رفتار و انگیزش در مورد دیگران در ارائه مراقبت های بهداشتی: یک آزمایش با دانشجویان پزشکی و غیر پزشکی

Other-regarding motivation is a fundamental determinant of public service provision. In health care, one example is physicians who act benevolently towards their patients when providing medical services. Such patient-regarding motivation seems closely associated with a personal sacrifice that health service providers are willing to make. Surprisingly, evidence on physicians' motivation is rare. This paper contributes to the literature by investigating prospective physicians', in particular, medical students', motivations and behavior. We measure the willingness to sacrifice own profit in order to increase the patients' health benefit. We conduct the same analysis for non-medical students. In a controlled incentivized laboratory experiment, participants decide, in the role of physicians, on the provision of medical services under fee-for-service or capitation schemes. Overall, 42 medical students and 44 non-medical students participated in five experimental sessions conducted between 2006 and 2008. We find substantial differences under both payment systems: compared to medical students, students of non-medical majors are less patient-regarding, less willing to sacrifice their own profit, and they state less motivation to improve patients' health. This results in significantly lower patient health benefits. Some implications for health care policies in light of physician shortage and for physician payment systems are discussed.

Other-regarding motivation is a fundamental determinant of public service provision, the importance of which has been emphasized in the public economics literature (e.g., Francois, 2000, Besley and Ghatak, 2005, Delfgaauw and Dur, 2007 and Delfgaauw and Dur, 2008). In particular, this literature assumes that most public service providers derive great personal satisfaction from helping other people. A prominent example is physicians who provide medical services for their patients. Early theoretical models in health economics accounted for the physicians' other-regarding motivation when caring for patients' health benefit (e.g., Woodward and Warren-Boulton, 1984 and Ellis and McGuire, 1986). At the same time, patient-regarding motivation seems to depend upon the degree of personal sacrifice associated with the act (Le Grand, 2003). Presumptions on human motivations are crucial for an effective public policy design. Policies based on the assumption that people are entirely other-regarding might have considerable negative consequences if, in fact, they are predominantly selfish. The same holds for policies that assume people to be selfish, particularly if that would result in suppressing their natural other-regarding impulses (Le Grand, 1997). Thus, assumptions on other-regarding motivations are important, for example, for designing optimal payment schemes for physicians (e.g., Jack, 2005, Choné and Ma, 2011 and Makris and Siciliani, 2013). Field evidence on the motivation of health care providers and physicians, in particular, is surprisingly rare. A survey study analyzing the behavior of dentists in the UK, e.g., finds heterogeneity in behavior that hinges on both self-interested and patient-regarding motivations (Taylor-Gooby et al., 2000). In our paper, we wish to contribute to filling this gap by applying an experimental economics approach. The rising number of experimental studies in health economics indicates the increasing importance of running controlled laboratory experiments for this research area (e.g., Levy-Garboua et al., 2008, Schram and Sonnemans, 2011, Ahlert et al., 2012, Ahlert et al., 2013, Buckley et al., 2012, Brosig-Koch et al., 2013a, Brosig-Koch et al., 2013b and Godager et al., 2013). In our experiment, we investigate whether prospective physicians, i.e., medical students (MEDs), show patient-regarding behavior and are willing to sacrifice their own profit in order to increase patients' health benefit as conjectured by Le Grand (2003). As our workhorse, we use a fully incentivized, controlled laboratory experiment equivalent to Hennig-Schmidt et al. (2011). Subjects decide on the provision of medical services under a capitation (CAP) or a fee-for-service (FFS) payment system in a stylized physician decision-making context. When paid by FFS, subjects receive a monetary reward for each unit of medical services provided. Under CAP, subjects receive a lump-sum payment per patient, independent of the number of chosen medical services. A post-experimental questionnaire collects information on subjects' motivations. Our experimental design allows us to identify patient-oriented behavior and, in particular, to quantify individual sacrifices – physicians' choices imply a tradeoff, as increasing a patient's benefit comes at the cost of foregoing own profit. Le Grand's (2003) assertion of a personal sacrifice being a major characteristic of other-regarding behavior in the context of public service provision, involves physicians sticking out in that respect, compared to other people. Therefore, we also run our experiment with students of various other majors, i.e., non-medical students (NON-MEDs). We analyze whether MEDs are more caring for patients than NON-MEDs by exhibiting a higher willingness to sacrifice their own profits and by taking into account the patient-optimal treatment to a greater extent. Knowledge about differences between MEDs' and NON-MEDs' motivations is essential to investigate Le Grand's (2003) conjecture. In case NON-MEDs do show less patient-regarding behavior, important implications for health policy design might arise in view of the present severe physician shortage in many countries. To enhance the supply of physicians, health care policies mainly follow two strands: first, fostering immigration of physicians from abroad and, second, increasing the admission rates to medical schools. Attracting physicians from other countries is not a sustainable policy option, as it transfers the problem from one country to another, and causes substantial negative effects for the source countries (Astor et al., 2005, Chen and Boufford, 2005, Mullan, 2005 and McDonald and Worswick, 2012). Increasing the admission rates to medical schools has therefore been widely promoted as an alternative policy option (Association of American Medical Colleges, 2012). However, extending the pool of potential future doctors, in particular to young people, who did not intend to become physicians from the outset, would result in quite negative consequences for the patients in case the former show less patient-regarding motivations and behaviors when payment schemes are given. In light of the increased variation in other-regarding motivations and behaviors, the implications for overall social welfare might be ambiguous, however (see, e.g., Makris and Siciliani, 2013). Our main findings are, first, that MEDs are willing to sacrifice a considerable part of their profits to increase the patient benefits. Second, MEDs and NON-MEDs differ substantially in their provision behavior. MEDs' decisions are more strongly motivated by the patients' benefit, leading to a lower tendency to overprovide (underprovide) patients under FFS (CAP) than observed for NON-MEDs. This difference can be explained by MEDs' higher willingness to sacrifice their own profit compared to NON-MEDs. Participants' stated motivations from the questionnaires corroborate the behavioral findings.

The present paper systematically analyzes behavior of prospective physicians (MEDs) and non-medical students (NON-MEDs) in a physician decision-making context. Our vehicle is a controlled laboratory experiment as introduced by Hennig-Schmidt et al. (2011). We find that MEDs not only show notable patient-regarding behavior, but also a substantial willingness to sacrifice their own profit to enhance the patients' health benefit. In an incentivized decision task, our experiment thus corroborates the conjecture in the public economics literature (Le Grand, 2003) that patient-regarding motivation is closely associated with a personal sacrifice providers – in this case, medical students – are willing to make. We found significant differences between MEDs and NON-MEDs in their decisions to provide medical services. Even though both overprovide under FFS and underprovide under CAP, the extent varies significantly between them; compared to MEDs, NON-MEDs are much more driven by profit-maximizing objectives, and they are much less influenced by the patients' health benefit. This is reflected in a significantly lower willingness to sacrifice their own profit. One might argue that these findings result from educational influences during MEDs' and NON-MEDs' studies rather than being due to students' different motivational predispositions when entering the university. In other words, 'indoctrination' rather than self-selection (Kirchgaessner, 2005) would cause our results. Our analysis, however, showed that this argument does not hold, as we do not find significant correlations between length of study and other-regarding behavior for both MEDs and NON-MEDs. Moreover, as early as in the beginning of their university education, the former are already more patient-oriented than the latter. Finally, in line with their behavior, a significantly lower percentage of NON-MEDs state in a post-experimental questionnaire that they are decisively influenced and motivated by the patients' health benefit. In the following, we discuss the implications of our results. We argued in the Introduction that increasing admission rates to medical schools in order to fight physician shortage faces the problem of adverse effects for patients' health if NON-MEDs are less patient-oriented than medical students under a given payment scheme. And this is exactly what we found. Reducing physician shortage by immigration is no sustainable policy option. It has resulted in substantial physician movements from lower-income countries to developed countries such as the United States, the United Kingdom, Canada, and Australia (Mullan, 2005 and McDonald and Worswick, 2012), causing a severe brain drain in the former with far-reaching consequences for patients' health (Astor et al., 2005, Chen and Boufford, 2005 and Connell et al., 2007). But developed countries also suffer a loss from physician movements, for instance from Germany to Norway or Switzerland (Mullan, 2005). Our results do not imply that all students who choose not to become a doctor in the first place are necessarily more self-oriented than students who did choose to become doctors. But a remarkable percentage of them are more self-oriented, and therefore drawing attention to this motivational difference is vital to us. This is the more so, as the importance of personality measures and motivations, in addition to ability and achievement measures in the medical admissions domain, have been increasingly recognized in recent years (e.g., Ziv et al., 2008, Kuncel et al., 2010 and Adam et al., 2012). To select the best future physicians, medical schools test for personality attributes to ensure that applicants can be expected to be of high integrity and morally oriented (Lowe et al., 2001 and Cleland et al., 2012). Our findings suggest that policy-makers, when extending the pool of applicants to medical schools, should develop (or enhance) and employ rigorous screening procedures with regard to personality measures and other-regarding motivations during the admission process. Our findings on patient-regarding behavior also relate to the recent literature on incentives for other-regarding providers in the public sector (see, e.g., Delfgaauw and Dur, 2007, Delfgaauw and Dur, 2008, Makris, 2009, Makris and Siciliani, 2013 and Liu and Ma, 2013). This literature implies that provider payment systems need to address the misalignment of incentives for (partially) other-regarding providers. As MEDs care for their patients (some even as perfect agents), some element of supply-side cost sharing is necessary to be implemented in the payment system in order to realize the first-best service volume (Ellis and McGuire, 1986). When providers are partially other-regarding, too high a degree of patient orientation could be distortionary. Makris and Siciliani (2013), for example, show that lower social welfare could result if the other-regarding motivation exceeds a certain (moderate) level. Moreover, Ellis and McGuire's (1986) model suggests that, due to our finding of lower other-regarding motivations of NON-MEDs compared to MEDs, the rate of supply-side cost sharing needs to be lower for the former. Whenever a third-party payer is constrained to offer the same payment method to all providers, differences across MEDs' and NON-MEDs' patient-regarding motivation would make it impossible to implement an optimal payment mechanism that motivates a first-best volume of medical services (Jack, 2005). In such a ‘one-size-fits-all’ payment scheme, the rate of supply-side cost sharing needs to be such that the more patient-regarding providers harvest rents in order to ensure the participation of the less patient-regarding providers (see also Godager and Wiesen, 2013). However, offering different modes of payment – including different degrees of supply-side cost sharing – can be an efficient way of sorting individuals according to their patient-regarding motivation (see Jack, 2005). Laboratory experiments are not very common in health economics research. Yet, in the context of our research agenda, they constitute a valid research method that provides particularly adequate features. For one thing, we compare behavior of MEDS and NON-MEDs under true ceteris paribus conditions as only one parameter is changed at a time: either the payment system or students' field of study. Moreover, the experimental method allows thorough robustness checks of our findings because different scientists can repeat the experiment under exactly the same conditions. Finally, physicians face a real tradeoff between their own profit and the patients' benefit, as decisions are incentivized by monetary payoffs. In addition, physicians have incentives to show patient-regarding behavior as the patient benefit is measured in monetary terms and physicians' decisions are consequential for the medical treatment of real patients outside the lab. Due to the latter features, our experiment appears to be a suitable tool for assessing personal quality and moral orientation, comparable to the hypothetical dilemmas embedded in the selection procedure described in Bore et al. (2005) and Cleland et al. (2012). To inform policymakers in health care further, an important task for future research is systematic (panel) studies on how medical and non-medical students' motivations, perceptions, and behaviors differ when they begin their university education, and how these change in the years to follow. Our findings on patient orientation may have important implications for further theoretical work on optimal physician payment schemes as well as for policy-makers seeking to implement such remuneration systems for physicians with different patient-regarding motives.