otherwise leave a market. In many instances, exit barriers in one market are caused by entry barriers in all other markets where the resource might be used. In States where no CON is needed to close down one service facility but where a CON is needed to offer another service with the same beds or equipment, facilities have less incentive to reduce unneeded services because it is more difficult to find alternative uses for the capital. High costs associated with converting the facility to another use--such as a school or a hotel--might also keep a facility functioning as a nursing home despite the corporation's earning a low or even negative rate of return on its investment. Under such conditions, profits may fall well below the normal rate of return without inducing firms to leave the market. 27 For more discussion of the impact of such restrictions, see M. Pollard and R. Leibenluft, "Antitrust and the Health Professions," Office of Policy Planning Issues Paper, Federal Trade Commission, June 1981. For empirical evidence, see R. S. Bond, J. E. Kwoka, J. J. Phalen, and I. T. Whitten, "Staff Report on Effects of Restrictions on Advertising and Commercial Practice in the Professions: The Case of Optometry," Bureau of Economics, Federal Trade Commission, September 1980. -25- Providers will not shut down operations as long as they can cover variable or shortrun operating costs--even though fixed costs are not completely covered. Competition among firms unable to easily exit an industry may be quite intense as each firm attempts to re- duce its losses. However, until the competition is intense enough to drive the weakest firms into bankruptcy, the aggregate level of industry profits cannot rise to the normal rate of return.28 In these cases, the rate of return required to attract new capital to the industry is substantially higher than the rate of return necessary to induce the closing of existing facilities. When the rate of return on capital is too low to attract new investment but too high to induce exit, providers generally are reluctant to replace worn-out facilities and equipment. Renova- tions and maintenance of the facilities may also be delayed. Regulatory schemes to raise prices and allocate market shares are often proposed to avoid the intense competition characterizing markets with exit barriers. While such schemes alleviate some symptoms of having excess resources in markets with exit barriers, these forms of intervention may worsen the underlying conditions because they reduce the incentive for firms to leave the market. 28 Bankruptcy is not always necessary. When exit barriers are not absolute, competition eventually drives firms with the best opportunities elsewhere to leave. -26- So-called "destructive" competition serves the function of appro- priately allocating resources by forcing some providers out of the market, albeit in a painful way. Concerns about jobs, access to health services, and community identification may lead municipal, State, and Federal officials to consider providing financial assistance to a failing health-care facility. Such assistance typically prolongs the exit process and increases the strain on other, healthier facilities. Rather than preventing the exit of the weakest firms, it may often be less costly in the long run to ease their exit and to work with other facilities to assure that the needs of special populations are met. Exit barriers are primarily an issue in urban areas in the Northeast and the North Central States, where facilities built in inner cities now serve populations of predominantly uninsured or marginally insured patients. The controversy between hospital officials and the surrounding community over the proposed closing of Brooklyn Jewish Hospital illustrates the extent of exit barriers facing near-bankrupt hospitals. In the Brooklyn Jewish case, the financially distressed hospital was kept in business as part of a consortium, with a Federal contract to cover the cost of caring for the Bedford-Stuyvesant residents not covered by Medicare and Medicaid. The bailout effort has raised considerable controversy, with supporters describing the project as a prototype for national health insurance and critics characterizing it as a - 374-115 0 - 82 - 3 : QL 3 series of stopgap measures.2? As with many nearly bankrupt hospitals, Brooklyn Jewish was not allowed to exit, although many of its services were reduced during the eventual consolidation. As in other cases, issues of access, quality of care, and the politics of the local community were central to the debate about whether the facility could be closed. Such issues must be analyzed in the context of the economic realities of the market, to insure that any exit barriers serve the longrun interest of the community. In many cases, impeding the exit of a financially distressed facility keeps capacity utilization at other facilities low. Thus, they continue to earn low rates of return and may also be brought to the brink of bankruptcy. As in the case of Brooklyn Jewish, the solution often involves reducing the level of services available in the community and increasing the level of reimburse- ment for serving the poor. When facility closings do arise, actions that hasten rather than thwart market adjustments can mitigate the sapping of financial strength of the providers remaining in the market. 29 "'Rescuing’' a Financially Distressed Hospital--The Story of Brooklyn Jewish," Nat'l Health Policy Forum, April 22, 1980. -28- C. Natural Monopoly 30 In perfectly competitive markets, each provider supplies only a small fraction of the total available quantity. This is possible because a small-scale producer operates at as low a cost as providers producing on a much larger scale. Other situations are possible. In some markets, only a few providers can operate at costs low enough to compete successfully in the market. In extreme cases, only one provider can achieve lowest costs. In such situations, a provider may achieve a monopoly "naturally"-- merely by virtue of the technology in the industry. While there is no exact answer to how many firms must provide a product before a market can be characterized as competitive, most economists agree that the likelihood of providers exploiting their market power diminishes rapidly as the number of firms increases.3l Even in health-care markets with relatively small numbers of firms, competition may operate sufficiently well that regulatory intervention would not improve the market outcome. However, in markets characterized as natural monopolies--where the incumbent firm is relatively unconstrained by any sort of competition--regulation may be able to improve resource allocation. 30 For general discussion of market power and its consequences, see Paul A. Samuelson, Economics, 9th ed. (New York: McGraw-Hill Book Company, 1973), Ch. 25: "Maximum-Profit Equilibrium: Monopoly." 31 1n recent experiments, Charles Plott has found that competi- tive outcomes can be obtained in markets with as few as four firms, when there are no industry practices facilitating collu- sion. See, for example, Charles Plott, "Theories of Industrial Organization as Explanations of Experimental Market Behavior," in Strategy, Predation, and Antitrust Analysis, ed. Steven C. Salop (Washington, D.C.: Federal Trade Comm'n, 1981), pp. 523-77. -29- 1. Identifying a natural monopoly . 32 For a natural monopoly to exist, cost conditions must be such that one provider can produce at a lower unit cost than can two providers splitting the output. In many markets, the costs of production decline as the number of units sold increases, but the competition from addi- tional providers usually results in lower prices for consumers. In contrast, the natural monopoly is distinguished by having costs that fall so quickly and over such a large range of output that the unfettered market would not support more than one provider. Because price competition is limited in many health-care markets and because cost-based reimbursement schemes make it possible for providers offering undesired services to stay solvent, market forces may not always leave a single provider in a health "monopoly". In addition, barriers to entry may create monopolies when the industry's cost structure indicates that more than one provider could exist. Because the number of providers may not be a reliable indicator in heavily regulated markets, identification of natural monopolies is more difficult in health care than in other sectors of the economy. When the service is a 32 Many of the issues discussed in this section relate to the general problem of identifying competitors. For more detailed guidance, see P. Areeda and D. Turner, Antitrust Law, vol. 2 (Boston: Little, Brown & Co., 1980); W. Landes and R. Posner, "Market Power in Antitrust Cases," Harv. L. Rev. 94(5) (March 1981):937-96; and K. Elzinga and T. Hogarty, "The Problem of Geographic Market Definition in Antimerger Suits," Antitrust Bull. 18 (Spring 1973):45-81. -30- natural monopoly, the market alone may not yield the most desir- able outcome. Therefore, identifying natural monopolies is important for assessing the appropriateness of competition. Duplication of resources is not evidence that competition is wasteful. In most circumstances, having more than one provider in the market increases competitive activity even though excess capacity may result. A second provider is only undesirable when the rise in unit costs offsets the benefits that flow from competition--such as lowering prices, improving quality, and encouraging innovation. Identifying a natural monopoly in health care is also compli- cated by difficulties in defining the relevant market. For exam- ple, a service that is a natural monopoly in a sparsely populated rural setting is not necessarily a natural monopoly in a dense urban setting. Whether a particular service is a natural monopoly depends in part on the density of the population it serves. Determining whether a health-care institution or service ‘constitutes a natural monopoly within a particular service area requires identification of both the minimum efficient scale on which that service can be provided and the relevant geographic service area.33 The minimum efficient scale for a service is the smallest scale on which service could be provided at a cost that 33 The relevant geographic area may or may not coincide with the geographic boundary of the HSA. 31 would allow a provider operating at that scale to compete with a large-scale provider of the same service. Institutional arrangements may alter the scale at which a provider can offer service. If the physical plant needed to perform a particular service also can be used to perform other services, the minimum efficient scale for each service sharing the plant may be reduced.34 Alternatively, there are cases in which separating services usually performed together may reduce the minimum efficient scale of some services without substantially increasing the minimum efficient scale of the others.35 34 ror example, aside from the quirks of current reimbursement schemes, achieving minimum efficient scale for hospital facilities used for ambulatory surgery may require fewer operations on ambulatory patients than would achieving the minimum efficient scale in a freestanding ambulatory surgical center. The minimum efficient scale for the number of operations on inpatients may also be reduced when facilities are used for ambulatory-surgery patients as well. Once the minimum efficient scale is calculated on the basis of several services sharing a facility, the argument that any of the services constitutes a natural monopoly may be substantially weakened. In this case, certificates of need might be granted more liberally for entry into the provision of each service. 35 For example, trauma centers and emergency rooms may be combined in urban areas. However, trauma centers require sophis- ticated equipment, which is needed by only a small portion of all emergency patients. When the number of emergencies in an area is great enough to support several trauma centers, they can operate as adjuncts to large emergency rooms, or the two types of facili- ties may operate independently. Where the market is large enough, whether the services are offered together or separately should not be of great concern to the planner, because competition can be maintained either way. However, in a rural area, where there are fewer emergencies in each town, trauma centers may be run most efficiently on a regional basis, using air ambulances, while every town may be able to support its own emergency service. In this example, cost is reduced, access is increased, and competition is enhanced by treating the trauma center as a natural monopoly on a regional basis and allowing emergency rooms to compete with one another as a separate service. SG While analysis should generally proceed on a service-by- service basis, at times it may be necessary to consider a cluster of services together. This is often appropriate when services are supplied using the same facilities, even though the services are not interchangeable from the patient's point of view. For example, it may be appropriate to consider "general acute-care services" together for the purposes of market analysis, despite the fact that patients are treated in individual departments of an acute-care hospital for different conditions. In addition to determining the minimum efficient scale in terms of units of service, the planner must determine how large a population it takes to support the service. Doing this requires the use of both epidemiologic evidence about the frequency of use of the service in the population and economic evidence on the willingness and ability of consumers to travel to receive the service. Emergency care that is not quickly accessible, for example, is largely irrelevant. On the other hand, individuals may go considerable distances to receive treatment for rare diseases from highly trained specialists, although other treatment is available nearer home. Thus, the relevant geographic market may encompass the entire United States (for some rare diseases) or be limited to only a few miles (for emergencies). Factors to consider in determining the geographic market area include (1) the importance of speed in receiving the service, (2) the importance to the patients of receiving a specific quality or variety of the service, (3) the density of the patient population, —3 3 and (4) the importance of the service to the patient. All these factors influence the distance consumers may be willing to travel if the price or quality or variety of the service offered by providers located in their vicinity is unsatisfactory. Also important is the physical mobility of providers. When the service can be performed in the consumer's home or in rotating clinics, a number of providers may be able to compete over a larger geo- graphic area.36 2. Entry in markets characterized as natural monopolies. In most markets, the desire on the part of a second firm to enter a market indicates the existence of excess profits and indicates that the market is probably able to support more than one firm. Therefore, when a second provider wishes to enter a market desig- nated as a natural monopoly, it usually signals that the designa- tion was probably incorrect. However, there are two reasons this "market test" may be an erroneous indicator in the health sector. First, third-party-reimbursement schemes may remove most risks associated with establishing a new firm in some parts of the health-care field. Entry is a useful indicator of market incentives only if an entrant providing an unneeded service has some chance of going bankrupt. If the entrant can add unneeded capacity or provide an inferior quality of care with complete 36 gee section IV.D. for questions to ask in identifying a natural monopoly. -34- assurance that it will continue to earn an adequate rate of return on its investment, that firm is not at risk and its willingness to enter the market is not a reliable guide to the true nature of market conditions. Such a situation may exist where the entrant receives guaranteed municipal loans to build the facility or when reimbursement schemes cover the cost of the service regardless of the number of patients served or of capacity utilization rates. 37 The presence of external subsidies, internal cross-subsidies, and cost-based reimbursement for health facilities thus can make it difficult to use entry as an indicator that a market does not constitute a natural monopoly. These factors also may mean that entry barriers--with their ensuing problems--are needed to protect the market, because without such barriers, subsidies attract firms to enter or to attempt to enter the natural-monopoly markets. 38 However, if third-party payers were to adopt prospective reim- bursement or other incentive-compatible reimbursement schemes, entry prohibitions presumably could be relaxed.39 Because 37 For discussion of such schemes, see sections II.E.2. and IV.A. 38 such subsidies must be available to a natural monopoly when high fixed costs and a large minimum efficient scale would pre- clude the provision of the service, were charges made on the basis of total unit costs. However, the social benefits of these sub- sidies in maintaining services would be defeated if any entrant could also receive these subsidies regardless of the social desirability of its committing additional resources to the market. See sections II.B., II.A., and III.B. for discussion of the problems these regulations create. 39 gee section III.B. for discussion of incentive compatible schemes. -35- entrants bear some risk under these schemes, regulatory control of entry to prevent excessive investment would probably not be needed. A second reason entry may be an erroneous indicator of whether the market constitutes a natural monopoly is that the incumbent may provide more than one service, while the new entrant proposes to provide only a subset of those services.40 In that case, it is necessary to examine the degree to which the services provided by the incumbent firm must be supplied together.41l For instance, if a new firm intends to provide only one service using a different, lower cost technology than an existing firm that uses a joint production process to provide both services, it is neces- sary to determine whether the second service also can be produced alone, and at what cost. Under some limited conditions, it may be appropriate to protect the natural monopolist from the "cream- skimming" entrants. However, planners should be wary of thwarting entry on this basis, because entrants often introduce innovative and cost-saving production and marketing techniques. Surgi- centers, adult day care, and home renal dialysis all provide fewer 40 por development of this theory, see J. Panzar and R. Willig, "Free Entry and the Sustainability of Natural Monopoly," Bell J. Econ. 8 (Spring 1977) :1-22. 41 a hospital typically provides a multitude of services that could also be performed by separate entities. For example, surgi- centers and birth centers each offer only a small portion of the services performed by a hospital, and in some cases they use a different technology. Some services might be provided solely outside the hospital context without affecting either the cost or quality of the remaining services provided by the hospital. -36- services than the facilities with which they compete, but their entry has forced other providers to improve their services in order to attract patients. In many cases, entry merely leads to more competition and less profit, but services to the "skim milk" portion of the market continue, albeit without cross-subsidization from the "cream." In natural monopolies, regulation of prices and cost, together with regulation of capital investment, may improve allo- cation of resources; but problems are inherent in these regulatory techniques.42 The health services most likely to be natural monopolies are those requiring large amounts of capital investment relative to the frequency of use. In the health sector, most natural monopolies probably occur in sparsely populated rural areas. In many cases, these same services can be produced competitively in more densely populated areas. Changes in tech- nology and in size of the population mean that competition may be appropriate in some services once considered natural monopolies, and new natural monopolies may be created. Even in those markets designated natural monopolies, the case for entry should be carefully examined when additional providers wish to gain access to the market.43 42 gee sections III.A. and III.B. for details. 43 gee section IV.D. for specific questions to ask in identifying a natural monopoly. wd D. Costly or Inaccessible Information44 In order to achieve an optimal resource allocation, partici- pants in a competitive market must have access to sufficient infor- mation regarding the existence and location of providers, the nature of the products they offer, the usefulness of their services, and the prices at which these services are offered. Absolutely costless and complete information--sometimes called "perfect" information--is an ideal achieved in few markets. Yet most markets function reason- ably well without regulatory intervention. This section of the paper addresses the ways in which costly or missing information can affect the functioning of markets.45 It also discusses the advantages and disadvantages of selected options for improving resource allocation: (a) gathering and disseminating information; and (b) directing the allocation of resources by restricting entry, regulating services characteristics, or setting prices. Various combinations of these techniques are possible but will not be discussed here. Because intervention is costly and may create other distortions, such action is usually desirable only when the distortions associated with the absence of information are severe. 44 For a nontechnical discussion of this subject, see S. Salop, "Parables of Information Transmission in Markets," in The Effect of Information on Consumer and Market Behavior, ed. A. A. Mitchell (Chicago: American Marketing Ass'n, 1978). For a technical presen- tation of the results presented in this section, see J. Stiglitz, "Equilibrium in Product Markets with Imperfect Information," Am. Econ. Rev. 69 (May 1979) :339-45. 45 For a discussion of missing information about quality levels, see section II.E.2.b. -38- 1. Costs and methods of gathering information. The costs of obtaining missing information depend on the value consumers place on their time, consumer preferences for gathering information versus engaging in other activities, consumers' analytical ability and understanding of the market, the measurability of the data to be collected, and the manner in which information is collected. The amount of information consumers gather depends on the costs of gathering information and the expected value of the information in choosing among products or services. When the cost of gathering information is low relative to the price of the product and consumers expect that the prices or other important characteris- tics of the product differ among providers, consumers may seek extensive information about all providers. In contrast, when providers are numerous or when information costs are high, consumers may be unwilling to expend time or money gathering information. Information can be obtained by experience or search. Experience is knowledge obtained from actual use of a service, either by the consumer or by his acquaintances. For example, information about physicians may be gathered by seeking treatment for minor ailments from different professionals or by gathering information about physicians from acquaintances. Because information gathered by experience is usually not exhaustive and is imprecise, its uses are limited. It may help to eliminate unacceptable providers, but experience has limitations as a means of identifying providers best suited to individual needs. -39- "Search" refers to the gathering of information prior to purchasing the service. Consumers may search by visiting or phoning providers, reading consumer magazines, paying attention to advertisements, or using the services of associations that provide information. In most situations, consumers combine several search techniques. For example, a consumer seeking the services of a psychotherapist may get names of professionals in the vicinity from a professional association or from the phone book. The consumer may then call providers on the list to find out whether they are accepting new patients and what prices they charge. Before deciding on a program of therapy with one provider, the consumer may also have an initial consultation with several psychotherapists in order to identify the one suited to his needs. In this example, as in most, each progressive level of search provides more information but also is more costly. Less costly forms of search typically are used to screen out clearly unsuitable providers before the consumer undertakes the next level of search. Because search is costly, consumers can be expected to search more thoroughly when they expect to establish a continuing relationship with a provider, where it is difficult to switch providers (such as in psychotherapy), or when the costs of an incorrect choice are great. 2. Effects of costly and incomplete information. When consumers lack information about the existence or usefulness of a service, their decisions may not reflect the best possible use of their resources. For instance, when consumers lack information -40- about product variations, they may not choose the form of a service best suited to their needs. When providers cannot easily communicate information about prices to attract new consumers, less price competition typically occurs and higher prices result.46 Conversely, more consumer search increases the incen- tives for providers to lower their prices and to provide services with characteristics in keeping with consumer preferences. Not all consumers have to search for competition to work. As long as some consumers seek out providers offering the best combinations of price and service characteristics, providers may compete for their patronage. If providers cannot distinguish uninformed consumers from the informed consumers, uninformed consumers share in the benefits of competition.47 Consumers consider only the benefit to themselves in deciding how much to search, however. They generally have no interest in the fact that their search for a provider offering the best price/quality trade- off helps others who have higher search costs. They therefore engage in less search than might be beneficial to consumers as a whole.48 46 For development of a model with these characteristics, see G. Butters, "Equilibrium Distribution of Sales and Advertising Prices," Rev. Econ. Stud. 44 (October 1977) :465-91. 47 When various services are available, uninformed consumers may not pick the providers best suited to their needs and preferences, even though the services they get will reflect a good quality/ price tradeoff. 48 por full discussion of these issues, see S. Salop, "Informa- tion and Monopolistic Competition," Am. Econ. Rev. 66 (May 1976) :240-5. -41- The effects of costly and inaccessible information on prices and other characteristics of services offered by providers depend on (1) how costly it is for consumers to gather information, (2) how many consumers gather information, (3) how much information they gather, and (4) the sensitivity of consumers' demand to changes in price and other characteristics. If information is not freely available, but can be gathered at low cost relative to the size of the cost of the service, the absence of costless informa- tion may have little or no effect on the market, because enough consumers will gather adequate information to assure the competi- tive outcome. Similarly, if consumers are extremely sensitive to the price and other product characteristics, no producer charging too high a price or offering products with undesirable charac- teristics can retain customers. In addition, consumers may gain knowledge about the service over time, and as long as consumers can switch providers, providers wishing to maintain long-term relationships with consumers have an incentive to provide services with a satisfactory combination of price and other service charac- teristics. Even when the bond between consumers and particular providers grows over time, the desire to maintain reputation and attract new customers usually gives providers an incentive not to unduly raise prices or lower quality.49 49 For further discussion, see B. Klein and K. Leffler, "The Role of Market Forces in Assuring Contractual Performance," J. Pol. Econ. 89 (August 1981):615-42. nl) Ge 3. Solutions to problems caused by costly and incomplete information.%0 In many situations, it may be less expensive for one agent to gather information and disseminate it than for each consumer to gather the information independently. In such cases, gathering information centrally can lower the cost of search to each consumer and thus lead to more search. Gathering information centrally can also overcome the problem of too little search caused by each consumer's ignoring the benefits that his search generates for others. Centralized search occurs or may be introduced in many situa- tions. Providers advertising their prices, locations, and the characteristics of their products centralize the provision of information in many markets. Examples of information that can be conveyed in ads include office hours, willingness to make house calls, participation in private insurance programs, credit terms, willingness to accept Medicare assignment, and languages spoken. In some cases, however, advertising by an individual provider is less useful to consumers than comparative information such as is offered by consumer magazines. Because health-care markets tend to be local, publication of such information usually would best be done on a local or perhaps regional level. Once 50 For a more detailed discussion of information and the design of information remedies, see H. Beales, R. Craswell, and S. Salop, "The Efficient Regulation of Consumer Information," J. Law & Econ., 24 (December 1981):491-539; also see Schwartz and Wilde, "Intervening in Markets on the Basis of Imperfect Information: A Legal and Economic Analysis," U. Pa. L. Rev. 27 (1979) :630. -43- 374-115 0 - 82 - 4 : QL 3 comparative information is made available to the market, all services providers may benefit. However, because information tends to be a "public good," profitmaking organizations have little incentive to provide such information unless dissemination can be limited to those purchasing the right to use it.>l HSA's are well situated to assist local groups in the collec- tion and dissemination of such information as prices of routine services, locations, and certifications of providers. Indices of other characteristics consumers may consider important, such as case-severity-adjusted death rates per procedure for each facility, are more costly to compute but may offer information consumers cannot gather on their own and may therefore be useful.52 Other information such, as "bedside manner," is more difficult to measure or to convey to consumers but may be collected through consumer survey.>3 Some regulations (such as bans on advertising or on the use of trade names) are commercial restrictions exacerbating the 51 yseful compilations of information (such as reference books) can find a market because they organize a mass of information in such a way that the consumer's cost of using the information is reduced. 52 Using the dissemination of such indices as an alternative to other forms of regulation is discussed in section III.C.4. 53 gee, for example, "How to Find a Good Washington-Area Doctor or HMO--And How Their Patients Rated Them," Washington Consumers Checkbook 3 (Fall 1980):1+. If information were perfect, licen- sing would not be needed to protect consumers from charlatans. -44- inaccessibility of information.54 other regulatory schemes (such as licensing) are designed in part to alleviate problems associated with the absence of costless and complete information. In many cases, regulations designed to alleviate information problems create other imperfections in the market. For example, licensure creates barriers to entry.55 In some cases, schemes disseminating information and relying on market forces may be able to improve the allocation of resources at less cost. When the market can be improved rather than supplanted, it is often costly and inefficient to rely on regulation alone. For example, administratively setting prices or regulating quality levels requires that the regulating authority have information about providers' costs and about consumer needs, preferences, and budgets. Often it is easier to convey information about providers to consumers and allow the consumers to make their own decisions based on their own needs and preferences than it is to gather information about both consumers and providers and attempt to centrally determine and enforce service characteristics.56 When the regulating authority has the ability to measure quality and to identify appropriate providers but the information 54 For an empirical study of the effects of limiting advertising, see R. S. Bond, J. E. Kwoka, J. J. Phalen, and I. T. Whitten, "Staff Report on Optometry," op. cit. 55 For a detailed discussion, see section III.D. 56 Using certification for this purpose is explored more fully in section III.D. -45- cannot be conveyed to consumers for some reason, regulation may achieve a better allocation of resources. In many cases, a combination of information dissemination and regulatory techniques may be the best approach. For example, informational remedies alone may not be effective for extremely complicated and dangerous products, such as drugs, where all the relevant information cannot be easily communicated to the average consumer. In the case of prescription drugs, one proposed solution, which has been tried for a few drugs, is to require that drugs be prescribed by physicians but that certain information about drug use and contraindications be provided directly to consumers.57 Costly and inaccessible information is associated with many problems in markets. As discussed in the next section, most problems associated with the ability of providers to vary quality occur only when there is inadequate information about the quality of services performed by particular providers. In markets where a spectrum of products serve similar functions, costly and incomplete information may lead to a mismatch between services and consumers. However, regulators rarely have sufficient information about consumer preferences to dictate the correct matching of consumers and providers. In addition, traditional regulatory techniques (such as review of entrant characteristics) are often 57 See "When Patients Know What's Good For Them," Perspective section, Regulation: AEI Journal on Government and Society, September/October 1981. -46- 58 1n such inadequate to impose such a solution on the market. situations, advertising or the public dissemination of information may promote a better match between consumers and providers. Thus, providing information often may be cost effective and produce greater consumer satisfaction with the health-care system than would be attainable by using more intrusive forms of governmental intervention. E. Heterogeneity of Products The perfectly competitive model assumes that identical products are offered by all sellers participating in a particular market and that consumers can choose among producers on the basis of price alone. This assumption is frequently violated in health care. When products differ, consumers must make more complicated tradeoffs between combinations of price and product charac- teristics offered by different sellers. when the products produced by different providers are not identical and different consumers prefer different forms of the product, the presence of more than one producer is desirable because competition among numerous sellers often results in diversity. Under these circumstances, consumers value the presence of many providers both because increased competition lowers prices and because the choice of products is greater. Thus, even when regulations restricting entry reduce investment in duplicative capacity and thus lower resource costs, consumers will 58 gee section III.B. for further discussion. -47- be worse off if the disadvantages of a smaller choice of providers offset the advantages of reduced costs. Other forms of market intervention--such as minimum-quality standards--may improve or may cause further deterioration in the mix of products in the market. In general, the more important product variety is to consumers, the less likely it is that entry restrictions can improve the allocation of resources in a particular market.59 Conceptually, the differences among products competing in the same market can be characterized either as quality differences or as variety differences. "Variety" refers to differences in pro- duct characteristics about which there is no general consensus as to the preferability of one product over another. "Quality" refers to variations that can be objectively characterized as improving or deteriorating the desirability of the product, ignoring differences in cost. Although variety and quality differences often occur together, these two types of product variations raise somewhat different concerns and are therefore discussed separately below. 59 In some cases, economies of joint production mean that a single firm can provide a variety of services more cheaply than the same mix could be provided by a number of different firms. In such cases, the decision to limit entry should still depend on the extent to which the single firm can be expected to produce a desirable mix of services in the absence of competition and the extent of the savings from having only one firm. Under some con- ditions, a single firm may have the incentive to provide a wider range of services than would a small number of competing firms. In many cases, however, a single firm protected from entry has less incentive to improve its products or provide its products at as low a price as it would were it faced with the possibility of competition. -48- 1. variety.60 Variations among competing products are ubiquitous in this economy. There is diversity in virtually every consumer good, from toothpaste to housing to automobiles. In each case, the technology for producing the good allows producers to vary some characteristics without improving or debasing the product in any objectively verifiable way. Examples of product variety include differences in color, taste, styling, and location. Consumers rarely agree unanimously that certain variations of these types are superior to all others. The characteristics of health-care services generally vary according to the providers offering them. Access to a particular variety of a service can be quite important to particular consumers. For example, the "bedside manner" of a physician may not make an objectively measurable difference in the care a patient receives; patients presumably disagree about which phy- sician has the "best" bedside manner. Nevertheless, many patients have strong preferences for the personal characteristics of their physicians and select physicians partly on that basis. Similarly, for some conditions, several modes of treatment may be available, with none clearly superior to the others in terms of morbidity, speed of recovery, or other objective indicators. For instance, consumers obviously have individual preferences for one form of birth control over another. Although methods of contraception 60 For a theoretical treatment of product variety, see A. Dixit and J. Stiglitz, "Monopolistic Competition and Optimum Product Diversity," Am. Econ. Rev. 67 (June 1977):297-308. -49- differ somewhat in efficacy and cost and in their suitability for use by individuals with particular health profiles, the choice of a method is often based largely on other, less tangible differ- ences, which vary widely among individuals. Similarly, consumers also have preferences for various forms of treatment for other specific medical conditions. Because variations in product or service characteristics matter to consumers, two products or services performing similar functions but differing in one or more characteristics are close but not perfect substitutes for one another in the marketplace. Examples of service and product variety in health care include physicians with varying amounts of postgraduate training and experience performing similar services; obstetricians and nurse- midwives both providing routine obstetrical care; similar surgical procedures performed in hospitals, hospital outpatient clinics, freestanding surgical centers, and ambulatory clinics; and drugs with different chemical origins available to treat the same conditions. When consumers prefer a particular product variety, they may continue patronizing the provider despite an increase in price or a reduction in quality. In some cases, consumers will only toler- ate a small price increase before switching to other products or other providers. In other cases, consumers' preferences for certain forms of health care may be quite strong. However, in a market without entry barriers, price increases only lead to -50- excess profits temporarily because new providers quickly enter the market. When there are sufficient economies of scale, the market may not be able to support enough sellers of each variety of the product to sustain competition among several sellers of identical products. Under such conditions, entry may compete away excess profits but price may remain above the level expected in a perfectly competitive market. Thus, when moderate economies of scale are associated with the production of any variety of the product, each provider will typically produce a different variety, but not all possible variations will be provided. Because competition in such a market is imperfect, each provider has a limited ability to raise its prices without losing customers. In such a market, total resource costs might be reduced by cutting back the number of product varieties and allowing each provider to serve more consumers. However, in comparison to the natural monopoly case, the product-variety analysis is complicated because any reduction in variety makes some consumers worse off. Imperfect competition with product variety also differs from natural monopoly because greater substitutability among the various varieties of the product limits the distortion in the allocation of resources. Because the distortion is smaller, it is less likely that regulatory intervention would achieve a better resource allocation. Even when the unfettered market provides far too little or far too much product diversity, there are three reasons regulatory -51- schemes may not be able to produce a better resource allocation. First, identifying the best mix of products requires information both about the ways in which products provided by various provid- ers differ and about the distribution of consumer preferences. Such information is typically quite difficult to obtain. Second, most regulatory schemes, such as certificate of need and other entry restrictions, are blunt tools for influencing the appropri- ate mix of providers because they allow regulators to deny entry or expansion requests but do not allow regulations to rationalize all resources in the market. As discussed in the section on entry barriers, such schemes also reduce the incentive for incum- bent providers to respond to consumer preference. 61 Third, markets operating without entry restrictions have a bias towards product variations for which demand is more elastic--for which price decreases call forth greater increases in the quantity demanded--and which have lower fixed costs of production.62 Product varieties strongly preferred by small groups may not be produced if providers are unable to obtain the benefits of supply- ing such products. If such products are deemed desirable by soci- ety, subsidies are often necessary to assure their availability. Local regulators typically lack the revenues necessary to provide 61 see section II.B. for discussion of entry barriers. 62 For technical discussion and proof, see A. M. Spence, "Product Selection, Fixed Costs and Monopolistic Competition," Rev. Econ. Stud. 43 (June 1976):217-35. BZ direct subsidies and instead must rely on cross-subsidy schemes, which create their own distortions. Not only is it difficult to use regulatory schemes to elimi- nate the distortions associated with product diversity; product diversity can also complicate the administration of regulatory schemes designed to correct other problems. For instance, certificate-of-need decisions may involve applications by providers offering varieties of the service not currently avail- able in the market. In such circumstances, information about consumer preferences is needed to make the tradeoff between cost savings and consumer satisfaction. For example, in a market with sufficient nursing-home beds according to demographic criteria, the entry of a facility serving a religious or ethnic group to which no existing facility specifically caters may increase the satisfaction of some consumers. It may also raise the costs per patient in all facilities. As planners are well aware, reviewing such a certificate-of-need application involves weighing the benefits to one group of consumers against the costs to the system as a whole. Other situations, while analytically identical to the nursing-home example, involve less obvious tradeoffs between increased costs and increased consumer satisfaction from greater product variety. The consumers who are made better off by the entry of the new variety of a service may be less readily identified when the CON application is by another hospital wo lB offering traditional medical care or by facilities for the use of chiropractors, psychiatric social workers, nurse-midwives, or an HMO. Nevertheless, the gains from diversity may still be great enough to offset the costs associated with increasing capacity. In other cases, when the vast majority of consumers have similar preferences, head-to-head competition among facilities providing services similar to those already in the market may on balance result in more consumer satisfaction than would the entry of a facility exactly meeting the preferences of a small group of consumers. 2. Quality.63 Products differ in quality when informed consumers would unanimously agree that some varieties of the product or service are more desirable than others. The importance of quality in health care is the major reason some health profes- sionals and consumers are reluctant to have nonmedical personnel involved in decisions affecting the health-care system. However, while economists and planners may not have skills enabling them to assess the quality of care a provider offers, the analytical tools of economics and planning can help those with the necessary medical knowledge to make decisions affecting quality of care. a) The cost/quality tradeoff. The fundamental message of economics regarding quality is that producing higher quality care 63 For a theoretical treatment, see A. M. Spence, "Non-Price Competition," Am. Econ. Rev. 67 (February 1977) :255-59; and E. Sheshenski, "Price Quality and Quantity Regulation in Monopoly Situations," Economica 43 (February 1976):127-37. i Ble efficiently is costly.64 Because producing higher quality care generally costs more, different levels of quality may be appro- priate for different conditions. For example, ignoring costs, CT scanners are widely regarded as the highest quality diagnostic tool for a wide variety of conditions. However, the high cost of these devices means that they should be limited to their highest priority uses. For the diagnosis of brain tumors, using scanners is generally regarded as well worth the cost, while lower cost, conventional technologies may be preferable for diagnosing patients likely to be suffering from an ordinary concussion. Were the price of the machines to drop, it might make economic as well as medical sense to use scanners routinely for a wider range of conditions. Similarly, the costs associated with training spe- cialists means that their services should be allocated to those most in need of their skills and that less highly trained medical personnel should be used to treat more mundane ailments. In both these cases and in countless others, the cost associated with pro- viding higher quality care forces a compromise: less quality is traded for more widely available medical care at a lower price. The diversity of medical conditions means that the same tradeoff is not appropriate in every situation. Higher quality care is most important where it makes the most difference. The 64 The assertion that higher quality care costs more assumes that the care at each quality level is being provided as cheaply as possible. When resources are used inefficiently, this may not be true. -55- diversity of medical conditions means that the preferences of fully informed consumers for different quality levels would vary, depending on their conditions and the expected benefits from higher quality care. If consumers had full information about the quality of different services, the analysis of differentiated quality would be fully analogous to the analysis of product variety.65 If consumers were fully informed about quality differences, problems with resource allocations by the market only arise if there are economies of scale associated with producing each level of quality. In that case, not all quality levels would necessarily be produced and competition among providers offering identical products or services might be limited. As in the case of pure variety, regulatory intervention is unlikely to improve the mix of services offered because the available tools are too blunt and too much information about providers and consumers is needed. b) Quality differences with imperfect information.66® The assumption of full consumer information is unrealistic in many health and non-health markets. Incomplete information about health-care quality occurs when consumers have little knowledge about their illness, when (as in emergencies) they cannot shop, 65 gee section I1.E.1. 66 The general problems associated with costly and incomplete information are discussed in section II.D. However, because there are many special problems associated with imperfect information about quality, these are discussed separately here. -56- when they lack understanding of the treatment, and when it is difficult to assess the contribution of the service to their recovery rate. When consumers lack information about the quality of the services offered by particular providers, quality levels offered in the market may differ substantially from those expected if complete information were available. Depending on the conditions in the market, imperfect information may result in higher or lower quality. Intuitively, most individuals find it plausible that lack of information would lead to a lower-than-optimal level of quality. When consumers cannot determine whether providers shade quality, reduced quality may occur. For instance, suppose consumers only know the average level of the quality of services on the market and cannot identify above-average providers. In that situation, they would only be willing to pay a provider for average-quality care because they could not verify that they had received care of a higher quality. If all resources are used efficiently, produc- ing higher-than-average quality is more costly. Therefore, pro- viders would not offer care of a higher quality because consumers would not be willing to pay for it. The withdrawal of higher- than-average care from the market would, in turn, lower the average quality in the market. This, in turn, would reduce the -57- amount consumers would be willing to pay. In such a market (called a "lemons" market), quality sinks to a low level.®7 In some markets, imperfect information is associated with higher-than-optimal quality. For instance, because consumers have limited information, they often trust physicians to make decisions about quality-related issues, including the determination of the tradeoff between cost and quality. From fear of malpractice suits or from professional pride, physicians may choose a level of quality that is higher than that which the fully informed consumer would pick for himself after balancing the benefits against the costs. The proliferation of medical tests may fit this model: each test conveys some information and thus, to some extent, may increase the quality of care; however, tests are costly, and the fully informed consumer might choose fewer tests than would his physician. 68 Nonprofit hospitals may also have an institutional bias in favor of providing care at too high a quality, given the cost. Because nonprofits do not pay dividends back to investors and instead use excess revenues as they see fit, they may choose to 67 For a full presentation of this model, see G. Akerlof, "Quality, Uncertainty, and the Market for Lemons," Q. J. Econ. 84 (August 1970) :488-500. As applied to health care, see B. Weingast, "Physicians, DNA Research Scientists, and the Market for Lemons, " in Regulating the Professionals, ed. Blair and Rubin (Washington, D.C.:American Enterprise Institute, 1980). 68 of course, insurance policies that obscure the full social cost of medical care also raise the quality of care demanded. See discussion in section II.F.1. GF overinvest in equipment and facilities, both because it increases physician satisfaction with the facility and because it fulfils an institutional desire for high-quality care. In some markets, proxies for quality exist. For example, in a market with incomplete information, attributes such as price or a professional's association with a particular facility may serve as a signal of quality and may allow the market to provide a more nearly optimal level of quality. Even though consumers typically lack such information, university medical centers and community hospitals may have information enabling them to judge the quality of care provided by some physicians. Each type of facility may decide to hire physicians who offer a particular level of quality because it is advantageous for the facility to compete in that segment of the market. University hospitals may wish to treat only rare and complicated cases and may therefore select highly specialized clinicians. Community hospitals have a comparative advantage in serving the bulk of the population. Thus, they may allow physicians to perform any procedure at which they are adequately trained. If hospitals specialize in different kinds of care, and consumers are aware of this specialization, the consumers may be able to use the physician's hospital affilia- tion to judge his competence in performing particular procedures. Patients with rare conditions would choose to go to university -59- 374-115 0 - 82 - 5 : QL 3 hospitals; patients with more routine ailments would choose to go to community hospitals, where they would be treated adequately.69 The market may also work well despite some consumers' having incomplete information. If a sufficient number of consumers have adequate information about quality, uninformed consumers can sometimes rely on a firm's general reputation without needing to fully investigate the quality of services on their own. In these cases, appropriate quality levels may be produced despite the absence of complete information.70 However, even if the market is producing the right mix of products, consumers with limited information may not be able to find the provider offering the product best suited to their needs. As a result, there may be a serious mismatch between consumer preferences regarding quality and the quality of services they consume. 71 Planners do not administer most of the regulations designed to alleviate problems associated with quality in the market. CON, entry restrictions, and rate-of-return regulations are clumsy 69 In a market where consumers were price sensitive and providers had an incentive to compete on price as well as other attributes of their service, the availability of services at different prices could facilitate this "sorting out" process. 70 For the theory of price as a signal of quality, see P. Nelson, "Advertising as Information," J. Pol. Econ. 82 (July/August 1974) :729-54. 71 gs, Salop, "Information and Monopolistic Competition," op. cit. en tools for addressing these problems. ’2 However, planners are uniquely able to provide a partial solution to problems caused by the lack of information regarding the quality of particular pro- viders. Planners can often measure quality more accurately than can individual consumers. In addition, planners have training that may help them to evaluate and report the information in an understandable format. While collecting the appropriate informa- tion may be too costly for each consumer to undertake individual- ly, it may be cost effective at an aggregate level. In such cases, planners can facilitate consumer education by publishing such vital information as services available in each facility and the death rates by facility for commonly performed procedures. Such information can improve the ability of consumers to make their own choices. It can also increase competition and quality in the health-care market. F. Pricing Distortions and Externalities Participants in a perfectly competitive market make their decisions in response to prices that reflect the true costs of the resources involved. When prices paid or received do not reflect the full cost of the service, externalities exist. In health care, externalities are generally associated with pricing distortions generated by health plans and other reimbursement schemes. Because the distortions in incentives are often 72 gee sections III.A., B., C., and D. for more discussion of these regulations. -61- associated with price and cost regulations rather than with underlying characteristics of the market, many of the problems created by pricing distortions are discussed in the section on regulation.’3 Issues of competition among health plans are covered in the questions at the end of the paper.’4 The main focus in this section is the effect pricing distortions have on the ability of a market to achieve appropriate resource alloca- tions and the ways in which the allocative efficiency of such markets might be improved. The basic goals of health plans and other reimbursement schemes (such as Medicare and Medicaid) are risk sharing, income distribution, and prepayment to lessen cash-flow problems. Ideally, these basic goals could be achieved without creating distortions in incentives. However, distortion-free schemes are often difficult to design and administer. Therefore, under these schemes, market participants often face prices that do not fully reflect real resource costs. In such cases, the decisions made by market participants may differ substantially from those they would make were prices not distorted. As with all other deviations from the characteristics of the perfectly competitive market, the effect of externalities on the appropriateness of using competition as an allocative mechanism 73 gee section III.A. 74 gee section IV.B., questions 1-7 and section IV.C., questions 8, 9, and 13. wily depends on the degree of the distortion. In addition, in many cases the pricing distortions created by health plans and reimbursement schemes also affect the allocations achieved under regulation. Thus, improving the incentives communicated by insurance and reimbursement schemes is desirable regardless of whether it results in greater use of competition in allocating resources. The analysis of the incentives created by third-party reimbursement schemes and other health plans is complicated, and the effects differ from plan to plan. Therefore, this section only discusses distortions associated with health plans and reimbursement schemes in general terms. While the sources of the distortions in providers' decisions and consumers' choices are quite similar analytically, the effects of health plans and reimbursement schemes on consumers and providers are treated separately here for clarity. The discussion of the distortions on consumers' choices is presented in terms of insurance schemes, in order to abstract away from issues of prepayment and the income redistribution effects of such Government-administered and Government-funded reimbursement schemes as Medicare and Medicaid.’5 The discussion of the distortions in producers’ 75 For discussions of access and income-redistribution issues, see L. A. Aday, "Economic and Non-Economic Barriers to the Use of Needed Medical Services," Medical Care 13 (June 1975) :447-56; R. G. Beck, "The Effects of Copayment on the Poor," J. Human Resources 9 (Winter 1974):129-42; and K. Davis, "Medical Payments and Utilization of Medical Services by the Poor," Inquiry 13 (June 1976) :12-135. -63- decisions is presented in terms of reimbursement schemes because the insurance aspect of third-party-payment schemes has little importance in understanding the effects on providers. 1. Health plans: the effect on consumers' choices. The primary purpose of health insurance is to protect consumers from the risk of unexpected health-care expenditures. However, when health plan payments depend not only on the consumer's medical condition but also on the treatment the consumer chooses, the consumer's incentives to choose the most cost-effective treatment may be diminished. The incentives built into many current schemes are a major reason that insured consumers often exhibit little concern about the costs of the health services they consume. Distortions occur because consumers do not receive the full bene- fit of the cost savings they generate by choosing lower priced providers, by limiting the number of tests received, or by shortening lengths of stay in health-care facilities. Under these schemes, consumers have little incentive to deny themselves any treatment (regardless of cost) that may generate even minimal benefits. This "overconsumption" of health services results not only from misinformation and the high value consumers place on health but also from the lack of proper incentives to curb unnecessary health expenditures. Theoretically, health plans could pay consumers solely on the basis of the condition diagnosed. For example, consumers might receive $150 for a simple fracture, $250 for appendicitis, -64- $30 for a strep throat, etc., depending on the average cost of treatment. By compensating the consumer for the expected cost of treatment, consumers would be insured against any financial hard- ship imposed by the illness. However, if reimbursement were independent of the cost of treatments actually received, consumers would have an incentive to shop for the provider offering the best price/quality combination, and every dollar the consumer saved would remain in his own pocket. 76 In contrast to this theoretical scheme, most health plans reimburse consumers on the basis of costs actually incurred. In other words, health benefits are only received when the consumer undergoes treatment. In many cases, the marginal out-of- pocket costs consumers incur are far less than the marginal resource costs of providing the services they consume. When health plans pay more for more expensive care, consumers' price sensitivity is reduced. Under some plans (such as HMO's), the only "payments" are in the form of treatments by a certain team of 76 one drawback of such a system is that it might not communicate the correct incentives for avoiding illness. When consumers have some control over their medical conditions, some individuals might not take sufficient care to avoid illness, or might even purpose- fully contract some conditions, if they were willing to forego treatment or thought they could get a "bargain" and pocket the savings. Insurance schemes paying on the basis of medical condi- tion might therefore choose to exclude conditions where consumers could "contract" the condition voluntarily. Another drawback is that it is often difficult to identify the severity of a condition at the outset. A system basing payment on diagnosis alone is therefore probably best suited to conditions that are easily diagnosed and treatment is fairly standard. (See section III.E. for further discussion.) -65- medical personnel. Were the payments not directly tied to the consumers' choices of treatment, they might choose to consume less costly medical care in order to spend the remaining benefits on other goods or services. Clearly, in situations where a plan pays 100 percent of the cost of treatment or where (as in the HMO) treatment is the only form of health-plan benefit, consumers have little if any incentive to exercise restraint in their health-care consumption.’ Except for the inconvenience of hospitalization or office visits and the discomfort of tests, the consumer has no incentive to limit his use of the medical care system.’8 Some plans set maximum payment levels per treatment.’9 When maximum payment levels are binding (i.e., set below the minimum cost for which treatment is available in the market), they commu- nicate the same incentives as payments not based on the cost of treatment, except as they affect the decision to seek treatment at all. The drawback to plans with maximum payment levels is that 77 of course, as certain types of HMO's have shown, health-care costs can be contained in such systems, despite their communicat- ing incorrect incentives to consumers. In the context of such HMO's, the provider's motivation for cost control comes from a competitive insurance market. Aside from queues created by the provider, the consumer himself has little or no incentive to restrict his use of HMO services. The HMO structure, however, provides an important paradigm for the allocation of health resources in the absence of full fee-for-services pricing. 78 For fuller discussion, see R. C. Auger and V. P. Goldberg, "Prepaid Health Plans and Moral Hazard," Public Policy 22 (Summer 1979) :353-98. 79 In some cases, the maximum level of payment depends on the history of charges by the provider chosen. -66- because they do not cover the full cost of treatment, such schemes do not insure consumers as completely against financial loss from illness as may be desirable. For some forms of treatment, such as outpatient care, con- sumers often pay a fixed percentage copayment of the cost of treatment received. Copayments communicate some incentive to consider costs, but the incentive is less than it would be were the consumer allowed to keep all the savings generated in shopping more carefully and in questioning the need for particular tests.80 A numerical example helps illustrate how schemes with partial copayments give the consumer less incentive to shop than do schemes with binding maximum payments. With 80-percent coverage, a consumer saves only $1.00 out of pocket by choosing a physician charging $50 over one charging $55; thus, he has less incentive to search out the lower cost physician than he would were he to receive a larger part of the savings. For instance, were the consumer to receive $50 or $55 from the third-party payer regard- less of the cost of treatment or a maximum of $50 towards the cost of treatment, he would have the full $5 incentive to pick the lower cost provider. Of course, the consumer might still choose 80 While the paradigm of consumer choice may not be appropriate in all health-care contexts, this mode of analysis illuminates the way in which current schemes blunt incentives. To the $55 treatment, were the style of treatment offered by the $55 provider preferred at least $5 more than that of $50 provider.8l In general, plans with copayments give consumers more incentives for search and for cost savings than plans covering the full costs of care. There is no hard and fast rule for how large a copayment is necessary to induce consumers to search. A small percentage copayment (e.g., 10 percent) on a large expenditure may be sufficient to induce the consumer to search for a low-cost provider, while a larger copayment (e.g., 30 percent) may not be sufficient when smaller expenditures or emergencies are involved.82 Maximum payment limits can also be effective at inducing search and cost savings, where these limits are low enough. However, low limits and high copayment rates may negate 8l Also see section IV.C., question 8 and diagram: "Comparison of demand for service with and without third-party reimbursement." 82 For evidence, see J. P. Newhouse, "The Impact of Cost Sharing: Results from the Rand Health Insurance Equipment"; G. R. Wilensky, L. F. Rossiter, and A. K. Taylor, "The Role of Money and Time in the Demand for Health Care"; and Richard M. Scheffler, "Deductions, Coinsurance and the Demand for Health Care: Evidence from the UMW Plan" (papers presented at American Economic Association meetings, December 1981). Also see E. B. Keeler, D. T. Morrow, and J. P. Newhouse, "The Demand for Supplementary Health Insurance, or Do Deductibles Matter?" J. Pol. Econ. 85 (August 1977):789-801; and E. B. Keeler, J. P. Newhouse, and C. E. Phelps, "Deductibles and the Demand for Medical Care Services: The Theory of a Consumer Facing a Variable Price Schedule Under Uncertainty," Econometrica 45 (April 1974):641-55. BB part of the purpose of health plans--the elimination of financial risk asociated with illness.83 In many instances, consumers' opportunity to shop may be limited by the emergency nature of the need for treatment, or by lack of information. As has been explained by Enthoven, 84 perhaps the best way to improve incentives in such situations is by — instilling competition at the point where the consumer chooses a health plan. Because such choices are not overshadowed by the aura of emergency that surrounds some health conditions, consumers can better gather information and compare benefits and costs of various plans rationally when making decisions at that juncture. Competition among health plans also gives health plans the incentive to encourage cost savings in order to reduce premiums. Competition may therefore lead health plans to structure benefits payments in ways that communicate incentives for cost savings to consumers. Because most health plans are purchased through the 83 For a fuller discussion, see R. Kihlstrom and M. Pauly, "The Role of Health Insurance in the Allocation of Risk," Am. Econ. Rev. 61 (May 1971):371-79, and R. J. Zeckhauser, "Medical Insurance: A Case Study of the Tradeoff Between Risk-Spreading and Incentives," J. Econ. Theory 2 (1970):10-26. For a discussion of a proposal mitigating the impact of cost-sharing on the poor, see L. Seidman, "Income-Related Consumer Cost Sharing: A Strategy for the Health Sector," in National Health Insurance, ed. M. Pauly (Washington, D.C.: American Enterprise Institute, 1980). 84 gee, for example, A. Enthoven, Health Plan: The Only Practical Solution to the Soaring Cost of Medical Care (Reading, Mass.: Addison-Wesley, Inc., 1980). Also see F. K. Ackerman Jr., "Competition and Regulation: The Consumer Choice Health Plan Alternative," Medical Group Management 4 (July/August 1980) : 58-64. BP consumer's place of employment, corporate decisionmakers have an important role in introducing competition into the health sector by encouraging the introduction of new plans and by offering employees a choice among plans.85 Analyzing the incentives communicated by health plans is crucial to evaluating the appropriateness of competition as a resource-allocating mechanism. The most appropriate comparison is always between the allocation achieved by the market under a competitive approach and that achieved under a specific regulatory scheme. When third-party payment schemes reduce consumers' price sensitivity, a regulatory scheme that restricts entry or invest- ment may be even less able to hold down health-care costs. Providers facing few competitors may have little incentive to reduce costs or to compete for consumers by improving quality or offering variety. Thus, in heavily regulated markets in which most consumers bear no out-of-pocket costs, not only may consumers have little incentive to search; providers may have little incentive to compete. In many cases, reforming reimbursement schemes to increase the consumers' price sensitivity could improve the resource allocations achieved in both heavily regulated and competitive markets. 85 For more discussion on offering a menu of plans, see section IV.B., questions 2 and 3. Also see C. C. Havighurst, "Health Insurers and Health Care Costs: Can the Problem be Paru of the Solution?" Health Communications and Informatics 5 (1979):319-38, and J. R. Holzgraefe, "Controlling Health Care Costs Through Commercial Insurance Companies," Duke L. J. 1978 (May 1978): 728-52. “PQ 2. Reimbursement schemes: the effect on providers’ decisions.86 For many health services, consumers do not pay providers directly; rather, providers are reimbursed by third-party payers. The reimbursement schemes administered by third-party payers fall into two major categories: (1) schemes basing payment on actual cost incurred, and (2) schemes basing payment on a prospectively determined rate structure. Although the exact distortions created by the schemes vary from plan to plan, schemes in each category share some general features. Cost-based schemes are primarily used to reimburse institutional providers. The incentives created by schemes based on the actual costs incurred by facilities are analogous to the incentives created by schemes fully reimbursing consumers for the costs of treatment. Under such schemes, providers have little incentive to reduce costs or to produce services efficiently, because all costs are paid. Even when consumers have an incentive to search for providers charging lower prices, providers may have little incentive to attract more consumers if costs are fully reimbursed regardless of the number of patients served. Virtually all cost-based reimbursement schemes divide costs into various categories of eligibility for payment. The simplest 86 For a fuller discussion of many of the issues presented here, see M. Pauly, "Efficiency, Incentives, and Reimbursement for Health Care," Inquiry 7 (March 1970):114-31. Specific regulatory schemes are discussed in greater detail in section III.A. -71- characterization is that costs may be "allowable" (i.e., reimburs- able) or "nonallowable" (i.e., nonreimbursable). A facility may appeal an unsatisfactory categorization for an expenditure, but the process is time-consuming and payments are often delayed. Because providers are only reimbursed for "allowable" costs, they have an incentive to avoid making "nonallowable" expenditures. These incentives may lead to distortions in production decisions by encouraging health-care facilities to make "allowable" expendi- tures rather than "nonallowable" expenditures, even if larger expenditures are required to perform certain tasks. For example, when administrative costs are "allowable" but television and telephone charges are not, facilities may incur administrative expenses for billing patients individually for telephone and television rentals that are as great as the "nonallowable" charges themselves. Similarly, designating advertising expenditures as "nonallowable" ray lead facilities to attempt to attract consumers through community outreach programs or by offering internally subsidized emergency or maternity services. While such services have value to those that receive them, such programs may be an inefficient (but reimbursed) way to compete for consumers. Typically, institutional providers are reimbursed by a variety of third-party payers, each of which only pays for the services provided to its own beneficiaries. The presence of multiple payers creates the need to determine the costs of facilities and equipment used by patients receiving different i types of care. A common solution is for third-party payers to reimburse providers on a "cost-plus" basis for those costs that are fully assignable. The "cost-plus" system only approximates the full costs actually incurred in providing services to particular patients.87 Often it results in a system of internal cross-subsidies among services. This creates two types of distortions. First, for services or procedures where the "cost-plus" reimbursement rate is greater than the total cost of providing the service, the facility receives extra revenue above and beyond its costs for every unit of the service it provides. When the extra revenue is a percentage of the assignable costs, reducing these basic assignable costs means that less additional revenue is received. For example, when providers receive 110 percent of the assignable costs for a service, an increase in capital costs of $50 allows the facility to collect $5 in extra revenue. Even if the same service could be performed at lower total cost using less capital and more labor, the provider may have an incentive to use the more capital-intensive technique because it is reimbursed at a higher markup. Such incentives inadvertently reward high-cost providers with higher revenue per unit of service and punish pro- viders who reduce reimbursable costs. 87 some State rate-setting agencies include a fixed rate of return on capital equipment in cost-plus charged. The distortions created by those particular cost-plus schemes are discussed in section III.A. wl] Bo Second, when a service (such as an emergency room) does not pay its own way, the facility has an incentive to discourage its use, unless it is indirectly used to attract patients to services with higher markups (such as laboratory tests). In such situa- tions, the cost-based reimbursement scheme can distort the individual provider's incentives both to provide services as efficiently as possible and to allocate resources to their best use within the health-care facility.88 The case for imposing restrictions on providers to prevent price gouging is intimately related both to the level of competi- tion among providers of a particular service and to the incentives consumers have to search among providers. The consumers' incen- tives to search are partly determined by the structure of health plan benefits. Although third-party payment, rate setting, and negotiations are an integral part of many health-care markets and may thwart competition, there are techniques that third-party payers can use to simulate some of the incentives of a competitive market. Unlike the cost-based schemes, these techniques require that the rate schedule be set in advance and not be directly influenced by the provider's actions. These schemes typically base reimbursement rates on the costs of an average or proto- typical provider rather than on the individual provider's actual costs. Because the revenues received are not influenced by costs incurred, such schemes give providers the incentive to perform 88 gee section IV.D. for more discussion. wie services as efficiently as possible. For example, under "prospec- tive" reimbursement schemes, the total level of reimbursement is announced at the beginning of the fiscal period.89 Because any surplus can be retained but deficits are not compensated, this system gives the provider an incentive to provide the service at the least cost. Ideally, reimbursement schemes can simulate the benefits of competition even in markets where large economies of scale limit the number of providers or where the emergency nature of the ser- vice means that consumer search cannot realistically be expected to occur. In practice, schemes such as prospective reimbursement can create distortions that depend on the base used for making payments. For instance, when the total level of reimbursements is set for a hospital's treatment of Medicare patients, that hospital has an incentive to provide those patients with lower quality or less service. When the reimbursement rate is tied to the daily census of patients, hospitals have an incentive to encourage longer stays. When it is based on admissions, the incentive is for more admissions of patients requiring less costly care. The severity of the distortions created by each of these schemes depends on the degree of the preexisting distortion and on the ability of the provider to manipulate the variable on which payment is based. For instance, although there is a great 89 gee section III.A. for a fuller discussion of prospective reimbursement. Wy 374-115 0 - 82 - 6 : QL 3 incentive to cut the amount of service to individual patients when the system sets the total level of reimbursement, it is usually thought that other factors prejudice hospitals towards providing too much care to too many patients. 90 A reimbursement scheme with a built-in incentive to cut the quality of care might be of greater concern if applied to long-term-care facilities, where quality has long been perceived as a problem.91 On the other hand, schemes that change the facility's incentive to retain patients have limited distortionary effects in the long-term-care setting where many patients are life residents. In contrast, length of stay can vary considerably for many acute conditions and therefore reimbursement schemes may greatly influence a facility's incentive to retain patients during the recuperation period. The widely used "usual, customary, and reasonable" system for reimbursing professionals communicates many of the same incentives as prospective reimbursement schemes. Because the reimbursement a professional receives is based on the individual's historical charges and those of other professionals in the area but not on the cost of performing the service, the UCR system communicates an incentive to produce each service efficiently. However, the system does not necessarily provide incentives to pass this 90 See sections III.A. and III.B. 91 see section III.C. When reimbursement schemes contain an incentive to shade quality, it may be desirable to monitor the care patients actually receive. Unfortunately, such monitoring systems are usually quite costly. See section III.E. -76—- savings on to the consumer or to the third-party payer. For example, in markets where low copayments reduce the consumer's price sensitivity, or where, for other reasons, there is little incentive to compete, the UCR system can allow prices to rise substantially above costs and may facilitate the maintenance of prices at supracompetitive levels.92 Redesigning reimbursement schemes appears to be a promising method for instilling greater competition and improving performance in many markets. The apparent success of many health maintenance organizations in holding down the total cost of health care for their members appears to come both from their having the same incentive for cost savings as providers facing a prospective reimbursement scheme and from their need to pass on those savings in the form of reduced premiums in order to compete for membership. 93 Unintegrated third-party payers--that is, entities that pay for but do not provide health care--can give health providers the incentives to pass on cost savings. For instance, such plans can use high copayments to increase the consumers’ price sensitivity. Alternatively, third-party payers can set a 92 For discussion of problems with UCR and other methods of reimbursing for professional services, see C. C. Havighurst, "Professional Restraints on Innovation in Health Care Financing," Duke L. J. 1978 (May 1978) :303-87. 93 see section IV.B. for more discussion of health-plan competition. For a fuller discussion of HMO's, see H. S. Luft, Health Maintance Organizations: Dimensions of Performance (New York: Wiley-Interscience, 1981). Also see H. S. Luft, "Assessing the Evidence on HMO Performance," Milbank Memorial Fund Quarterly 58 (Fall 1980) :501-36. we maximum reimbursement rate based on the average cost of services in the market.%4 Of course, high copayments and maximum- reimbursement-rate schemes work best in markets where consumers have the time and information needed to shop among providers. In such markets, reforming reimbursement schemes can also increase the ability of new entrants to compete by offering lower cost services or by providing services more in keeping with consumers’ needs and preferences. In markets such as natural monopolies, where competition is not workable, prospective reimbursement schemes can nevertheless simulate the incentives for efficient production inherent in the competitive market and can thus improve the allocation of resources even in the absence of competition. Analyzing the incentives created for providers by a reim- bursement scheme is one of the most difficult parts of evaluating the potential for using a competitive approach to a particular market. Each reimbursement scheme communicates a different set of incentives to market participants. In addition, reimbursement schemes that thwart incentives for cost reduction under competi- tion may also thwart incentives for cost reduction when entry and capital investment are restricted. The presence of cost-based 94 For a discussion of these and other schemes, see D. W. Kallstrom, "Health Care Cost Control by Third-Party Payers: Fee Schedules and the Sherman Act," Duke L. J. 1978 (May 1978) :645-97. -78- reimbursement schemes alone does not indicate whether a competi- tive or a regulatory approach results in a more desirable alloca- tion of resources. Regardless of which approach to resource allocation is used, the incentives for cost reduction and appropriate innovation may be substantially improved by third- party payers' adopting schemes basing payments on the costs of an average provider or on prospective (expected) costs, rather than on the individual provider's actual costs for services performed. As the communication of appropriate incentives to providers improves, the assessment of whether competition or regulation can be expected to achieve better resource allocations may change. However, even in markets where a competitive appproach is not expected to be appropriate, adopting reimbursement schemes that communicate the correct incentives may substantially reduce costs and may help to improve the performance of health-care providers. -79- III. REGULATIONS AND THEIR INTERACTION WITH COMPETITION In ascertaining the potential for taking a competitive approach to a particular health-service market, planners should consider both the resource allocations expected under viable regulatory alternatives and the effect of existing regulations on the functioning of the market. Regulations are pervasive in this sector and largely define the structure of the health-care industry. For the most part, this regulatory environment must be taken as given at the local level. Planners make recommendations affecting the administration of only a small portion of these regulations, and private decisionmakers can affect the regulatory environment only through the political process. Federal regulations, such as those affecting prescription drugs or the overall level of demand for health care through the income-tax code, also have a substantial impact on most health- care markets. As such, these regulations do not directly affect the appropriateness of the competitive approach to any particular local service market. Because the effects of these regulations are far beyond the scope of allocating local health resources, regulations that affect the entire industry are not discussed in this paper. Other regulations, such as professional licensure and standards for institutional quality, also are typically outside the control of the local planner and must be taken as "givens." However, these regulations intrinsically affect specific markets Ye with which the planner and the corporate decisionmaker are concerned. Because the particulars of many of these regulations may differ from State to State and because their impact differs from market to market, local planners and private decisionmakers need the tools to analyze the impact of these regulations. Analyzing such regulations is a major aspect of determining the potential for introducing competition into a particular market. A third class of regulations--those controlling capital investment--is directly within the purview of the local planner. For example, certificate-of-need recommendations are one way planners can implement a competitive approach in a particular service market. Because changes in the regulatory environment may affect the appropriateness of a competitive approach, regulatory reform is probably the most significant factor affecting the use of market forces in health care. As regulatory reform proceeds, planners and others concerned with the performance of the health sector must continually reevaluate the competitiveness of individual markets. It also means that the major instrument for improving market performance is outside the planner's direct control. Much has been written about regulation, its efficacy in attaining its goals, and its effect on the market. Because extensive discussions of this topic are available elsewhere, the treatment here is limited to a basic overview of regulations wi] directly affecting local health-care markets.95 This section of the paper treats five major classes of regulation: price and cost regulation, capital-investment regulation, institutional-quality regulation, professional-quality regulation, and regulations governing provision of care. For each class, the paper discusses the goals of the regulations, the mechanisms used, the theory of how the mechanisms achieve their goals, the theoretical and empirical evidence as to their efficacy, and the unintended side effects of these mechanisms on the market. Because the particu- lars of many of these regulations differ from State to State, these discussions are fairly general. To aid the planner or local decisionmaker in investigating the effect of the regulatory environment on particular local markets, the paper discusses the relationship of each class of regulation to competition. This discussion focuses on whether these regulations enhance or thwart the competitive process, whether alternatives to current regula- tory schemes exist, and how the use of a competitive approach affects the need for these regulations. 95 Two excellent books for the reader interested in exploring the topic further are Patrick O'Donoghue, Evidence About the Effects of Health Care Regulation (Denver, Colorado: Spectrum Research, Inc., 1974); and Regulating Health Care: The Struggle for Control, ed. Arthur Levin, Proceedings of the Academy of Pol. Sci. 33 (1980):4. Along with a paper presented by B. Steinwald and F. Sloan, "Regulating Approaches to Hospital Cost Containment: A Synthesis of the Empirical Evidence," at the AEI Conference on Health Care (1980), these books provide valuable guidance to the empirical literature. The author relied heavily on these sources in developing this section. pC A. Regulation of Costs and Prices Price and cost regulations have traditionally been prescribed to control the allocation of resources in situations where market imperfections are so extreme as to preclude the use of the unencumbered market to allocate resources. Such regulations have been applied to natural monopolies in other sectors, such as electric utilities, to seek to limit firms' ability to extract excessive profits and to insure consumers greater access to prod- ucts. As discussed previously, natural monopolies are rare in health care. The classic justification for price and cost regula- tions is thus absent in this sector. Comprehensive third-party payment, high search costs, and lack of consumer information can give providers of health-care services the power to raise prices above the level expected in more competitive markets. However, price and cost regulations do not necessarily improve the market's performance in such circum- stances. Despite the lack of evidence that such regulations improve the allocation of resources, price and cost regulations have nevertheless been imposed on many health markets that cannot be characterized as natural monopolies. In many cases, the impe- tus for administratively setting prices stems from the role of the Federal and State Governments as third-party payers. While price and cost regulations may reduce the total revenues expended for health-care programs, these regulations may also distort the allocation of health resources and interfere with the market's ability to function. -83- In otherwise "healthy" markets, the imposition of price and cost regulations blunts market signals, thwarting the competitive allocation of resources. Of all forms of regulation currently applied to the health-care sector, cost and price regulations are among the most antithetical to the competitive approach. In addition, administratively setting prices requires that regulators have access to information not needed when prices are set by market interactions. This makes price and cost regulations quite burdensome to administer. Incentive-oriented reimbursement schemes can reduce the need for these regulations. By allowing providers to make decentralized pricing decisions, these schemes can generate the benefits of a competitive resource allocation. Such mechanisms are not merely a theoretical possibility. For example, reimbursement systems based on prices charged by representative providers can both allow decentralized pricing decisions and communicate the incentive for providers to perform health-care services as efficiently as possible. 1. The purpose of cost and price regulations. Theoreti- cally, the main purpose of cost and price regulations is the control of profit-taking and price-gouging by health-care monopolists and the control of monopolylike behavior by providers in health-care markets where third-party payment gives consumers little incentive to shop. General economic research shows that profit-maximizing monopolists find it in their interest to raise prices and reduce output relative to the socially optimal level. we Ble Similar behavior occurs among firms serving consumers with high search costs or with inadequate incentives to shop. 96 Many health-care facilities are operated on a "nonprofit" basis rather than on a "profit maximizing" basis. While these facilities may not raise prices as far above costs as other monopolists, they typically exhibit less vigilance in controlling costs.?7 some forms of price and cost regulations may induce more efficient use of inputs by these "nonprofit" providers.98 Price and cost regulations have also been used to promote internal cross-subsidization of services. Such internal subsidization of one service by another usually results in a less desirable allocation of services than that which could be achieved through a program of direct subsidies. However, rate- setters often have limited access to general tax revenues and may therefore rely on cross-subsidization as a way to achieve their goals.9? 96 For discussion of search and the effects of imperfect infor- mation, see section II.D. 97 For fuller discussion, see J. E. Harris, "The Internal Organization of Hospitals: Some Economic Implications," Bell J. Econ. 8 (1977):467-82. For another view, see M. V. Pauly and M. Redesch, "The Not-for-Profit Hospital as a Physicians’ Cooperative," Am. Econ. Rev. 63 (March 1973):87-99. 98 For evidence on these points, see Lave et al., "A Proposal for Incentive Reimbursement for Hospitals," Medical Care 11 (March/April 1973):79-89. 99 For discussion of cross-subsidization and empirical evidence, see J. E. Harris, "Pricing Rules for Hospitals," Bell J. Econ. 10 (Spring 1979) :224-43. wd By Price and cost regulation also may be used to control the quantity, variety, and quality of services patients receive. Regulations excluding certain costs from the rate base may induce providers to offer a particular variety of the service. Regula- tions setting reimbursement rates per hospital day or per illness indirectly affect the amount of care patients receive, especially in situations where providers determine the amount of the service patients consume. 2. Types of cost and price regulations.l100 a) Reimbursement based on the individual provider's costs. Regulatory schemes based on the individual provider's costs seek to reduce the provider's profits and to pass all reductions in profits and cost savings on to consumers or (in the case of publicly funded programs) to taxpayers. However, such schemes are costly to administer and tend to create incentives for inefficient behavior by providers. These cost-based reimbursement schemes-- including the Medicaid programs in most States--are conceptually similar to public-utility rate-of-return regulation. The vast literature on the problems of public-utility regulation may be applicable to cost-based health-care reimbursement schemes as 100 por a detailed discussion of specific programs, see P. L. Joskow, Controlling Hospital Costs: The Role of Government Regulation (Cambridge, Mass.: MIT Press, 1981), ch. 6, pp. 100-138. -86- well.lOl gynder such schemes, regulators determine the nature of "allowable" costs and set reimbursement rates that allow an "ade- quate" rate of return on capital investment. The object is to force the provider to increase his output to the level that would prevail in a competitive market. This is accomplished by limiting the provider's earnings to a "normal" or competitive rate of return and requiring that the provider serve all consumers. Because the rate of return is earned only on investment in capital, these schemes communicate an incentive for providers to substitute capital for labor in producing this level of output.102 While programs to regulate capital investment may mitigate this effect, inefficient uses of capital may be difficult to pinpoint, especially when they are adopted as industrywide practice. In a recent paper, two Maryland rate-setters argue that dividing the responsibility for investment and rate-setting decisions among State agencies actually weakens the ability of rate setters to audit and prevent inefficient investment by providers.103 101 por general discussion, see Alfred Kahn, The Economics of Regulation (New York: John Wiley and Sons, Inc., 1970). 102 por more detailed discussion, see H. Averch and L. L. Johnson, "Behavior of the Firm Under Regulatory Constraint," Am. Econ. Rev. 52 (December 1962):1053-69; W. Baumol and A. Klevorick, "Input Choices and Rate of Return Regulation," Bell J. Econ. 1 (Autumn 1970):162-90; E. Sheshinski, "Welfare Aspects of a Regulatory Constraint: Note," Am. Econ. Rev. 61 (March 1971):175-78. 103 Harold A. Cohen and Carl J. Schramm, "A Model for Resolving Planning Rate-Setting Conflicts," AEI Conference on Health Care, September 25-26, 1980. -87- Designing rate schedules that do not contain incentives for providers to alter the quality or variety of the service provided is considerably more difficult in health care than for traditional public utilities. Public-utility regulators can more easily measure kilowatt/hours or cubic yards than health-care regulators can measure "health". Unlike electric utilities, which have little if any role in advising their consumers on how much electricity they should purchase, health-care providers help determine whether the patient needs treatment. To some extent, this may enable health-care providers to maintain revenues despite reductions in the rates at which services are reimbursed. b) Reimbursement based on costs of prototypical provider. Regulatory schemes based on the costs of a prototypical or average provider are relatively new. Under such prospective reimbursement systems, costs exceeding the announced reimbursement rate are not covered by the third-party payer. As a result, the provider is at risk and has an incentive to reduce those costs as effectively as possible. To the extent that surpluses can be retained, providers have an incentive to reduce the amount of resources they use in serving their consumers, regardless of whether their costs are above or below the average. While prospective reimbursement communicates the correct incentives to proprietary institutions, there is some doubt about the desirability of such schemes in the case of nonprofit providers whose costs are below those of the prototype. Such -88- institutions cannot convert surpluses into profits; they can, however, use them to reduce their need for philanthropy, to cross-subsidize other services, to invest in additional equipment, to raise salaries, or to lower deficits. While some of these outlets for extra funds have no effect on efficiency, others may be undesirable.l104 In the case of nonprofit institutions, prospective reimbursement may simply drive the costs of both high- and low-cost providers towards the average reimbursement rate. In a study of hospitals reimbursed under prospective-reimbursement plans, Lave found that costs of nonprofit providers with costs below the prospective reimbursement rate rose faster than did the costs of nonprofit providers with costs above the prospective reimburse- ment rate.l05 While there are other explanations for this phenome- non (such as lags in cost increases between groups of hospitals), it may well be that prospective reimbursement programs communi- cate incentives for previously low-cost providers to raise their costs and for previously high-cost providers to lower theirs. The same study presented evidence that prospective reimbursement 104 ag in other sectors, access to long-term financing ultimately limits investments in the health sector. However, many experts believe that under current reimbursement schemes, the money markets only play a residual role in allocating capital in the health sector. For an alternate view, see R. Posner, "Certificate of Need for Health Care Facilities: A Dissenting View," in Regulating Health Facilities Construction, ed. C. C. Havighurst (Washington, D.C.: American Enterprise Institute, 1974). 105 pave et al., op. cit. -89- effectively communicates incentives to all proprietary institu- tions to reduce their costs, whether they are above or below the average. Thus, based on these empirical results, prospective reimbursement appears to be more appropriate for sectors such as nursing homes, where the majority of providers are proprietary institutions, than for sectors such as hospitals, where nonprofits often predominate. The "usual, customary, and reasonable" (UCR) method of reimbursing physicians shares many of the features of prospective reimbursement. The major flaw of UCR--that physicians with fees below the maximum reimbursement rate can easily raise their prices when consumers are not price sensitive--is also shared by prospective reimbursement plans where facilities are allowed to retain surpluses. 3. Effects of cost and price regulation. There are numerous empirical studies on the effectiveness of cost and price regula- tion. While many of these studies suggest that the regulations may lower health-care costs, most studies do not examine other effects of the programs on the market.106 Most price- and cost- regulation programs were instituted with the express purpose of controlling the rise in health-care costs. But neither the mechanisms spurring the high rate of health-care cost inflation 106 gee B. steinwald and F. Sloan, Regulatory Approaches to Hospital Cost Containment: A Synthesis of the Empirical Evidence, AEI Conference on Health Care (1980), for a concise overview of recent studies in this area. -90- nor the mechanisms by which the programs reduce the rate of increase are well understood. One theory of the recent run-up in health-care prices is that the health sector, which was subject to price controls under the Economic Stabilization Program (ESP) until April 1974, is still catching up from the distortions in prices created under that program. This theory is supported by several studies of the effects of the ESP price controls. Most of these studies found that ESP lowered the rate of inflation in the health sector during the period of the program.107 Some of these studies also indicate that relaxation of controls was followed by increases in health- care prices that exceeded the general rate of inflation.108 Another, less well formulated, theory of health-care infla- tion explains the increase on the basis of the prevalence of insurance, the physician's role in determining the mix of services 107 Strong negative effects are reported by S. Altman and J. Eichenholz, "Inflation in the Health Industry--Causes and Cures," in Health: A Victim or Cause of Inflation? ed. M. Zubkoff (New York: Prodest, August 1976); and by F. Sloan and B. Steinwald, "Effects of Regulation in Hospital Costs and Input Use," J. Law & Econ. 23 (April 1980):81-107. No effect was found by P. Ginsburg, "Impact of the Economic Stabilization Programs on Hospitals: An Analysis with Aggregate Data," in Hospital Cost Containment, ed. M. Zubkoff, I. Raskin, and B. Hanft (New York: Prodest, 1978), pp. 293-323. 108 A major problem with empirical work on rising health costs is that technology has also changed rapidly. It is difficult to determine whether the change in costs is associated with an equally valuable improvement in the quality of care. For a dis- cussion of this problem, see M. S. Feldstein, "Quality Change and the Demand for Hospital Care," Econometrica 45 (October 1977): 1681-1702. mG Lows 374-115 0 - 82 - 7 : QL 3 patients consume, the high rate of technology expansion, and the high costs of consumer search. This theory, however, fails to explain why health-care costs are not even higher than they now are.l109 A Congressional Budget Office study found that private, voluntary cost-containment programs had only a slight and statistically insignificant effect on health expenditures, while mandated State rate-setting programs had a sizable effect.110 The type of program adopted in the State appeared to have little relationship to the State's prior history of health-care cost increases. Several other studies also report evidence that State- mandated programs have helped hold down the rate of health-care cost increases.lll Some studies report that the volume of services consumed has increased more rapidly in States where programs have effectively 109 For a more detailed discussion, see K. Davis, "Theories of Hospital Inflation: Some Empirical Evidence," J. Human Resources 8 (1973):1981-2010. 110 congressional Budget Office, Controlling Rising Hospital Costs (Washington, D.C.: Government Printing Office, September 1979). 111 American Hospital Association, Reimbursement Survey (Chicago: American Hospital Association, October 1979); Health Care Financing Administration, Research and Demonstrations in Health Care Financing (Washington, D.C: Government Printing Office, 1980); F. Sloan, "Regulation and the Rising Cost of Health Care," reported in F. Sloan and B. Steinwald, Regulatory Approaches to Hospital Cost Containment, op. cit. -92- restrained cost increases on a per unit basis.ll2 Increases in total services provided decrease the overall savings from the cost-control programs but do not necessarily indicate a problem with the regulatory programs. The result is compatible both with traditional models of regulation and with models which assume that providers decide how much treatment consumers receive. Traditional models predict that price regulations force pro- viders to act like firms in a competitive market. To maximize profits at the "normal" rate of return, they must increase their supply of services to the level that would be found in a competi- tive market. In the case of a profit-mazimizing monopolist, the increase in the volume of services supplied is a desirable result because monopolists ordinarily create distortions by producing too little volume. Other models predict that, in the face of regulatory con- straints, health providers can maintain overall revenues by decid- ing that each patient will consume more services.113 Were these models accurate, such an increase in volume might be a perverse side effect of price and cost regulations. But it is difficult to determine whether this phenomenon is occurring without testing 112 p, s. salkever, Hospital Sector Inflation (Lexington, Mass.: D.C. Heath, 1979); F. Sloan and B. Steinwald, "Effects of Regulation on Hospital Costs and Input Use," J. Law & Econ. 23 (April 1980):81-109. 113 For more on this theory, see The Target Income Hypothesis and Related Issues in Health Manpower Policy, U.S. Department of Health, Education and Welfare, Public Health Services, Bureau of Health Manpower, DHEW Publication No. (HRA) 80-27, January 1980. YY more detailed models of provider behavior and how price and cost regulations change their incentives. There are numerous explanations of how price and cost regulation might lower the rate of health-care inflation at least in the short run. Included among the possible mechanisms for inducing these savings are increased productivity through more efficient use of resources, decreased input prices through more effective negotiations with input suppliers, changes in case mix to less critically ill and therefore less "costly" patients, reductions in the quality of care, and substitution of ambulatory care for institutional services. Cost reductions achieved through some of these mechanisms have unintended side effects that may or may not be desirable in the long run. Some cost-saving mecha- nisms, such as delayed maintenance or delayed salary increases, cannot be reproduced annually and may perversely affect the quality of care even in the short run.114 The longrun effects of cost and price control schemes may depend on their relationship to capital expenditure controls, which are discussed later in the paper.l15 Ultimately, the effectiveness of various price- and cost-control mechanisms may 114 For a discussion of particular programs and the mechanisms by which they work, see R. Buttisella and S. Eastaugh, "Hospital Cost Containment," in Regulating Health Care: The Struggle for Control, pp. 192-205. 115 cohen and Schramm discuss the relationship between rate setting and capital investment in their paper (cited previously). “Od depend on the other regulations in the market as well as on the underlying market structure. 4. Alternatives to price and cost regulation for institu- tional providers. Administratively setting prices and designing rate schedules without perverse incentives is time-consuming, costly, and unwieldy. Such regulatory efforts involve gathering massive amounts of information and second-guessing the decisions and behavior of institutional providers. In the literature of public-utilities regulation, several alternatives have been suggested which require less intervention and which communicate incentives for providers to act in a socially desirable manner. The advantages and disadvantages of two of these schemes--public ownership and "franchising"--are discussed below. Public ownership, one alternative to regulating natural monopolies, is relatively common in health care. Ideally, this alternative works by imbuing the managers of a publicly owned health-care facility with the incentive to act in the public interest. In practice, public ownership in health care often results in inefficient management. Consequently, the desirability of public ownership as an alternative to regulating natural monopolies remains a controversial issue. Moreover, because there are relatively few natural monopolies in the health sector, the usual rationale for public ownership as an alternative to price regulation is not present in most instances. -95- Another alternative to price and cost regulations is to allow private entities to bid for the right to supply the market .116 Proposals can be evaluated on the basis of price and quality guarantees, with the firm offering the best combination being awarded the contract or "franchise". This "franchising" alter- native is not a general substitute for other forms of natural- monopoly regulation but can provide a less costly and less intrusive alternative in some situations. It is probably best suited to services like Meals on Wheels, where there may be several potential suppliers but where the service might be provided most efficiently by one firm in each geographic area. In designing alternatives to price and cost regulation, the goals to be accomplished must be carefully articulated. In many cases, price and cost regulations seem to be designed solely to lower governmental and private expenditures for health care. As discussed above, 117 large and increasing expenditures may result from the prevalence of payment schemes that mask the true costs of services. When the source of the problem is the predominance of third-party payment, price and cost regulations are almost always an inappropriate solution to the problem of large health- care expenditures because these regulations further distort the 116 Rg. pemsetz, "Why Regulate Utilities?" J. Law & Econ. 11 (1968) :55-65; O. Williamson, "Franchise Bidding for Natural Monopolies," Bell J. Econ. 7 (Spring 1974):73-101. 117 gee section II.F.1l. -96- allocation of resources. In many markets, competitive solutions may be the best remedy. It is to these we now turn. 5. Interaction of price and cost regulation with competi- tion. Applied to otherwise "healthy" markets, price and cost regulations blunt the ability of providers to compete for custo- mers. When imposed on most markets, price and cost regulations have distorting effects and few benefits. Nevertheless, price and cost regulation does not entirely stifle the ability of providers to compete. For example, when regulations control the prices providers charge, providers often compete by offering higher quality service or by tailoring their service to the needs of a particular segment of the population. While such competition is less efficient than competition based on price as well as quality and variety, competition in quality and variety alone may be beneficial. In some cases, this type of competition results in a better mix of services than would be achieved if all competition were stifled. In addition, it is often even more difficult to administratively determine the optimal quality level and to quell all quality competition than to administratively set prices. In some States, rate-setting commissions and other Government agencies are currently reexamining their reimbursement systems and attempting to design systems that better communicate incentives for providers to compete and to provide services as efficiently as possible. The prospective-reimbursement programs recently adopted in Maryland and some other States, for instance, - allow market forces to play a larger role in the allocation of resources. One feature of the Maryland plan is that providers are put at risk for their decisions by the State's firm commitments not to pick up the debt of failing providers and to allow providers to keep for their own use the bulk of any surplus generated.l18 For some health services, the adoption of incentive-based reimbursement schemes means that the presence of third-party payers has a less distorting influence on the market. In these cases, the use of a more competitive approach to the health-care market can go hand in hand with the introduction of new methods of price regulation and innovative reimbursement systems. Not only can regulatory reform allow for a more competitive approach, but the increased competition can also eliminate the need for rate-setting commissions to administer prices. Allowing entry into a market can reduce the need to control monopoly profit-taking and price-gouging because competition among pro- viders drives prices down. Allowing increased competition for institutional services like kidney dialysis, long-term care, and obstetrical care may eliminate the necessity for price or cost regulations in these markets. Price competition rests on the incentive, knowledge, and ability of consumers to shop among providers. In removing price 118 gchramm and Cohen, op. cit. For more discussion on this point, see section II.F. 2. -98- and cost regulations and replacing them with a competitive approach, rate-setters and health planners should work with third- party payers, employers, and State legislators to assure that, where it is feasible, consumers have adequate incentives to search. While the replacement of price and cost regulation with a competitive approach may not be possible in health service markets which constitute natural monopolies, 119 there are many markets where the redesign of reimbursement systems and the increased potential for taking a competitive approach go hand in hand. B. Regulation of Capital Investmentl20 Expenditures for capital investment and facilities constitute a large share of total health-care costs. Most of these costs are associated with services provided by the hospital sector, and thus regulations of capital investment in health care are often designed to have their major impact on that sector. The primary vehicle for control is certificate-of-need legislation, requiring health-care institutions to document community "need" for proposed expansion. The role of government under certificate-of-need programs, regional health-planning programs, and Section 1122 reviews is basically restrictive, contrasting sharply with the 119 gee section II.C. for discussion of natural monopolies in health care. 120 For a classic discussion of this topic, see C. C. Havighurst, "Regulation of Health Facilities and Services by Certificate of Need," Virginia L. Rev. 59 (October 1973):1143-1242. -99- stimulative role taken by the Federal Government under the Hill- Burton hospital financing program.l121 Whereas Hill-Burton sought to stimulate investment after World War II when underinvestment in hospital construction and slow adoption of new technology were perceived to be problems, the controls introduced since the enactment of Medicare and Medicaid in 1965 have sought to limit investment and to slow the proliferation of technology .122 1. The purpose of current regulatory schemes. Programs regulating health-care investment are designed to constrain expenditures on capital equipment and to assure that approved investments are allocated rationally within a planning area. The premise behind these regulations is that the market does not lead to the "correct" level of capital investment in health. Often, the data cited to argue for such regulations are statistics on the growth in health-care investment and expenditures relative to the growth in national income. By themselves, such data are not evidence of overinvestment. A large proportion of the investment in this sector in the past 15 years may have been necessary to satisfy the burgeoning demand for health care (a demand that 121 Hill-Burton Hospital Survey and Construction Act (August 13, 1946), P. L. 79-725, 60 Stat. 1040. (It has since been repealed.) 122 1n some States, the subsequent efforts to control reimburse- ment levels have created problems for health providers seeking to borrow money from private lenders and in the capital markets. While poor credit ratings can limit investment in the health sector, these ratings may reflect the market's perceptions of the rate-regulating authority rather than the underlying desirability of investment in the health market. For more on this point, see R. Posner, op. cit. -100- resulted from the introduction of Medicare and Medicaid) and to incorporate rapid changes in technology into existing health facilities. Private third-party-payment schemes requiring little or no copayment for inpatient care also have increased the demand for health care and have further encouraged investment in tech- nology and in capacity. While the degree of overinvestment is unclear, the amount of investment in the health-care sector appears to be larger than that which would have occurred in a market where providers bore the risk of their investment decisions. There are several explanations for the phenomenon of over- investment in the health-care sector. These explanations are relevant because the effectiveness of a regulatory scheme almost inevitably is related to the ability of the scheme to correct or override the underlying cause of the problem. Thus, the efficacy of schemes designed to correct overinvestment depends on the match between the cause of the problem and the approach taken to correct ie. a) Rate-of-return regulation and increased investment.123 Overinvestment in health care is often attributed to the form of reimbursement that predominates in the health-care sector. As was previously discussed, 124 under most reimbursement schemes, the 123 gee Schramm and Cohen, op. cit., for more detail. 124 gee section IV.A. -101- amount of revenue a facility receives partly depends on its capital expenditures in a previous period. These cost and price regulatory schemes may give health-care providers two types of incentives to increase capital investment, one of which brings the facility closer to the efficient level of service provision, while the other results in overinvestment. The first type of incentive for increased investment is that when an institutional provider has market power, it invests in too little capital because it serves too few consumers at too high a price. Thus, in such situations, increased investment represents an increase in efficiency. This situation applies only to those markets where there are a small number of providers. The second incentive for increased investment is that, in some cases, the rate of return that health-care facilities receive is based only on capital investment and not on expenditures for operating costs. Cost and price regulations typically constrain the facility's earnings below the level of revenue it could receive in the absence of the regulations. This means that if the facility invests in more than the optimal level of capital to produce a given level of output, it receives more revenue because the regulations allow it to earn a positive rate of return on investment and because it is still not charging as high a price as it would charge in the absence of regulation. The result is an efficiency distortion caused by the provider's using too much capital to produce a given level of output. This effect, known as -102- the Averch-Johnson effect, has been documented in other regulated 125 Regulation of the costs and prices is usually industries. combined with review of capital investment to prevent this form of overinvestment. b) Reasonable cost reimbursement and increased investment. A second reason for overinvestment in health care relates to another aspect of reimbursement for health-care services--the notion of "reasonable" costs. In some States, health-care facilities are reimbursed for all expenditures deemed "reasonable," in the sense that they are both necessary for patient care and not exorbitant. Under such schemes, certain classes of expenses--such as private nurses, television sets, or community education--are routinely disallowed. Without capital investment regulation, the incentive for overinvestment occurred because the need for a particular piece of diagnostic equipment or additional suite of beds was not closely examined. It is partly in response to this bias in the system that certificate-of-need was developed. c) "Technology" competition and overinvestment.l26 Those concerned with resource allocation in the health system frequently express concern that "technology competition" may be another cause 125 por a recent theoretical presentation of this model, see W. Baumol and A. Klevorick, "Input Choices and Rate of Return Regulation," Bell J. Econ. 1 (1970):162-90. 126 por further discussion, see L. B. Russell, "Regulating the Diffusion of Hospital Technologies," Law and Contemporary Problems 43 (Winter/Spring 1979) :26-42. -103~ of overinvestment in health care. Although higher quality health care is not always associated with more technology, consumers may lack sufficient information to judge for themselves whether more technology means higher quality care for their specific condition. Because insured consumers often pay no more for more costly forms of care, they have little, if any, incentive to choose less costly providers. If consumers, or more typically their physicians, choose institutional providers on the basis of which one has the latest model equipment and the most extensively equipped labs, investment in technology becomes a necessary component of the competition for patients. In some cases, choosing a hospital on the basis of its equipment is appropriate because consumers receive real benefits from patronizing the provider with the most sophisticated equip- ment. In other cases, however, consumers or their physicians may employ technology as a general signal of the quality of care they will receive, even though it may bear little relationship to the facility's quality with respect to a particular procedure. For example, because the patient may lack other criteria or relevant information for making a judgment, a patient entering a hospital for a relatively simple procedure may choose the hospital with a reputation for having the best equipped testing facilities. The hospital in fact may have no better facilities for the procedure in question than would a small community hospital--and may even provide lower quality general nursing care. However, -104- because this information is not readily available to consumers, the small community hospital may be forced to expand its invest- ment in technology in order to attract patients.127 The absence of widely available and costless information about institutional quality on specific procedures is a factor in consumers' using such "signals" .128 The capital equipment a hospital has not only attracts patients; it is probably even more important for attracting high- caliber staff. The quality of the staff, in turn, is important for attracting patients. As soon as one facility in the region has a new diagnostic device, other facilities may feel pressure from their staffs to obtain the new equipment. In some cases, this leads primary and secondary facilities to acquire equipment that is needed primarily for delivering tertiary care. It is important to note that overinvestment in technology does not necessarily imply that stifling all technology competi- tion is desirable. Competition increases the diffusion of tech- nology and spurs new technological advances. Without any technology competition, such advances may be diffused too slowly, and there may be an underinvestment in technology, which is also undesirable. 127 por the theory underlying this argument, see A. M. Spence, "Job Market Signalling," Q. J. Econ. 87 (1973):355-74. 128 por more on this point, see section IV.D., question 26. 105m When lack of consumer information and price sensitivity distort the market in such a way that overinvestment results from competition on the basis of technology, regulation of investment obviously does not alleviate the underlying problem. Moreover, such regulations may have unintended adverse effects even in the short run. For example, it may put those already possessing the technology or those allowed to acquire it at an advantage rela- tive to those denied the opportunity. This may give "privileged" facilities a competitive advantage in attracting physicians and patients. There are more direct solutions to this problem than regulating investment. For instance, providing information about hospital performance on particular procedures could allow con- sumers to choose among hospitals on the basis of more complete information and thus eliminate the need for costly signaling.l29 Another solution would be to alter reimbursement schemes so that hospitals with less extensive technology can compete on the basis of price and so that consumers receive a benefit from patronizing providers offering both lower levels of technology and lower prices.130 129 por the theory of when and how such information should be provided, see H. Beales, R. Craswell, and S. Salop, op. cit. Also see sections II.D., III.C., and III.D. 130 gee section II.F.l. and section III.A. -106- d) Roemer's Law and overinvestment. Roemer's Law, "a bed built is a bed filled is a bed billed," or that supply creates its own demand, is believed by some to hold in the hospital indus- try.131 When hospital beds are scarce, treatments not requiring hospitalization can be used for many conditions and lengths of stay can be shortened. According to Roemer's theory, as the capacity of the hospital sector increases, patients with less and less serious conditions are given inpatient care. The phenomenon of supply increasing generating demand occurs, in a sense, in almost any market. The usual mechanism is for the supply increases to work through the price system. When supply increases, price falls, and the quantity demanded increases. By contrast, in health care, such price effects are often limited. Physicians and hospitals have incentives to overutilize hospital beds, and as more capacity is built, some experts theorize that more demand for these beds is generated even though there is no 131 M. S. Roemer, "Hospital Utilization and the Supply of Physicians," J.A.M.A. 178 (December 1961):933-89. For more recent evidence supporting this view, see A. Gettelsohn and J. Wennberg, "Small Area Variations in Health Care Delivery," Science 182 (December 1973):1102-08. For evidence that this effect may be weaker than previously thought, see M. S. Feldstein, "The Supply and Use of Inpatient Care," in Economic Analysis of Health Service Efficiency (Amsterdam: North-Holland Publishing Co., 1967). -107=- 374-115 0 - 82 - 8 : QL 3 price decline.132 Because there are few out-of-pocket costs to the consumer for receiving inpatient care rather than outpatient care or for recuperating in the hospital rather than at home, it is reasonable to suspect that consumers will utilize additional capacity until the perceived value to them of using these addi- tional hospital resources is nil.133 1p addition, because con- sumers rely on the judgment of physicians to determine the need for hospitalization and the length of stay, it is possible that, partly in response to institutional pressure, physicians fill up more than the socially optimal number of hospital beds.134 Roemer hypothesized that, together with the third-party-payment system, the incentive for overutilization is strong enough to insure that additions to capacity lead neither to lower prices nor to empty beds. If this theory is correct, it means that unregulated investment in the hospital sector combined with current reimburse- ment schemes would probably result in far too great a level of 132 of course, most people do not choose to stay in hospitals and wish to avoid lengthy stays. Nevertheless, lengths of stay for some procedures vary widely, particularly cross-nationally. For many conditions, Europeans use longer stays and less surgery than Americans. The amount of time spent in the hospital is most often determined by the physician, who has the discretion to shorten stays if bed space is needed. 133 Despite Roemer's argument, there is a clear trend to shorter hospital stays. One explanation is that utilization review programs mean doctors are watched carefully with respect to length of stay. Challenges on admissions appear less frequent, however, and there also appears to be a trend toward more numerous stays. 134 of course, the ability of physicians to determine lengths of stay also means physicians can choose to shorten stays if given the correct incentives. -108- hospital capacity. As will be discussed below, the empirical evidence does not support this theory. 2. Programs to control capital investment.l35 state certificate-of-need (CON) programs mandated by Public Law 93-641 and Federal Section 1122 reviews mandated by Public Law 92-603 are the basic mechanisms by which capital investment in health- care institutions is regulated. In some States, CON functions much like a building permit, and denial of a certificate of need may effectively bar entry. In most States, Section 1122 review is required only for providers seeking Medicare or Medicaid reimbursement.l36 Although the details of CON differ slightly from State to State, the conceptual underpinnings of CON and Section 1122 review programs are the same. The motivation for the regulations is the belief that the unconstrained market results in too much invest- ment in health care. The programs operate by reviewing proposals by existing health-care facilities and by potential entrants for 135 programs differ considerably from State to State. Therefore, this section only contains a cursory discussion of the major types of programs. For a description of programs in each State, see P. L. Joskow, Controlling Hospital Costs: The Role of Government Regulation, op. cit., ch. 5. 136 some States combine the Section 1122 reviews with the certificate-of-need process. -109- substantial new investments in hospital-bed capacityl37 and for purchases of expensive capital equipment .138 In some States, pro- posed hospital closings or reductions in beds are also reviewed. In theory, the regulatory programs constrain supply to the level of the community's "need" according to demographic and epidemiologic criteria and allocate the capacity according to "access" criteria. In practice, the lags between proposals and completion of construction can be so great that overcapitalization or undercapitalization in a market may be exacerbated rather than alleviated by these regulatory controls. As discussed in the next subsection, the predominance of the evidence supports the conclusion that HESS SOHENES have little, if any, effect in controlling costs. Private third-party payers negotiate reimbursement rates on their own. Some, such as Blue Cross, have developed private programs quite similar to the Federal and State investment review programs .139 There is some evidence that these private programs 137 con applications must be filed for changes involving 10 or more beds. 138 The Federal standard is $150,000. Some States have a lower limit, usually $100,000. 139 1n National Gerimedical v. Blue Cross of Kansas City, the Supreme Court ruled that where the State has not adopted an HSA's determination regarding "need," a private insurer who relies on that ruling in its reimbursement decisions is not immune from antitrust scrutiny. However, this does not mean that the private insurer has necessarily violated the antitrust laws. That issue is currently before the lower court. National Gerimedical Hospital and Gerontology Center v. Blue Cross of Kansas City et al., No. 80-802. Supreme Court of the United States, 49 U.S.L.W. 4672, 1981-1 Trade Cas. ¥ 65125, June 15, 1981. ~110- are more successful than those used by State and Federal Government. The evidence on the efficacy of public and private programs is discussed more fully below. 3. Effects of capital investment regulations. Because the primary purpose of capital investment regulation is to constrain entry by new firms and expansion by existing firms, empirical analysis of the effect of these regulations should address two issues. First, it is important to determine whether the regula- tions do in fact constrain investment. Second, the effects of entry and expansion restraints on the ability of the market to allocate resources efficiently should be ascertained. Unfortun- ately, virtually all research on the effects of these regulations is limited to the first issue. These research findings show little evidence that regulation constrains investment. a) The effect of certificate-of-need programs. An American Hospital Association survey of hospitals reported that 53 percent of those hospitals discontinuing capital expansion plans in 1978-79 did so because CON was denied or was too costly and uncertain to undertake.l40 Despite this affirmation of the importance of CON in the hospital administrator's investment decisions, most statistical studies discussed in this section show little if any significant effect of CON on the total level of investment. 140 American Hospital Association, Reimbursement Surve, (Chicago: The Association, October 1979). -11)- When studies conducted in the early 1970's showed no effect, a frequently offered explanation was that the programs had not been in place long enough to have much impact. Today, with programs having been in place as long as 17 years in New York and for 10 or more years in many other States, there is still scant evidence that CON holds down costs.l4l One study by Sloan and Steinwald presents some evidence that programs focusing on beds alone have more effect than those also reviewing service expansion and purchases of relatively inexpensive equipment. However, the study concludes that the effect of these narrowly focused programs is still very weak. 142 This study also showed that costs tended to rise significantly just before CON was put into effect. The authors explain this as "investment in anticipation of regulation." It is difficult to differentiate their theory from another explanation of the phenomenon--that State legislatures vote for CON after a surge of investment activity. Thus, the decline in additional investment following the adoption of CON laws may be "natural" and not the 141 Fp. sloan, "Regulation and the Rising Cost of Hospital Care," 1980 (mimeographed); F. J. Hillinger, "The Effect of Certificate of Need Legislation on Hospital Investments," Inquiry 13 (June 1976) :187-93. 142 fp. sloan and B. Steinwald, "Effects of Regulation on Hospital Costs and Input Use," J. Law & Econ. 23 (April 1980):81-109. -112m- effect of enacting CON laws. Joskow's findings that capital- investment regulations are most successful in States with higher than average costs supports this alternative hypothesis.143 Authors of another study report evidence that CON laws con- strain expansion of bed supply but result in higher levels of in- vestment in assets per bed. They report that CON laws have no net effect on the overall level of investment.l44 studies of individ- ual ancillary hospital-based services also find no effect.145 b) The effect of Section 1122 review programs. Section 1122 review programs for Medicare reimbursement fare little better than CON programs in studies of their effectiveness. The American Hospital Association reported that only 3 percent of hospitals in States with Section 1122 review programs had costs disallowed or reduced under the program.l146 While 41 percent of hospitals’ indicated that cutbacks in their expansion plans were influenced by the existence of Section 1122 review programs, this survey evidence is unsupported in statistical studies comparing States with and without Section 1122 review. One such study reports no 143 p, 1. Joskow, Controlling Hospital Costs: The Role of Government Regulation, op. cit., p. 168. 144 p, sg. salkever and T. W. Bice, "The Impact of Certificate of Need Controls on Hospital Investment," Milbank Memorial Fund Quarterly 54 (Spring 1976):185-214. 145 bp, L. Joskow, Controlling Hospital Costs: The Role of Government Regulation, op. cit., pp. 162-68. 146 american Hospital Association, op. cit. -113- effect.147 Another reports mixed results, although the author claims that Section 1122 review on the whole is more effective than CON laws.l48 c) The effect of private programs. The American Hospital Association survey reports that only 22 percent of hospitals altering capital expansion plans did so in response to Blue Cross approval requirements and that only 2 percent experienced cost disallowances or reductions under these plans.149 However, a statistical study of the Blue Cross program reports a small but significant effect of the program on the average cost of admissions and on bed expansion.l50 Although the results of this study are not conclusive, these results support the contention that the market gives third-party payers a greater incentive to come up with an effective system to control capital investment than can be realized through the inevitable compromises inherent in legislative and bureaucratic processes. More research is needed to document the success of the Blue Cross program. d) General effect of capital investment regulations. Overall, there is little evidence that CON, Section 1122 review, 147 Fp. sloan and B. Steinwald, "Effects of Regulation on Hospital Costs and Input Use," op. cit. 148 Ff, sloan, "Regulation and the Rising Cost of Health Care," op. cit. 149 American Hospital Association, op. cit. 150 Ff. sioan and B. Steinwald, "Effects of Regulation on Hospital Costs and Input Use," op. cit. -114- or private-investment review schemes significantly hold down the level of investment. There is some evidence that these programs, together with cost-based reimbursement schemes, skew investment towards capital equipment expenditures and away from increases in the number of beds.l3l There is also evidence that these programs retard the investment process.152 In a period of overall infla- tion, such lags may cost more in added construction costs incurred at the later date than they save through postponing the opening of new facilities. One plausible explanation of why the public programs have failed to achieve appreciable results is that these programs put local planners in a difficult political position. The system largely depends upon the decisions of local planners as to how Federal money should be spent. However, these planners must work closely with community members to see their plans carried out. Many of the political conflicts involved in health planning are eased by granting CON's liberally. The incentive to recommend that CON requests be granted is even greater when a large propor- tion of the cost is borne by the Federal system. In some cases, denials of "need" mean the loss of Federal revenues to other communities. While most planners attempt conscientiously to fulfill their mandate, the pressures they face work against strict application of these regulations. 151 1pia. 152 71pia. -115- A further problem is that planners are directed to use nationally established demographic and epidemiologic criteria.l53 This creates two problems. First, as Paul Joskow has shown in his work, a given level of average capacity utilization implies a higher probability that patients will be turned away in rural areas than in urban areas.l54 This result suggests that to guar- antee patients throughout the country the same access to health facilities, the target rate of capacity utilization should depend on the population of the region. Second, using demographic and epidemiologic criteria rather than engaging in an analysis of the economic viability of the program can result in efficiency distor- tions when the demographic and epidemiologic criteria do not match the dictates of market forces. In addition, "market" viability alone may also be an inadequate test, especially when price is not set by market forces. As Cohen and Schramm (two Maryland regula- tors) point out, a requirement of "financial viability" is often 153 gee 42 usc, § 300n-1 (1981). 154 This useful insight is simply an application of the "law of large numbers" from probability theory: as the sample size grows large, the mean of the underlying distribution is more likely to be observed in the sample. As applied to hospitals, this means that for the same probability of conditions requiring hospitali- zation and the same number of beds per thousand, the less likely it is that there will be excess demand in a large city than in a small town. For development of this model, see P. L. Joskow, "The Effects of Competition and Regulation on Health, Bed Supply, and the Reservation Quality of the Hospital," Bell J. Econ. 11 (Autumn 1980) :421-47. -116- meaningless because cost-based reimbursement makes any agency- approved plan economically viable.l55 as long as the dominant third-party payers have a policy of covering the facility's capi- tal costs, the capital market should be willing to provide the funds.156 under such conditions, ordinary market-based tests of capital investment proposals fail in this industry.157 Finally, it may be that despite their negligible effect on overall health-care costs, these programs help discourage new entrants to the health sector. For instance, in some cases exist- ing hospitals have been able to get CON's to expand their facili- ties, while HMO's in the same region have had difficulty getting a CON to build completely new facilities. While such decisions may result in fewer total hospital beds, they may also result in less competition. For example, without a CON, the HMO is forced to negotiate with existing institutions for use of their facilities. This makes entry by HMO's more difficult; in some areas where there is substantial professional hostility to the HMO concept, it may effectively thwart entry. The empirical literature sheds little light on this point. However, if these programs raise significant barriers to new entry while not limiting expansion by 155 cohen and Schramm, op. cit. 156 The recent problems some hospitals have experienced in the capital markets probably indicate that financial analysts doubt that State and private third-party payers can long continue this policy. 157 por an alternative view, see R. Posner, "Certificate- of-Need: A Dissenting View," op. cit. -117- incumbents, they may eliminate the already limited incentive that existing providers have to improve their services in order to attract consumers. Regulation of capital investment is inextricably tied to cost and price regulation. Like cost and price regulation, regulation of capital investment has its place in markets characterized as natural monopolies. However, such markets constitute a tiny fraction of all the health-care markets to which such regulations are now applied.l58 Controls on investment may also be needed in markets where providers bear no risk for their investment deci- sions. Reforming reimbursement schemes is a far more direct solu- tion to that problem. When such reforms are enacted, many of the current controls on investment may no longer be needed. Regulation of capital investment is sometimes proposed to restrict excessive investment by hospitals as part of the "tech- nology war" used to attract patients and physicians. The problem, however, stems from consumers and physicians using tech-nology as a proxy for the quality of the hospital in the absence of other information and from a general lack of price competition. The effect of restricting investment is to give an advantage to some facilities over others. As discussed earlier, more direct solu- tions to the problem of excess investment in technology would be to provide consumers and their physicians with information about other indices for quality on a procedure-by-procedure basis and to 158 gee section II.C. -118- revise reimbursement schemes. With better information, consumers and their physicians could choose the health-care facilities best suited to their particular needs. Such information also might spur competition among hospitals to increase their overall level of quality, especially in those areas most important to patients and their physicians.139 4. Capital-investment regulations and the competitive approach. Regulation of capital investment may be appropriate for restricting excessive investment by natural monopolies subject to rate-of-return regulation and by firms operating in markets where reimbursement schemes remove all risk associated with unnecessary investments. As has been discussed above, natural monopolies are a rare phenomenon in the health-care sector.160 For the few ser- vices in particular geographic markets that can be characterized as natural monopolies, regulation of prices and capital investment might be combined to force the firm to provide the socially desirable level of output. In such situations, the regulatory process may be used as a surrogate for the competitive market to achieve a better allocation of resources. Even in these cases, however, regulations may create additional distortions.161 159 See section II.E.2. and section III.C. 160 gee section II.C. 161 1p addition, some critics charge that the split between regulation of costs by one agency and determination of "need" by another agency may result in an uncoordinated process incapable of achieving desirable results even for natural monopolies. See H. Cohen and C. Schramm, op. cit. -119~ In markets where capital investment regulation is designed to constrain investment because market signals have been blunted by reimbursement schemes, restricting investment through regulation is clearly a less desirable solution than reforming the reimburse- ment system. One characteristic of using regulation of investment rather than reform of reimbursement schemes is that it places incumbent firms at an advantage over potential entrants. Despite efforts to allocate investment in the best possible manner, planners typically only review proposals for additions or changes and cannot reallocate equipment already purchased or services already established. This means that the ability of capital investment regulations to generate the best allocation of resources is extremely limited.l62 The empirical evidence raises doubts as to whether these programs even effectively lower costs. The empirical evidence indicates that CON and Section 1122 review programs may skew investment into areas less closely reviewed by planners, while the overall level of investment stays approxi- mately the same.163 The resulting allocation of investment may 162 por instance, planners often face the dilemma of whether CT scanners owned by private physicians who deny access to others should count against the allotment designated for their areas. While they may sometimes be able to negotiate open-access policies for new acquisitions, reallocations and imposition of such conditions on those already possessing the equipment before the adoption of the regulation may be impossible. Given these difficulties, it is unlikely that regulation of capital invest- ment can achieve its goal of efficient resource allocation in markets where reimbursement schemes blunt market signals. 163 sloan and Steinwald, "Effects of Regulation on Hospital Costs and Input Use," op. cit. =-120- constitute a less efficient allocation of resources than would occur in the absence of regulation. In addition to their apparent inefficacy in controlling costs, regulations aimed at capital investments also may thwart the competitive process. By definition, CON and Section 1122 review programs constitute barriers to entry, and thus they limit the ability of competition to respond to market forces. In some States, where CON is needed to close facilities, these regulations also create barriers to exit.l64 Health planners and State regulators often see the unused capacity of a facility driven out of the market by the competitive process as a waste of resources. However, the ability to enter and exit an industry is essential to competition. The process of competition that leads some firms to fail and creates some unused capacity may provide benefits that far exceed the costs associated with investment unneeded according to demographic and epidemiologic criteria.l65 Even where the regulations have no perverse effects on the allocation of investment, the CON requirement itself raises the cost of investment. The American Hospital Association survey documented that health-care facilities are concerned about the costs of the administrative process.l66 In addition to the 164 gee section III.B.2. for discussion of exit barriers. 165 gee sections II.B.l1. and II.B.2. and II.C. for more on this point. 166 American Hospital Association, op. cit. -]2y- monetary costs, the regulations add uncertainty to investment decisions. If, as is likely, such costs are felt more severely by new entrants than by established providers, entry by outsiders may be further deterred. The strict application of CON and Section 1122 review is only justified in service areas where market forces are inoperable. Some States, such as Colorado, are experimenting with applying such regulations less strictly to markets where consumers have the ability and some incentive to shop or where new entrants who might be excluded by these regulations may potentially offer innovative forms of services. Long-term care, kidney dialysis, ambulatory surgery, and ordinary obstetrical care are examples of health services that are normally provided in institutions and that have the potential for becoming more competitive.l167 The possibility of relaxing the structures on investment for these and other services should be examined in each geographic area. 167 gee section IV for a sample checklist of criteria for identifying markets where competition may work. -}22~ C. Regulation of Institutional Quality Institutional licensure, certification, and accreditation are regulatory mechanisms designed to ‘assure minimum levels of quality. The arguments for regulating quality in health-care institutions include consumers' difficulties in assessing quality, the importance of competently performed services, and the possibility of irreversible harm resulting from low quality care. To the extent that these regulations reduce problems in the market that are associated with the absence of costless and freely accessible information, they may allow the adoption of a competi- tive approach without fear that unscrupulous providers can take advantage of consumers. All of these regulatory programs tend to measure quality by using input standards rather than actually measuring the quality of services provided. This approach rests on the assumption that higher inputs lead to better quality of care. While some rela- tionship between inputs and quality of care may exist, recent empirical studies discussed below show that some regulatory programs raise the cost of providing health services without significantly increasing the quality of care patients actually receive. Because "quality" regulations usually dictate the inputs providers must use to perform services, these regulations also may restrict the variety of services offered and create barriers to innovation in the provision of health services.l168 Therefore, 168 gee section II.E.- for the distinction between quality and variety. -123- 374-115 0 - 82 - 9 : QL 3 while some form of quality regulation may be appropriate, it is also important that these regulations not be unnecessarily strict. 1. The purpose of institutional quality regulations. Licensure, certification, and accreditation all address problems associated with low quality in markets where consumers sometimes have difficulty assessing the quality of the providers from whom they receive care. The goal of these regulations can be separated conceptually into two levels: first, protecting consumers from clear risks to health and safety; and second, differentiating acceptable, but low-quality, providers from high- quality providers. The impetus for regulation at the first level is that very low quality providers can potentially inflict great harm on consumers. 169 Assuming that a minimal level of care can be identified, strong arguments can be made for banning providers that do not meet that standard. Although a facility not meeting the standard may quickly earn a reputation for harm which drives it out of business, the harm that it can cause in the interim may be so great as to justify barring it from the market altogether. 169 1n the not-too-distant past, unsanitary conditions caused many secondary illnesses in patients. For instance, "childbed fever"--a killer of many 19th-century women--was caused primarily by unsanitary medical practices. -124- Above this minimal level of care, some consumers may be willing to trade off quality against cost. Before receiving the service, however, consumers may not be able to distinguish high- quality providers from those barely meeting levels of accept- ability. Even after the service is provided, consumers may be unable to determine whether a speedy recovery resulted from high- quality care or from the natural disease process in their parti- cular case. Situations of this type provide the impetus for information schemes. In many cases, centrally gathered informa- tion about institutional quality may help consumers make rational decisions. Accreditation is one mechanism for providing such information. In general, given the costs of measurement, the more finely graded the measurement system, the better such information helps the market function.l70 The underlying assumption of all regulations addressing institutional quality is that consumers, or their physicians, lack information needed to choose among facilities and that the regula- tory agency can improve the allocation of resources by providing this information or by barring low-quality providers.l71 A related assumption is that reputation alone does not operate to provide consumers with sufficient information to choose among providers. 170 gee Beales, Craswell, and Salop, op. cit., for a full dis- cussion. 171 gee section II.D. for a full discussion of information issues =125~ 2. Schemes regulating institutional quality. a) Licensure.l72 Licensure refers to State laws that require facilities to meet minimum standards in order to operate.l73 Most States license hospitals and nursing homes. The application of these regulations to clinics and other free-standing centers differs from State to State. Review of licensure programs shows that regulations primarily address easily specified standards pertaining to inputs used in construction, maintenance, and sanitation of the physical plant. Standards relating to patient care itself are often vague. b) Certification. Certification programs are voluntary Federal and State programs under which hospitals and nursing homes must meet certain standards in order to get a certificate. Such certificates are not a legal prerequisite for operation, but they are quite important, because certification is necessary to qualify for reimbursement under the Medicare, Medicaid, and other Federal programs. These programs are designed to raise the level of care received by beneficiaries of the reimbursement programs. Standards set for attainment of certification exceed licens- ing requirements in some States. In recent years, however, 172 por a thorough, if somewhat dated, discussion, see H. G. Fry, "The Operation of State Planning and Licensing Programs, " Hospital Monograph Series 15 (Chicago: American Hospital Association, 1971). 173 Although the Federal Government never enacted a formal licensing program, during the heyday of Hill-Burton the almost universal participation of facilities in this program for con- struction, expansion, and renovation created a de facto licensing program. -126- several States have modified their licensing requirements to incorporate Medicaid certification standards. Medicaid is a dominant form of payment for nursing-home patients. Therefore, even in States where licensure requirements are lower than certification requirements, the nursing homes not meeting the Medicaid standards typically are eligible to serve only a small proportion of the patient population.l74 Despite the presence of the Medicaid certification programs, the level of quality of care in nursing homes has repeatedly raised concern.175 Limited enforcement funds and standards measuring minimum input levels rather than care levels may contribute to these problems. c) Accreditation.l7® Accreditation programs are voluntary quality-measurement programs sponsored by private organizations 174 por evidence from one State, see Leon S. Pocinki, A. Doerman, and P. Schroedel, "An Evaluation of the Operating Costs and the Cost Impacts of Regulations on Nursing Homes in the State of Ohio," GEOMET HF-72 (Cleveland: Federation for Community Planning, September 1972). 175 gee GAO, Many Medicare and Medicaid Homes Do Not Meet Federal Fire safety Requirements, DHEW, Report to Congress, March 18, 1975; and GAO, Analysis of Proposed New Standards for Nursing Homes Participating in Medicare and Medicaid, Report to the Chairman of the Select Committee on Aging, U.S. House of Representatives, February 20, 1981. 176 while not a regulatory program, accreditation is discussed here because it is sometimes studied empirically along with institutional licensure and certification. For a thorough discussion of the history and recent changes in accreditation, see John E. Affeldt, "Voluntary Accreditation," in Regulating Health Care: The Struggle for Control, ed. Arthur Levin, op. cit., pp. 182-91. -127- that set the standards and verify compliance by applicant facili- ties. The program with the largest impact on health-care institu- tions is administered by the Joint Commission on Accreditation of Hospitals (JCAH). Accreditation provides consumers with the information that the facility meets certain standards. These standards concentrate primarily on inputs, on the presumption that inputs and quality of care are correlated. Accreditation serves as a simplified index of the quality of care that patients can expect to receive in a facility. Because the JCAH standards have been adopted by the Federal Government as the certification standards hospitals must meet to qualify for Medicare reimbursement, this private accreditation program is now indistinguishable from a certification program. Acquiring accreditation is not useful merely for communicating to consumers that the facility attains certain standards; it is necessary for receiving reimbursement for a large number of patients. 3. Empirical evidence on regulation of institutional quality. Studies of institutional licensure, certification, and accreditation tend to call into question the ability of these programs to achieve their goals. Partly as a result of Medicaid- related scandals in the early 1970's, many empirical studies have focused on the efficacy of nursing-home standards. -128- Some of the early inquiries are little more than "asposbe of the failure of regulation to assure patients’ safety.177 However, statistical studies, as well as anecdotal evidence, suggested that problems in nursing-home quality were serious and widespread. Braverman, 178 for example, reported lax enforcement of both State licensure laws and Federal certification standards. He also documented the wide variability among State laws regulating nursing-home quality. Partly in reaction to these studies, several States signifi- cantly upgraded their enforcement of nursing-home standards. The Federal Government upgraded its standards for nursing homes par- ticipating in Medicaid and Medicare in 1974 and revised these standards again in July 1980. However, a recent Government Accounting Office report remained skeptical about the efficacy of the new certification requirements.179 Some of the problems noted in the GAO study, such as failure to meet fire-safety standards and low nurse-to-patient ratios, could be addressed by enforcing existing input standards. Others, such as physician inattention to institutionalized patients and high levels of untreated adverse 177 por an example of a 1970's expose of the” nursing-home industry, see A. Mendelson, Tender Loving Greed (New York: Alfred A. Knopf, 1974). 178 3. Braverman, Nursing Home Standards (Washington, D.C.: American Pharmaceutical Association, 1970). 179 gee GAO (1981), op. cit. -129- Lo drug reactions, require an approach addressing the actual provision of care. These standards are once again under review. Like the GAO study, this HHS review is expected to call into question the work- ability of the current schemes regulating quality of care in nursing homes. Although annual or biennial review of patient records as well as inspections of facilities may be the only way to assure compliance with the standards, these techniques are costly.180 Frequent recertification or reaccreditation by the Federal Government, or JCAH, is also costly and burdensome. In addition, the incentives for compliance are weak because the Federal Government typically denies reimbursement only in extreme circumstances. Some critics suggest that the quality problem is best addressed by raising reimbursement rates and facilitating the transfer of patients when they or their families are dissatisfied with the qualify of care provided. Others suggest that Medicaid reimburse higher quality homes at a higher rate to give facilities a positive incentive to improve quantity without relying on patient movement.l181 Although the issue of quality of care in hospitals has not been as controversial as that in nursing homes, several studies suggest that hospital certification and accreditation standards 180 gee IV.E. for evidence on this point. 181 y. pollak, "Long Term Care Facility Reimbursement," in Altering Medicaid Provider Reimbursement Methods, J. Holahan et al. (Washington, D.C.: The Urban Institute, June 1977). =130~- bear little relationship to the care patients actually receive. One study found that the JCAH standards have low correlations with other measures of hospital efficiency or quality of care.182 Another study showed that JCAH-accredited hospitals vary widely in their quality of care, as measured by various other indices.183 Finally, the level of quality as measured by a case-severity- adjusted index of death rates appears to differ only slightly between accredited and nonaccredited hospitals.l184 One reason for this may be the lax enforcement of standards, found in another study of hospital standards.185 Although these studies of hospital certification and accreditation programs are slightly dated, they suggest that input-based standards may do little to raise the quality of care patients receive. Rather, their principal effect may be to impose compliance costs on facilities wishing to qualify for State licensure and Federal reimbursement programs. Since the early 182 pg. Longest, "Administrative Coordination in General Hospitals," Georgia State University School of Business Administration, cited in O'Donoghue, Evidence About the Effects of Health Care Regulations, op. cit. 183 J. Neuhauser, The Relationship Between Administrative Activities and Hospital Performance (Chicago: University of Chicago Press, 1971). 184 uM, I. Roemer, T. Mostafa, and C. Hopkins, "A Proposed Quality Index: Hospital Death Rate Adjusted for Case Severity," Health Services Research 3 (Summer 1968) :96-118. 185 y. Worthington and L. H. Silver, "Regulation of Quality of Care in Hospitals? The Need for Change," Law and Contemporary Problems 35 (Spring 1970) :305-33. -131~- 1970's, the major focus of Federal efforts in regulating the quality of hospital care has been on programs reviewing the provision of care through medical audits. Although these programs (discussed later in the paper) are primarily aimed at reducing costs, they also incorporate procedures that could be used to regulate quality of care. 4. Alternatives to the current system. Currently, there is no clear delineation among the uses of licensure, certification, and accreditation in assuring institutional quality. In some States, licensure standards are lower than certification or accreditation standards, are loosely enforced, and may do little to assure that facilities provide the minimum acceptable level of care. In other States, the minimum standards for licensing some types of facilities exceed the levels required for certification under Federal reimbursement programs and thus may bar providers offering a low, but acceptable, quality from the market .186 One reason the delineations among licensure, certification, and accreditation tends to be unclear is that some States and the Federal Government have adopted the standards of accrediting bodies for purposes different from those for which they were originally designed. This reflects both the difficulty State and Federal regulators have in defining their own standards and the confusion as to the purpose of particular standards. 186 5. Affeldt, op. cit. -132- While the JCAH attempts to define standards of optimal prac- tice on the basis of medical and economic criteria, the stated purpose of licensure is usually to define the minimally acceptable quality level for facilities operating in the State. By adopting the JCAH standards as licensing requirements, States legally define variations from optimal practice as unacceptable. However, a level of operation that is minimally acceptable from a health and safety perspective may differ from the optimal level of opera- tions. The failure to distinguish between these two levels may increase costs and limit consumer choice by imposing costs on all facilities that some would not voluntarily choose to spend. Of course, the value of this choice depends on whether consumers or their physicians have access to information about quality differ- ences and whether providers have the incentive to pass the cost savings on to consumers or third-party payers. One alternative to the current system is to identify minimally acceptable levels of operation and require that only those levels be met for licensure. Another problem pointed out in the discussion of empirical work is that the standards may bear little relationship to the level of care patients actually receive. Other indices, such as case-severity-adjusted death rates, may provide better information about the care patients can expect. Measuring care according to such indices and publishing the results on a facility-by-facility and procedure-by-procedure basis is an alternative to the current -133~ system of regulating institutional quality. A major advantage of this approach is that publishing indices could help focus hospital competition on the actual quality of care delivered. The major drawback to this approach is that the preparation of such indices may be costly and technically complicated. Fre- quent updating is also needed to give facilities the incentive to improve quality. Although studies on the use of these indices are being done on a pilot basis in California and at Johns Hopkins University Medical School for the State of Maryland, no State has yet adopted an information program based on quality of care patients receive.187 If use of these indices proves to be cost effective, it may represent a significant alternative to the current system of regulation. 5. Interaction of competition with institutional-quality regulations. The programs regulating institutional quality may aid or may thwart competition. When the regulations effectively bar institutions whose quality is so low as to be injurious to consumers, the regulations improve the market's ability to function because consumers are assured that they do not face an unreasonably high risk of harm by entering these facilities. Programs such as accreditation, which provide consumers with information about the quality of care patients receive in particular facilities, also may improve the functioning of the 187 Results of these pilot programs have not yet been published. The information contained here comes from private conversations with health planners and project participants. -134- marketplace. Such programs allow consumers to differentiate among facilities without barring competitors from the market and encourage facilities to compete by raising their quality. Programs regulating institutional quality can thwart competi- tion by setting standards that exceed the level of quality for which fully informed consumers would choose to pay. Excessively high standards raise costs and may exclude some potential entrants. By narrowly prescribing the ways in which the standards may be met, these programs may also make it difficult for facilities to use new construction or maintenance technology. In addition, the lag in revising standards may make it difficult to substitute allied health personnel for physicians or nursing staff, even where a new technology eliminates the need for as many highly trained personnel. D. Regulation of Personnel Regulations affecting health personnel are usually taken as givens by health planners and corporate decisionmakers. Neverthe- less, because these regulation profoundly affect the structure and efficiency of health-care markets, they may be important in analyzing the potential for competition in many health services. For instance, State law may bar potential providers such as nurse- midwives, pharmacy clerks, or physicians' assistants from perform- ing certain procedures. In States with restrictive laws, entry by alternative providers cannot occur, despite the adoption of an otherwise competitive approach to the market. Such laws prevent -135- providers from offering services that they are otherwise qualified to perform and thus thwart the ability of market forces to improve the allocation of health resources. In some States, reform of laws governing professional practice may be important in infusing competition and improving the allocation of resources into parti- cular health-services markets. Likewise, State laws governing licensure of professionals may greatly affect the ability of third-party payers to reduce health costs through the adoption of incentive schemes designed to promote efficient use of health resources. State laws may also affect the ability of health-care facilities to use physical and human resources as efficiently as possible. A vast literature exists on regulation of the professions, especially those involved in the delivery of health care. Recently, the literature on licensure was reviewed and discussed extensively in an American Enterprise Institute Conference volume, 188 and the literature on corresponding private regulatory schemes has also been reviewed and discussed recently.189 Because these timely pieces are available, discussion of these issues here will be brief. 188 Occupational Licensure and Regulation, ed. Simon Rottenberg, (Washington, D.C.: American Enterprise Institute, 1980). 189 M. pollard and R. Leibenluft, "Antitrust and the Health Professions," Office of Policy Planning Issues Paper, Federal Trade Commission, 1981. -136- 1. The purpose of regulating health-care personnel. The stated objective of most regulation of health-care personnel is to assure that those providing services meet minimum levels of competence. The rationale for governmental intervention is similar to the rationale behind the regulation of institutional quality: consumers lack adequate information to assess the competence of health-care personnel and therefore cannot protect themselves against providers of dangerously low quality. 2. Programs addressing the qualifications of health- care personnel. The major programs for regulating the qualifications of health-care personnel are licensure and certification schemes.l190 a) Licensure. Licensure programs are administered by the States. Generally, they require that a person meet minimum educa- tional or competency standards in order to practice a particular profession. The performance of certain procedures and the prescription of certain treatments are often limited to the licensed members of particular professions and, in some cases, to those under their supervision. In recent years, national licensure testing programs, such as those administered by the National Board of Medical Examiners, have been developed. States may participate in these national programs, but they of course retain the right to establish their 190 accreditation is excluded from this discussion because technically it refers only to programs for verifying institutional quality. -137- own standards and procedures. States also differ in their policies toward accepting licenses granted by other States and in their policies regarding relicensure and continuing education. b) Certification. Certification is a voluntary credential- ing program operated by private organizations or by Government agencies. It verifies that a person meets the standards of the certifying board. The standards may involve competency require- ments, or educational requirements alone. Certification often involves the designation of a title, such as "board-certified" surgeon, for individuals meeting the standards. Uncertified personnel are prohibited from using the title but may offer the same services to the public.l191 3. Evidence on the effects of regulation of personnel. Proponents of regulation of personnel hold that these programs are necessary for the protection of the public. Critics of the regulations charge that the system creates unnecessary barriers to entry and to geographic mobility, and therefore results in higher incomes through higher prices. 192 Most experts in the field agree that the truth lies somewhere between the two positions, with the 191 por institutions, "accreditation" is analogous to "certifica- tion" of professionals. In the institutional context, "certifica- tion" is a term reserved for programs under which facilities verify that they meet the standards required by various reimbursement schemes. (See section III.C.2.). 192 por a statement of the theory and an example of the empirical work by the best-known critic of regulation of personnel, see G. Stigler, "The Theory of Economic Regulation," Bell J. Econ. 2 {1971)33=21. -138~ regulations both assuring some degree of quality and raising the income of practitioners.193 Most of the empirical work in the area has focused on the effects of interstate mobility and reciprocal licensing on income and prices. Empirical studies of personnel with different train- ing providing equivalent services are more difficult because the comparability of quality is nearly always an issue. A recent study by Shepard found that State dental laws with more restrictive reciprocity policies were associated with less interstate movement of dentists.l94 In addition, in States with more restrictive barriers to entry, dentists charged higher prices and earned larger incomes. For example, in 1976, the price of dental services was on average 12 percent higher and income was 15 percent higher in jurisdictions without reciprocity. In 1976, the overall annual cost of these restrictions was estimated to be $700 million. A similar study confirmed the same effect among lawyers but noted that restrictions on mobility only add marginally to the major barrier--licensure itself.195 while the results of this second study cannot be applied directly to health-care personnel, 193 For a summary, see Thomas Moore, "The Purpose of Licensing," J. Law & Econ. 4 (1961):93-177. 194 1, ghepard, "Licensing Restrictions and the Cost of Dental Care," J. Law & Econ. 21 (1978):187-201. 195 pg, p. Pashigian, "Occupational Licensing and the Interstate Mobility of Professionals," J. Law & Econ. 22 (April 1979):1-25. =-139- the results are suggestive of effects that might be found in the health-care context. In contrast, a historical study by William White of licensure of nurses showed an initially positive effect of licensure on wages in the 1950's, but no effect in the 1960's and 1970's.196 The author postulates that over time, enough people were attracted to the profession to eliminate the high wages. The author predicts that in times of shortage, nurses in States with stricter licensing laws will receive relatively higher wages and that the duration of shortages in those States will be lengthened by restrictive licensure laws. In his study of physicians, Leffler similarly found that even without reciprocity, licensure only results in higher incomes and prices during periods of shortage and that as the supply of physicians catches up with the demand, this effect disappears.197 However, the periods of shortage appear to have lasted for decades at a time in some States and cannot be discounted as having minimal economic impact. As Boulier points out, lowering barriers to mobility not only results in lower prices and incomes in those jurisdictions where there would otherwise be a shortage; it may also raise incomes and prices in jurisdictions from which 196 wy. white, "Mandatory Licensure of Registered Nurses: Introduction and Impact," in Occupational Licensure and Regulation, ed. S. Rottenberg (Washington, D.C.: American Enterprise Institute, 1980). 197 Kk. B. Leffler, "Physician Licensure: Competition and Monopoly in American Medicine," J. Law & Econ. 21 (1978):165-86. =-140~- licensed personnel would have difficulty moving, had reciprocity not been granted.l198 In addition to empirical work on the effect of licensure on prices and income, several studies document the history of licensure within particular professions. For example, Paterson relates the reactions of members of other licensed occupations-- dentists and hygienists--to the development of a new occupation-- denturism--and the motivation behind the initial push for licensure.l9? This study and other histories of licensure support the theory that licensure typically results both from concerns about quality and from the desire on the part of members of the profession to restrict entry. Finally, some studies attempt to assess the quality of care provided by different professional groups. In a recent study, Appelbaum and Scheffman empirically tested the performance of Canadian dentists, dental nurses, and hygienists in well-defined tasks.200 The results indicate that on average, dental auxiliaries performed as well as, or in some cases better than, dentists at tasks they are legally prohibited from performing 198 pg, Boulier, "An Empirical Examination of the Influence of Licensure and Licensure Reform on Geographic Distribution of Dentists," in Rottenberg, op. cit. 199 gee, for example, S. Peterson, "Legalized Denturism: Consumer Demand, Public Health, and Legislative Response," Northwestern U. L. Rev. 74 (1979):97-121. 200 Eg. Appelbaum and D. Scheffman, "Occupational Licensing and Quality," University of Western Ontario Working Paper, 1980. -141- under Canadian licensure laws. Two drawbacks of the study are that the data from an experimental setting may not be representa- tive of actual practice and that the results of the study cannot be applied outside a small number of routine dental health pro- cedures. Nonetheless, the study calls into question the assump- tion that performance of services involving skill and discretion should necessarily be restricted by licensure to professionals such as physicians or dentists, who receive a high level of train- ing in many areas. It also raises the possibility that care of equivalent quality could be performed at lower cost, were less highly trained personnel allowed to provide certain services. 4. Alternatives to regulating personnel. The major alterna- tive to licensure of personnel is certification. Certification provides information about the provider's level of competency or education but does not directly restrict consumer choice. How- ever, when personnel must have access to health-care facilities in order to practice their trade and certification is used as a general requirement for obtaining access privileges in health facilities, provision of services by uncertified personnel may be restricted. In cases where consumers rely on facilities to select personnel for them, and where fully informed consumers would choose to have certified personnel perform the service, the ability of facilities to limit access may not have any negative effects. However, in some cases, lack of privileges in existing facilities, combined with regulatory controls on the entry of -142- additional institutional providers, may effectively bar uncertified personnel from providing a service fully informed consumers would desire. Requirements by third-party payers that treatment be provided by certified personnel in order for the consumer to be reimbursed can also affect the competitive position of uncertified personnel. Even though treatment by certified personnel may be more expen- sive, insured consumers have strong incentives to seek care from personnel recognized as legitimate by their insurer. When third- party payers require certification for reimbursement, it can effectively bar uncertified personnel from the market. Other modifications of licensure are programs requiring relicensure and continuing education. These modifications address the criticism that licensure only assures a certain level of quality among those entering the profession. The effectiveness of these programs has not yet been determined. Reciprocity among States allows greater mobility and prevents shortages from developing in individual jurisdictions. While not an alternative to licensure, it does reduce some of the barriers to entry and exit among licensed personnel without addressing the relative qualifications of personnel training for different professions. Restrictions placed on personnel with different types of licenses should be loosened in areas where studies show that services can be competently performed by professionals with train- ing on only a narrow range of procedures. As the Appelbaum and -143- Scheffman study points out, some restrictions on allied health personnel may not be justified if the performance of personnel holding different types of licenses is carefully examined. 5. The interaction of regulation of personnel with competi- tion. As with programs regulating institutional quality, programs regulating personnel may either aid or thwart competition, and for much the same reasons. When the regulation effectively bars providers totally unqualified to provide services to consumers, the regulations improve the ability of the market to function because the consumers' risk of serious harm is reduced. Like hospital accreditation programs, personnel certification programs can be used to provide consumers with information about the special training and therefore (presumably) about the competence of an individual practitioner. Certification programs make it easier for individual practitioners to communicate information about their training to the market and thus may make it easier for consumers to select providers with skills most suited to their needs. As discussed above, licensure programs thwart competition when they create entry barriers unjustified by quality considera- tions. When such programs prevent otherwise qualified personnel from providing a service, they typically generate the high prices -144- and incomes associated with an artificially induced scarcity of providers.201 Certification is generally less restrictive than licensing, but to the extent that it is adopted as a criterion for access to facilities or for reimbursement, it too can thwart competition and raise costs. Uncertified personnel may face the choice of not providing a particular service, creating their own institution from which to provide the service, or incurring the additional training costs needed to gain certification. To the extent that informed consumers would choose to use personnel who had not attained the certification standards, the restriction of their practice may unnecessarily raise health-care costs. E. Regulation of the Provision of Health care202 Regulations governing the provision of health care include medical audits, utilization review, and claims review. These regulations have two--sometimes conflicting--purposes. They are intended to decrease medical-care costs by reducing unnecessary or inappropriate utilization of facilities and to raise the quality of care by reviewing the care patients actually receive. Planners 201 For an alternative view--that physicians with more patients charge lower prices--see DHEW, The Target Income Hypothesis and Related Issues in Health Manpower Policy, op. cit. 202 The programs discussed in this section have been studied annually by the Health Care Financing Administration (HCFA). For further information, see the last publicly available HCFA report, Professional Standards Review Organization 1978 Evaluation, U.S. Department of Health, Education and Welfare, HCFA-03000, January 1979 (publicly released in January 1980). -145- have no direct role in designing or administering these regula- tions. However, the goals of reducing costs and increasing quality are shared by regulatory schemes with which planners are involved. To the extent that regulations governing the provision of care successfully achieve these goals, planners could take a more competitive approach to the market by, for example, inter- preting CON requirements less strictly. Unlike regulation of price and cost, institutional quality, capital investment, and personnel qualifications, regulation of the provision of care focuses on procedures, treatments, and lengths of stay recommended by individual physicians for indivi- dual patients. Working at this microcosmic level, these schemes attempt to assure that patients receive appropriate treatments and that facilities are appropriately utilized. At least in theory, regulations administered at the level of the individual physician's interactions with patients can better allow exceptions fitting particular medical circumstances. In practice, these schemes appear to be quite costly to administer, and it is not clear that they represent a cost-effective method for reducing costs or enhancing quality. 1. Types of regulations governing provision of care. a) Medical audit. Medical audits are federally mandated review programs administered by Professional Standards Review Organizations (PSRO's), in accordance with Public Law 92-603.203 203 42 y.s.c., § 1320c-4. -146- PSRO's are nonprofit corporations, typically linked to local medical societies. PSRO's establish criteria for appropriate care. These criteria are used in reviewing the care patients actually receive. During a medical audit, hospital records are reviewed to determine whether individual physicians' practices deviate from standard practice, controlling for case severity and other factors. Local PSRO's conduct medical audits. While these local groups have the authority to adopt their own criteria, many use model criteria developed by specialty boards such as the American College of Surgeons. An assumption underlying the PSRO medical audit program is that reviewing the actual provision of care will assure that patients receive optimal levels of care, and the result will be reduced health-care costs.204 This expectation--that medical audits can not only improve quality of care but also will reduce costs--is based on the assumption that deviations from optimal levels of care result in overutilization of facilities. However, if bringing current practice patterns into line with optimal prac- tice requires more tests and increased utilization of facilities, overall health-care costs could rise rather than decline as a result of medical audits. In fact, the ability of the PSRO medical audit program to achieve its goals has been questioned by 204 yeFa, op. cit., "Background: PSRO Program Mandate," p. 1. -147- 374-115 0 - 82 - 10 : QL 3 many experts. As discussed later in this section, the costs of the program appear to exceed the savings it generates. b) Utilization review. Utilization review is a mandatory self-administered program for hospitals receiving Medicare and Medicaid reimbursement.205 Under utilization review, patient case histories are reviewed to determine the necessity of admission, length of stay, and services received. As part of the program, the conditions of patients hospitalized for extended periods must be reviewed periodically to establish the necessity of continued acute care. The goal of this review process is to reduce over- utilization by assuring that patients are transferred out of acute-care facilities as soon as it is feasible. In some States, the federally mandated and hospital-administered utilization review programs are supplemented by State-administered programs. Some Blue Cross plans also conduct utilization reviews of member hospitals. c) Claims review.206 Blue Cross and several other private health-insurance companies review reimbursement claims for appropriateness of treatment. Reimbursement for tests not deemed necessary for assessment of the stated medical condition can be 205 ytilization review is required for Medicare participation under the original Medicare law (P. C. 89-97), 42 U.S.C. 1395(k), and for Medicaid under the 1967 Social Security Amendments (P. L. 90-248), 42 U.S.C. 1396(a). 206 claims review is, of course, not a regulation but a policy firms adopt to keep costs down. It is discussed here because empirical studies of utilization review and medical audit programs sometimes consider claims review as well. -148- denied under these programs. These programs not only reduce the expenditures by the third-party payer; they also increase the consumer's incentive to limit care received to medically necessary treatment by making the consumer liable for payment of denied claims. Claims review has been widely applied in the dental area, with some dental insurance plans requiring both pretreatment screening and retrospective review. The extent and success of claims review in dentistry has been attributed to the fact that symptoms and origins of disease are more easily determined and treatment is more systematically related to clinical conditions in dentistry than in medicine.207 2. The effect of regulations governing provision of care. Because both assuring appropriate quality of care and reducing costs through elimination of unnecessary utilization of facilities are goals of the programs regulating provision of care, a full examination of the effectiveness of these programs would entail both studying the effect of compliance on quality and analyzing the relationship between these programs and health-care expendi- tures. While the medical audit and utilization review procedures were being developed, several studies of the correlation between 207 For a fuller discussion of regulation of provision of care in dentistry, see Warren Greenberg, "Provision-influenced Insurance Plans and Their Impact on Competition: Lessons from Dentistry," AEI Conference on Health Care, September 25-26, 1980. For a review of the literature and extensive references, see I. L. Praiss, K. A. Tannenbaum, and C. A. Gelder-Kojan, "Quality of Dental Care--The Role of Third-Party Payers: A Literature Review," Medical Care Review, December 1978, pp. 1211-33. -149- compliance with the standards and other indices measuring quality of care were conducted. The studies showed that medical audits correlated well with some indices ‘of quality and poorly with others.208 However, little effort has been made to study syste- matically whether medical audits generally raise the quality of care patients receive. Another unanswered question is whether improvements in quality are sufficient to justify the costs of the program. Considerably more effort has been put into studying the effect of these programs on utilization rates. An early study of a precursor to medical audits in the State of New Jersey showed that such programs initially affected patterns of practice and utilization rates, but by the end of the first year, the effects had largely disappeared.209 A more recent study of similar programs by Sloan and Steinwald leads them to conclude that the programs that predated the adoption of federally mandated utilization review had little if any effect on hospital costs. 210 Sloan and Steinwald also studied the State-administered Medicaid utilization review programs and several similar programs conducted by private insurers. They found that evidence on the 208 gee Patrick O'Donoghue, Evidence About the Effects of Health Care Regulation, op. cit., ch. 5. 209 p. Rr. Bailey and D. C. Riedel, "Recertification and Length of Stay," Blue Cross Reports 6 (1968):1-10. 210 gloan and Steinwald, Insurance, Regulation, and Hospital Costs (Lexington, Mass.: D.C. Heath, 1980). -150- effect of these programs on costs was very sensitive to minor changes in the technical specifications of statistical tests.211 The authors therefore concluded that they could not put any degree of confidence in their results. Utilization review programs conducted under the auspices of PSRO's have been studied extensively.212 In the recent American Hospital Association survey, 28 percent of hospitals located with- in jurisdictions with active PSRO's reported disallowances of Medicaid, Medicare, or Blue Cross claims, due to utilization review. In contrast, only 21 percent of hospitals in areas with- out active PSRO's reported disallowances.213 while this indicates a somewhat higher number of disallowances in areas with active PSRO's, the savings from terminations and disallowances may not be large enough to outweigh the costs of administering the program. The best-known studies of the impact of PSRO's have been conducted by the Health Care Financing Administration as part of its annual evaluation process.214 These studies show. considerable variation among States in the impact of utilization reviews. Nationally, PSRO's are reported to have achieved insignificant 211 gloan and Steinwald, "Effects of Regulation on Hospital Costs and Input Use," op. cit. 212 gee, for example, HCFA, op. cit. 213 American Hospital Association, op. cit. 214 gee HCFA, op. cit. -151- reductions in length of stay. There is some evidence that utili- zation review has had a significant impact for certain diagnoses and procedures, but even in these cases it is not clear that the program generates savings which exceeded than its costs. Studies conducted by the Congressional Budget Office and the General Accounting Office concluded that many PSRO programs are not cost effective.215 3. Alternatives to these regulations. A major feature of the PSRO program is the development of detailed treatment criteria. PSRO criteria and standards for care appear to be more successfully implemented for some diagnoses and procedures than for others. One alternative to the current across-the-board approach is to concentrate PSRO activities on those conditions and and procedures where ease in pretreatment verification of need and standardization of treatment would allow them to be applied most easily. A second alternative is to integrate the criteria into reeducation and recertification programs and to disentangle the adoption of such standards from processing reimbursement claims. The desirability of using reimbursement as a lever for enforcing the adoption of these standards is particularly unclear because the criteria are designed to define "optimal" practice, with 215 congressional Budget Office, "The Impact of PSROs on Health Care Costs" (Washington, D.C.: Government Printing Office, January 1981); and General Accounting Office, "HHS Should Improve Monitoring of PSROs," Document HRD-81-26 (Washington, D.C.: Government Printing Office, 1981). -152- respect both to medical criteria and economic considerations. This suggests that the criteria should be continually revised as the relative resource costs of alternative techniques change and that the enforcement of economically obsolete standards results in patients receiving economically inefficient (although not medically harmful) treatment. Incorporating the standards into a voluntary and (presumably) more flexible system may therefore be beneficial. A major issue is whether the federally mandated program should be continued at all in light of its apparent high cost and minimal effect. This issue is currently under consideration at the Department of Health and Human Services, where proposals for the elimination of the federally mandated program are actively being reviewed. Even if these programs are eliminated, however, it is likely that many States will retain their own utilization review programs. 4. Interaction of provision-of-care regulations and competi- tion. To the extent that regulation of the provision of care assures patients that providers offer "safe" and appropriate services, the resulting reduction in unnecessary medical risk and overutilization can improve the performance of the health sector. However, the overall effect of these regulations on the market depends on the costs associated with these improvements in quality and efficiency. These costs include both administrative expenses and any anticompetitive effects of the programs in possibly -153~- prohibiting procedures that fully informed consumers would choose to have performed. The desirability of setting detailed design standards for the treatment of medical conditions depends on the degree to which undisputably superior procedures can be identified. This superiority depends not only on the probability of a successful medical outcome but also on the relative costs of different procedures and on the variability in preferences for different forms of treatment among informed patients. Regulation of the provision of care probably does not reduce the suitability of using a competitive approach to a particular health-care market and might even increase it were these programs successful. Private claims review programs appear to be marginally more success at controlling costs. To the extent that these programs help third- party payers communicate incentives for cost savings and quality enhancement, these private programs may increase the appropriate- ness of a competitive approach in some markets. -154- IV. QUESTIONS ABOUT THE MARKET: A SUMMARY A. Introduction The purpose of this summary is to help planners and others concerned with the performance of the health sector assess the potential for taking a competitive approach to particular health- service markets. This is done through a series of questions designed for use in analyzing a market. As this paper has discussed, analyzing health-care markets is complicated because certain market characteristics and regulatory structures may cause reactions different from those expected in most markets. To help the reader understand the extent to which these market characteristics may thwart the ability of competition to achieve a desirable resource allocation, this paper examined the distor- tions associated by natural monopolies, exit and entry barriers, costly or inaccessible information, the availability of several varieties or quality levels of a service, and third-party-payment schemes. It also examined five classes of regulations, their efficacy and their interaction with the competitive process. While planners and corporate decisionmakers typically have little influence over these regulations, understanding their effects is important. Despite the presence of these market characteristics and regulatory structures, there are two reasons why taking a competi- tive approach may improve the performance of many health-care markets. First, minor deviations from the characteristics of the -155- competitive norm typically result in correspondingly minor distor- tions in the allocation of resources under a competitive approach. In many health-service markets, distortions from the competitive norm are present only to a minor degree. In such circumstances, regulatory intervention can rarely, if ever, result in an alloca- tion of resources superior to that of the competitive process. Second, many of the characteristics of the health-care sector which impede competition from achieving desirable resource allocation also thwart the regulatory process. In both these situations, regulatory schemes may do little to improve the outcome of a competitive process. Clearly, the extensive regulatory structure currently in place does not begin to eliminate economic problems in this sector. Indeed, many of the regulations themselves cause further distortions in the market. As discussed earlier in this paper, the characteristics of the market and the desirability of regulations interact in a number of ways: ® In some health-care markets, the characteristics that may prevent competition from achieving a desirable allocation of resources are created by existing regulatory schemes. For instance, certificate-of-need regulations create barriers to entry and exit. -156- The ability of regulatory process and the competitive process to achieve desirable allocations of resources both may be adversely affected by the same market characteristics. For instance, when consumers lack important information about providers, regulators may do no better than the market at matching patients to the providers best suited to their needs. Some forms of regulatory intervention can improve the ability of the competitive process to allocate resources and can decrease the need for other, more intrusive forms of regulatory intervention. For instance, in some situations, providing consumers with improved information about providers can reduce the need for extensive price and quality regulation. Under some conditions, such as those associated with natural monopoly, regulatory schemes can improve the allocation of resources. Under these conditions, it is important to match the regulatory schemes to the needs of the market. The matching of regulatory schemes with the characteristics of the market is a general theme of this papers. This must be done by examining health-care services on a market-by-market Because regulatory structures differ from State to State and because demand for services also differs by geographic region, -157- competition analysis must be done at the local service level. This means that a paper written for a national audience can serve only as a general guide for conducting this analysis. By its nature, material of the sort covered in this paper is difficult to summarize. There are virtually always exceptions to any simply stated conclusions. Further, the purpose of the paper is not to enumerate a handful of lessons to be applied throughout the health sector; rather, it is to encourage the analysis of local service markets. In that spirit, the remainder of this section contains 31 questions about the market. These questions not only summarize the material in the paper; they also can be used by planners or decisionmakers who undertake a market analysis. These questions also may help those who do not conduct the actual analysis them- selves but who rely on such an analysis in making their decisions. Towards that end, each question is followed by a discussion of the question's importance and of how typical answers affect the appropriateness of taking a competitive approach to the market. The questions about the market are divided into three parts. The first seven questions discuss the role of third-party payers in stimulating competition in the service market. The next nine questions discuss the service from-the consumer's perspective. These questions focus on the availability of information, the consumer's discretion in receiving the service, and the consumer's ability to avoid financial risk. The last 15 questions address -158- the competitiveness of the market-providers, including both professionals and facilities. Here, entry and exit, quality and variety, and identification of natural monopolies are discussed. Existing regulatory structures are mentioned where applicable. As a guide to analyzing the market, this list is not exhaustive, and readers are encouraged to formulate more questions in order to fit the guide to their own needs. Because each service market is unique, the questions asked may vary from area to area. As competition analysis becomes more common, HSA's and other groups involved in health-sector planning can provide guidance to one another in developing techniques for market analysis and in designing procompetitive approaches to resource allocation. B. Reimbursement Schemes Many health-care services are covered, at least in part, by private health plans or by public reimbursement schemes. Studying the health financing market is important because competition in the financing market affects the competitive performance of health-care markets. Studying the health financing market is also important to corporate decisionmakers because it is often their best instrument for improving the performance of the health sector. Because health plans are chosen in advance of medical need, the consumer's decisions in this market are not shaded by the overtones of crisis or emergency that might impede their ability -159- 374-115 0 - 82 - 11 : QL 3 to weigh or interest in weighing alternatives. Reimbursement schemes are designed to shield consumers from directly bearing the costs of health care. This allows consumers to limit the financial risk associated with illness. However, when the cost of care is at least partly covered by a reimbursement plan, consumers have less incentive to shop among similar services on the basis of price or to consider alternative, lower cost, forms of treatment. Although consumers have little incentive to shop when cover- age is extensive, private third-party payers have incentive to keep reimbursements down when they must compete for subscribers by offering lower premiums. In a competitive market, third-party payers have the incentive to induce efficiency, to negotiate for lower rates, and to encourage the selection of low-cost providers because these efforts lower the costs of the health plan. The more competition exists among private third-party payers, the greater are the reimburser's incentives to perform these market- disciplining functions. To some degree, the viability of a competitive approach to a particular health service therefore depends upon the extent of third-party coverage, the degree of competition among third-party payers, and the efficacy of this competition in creating incentives for cost and utilization control. -160- 1. Is there more than one major private health plan available in the area? The more competition a health plan has from other plans or from HMO's, the harder that company must work to keep down costs in order to attract customers. Compared to companies with monopoly power, health plans facing competition have greater incentives to negotiate rates aggressively and to pass these savings on to consumers and to their employers. 2. How do these plans compete for customers? Do employers pick a single plan for their employees or do they typically offer several plans and allow the employees to decide? An employer's selecting a single plan differs from an employer's offering a choice to employees in that the employer selects the best plan for a large group of people, while employees select the plan best suited to their own families' needs. Because employees have superior information about their own medical needs and their willingness to pay for coverage, they can make tradeoffs between premium cost and coverage better than the employer can. While offering a choice of plans may raise the employer's administrative cost, it is also likely to result in greater overall employee satisfaction. Offering employees a choice of plans can also give individual employees an incentive to trade the extent of health coverage against the costs. In many instances, it is more difficult for an employer to make that tradeoff unilaterally. -161- 3. If employees are offered several plans, do the premiums paid directly by the employee reflect the differential costs of the plans? Unless the employees pay a higher price for a more comprehen- sive plan, they have no incentive to pick a "bargain" plan that provides limited coverage at a lower cost. When employees do not pay the cost of their health-care coverage directly, insurance companies have less incentive to offer a wide range of options at varying rates. Because Federal tax laws make it less costly for employees to receive employer-paid health benefits than to pay the premiums directly, some employers have adopted programs wherein employees opting for less expensive health plans are compensated with other (tax-free) fringe benefits. 4. Do the plans in the area provide coverage for services performed by the same group of providers, or does provider participa- tion differ from plan to plan? Plans that offer significant alternatives in terms of provider participation and style of treatment give consumers a broader range of medical choices. In this sense, a menu of health plans including group-practice HMO's in addition to more traditional plans offers a choice of practice styles as well as premium and copayment levels. Where alternative styles of practice are attractive to consumers, offering such plans can increase consumer satisfaction with the health-care system. -162~- However, there are many customers whose needs are best met by a conventional plan or an independent practice association. Never- theless, because competition among plans creates pressure for each to improve, all consumers are best served by having several conventional plans as well as HMO's all competing for the member- ship of individual premium-paying consumers. The more the health financing market departs from this description, the more closely it must be examined to determine whether there are competitive forces at work. 5. How is the health plan organized? What is the role of professional societies? The FTC study "Physician Control of Blue Shield Plans" by David I. Kass and Paul A. Pautler (November 1979) and subsequent research by these authors21® found that Blue Shield plans where local medical societies select at least one board member are associated with significantly higher reimbursement ceilings. In addition, when medical-society-influenced Blue Shield plans are merged with a local Blue Cross plan receiving hospital discounts, Blue Shield appears to reimburse physicians at a significantly higher rate. The authors, therefore, hypothesize that Blue Cross 216 gee David I. Kass and Paul A. Pautler, "Physician and Medical Society Influence on Blue Shield Plans: Effects on Physician Reimbursement," in Health Care: Professional Ethics, Government Regulation, or Markets, ed. Mancur Olson (forthcoming 1981) and Keith B. Anderson, David I. Kass, and Paul A. Pautler, "Physician Control of Blue Shield: A Further Discussion of the Issues," Bureau of Economics memorandum to the Federal Trade Commission, January 1981. -163~ plans receiving hospital discounts may be used to subsidize Blue Shield's payments to physicians in such markets. This study has raised considerable controversy, with some researchers reporting similar results but others reaching different conclusions.217 6. Do the third-party payers bargain aggres- sively with providers? How do they determine the rates at which they reimburse providers? Third-party payers only provide a market-disciplining force force to the extent that they play an active role in the market. Some HMO's, for example, have provided a market-disciplining force in the hospital market by negotiating flat-rate contracts with hospitals. Extensive appropriateness review by dental 217 positive associations between medical-society control of Blue Shield plans and higher Blue Shield reimbursement rates were found by Frank Sloan, "Physicians and Blue Shield: A Study of Effects of Physician Control," presented at DHEW conference, Washington, D.C., February 27, 1980, and by Richard Arnould and David Eisenstadt, "The Effects of Blue Shield Monopoly Power on Surgical Fees: A Theoretical and Empirical Analysis," presented at American Enterprise Institute Conference, Washington, D.C., September 26, 1980. For research yielding contrary results, see William Lynk, “Regulatory Control of the Membership of Corporate Boards of Directors: The Blue Shield Case," J. Law & Econ. 24 (April 1981): 59-174. -164- plans is another example of the active role third-party payers can take.218 7. Do third-party payers base payments on "costs" or charges by the individual provider, or do they base their payments on an areawide average of providers' costs? Because individuals have less incentive to shop on the basis of price for services with extensive insurance coverage, third- party payers must police the market by setting limits on the ability of inefficient or high-cost providers to operate successfully in the market. When reimbursement is based on each provider's own actual costs, cost reductions mean the provider receives less revenue, while cost increases lead to increases in revenue. Under such schemes, providers have little incentive to cut costs. In contrast, in reimbursement schemes based on an industrywide cost average, revenues do not vary directly with changes in the individual provider's costs. Although problems of comparability may make such programs more difficult to administer for institutionally based services, such programs give each provider more incentive to control its own costs. The usual, customary, and reasonable system widely used to reimburse 218 providers' collective attempts to resist such aggressive efforts by third-party payers were found to be illegal by a Federal Trade Commission administrative law judge. In Indiana Federation of Dentists, Docket No. 9118, Initial Decision, March 25, 1980, Aetna used dental X-rays to review benefit claims. Concerted efforts by the Indiana Federation of Dentists to refuse to submit such X-rays was found to be an illegal boycott. The case is currently on appeal to the Commission. -165- physicians is an example of a system where reimbursement levels depend on the charges of all providers in the area. While this system communicates incentives to perform services efficiently, it does not communicate incentives to keep prices low unless consumers' demand for the services of individual providers is price sensitive. C. The Consumer: How Price Sensitive Is Demand for the Service? Almost all markets depart somewhat from the ideal of perfectly informed consumers choosing among many sellers. However, in most markets, competitive forces still predominate. The market for a particular health-care service will be more or less competitive depending upon the level of insurance coverage, search costs, information availability, and the degree of discretion the consumer has in purchasing the service. 8. Is the service in question customarily covered by health plans? What part of the costs is typically paid by the consumer? In general, the higher the proportion paid by third-party payers, the higher the price consumers will be willing to pay for the service.219 Conversely, third-party payment tends to increase the quantity of the service consumers demand. Even when there are many providers, third-party coverage may increase both the quantity and the cost of the service consumed because higher 219 1p this discussion, "price" refers to the payment by the third-party payer plus the payment by the consumer. -166- cost providers may enter the market to satisfy the increased demand. These effects are illustrated in the diagram entitled "Comparison of demand for services with and without third-party reimbursement. "220 The more fully third-party payers cover a service, the less incentive consumers have to search on the basis of price and the more closely planners must examine the health-plan market to determine the viability of a competitive approach to the service. 9. Are alternative forms of the service treated differently by third-party payers? When third-party payers reimburse beneficiaries at varying rates for equivalent services provided by various types of providers or for services performed in different settings, the consumer's choice among providers may be distorted. For example, in an area where there might be significant consumer demand for nurse-midwifery services, third-party-payer coverage of obstetric- ians and exclusion of nurse-midwives could distort demand in favor of the more expensive form of obstetrical care for uncomplicated births. Similarly, more extensive coverage and lower copayment rates for inpatient services gives consumers an incentive to use inpatient services despite the availability of less costly out- patient care for some of the same medical conditions. Unless the consumer's out-of-pocket costs are higher for more expensive forms 220 por further discussion, see section III.H.1l. =167~ of the service, the postreimbursement costs consumers face mask the true cost differential. This does not mean that health plans should necessarily pay the same amount for all methods of treatment. Indeed, regulatory schemes requiring equal payment of all qualified providers and even regulatory schemes that require third-party payers to offer coverage of all providers may distort the allocation of resources if consumers have insufficient information or incentive to choose providers who perform quality services as efficiently as possible. 10. Is the service purchased on an emergency basis, or is "shopping" possible? The more immediate the need for treatment, the less ability the consumer has to search among providers. In true emergencies, consumers obviously will tend to choose the most immediately available or most familiar provider, as long as some quality threshold is met. When (as in emergencies) consumers do little, if any, contemporaneous comparison of the prices or qualities of services offered by various providers in the market, each provider has some price power over consumers seeking its services. In contrast, when consumers have an incentive and are able to "shop," no one provider has such power unless it has a monopoly in that market. In general, even if consumers would be willing to pay a very high price for the service were no alternatives available, -168- DIAGRAM: COMPARISON OF DEMAND FOR SERVICES WITH AND WITHOUT THIRD PARTY REIMBURSEMENT Price (including TPR) RS Supply Curve Price with TPR = p , Price without = P 3 TPR Demand with TPR Demand without TPR q q, = quantity with quantit TPR o quontity without TPR -169- they will not pay an exorbitant price to any individual provider when they have a chance and an incentive to shop. Consumers generally tend to be more willing to shop the more easily they can gather information about the prices and qualities of services offered by various providers. The more shopping consumers do, the more incentive providers have to compete on the basis of price and quality. Routine obstetrical and gynecological care, well-baby care, and psychiatric care are examples of services which lend themselves particularly well to such shopping. A competitive approach can work particularly well in such markets. 11. How much discretion does the consumer have in receiving the service? The more choice consumers have as to whether or not to receive a service, the less power an individual provider has to raise his price or to offer an unsatisfactory quality of care. When a consumer can satisfy his needs adequately without receiving a particular form of care, a provider of a particular service must offer consumers a satisfactory price/quality tradeoff, even when there is no other provider offering an identical service in the market. Physical exams, preventive dental care, and some types of minor surgery are examples of services where the typical consumer has some discretion in whether to receive the service and how quickly the service is needed. -170- 12. Are consumers who use the service likely to self-select into plans offering good coverage of that service, or are consumers unlikely to identify themselves as users of the service in advance? In some cases, such as outpatient psychiatric care and certain types of dental care, consumers of a particular service identify themselves prior to selecting a health plan and select plans partly on the basis of the coverage of that service. In such cases, coverage for that service ceases to be "insurance" in the ordinary sense. When consumers not interested in that service typically select other plans, the health plan's coverage of the service becomes a method of prepaying the cost of service--with one important difference: when the cost of the service is covered by a third-party payer rather than being paid directly by consumers, consumers have less incentive to shop among providers. This makes it more difficult to control costs than it would be if most market participants purchased the service without the involvement of a third-party payer. In such cases, the adoption of payment schedules with large copayments is a particularly appropriate way of inducing shopping behavior. 13. Is coverage of the service treated as a proxy for overall plan quality by consumers? It is difficult for consumers to compare several plans with small variations in coverage for a large number of services. In -171- such situations, consumers may base their choice of a plan on a few features, which they use as proxies for the quality or compre- hensiveness of the plan. When a particular service is commonly used as a proxy for overall plan quality, third-party payers have less ability to change copayment rates and reimbursement limits for that service than would otherwise be the case. For example, few consumers need home health care, but many may consider 100 percent coverage of such care a signal of a high-quality plan. Because relatively few consumers use the service, a 20-percent copayment may not generate enough savings to allow a large enough reduction in premiums to attract consumers to what they perceive to be an overall lower quality plan. Although the copayment would be expected to lead consumers to shop more carefully and might substantially lower the cost of this particu- lar service, when the service is used as a proxy the third-party payer stands to lose more in the way of membership by changing reimbursement patterns than it stands to gain from the cost savings induced by higher copayments. By improving consumer information about health plans and by making the comparison among plans easier, the consumer's reliance on such "signals" of overall plan quality can be reduced. Improved information also reduces the third-party payer's incentive to "advertise" by offering such coverage. Thus, by improving the information about health plans, the planner may increase the ability of a competitive approach to achieve an allocation of health resources which reflects both consumer preferences and the true costs of care. -172- 14. Who selects the provider of the service? The further consumers are removed from selecting the provider of a service, the less discretion they have over its purchase and (in general) the less price sensitive their demand for that service will be. In many instances, consumers may be effectively tied to a particular secondary provider by their choice of a primary provider. Consumers may choose the primary provider partly based on its associations with secondary providers, but this is not always the case. For example, in choosing a provider, consumers may ask about the hospital at which a surgeon, obstetrician, or internist has privileges. On the other hand, in choosing an internist, few consumers consider a physician's associations with particular laboratories or radiologists. When a relationship between the primary and secondary provider exists but is not considered by consumers in choosing the primary provider, the secondary provider has some power over price and quality with respect to those consumers. Thus, the competitiveness of the market for anesthesiology depends partly on the competitiveness of the market for surgeons and hospitals and on the incentives these providers have for selecting the anesthesiologist best suited to the patient's economic and medical interests. -173- 15. How easily can consumers obtain informa- tion about various providers? When search costs are high (for example, when consumers must have a formal consultation with each provider), consumers are less likely to obtain such information and are therefore less likely to search out the providers offering the price/quality combination best suited to their needs. In determining whether the costs of search can be lowered, the following questions should be considered. Could the information be centrally gathered and disseminated, and at what cost? Do providers have an incentive to advertise, if not otherwise restricted? How easily can consumers obtain and assess nonprice information about different providers? Can such information be disseminated at low cost? The more consumers know about providers and the more easily they can act on that information, the better the market is able to police itself. Some information, such as education, certifica- tion, or statistics on morbidity rates and their relationship to frequency of service performance, although not currently avail- able, might be easily disseminated to consumers. Other informa- tion can be gained only through experience with a particular provider. Still other information, such as that needed to decide among drug therapies or types of surgical procedures, cannot easily be summarized for consumers and medical training is required to understand the the tradeoffs involved. The less -174- consumers can understand and compare the differences between providers, the less consumers are able to protect themselves from low quality or unscrupulous providers and the better is the case for regulatory intervention. 16. How high are the costs of switching providers? What are its implications for the market? In many situations, a provider who is not a monopolist stands to lose current customers if experience shows that the price is too high or that the quality is too low. For example, consumers can easily choose to patronize another laboratory, pharmacy, or optometrist in the future if service is unsatisfactory. Even in the case of surgeons, where the effects of low-quality care may be irreparable, consumers rarely face significant costs for using different surgeons for different operations. In other situations, such as long-term care or psychotherapy, there may be substantial costs to switching providers. In these cases, the provider has substantial power to raise its price or lower its quality before the consumer is willing to undertake the costs of switching. Many medical services fall into an intermediate range, where a moderate price increase may be tolerated before a switch in providers is made. In the case of primary physicians, there is little cost in transferring records to a new provider. However, the physician-patient relationship, including interpersonal -175- 374-115 0 - 82 - 12 : QL 3 information not contained in the medical record, is a valuable part of the long-term and ongoing patronage by a consumer of a single provider. The higher is the cost of switching providers, the greater is the provider's power to raise price or lower quality while retain- ing current consumers. In most cases, medical ethics, concern for reputation, and the desire to attract additional consumers keep providers from exploiting current patients. However, in extreme cases, such as those involving institutionalized patients, the market does not always protect consumers once a provider has been chosen. In these extreme cases, regulatory intervention may be necessary to assure that a provider does not drastically lower quality or raise prices once the patient is locked into a rela- tionship with that provider. Under these circumstances, regula- tory safeguards can improve the ability of the market to operate, because consumers can purchase the service without fear of unscru- pulous providers. This makes the competitive approach more applicable because such safeguards allow ethical providers to compete more easily. D. The Competitiveness of Providers. A basic component of a competition analysis is identifying current and potential providers of a service and the incentives they face. In some cases, several types of providers or technolo- gies can be used to provide similar health services. When consu- mers consider these groups to be acceptable substitutes, membership in different professions or use of differing techniques -176- should not prevent their inclusion in the same economic market. Similarly, providers possessing equipment, facilities, or training that could quickly and inexpensively be altered to provide the service in question also may represent a competitive force in the market. The number of providers included in the market depends not only on the types of facilities, equipment, and professional training considered as substitutes; it also depends on the geographic boundaries of the market. In health care, the relevant geographic market area may vary considerably, depending on the speed with which consumers must receive the service, the importance they attach to variety and quality, and their ability or willingness to travel. Also relevant is the speed with which providers currently offering the service in other geographic areas can shift their resources to a new location. When there are few impediments to entry, when the service is not a natural monopoly, and when there are many potential pro- viders, competition may function effectively despite there currently being only a small number of providers currently in the region. However, in many health-care markets, third-party- payment schemes, cross-subsidy schemes, and various regulatory restrictions thwart the communication of economic incentives. these markets, the first step in adopting a competitive approach is to identify and remove these impediments. In this manner, planning techniques can be used to increase the competitiveness of markets. -177- 17. How many providers in the area can deliver this service? Can the market sustain more than one provider of minimum efficient scale? The more providers offer the service in an area, the more likely it is that competition operates effectively. However, when the cost of providing each unit of the service falls substantially as the quantity supplied increases, having more than one provider in the market does not necessarily lead to lower prices. In such markets, there is a tradeoff between the higher cost of providing each unit of the service and the decrease in market power as the number of providers increase. In markets characterized as natural monopolies, prices are higher with several providers than with one provider because the increase in competition is not sufficient to offset the increase in costs. In the rare cases where natural monopolies exist, regulatory intervention may be needed to restrain the provider's power over price. In some cases, the benefits of competition can be gained in Batural-sonopoly settings by using franchise bidding schemes or by basing compensation of the natural monopolist on the costs of similar providers in other markets.221 221 gee section III.B. for full discussion. -178- 18. How large a capital investment is required to provide the service? Services characterized by small fixed costs or small minimum- capacity levels relative to the market are unlikely to be natural monopolies. As explained in the text, natural monopolies are the only situation in which economists would presume that a competi- tive approach is inappropriate. 19. Is the capital physically mobile? Can it be sold to other providers in other areas if there is excess capacity in one market? Can the capital be used outside the health-care industry? The more uses there are for the equipment and the more physi- cally mobile it is, the less likely it is that any particular provider can exercise power over price for any significant period of time; entry will quickly eradicate any excess profits. When capital can be easily moved or can be used in other industries, it is unlikely that "cutthroat" competition can develop; providers may always sell out or move to another market. 20. How easily can new providers enter the market? Is a special license required to provide the service in the area? Can providers from other areas easily enter this market? How specialized is the capital or training required to become a provider? Are providers with training in short supply in the service area? The region? The State? The Nation? The fewer the barriers (such as special licenses) that exist between geographic areas, the less likely it is that some regions -179- will have gluts while other have shortages.222 However, the more specialized the training or other investment necessary to become a provider, the more likely it is that unanticipated increases in demand will lead to temporary nationwide shortages, regardless of interstate barriers. The duration of these shortages depends on the length of time it takes to develop new capacity. For example, it takes 8 to 10 years to significantly increase the number of physicians but only 3 to 5 years to build and equip new hospitals. During a shortage, prices may rise above the competitive level and high rates of return may be earned by providers in competitive markets. These profits serve to entice new providers into the market and do not mean that a competitive approach is inappropriate. The best way to speed the return of prices to the competitive level is by removing barriers to entry. 21. If capacity can be used for several purposes, how is it allocated among these uses? In a competitive market, providers have the incentive to use the capacity for its most valuable uses first. In health-care markets, third-party reimbursements may distort this incentive. Reimbursement schemes may also create an incentive for providers to invest in too much costly equipment, if investment controls are removed. For the competitive approach to be used successfully in 222 gee sections II.B.l. and II.B.2. and III.D. for discussion of mobility barriers and licensure. -180- a market where capacity can be used for several purposes, reimbursement rates should be set so that medical conditions that could be diagnosed or treated using a lower cost technology are only diagnosed or treated on more costly equipment when excess capacity exists. The reimbursement rates should also be designed so that providers have no incentive to build additional capacity to serve these low-priority users. In general, third-party payers give providers the incentive to use capacity appropriately by reimbursing different uses of the capital at different rates. For example, high-priority uses can be reimbursed on the basis of average total cost, while low- priority uses can be reimbursed on the basis of variable cost. This gives the provider the incentive to allocate capacity to its most valuable use first.223 22. Who bears the cost of investment in excess capacity? If the provider is not at risk when it adds unnecessary capacity to the market, the market alone will not limit supply. The argument that the poor credit rating of many hospitals provides a break on unnecessary investment does not apply to cases where reimbursement schemes assure an adequate rate of return on unnecessary capacity.224 223 The fine points of such pricing schemes are discussed in the literature on peak-load pricing. See, for example, M. A. Crew and P. R. Kleindorfer, "Peak-Load Pricing with a Diverse Technology," Bell J. Econ. 7 (Spring 1976) :207-31. 224 For an example of this argument, see R. Posner, op. cit. -181- When reimbursements are based on the individual provider's cost rather than on an industrywide average, the market will not provide incentives for inefficient providers to leave the market and for efficient providers to remain. A similar situation exists when reimbursement rates allow providers to subsidize unnecessary investments by increasing the price of other services. For competition to work, reimbursement plans should (as far as possible) attempt to reward facilities that face high demand and attain low costs and should avoid subsidizing high-cost facilities or services (especially those consumers avoid using). 23. How are rates for the service determined? The more freedom providers have to set their own rates with- out "guidance" from an association of providers and the more actively consumers or third-party payers negotiate, the more likely it is that the competitive price will prevail in a market with a large number of providers. For example, third-party payers may take an active role by unilaterally using a reimbursement formula based on average charges in the area. They may also negotiate contracts that are based on quantity discounts with providers. In both cases, the provider has an incentive to perform services as efficiently as possible. 24. Do providers in the area compete over price or nonprice attributes of the service? -182- In the absence of price competition, which may be precluded by reimbursement practices, providers may compete for customers on quality attributes, such as hospital food or advanced technology. Such competition shows that consumers or their agents compare the services offered by different providers and suggests that price competition might work as well, were incentives properly structured. Where price competition cannot be introduced, it is not necessarily desirable nor administratively feasible to eliminate quality or technology competition because it is often difficult to identify and to enforce the optimum quality level. 25. Who selects the provider delivering the service? Does the agent selecting the provider have an incentive to negotiate rates, or is the charge passed through to consumers or their health plans? The more removed the decisionmaker is from the source of pay- ment, the less important price competition is likely to be. For example, drug manufacturers focus their competitive efforts on the physicians who select prescription drugs. Not incidentally, little of their promotional material focuses on the product's price. Emphasizing only the nonprice attributes of prescription drugs does not necessarily reflect consumer preferences. Indeed, many consumers pick pharmacies on the basis of price, and in States with generic-substitution laws, many consumers ultimately pick the drug manufacturer on the basis of price as well. How- ever, the physician, acting as the consumer's agent, does not -183- necessarily have an incentive to consider price in selecting among manufacturers' brands. Similarly, physicians may select a laboratory on the basis of turnaround time or mutual referrals. While the physician may include price as one of many factors in recommending a lab, the physician may have little incentive to gather comparative price information on the consumer's behalf. In these cases, as in others, there is less price competition than there might be if consumers had the incentive as well as the knowledge to pick providers for themselves, or if physicians were given greater incentives to compare prices on the consumer's behalf. Prepaid health plans are one institutional structure that provides these incentives for physicians. Improving the incentives faced by the agent who picks the service allows the unfettered market to allocate resources more efficiently. 26. 1s the presence of certain equipment used by consumers or by other provider- associates as a proxy for quality? When a certain type of equipment or service is used as a signal for quality, a provider has an incentive to invest in that equipment or service, whether or not that service is self- sustaining. Absence of the quality-signaling equipment may cause the hospital to lose consumers, whether or not those consumers have conditions requiring use of that equipment. For example, in markets where computerized tomographic (CT) scanners are used as signals of general hospital quality, hospitals may wish to -184- purchase CT scanners whether or not the demand for the scanner itself would be sufficient to make its presence necessary in the hospital. One effect of regulations restricting the proliferation of equipment used as a signal of quality is to put facilities denied such equipment at a competitive disadvantage disproportionate to the absence of the equipment alone. A way to combat this unintended effect of such regulations is to disseminate informa- tion about institutional sharing arrangements and other informa- tion about insitutional quality so that the presence of such equipment ceases to be a signal. This is an example of how an nformational remedy can be used in place of certificate-of-need regulations to limit excessive investment in technology. 27. 1s there an active claims review program for this service? The more actively third-party payers monitor the service to prevent unnecessary use, the less likely it is that "Roemer's Law" (hospital bed supply creates its own demand) can operate.225 Claims review works most successfully in areas such as dentistry, where diagnoses can be easily documented and where treatment for diagnosed conditions is fairly standard. 225 gee M. I. Roemer, "Bed Supply and Hospital Utilization: A National Experiment," Hospitals 35 (1961):998-1003, for a statement of this "law." -185- 28. Is the market characterized by significant cross-subsidization between services? Do entrants have these sources of cross- subsidization? What is the rationale for cross-subsidization? There are many situations in health care where users do not pay the full cost of the services they use. In these cases, the cost of these services is borne by the users of other services. In some cases, the "subsidy" makes sense. For instance, consumers of obstetrical services pick their hospital or birth center partly on the basis of emergency services available, in the event that problems arise. Because each user of the obstetrical unit had some chance of using the neonatal intensive-care unit, spreading the cost of neonatal intensive-care units across all users of the obstetrical service is a form of risk sharing. All individuals paying to support the service have some likelihood of benefiting from its existence. In other cases, the cost of one service is added to the overhead for another even though there is no logical or clinical relationship between the two. For example, when users of a hospital laboratory subsidize emergency medical care, they benefit no more from this involuntary contribution than do members of society in general. Although such cross-subsidies may be the hospital's only available instrument for achieving to achieve a desirable social goal (such as care of the indigent), cross- subsidies create distortions in the quantity demanded of both subsidized and subsidizing services. -1B6~ Cross-subsidies also complicate the issue of entry. An entrant wishing to provide a service that is usually subsidized cannot compete effectively if it does not have sources of cross-subsidy similar to those of its competitor. In other instances, an entrant may be able to provide a service more cheaply than existing providers only because its prices do not include cross-subsidies for other (unrelated) services. Thus, cross-subsidies make it more difficult to use prices and competition to allocate resources. 29. Who are the potential entrants capable of providing this service? Who finances new entry? Who finances expansion by current providers? When the service is not a natural monopoly, the threat of entry can often keep prices down and quality up, whether or not entry actually occurs. For entrants to keep price and quality in line, they must not be at a serious disadvantage in relation to incumbents. In part, this depends on whether potential entrants have equal access to financing. When incumbents have access to financing at below-market rates through municipal revenue bonds or cross-subsidy schemes while entrants must rely on normal channels for their financing, entrants may be put at a comparative dis- advantage unrelated to their own efficiency in providing the service. The cost of municipal revenue bonds and cross-subsidy schemes to finance incumbents but not entrants is borne twice by the public--once when the difference between subsidized and market -187- rates is paid and once again when giving the incumbent below- market financing results in a less competitive service market. 30. Might new entrants provide an alternative form of care not currently available in the market? Are such potential providers precluded in any way from entering the market? Birth centers, surgicenters, physician's assistants, and midwives are examples of alternatives that may be able to provide certain types of care at lower cost than traditional providers of equivalent services. However, certificate-of-need procedures and licensing laws may hinder or discourage the entry of such alterna- tives. The lack of equivalent access to financing from public sources (discussed in the previous question) may also make it more difficult for alternative providers to enter the market. 31. Do indicators of consumers' preferences for particular providers or particular forms of the service exist? Do providers preferred by consumers have an incentive to expand? Do other providers have an incentive to contract? Certificate-of-need programs tend to take the view that "a bed is a bed." However, from the consumers' point of view, this may be far from accurate. Few service markets are characterized by uniformly excess demand or by excess capacity throughout the system. Instead, there are often queues for the services of one provider of a service, while another provider could still treat more consumers. Tightly booked schedules, crowded waiting rooms, -188- waiting lists, and other queues are important signals of consumers' preferences. In a competitive market, providers of more highly desired services would increase price or capacity, while providers with less highly desired services would cut price or capacity. How- ever, in the health-care context, providers reimbursed on the basis of cost have no incentive to respond to demand in this way. Further, even when the correct reimbursement incentives exist, there are often institutional constraints, such as CON require- ments, which prevent the market from responding to consumers’ demands. In such cases, a CON process that takes actual demand patterns into account can increase consumers' satisfaction with the system. Maximum reimbursement limits can also blunt a provider's incentives to expand, despite consumer queues. Increased sensitivity of third-party-payment schemes to capacity utiliza- tion and patterns of consumer preferences (in markets not chara- cterized as natural monopolies) can give those providers preferred by consumers the incentive to expand and others the incentive to contract. Copayments can insure that consumers are cognizant of the actual costs associated with selecting higher quality and more costly providers. -189- U.S. GOVERNMENT PRINTING OFFICE : 1982 O - 374-115 . TT oT Tw : a YLT > TELE, 5 A = 176866 U. C. BERKELEY LIBRARIES C041L52537