22385 2: African Traditional Healers: 9 The Economics of Healing raditional healers are a source Economic theory _ rT of health care for which Afri- T cans have always paid. Even When consumers purchase a ser- vice where quality matters but cannot with the expansion of modern be directly observed or evaluated, r medicine, healers are still popular economic theory raises the possibility and command fees exceeding the av- of a market failure. Health care is a erage treatment cost at most modern _____________ practitione Are tadtion heaers classic example of this type of failure, practitioners. or aretheala- commonly referred to as agency. Pa- mirace wokersor ae thy chrla-tients choose to visit physicians pre- tans? Clearly either view is too ex- tient chse physicians pre 1.~ ~ ~ ~ ~tee Trdtoa helrsae,o cisely because physicians know more *. _ ~~~~treme. Traditional healers are not perfect: although we have all heard than patients do. The patient cannot stories of miraculous cures, the evaluate what the doctor is doing for stories of miraculous cures, the health status of the average African her sake, nor can she infer physician peasant cannot be reconciled with effort from the outcome. Patients are omnipotent healers. Nor, however, often cured despite poor quality care, or fail to be cured despite expert can they be charlatans: people can- not be continuously ignorant about care. If patients cannot tell what doc- the capacity of someone who lives tors are doing, then doctors have fewer , , ~~~~~reasons to exert extraordinary effort on among them. This article advances a view of traditional healers that relies the patient's behalf. Thus, although pa- tients are willing to pay for high-quality on neither supernatural power nor manifest ignorance. It suggests that care and physicians are able to provide healers remain popular despite abun- it, the market may fail and high-quality No. 32 dant modern medicine because they health care would not be delivered. May 2001 have wisely used an important eco- nomic contract to the mutual benefit of their practice and the population 1K Notes reports perlodnl , o Indigenion nwd (l 110ntvi*tl they serve. -in Sub-aharan nAfica. lt puilsbdd by thie Afica Rglons Kowegeaid While the contrasts between tradi- LeaningCenteras part of an ling tional medicine and modern medi- 1Kpartnesween teWr -Bank, conimunites,,NGO)s, 'develop.. cine are many, the article focuses on mert institutions andmlilatea the differences in the way traditional organizatiorns. The views e in this article rethoefthauor and modern healers are paid. An im- .,am X t he - andshouild not be a4ttribuei t portant element of their practice has World Bnk Grup or, its prners in been previously ignored: traditional this iiiative. A wpage Qn 1K Js avaible at -/w.worldorafr/ healers receive the bulk of their pay- Wdefault.htm ment only if the patient is cured. 2 The standard economic solution is to force the agent cured and not cured are inherently subjective and subject to (the person performing the unobservable services) to face deliberate misrepresentation. the loss or gain of his/her actions. Thus, salespersons are paid on commission, and CEOs receive stock options. The health counterpart is the pay-only-if-cured or outcome- Eoncmic practices of traditicnal healers in Africa contingent (since payments are contingent on outcomes) contract. If patients paid their doctors only if they were This "ideal" but previously unobserved economic contract cured, or more if they were cured than if they were not, is exactly the method by which traditional healers in Africa the economic cost of agency would be reduced, do business. Patients pay traditional healers more if they are cured than if they are not cured. The results of interviews Despite the fact that health care suffers from a problem with healers in Cameroon, Tanzania, and Ethiopia reported widely identified in theory, health economics as a whole in a publication by this author are summarized here. The has dismissed the economic solution to the problem as in- healers spoken with received an initial payment and, in addi- applicable to health care. Non-contractibility is one of the tion, negotiated with the patient over a payment to be made most widely cited reasons for dismissing this solution, in the future. In all cases, if the treatment did not result in Contractibility means that agreements about payments improvement of the condition, the patient paid nothing be- are enforceable. The outcome-contingent contract is con- yond the initial payment. sidered non-contractible because patients can simply lie about the outcome of treatment and the doctor would be When asked about the practice of "cursing" patients for unable to prove his case to a third party. The definitions of non-payment, many healers were adamant that they never engaged in the practice, though almost all admitted that their ancestors or, specifically, parents had done so. The practice traditionally operated as follows: when a patient IK N otes refused to pay, the healer would either invoke a curse on would be of interest to: the patient or revoke the cure. This practice invokes near- universal fear in rural populations, and most non-healers Name believed that if they failed to pay they would be cursed. All In,-stitution healers told stories of patients leaving without paying and then returning, sometimes years later, begging to be al- Address lowed to pay. Patients believe cursing is still practiced, al- lowing healers to wait until after the treatment to collect payment without fear that the patient will refuse to pay. The outcome-contingent contract means that healers have strong incentives to provide quality care even if the __ l __* _ patient cannot evaluate or observe this quality. But what of the non-contractibility concern? How can traditional healers agree to accept payment when the patient could lie about the outcome and refuse to make any payments? Two reasons are advanced: first, healers live in the same village as many of their patients and can therefore observe _ ;1; !S . ' S IS a. ' a, what patients are doing; second, patients believe that healers know whether they are cured. 3 In contrast, outcome-contingent contracts are not used evaluated on tests ordered, treatments prescribed, or advice in the practices of modern medicine anywhere in the given but will not be evaluated on whether the patient is world. Instead, patients pay a fixed fee for all services de- cured (or in this case, whether his/her symptoms are allevi- livered whether they are cured or not. No one expects this ated). If the patient refuses to quit smoking s/he will not be type of payment scheme to deliver quality. Instead, the as- cured, but this does not matter to the modern physician. sumption is that quality care is assured by restrictions on The traditional healer, on the other hand, has no chance the activity of practitioners that come from other sources of being paid if the patient does not quit smoking. If he is (including, but not limited to medical associations, refer- unable to convince the patient to quit, he could refuse to ral networks, hospital networks and direct regulation of the practitioner.) In Africa, physicians at modern facilities take-thcae, t thecer least,frefue totacep the come-contingent contract. The difference between the such as government or not-for-profit health systems (pri- two types of contracts emerges not from the amount of marnly church-operated) are regulated through hierarchi- medical effort finally delivered, but from the difference in cal supervision, where physicians evaluate the quality of the relationship between the efforts of practitioners and other physicians and employment status or bonuses de- the efforts of patients. Because outcomes (and not inputs) pend on the result of the evaluation. Since the payment (or ellbeig) o th prctitone deends no onthe matter, a traditional healer cares more than the modern (or well-being) of the practitioner depends, not on the dco bu h cin htptet ae doctor about the actions that patients take. outcome, but the effort s/he exerts, this is referred to as an effort-contingent contract. Note that in order to regu- late another doctor you need to be at least as well trained as s/he is: only doctors can regulate other doctors. Patient perception of the practice of healers Economic theory suggests that the contract used by healers will lead them to provide high-quality effort in Comparing outcome-contingent health care. Modern practitioners may also provide high- to effort-ccntingent c:ntracts quality effort, but they will not work as well with patients in situations where patient effort is important to outcomes. The outcome-contingent contract of traditional healers gives them good reason to exert effort in the treatment of How do patients use traditional healers? If the theory is patients even if the patient cannot evaluate what the correct, traditional healers should have an advantage healer is doing. However, if a modern physician practices (holding other factors constant) when patients suffer from in an organization that cares about quality, s/he will also illnesses that require both medical and patient effort. have a good reason to produce high-quality care; her/his They will not have an advantage in situations in which spe- employers will insure that s/he does. Thus, while it would cial equipment or skills are required. Contrast the ex- appear that the contract of the traditional healer does not ample of asthma with malaria and appendicitis. Malaria is offer any additional benefit over a well-implemented con- comparatively easy to diagnose and does not require ex- tract at any modern provider, this is not quite true. tensive effort on the part of either physicians or pa- tients - instead, it requires widely available medication. The fact that healers contract on outcomes rather than Patients should visit the least expensive provider who has effort has important repercussions on their practices. Take, for example, a patient with asthma who smokes. A the appropriate medicine without concern for quality. Ap- modern doctor treating this patient can be evaluated pendicitis, on the other hand, requires surgery - a service for which hospitals are much better equipped than tradi- based on what s/he does for this patient by another physi- tonai he spits should seek skideatment. cian who knows what s/he is supposed to do. S/he can be 4 Indeed, across Africa there are consistent patterns in the with the right incentives to provide high-quality care, and choice of health care practitioner according to illness condi- patients behave as if they are aware of the implications of tion: certain conditions tend to lead to visits to certain prac- this contract. Importantly, the magical or mysterious ele- titioners. This author tested these patterns to see if they fit ments of their practices are essential to the proper function- the patterns that would be predicted by theory. In order to ing of this contract. Without the belief that healers know the do this, each individual illness condition is examined by phy- outcome of treatments, the outcome-contingent contract sicians who evaluated the degree to which the outcome de- would be non-contractible. Thus, although this contract ap- pends on the efforts of the practitioner and the efforts of the pears to be very successful for healers, it cannot be widely patient. adopted. In the southwest province of Cameroon, patients are more Traditional healers, far from duping a gullible population, likely to visit a mission facility over a government facility behave as if they are rational and serving a rational popula- when they suffer from conditions that require substantial tion. As long as modern medicine is delivered in a context in amounts of medical effort. Since mission facilities are well which quality is uncertain, traditional healers will continue regulated compared to government facilities, this makes to attract patients. sense; patients seek and are willing to pay for this particular measure of quality only when they deem that it really mat- ters. In addition, as would be predicted by economic theory, patients are more likely to visit a traditional healer (even over the high-quality mission facility) when they suffer from Selected references a condition that requires large amounts of both medical and Leonard, Kenneth L., 'African Traditional Healers and Out- patient effort. Patient behavior follows the patterns pre- come-Contingent Contracts in Health Care," mimeo (avail- dicted by an economic understanding of the payment scheme able on-line at www.columbia.edu/- kl206/research.html), used at traditional healers. Columbia University 2000. -, "Incentives and Rural Health Care Delivery," in David Conclusion K. Leonard, ed., Africa's Changing Markets for Human and Animal Health Services, London: Macmillan, 2000, chap- While traditional healers perform many roles in their soci- ter 4. eties, this article focuses on one important feature of their practices: they use an economically rational tool in their "When States and Markets Fail: Asymmetric Informa- tion and the Role of NGOs in African Health Care," mimeo practices and their behavior and the behavior of patients re- flects the use and benefits of this tool. The pay-only-if-cured (available on-line at www.columbia.edu/- k1206/ or outcome-contingent contract provides traditional healers research.html), Columbia University 2000. This article was written by Kenneth L. Leonard, Department of Economics, Columbia University, Mail Code 3308, 420 W 118th Street, New York, NY 10027; e-mail address: KL206@columbia.edu