Incentives and Performance of Healthcare Professionals
Abstract and Keywords
Economists have long regarded healthcare as a unique and challenging area of economic activity on account of the specialized knowledge of healthcare professionals (HCPs) and the relatively weak market mechanisms that operate. This places a consideration of how motivation and incentives might influence performance at the center of research. As in other domains economists have tended to focus on financial mechanisms and when considering HCPs have therefore examined how existing payment systems and potential alternatives might impact on behavior. There has long been a concern that simple arrangements such as fee-for-service, capitation, and salary payments might induce poor performance, and that has led to extensive investigation, both theoretical and empirical, on the linkage between payment and performance. An extensive and rapidly expanded field in economics, contract theory and mechanism design, had been applied to study these issues. The theory has highlighted both the potential benefits and the risks of incentive schemes to deal with the information asymmetries that abound in healthcare. There has been some expansion of such schemes in practice but these are often limited in application and the evidence for their effectiveness is mixed. Understanding why there is this relatively large gap between concept and application gives a guide to where future research can most productively be focused.
Incentives in Healthcare
The literature on the economics of healthcare abounds with references to incentives. It is implicit in most instances that what are being referred to are financial inducements or rewards designed to elicit some action or choice, and that interpretation is followed here. While financial payments are intrinsic to exchange the linkage to incentives seems much more common in relation to healthcare. Why is that?
Since Arrow (1963) economists have viewed healthcare as distinct from other areas of economic activity on account of a combination of characteristics. Healthcare is provided by individuals or organizations that are experts relative to the patients they treat, and so there is asymmetric information. These experts are consulted precisely because they have knowledge that their patients do not; thus the quality of the services provided by these experts is difficult to assess in advance or even after services have been consumed. They are thus neither search nor experience goods (Nelson, 1970). Furthermore, healthcare is often provided on the basis of insurance so that the recipient does not pay the full cost of the services they receive, a feature that is referred to as third-party payment.
Taken together these features of healthcare mean that the incentives implied by fixed price exchange, wherein the consumer assesses the value of a service against the alternatives on offer, will be muted or absent in healthcare. It, therefore, becomes relevant and important to consider more fully the role of payment methods. In essence financial incentives can be considered a consequence of the conditionality placed on a financial transfer, which is the approach adopted here. A simple price mechanism conditions the transfer of payment only upon the delivery of service. Payment could, however, be conditioned on, for example, an independent assessment of the value of that service or delayed pending proof of that value. These conditional payments might induce the supplier to take more care or deliver a better quality service.
The literature on incentives in healthcare delivery is vast and encompasses arrangements for paying, among others, healthcare providing organizations (hospitals and clinics), the providers of healthcare insurance, suppliers of healthcare equipment and drugs, and individual healthcare professionals (HCPs). Henceforth, the focus is exclusively on the last of these. In common usage HCPs include doctors of many different specialties, dentists, nurses, therapists, and other care providers, but following much of the literature in economics we will primarily focus on primary care (general practitioner) doctors and dentists. These types of HCPs are very often either self-employed sole practitioners or in partnerships, so those are the dominant organizational structures considered here. There is an extensive and much more general literature on the motivation of employees within larger organizations (Oyer & Schaefer, 2010; Roberts, 2010) that draws upon many of the same conceptual ideas and methods that will be described.
One reason why the HCPs who have been studied are seldom employed in the conventional sense of being obligated to supply their time in return for a fixed wage may relate to the definition of a profession, which often includes requirements not only of specialized training but also of independence from financial obligations to others. This acts as an important caveat. The idealized view of a profession is that its members will supply services in a disinterested and objective way, for direct remuneration but independently of other business or financial gain and so without obligations to shareholders or even business partners. The idealized view suggests that incentives of the kind examined by economists may not have much impact, and it is important to accept at the outset that financial incentives may be a small part of what determines the behavior and performance of HCPs. Professional ethics, duty, and obligations along with an inherent motivation to serve their clients or the greater public good will all have a role to play (Andersen, 2009; Ash & MacLeod, 2015; Andersen & Pedersen, 2012).
How HCPs Are Paid
Having established that incentives derive from payment mechanisms it is useful to characterize the different mechanisms that typically apply to HCPs. By far the most dominant of these until recent decades (and still dominant in regard to dentists in many countries and doctors in some countries) is termed fee-for-service (FFS). This is an arrangement that is familiar to anyone who has used the services of an expert or professional and entails the service provider billing the client for the time and other expenses incurred on their behalf. In regard to the conditionality of payment this arrangement comes closest to a simple price mechanism and is analogous to the way in which many economic goods are transacted.
In the context of third-party payment there are two common alternatives. Under capitation the HCP is paid a sum for each patient under their care, independently of the time or other effort that is expended on each patient. The capitation payment is often determined in relation to the anticipated healthcare needs of a patient, or more usually the population of patients that the HCP serves. This calibration of the capitation payment is termed risk-adjustment (McClure, 1984). Under what is termed salary the HCP becomes analogous to an employee and receives reimbursement for their time, independently of either the treatments they deliver or the number of patients they are caring for. For the reasons previously set out it is best to regard a salary payment as analogous to, rather than the equivalent of, employment, because these payments can be made to the HCP as a self-employed individual.
The mechanisms under which HCPs are paid have proliferated in recent years to include combinations of FFS and capitation with sometimes the HCP able to choose or select the proportionality of the different elements. A further development has been to add specific components of payment to either existing FFS or capitation payment where the HCP receives an additional payment for carrying out a particular procedure or test on a type of patient deemed to be at risk. These latter arrangements are often termed pay-for-performance (P4P) (Eijkenaar, 2013; Greene, 2013; Kantarevic & Kralj, 2013; Sherry, 2016). More discussed than implemented, there has most recently been interest in linking payment to the outcome of treatment reflecting a long-running aspiration to focus on what healthcare treatment achieves (Epstein, 1990; Leonard & Zivin, 2005).
Conditionality of payment determines what needs to be established in order for payment to be claimed, so that these different payment systems have differing information requirements. This can be thought about in terms of what evidence would have to be offered to a court in order to establish that payment was due. Under FFS HCPs would have to establish what treatments they carried out, under capitation it would be necessary to establish how many and what kinds of patients were under their care, and for salary it would be necessary only to establish that they were available to be consulted.
The Link Between Payment and Performance
When discussing the performance of HCPs there are number of elements to consider. The overall level of service is one key concern, that is, whether the HCP delivers an appropriate number and mix of treatments. In healthcare as with many services the concept of quality is important but elusive and clearly a critical aspect of performance. The fact that quality might itself have many elements, such as the general care the HCP takes, the degree of explanation they offer, the physical environment they establish to deliver treatment, and so on, gives rise to particular concerns for incentives that are discussed under the heading risks and unintended consequences. Where the HCP also refers patients to other specialists the frequency and appropriateness of referral is another aspect of performance.
Relating these different aspects of performance to financial incentives requires establishing how HCPs might choose to vary their behavior in response to different payment mechanisms. One approach to this is essentially based upon intuition and introspective reasoning. There is a long-established and very well-documented narrative in regard to FFS. Under FFS HCPs can increase their payment by carrying out more treatment; thus, this remuneration method has long been associated with an incentive to treat more. This concern forms the basis for a long-running debate concerning whether HCPs over-treat their patients on account of FFS, especially when they come under pressure on their income due to competition. This has spawned an area of inquiry in health economics termed supplier-induced demand (Evans, 1974; Labelle, Stoddart, & Rice, 1994). The impact of FFS on other aspects of performance is less intuitively clear. For example, it is possible that the volume of treatment is closely related to its effectiveness, in which case FFS might be conducive to quality, but on the other hand it may be that delivering more treatments quickly compromises quality. Referring patients to specialists might entail forgoing treatment that could be charged for, but on the other hand it might free up time to treat more patients and so on.
Regarding capitation and salary arrangements the reasoning is often the reverse of that for FFS. The concern is then that HCPs might under-provide treatment and skimp (Ellis & McGuire, 1986) or stint (Newhouse, 1998) on healthcare. These concerns in part explain the focus on risk-adjustment of capitation payments. Again, the intuition regarding other aspects of performance is less clear cut, although it is often argued that capitation and salary might induce increased referrals to specialists as a means of off-loading the work required to treat patients (Iversen & Lurås—chapter 26 in Jones, 2012). Against this backdrop the developments in payment mechanisms can be seen as responses to the perceived failures of existing mechanisms to provide appropriate incentives.
Economics has made considerable contributions in terms of subjecting this kind of intuitive reasoning regarding the impact of incentives on performance to both rigorous critical appraisal and empirical investigation. One approach that is taken, and exemplified in countless academic contributions, is to consider a utility maximizing HCP choosing some dimensions of performance and to consider how different payment mechanisms will modify the optimizing decisions. This approach typically highlights the fact that even intuitively plausible dependencies of performance on payment often require stringent assumptions in order to hold unambiguously. It therefore suggests that empirical investigation is necessary to establish whether such effects operate in practice.
This is a prolific area of literature. Some starting points on conceptual models are the literature on supplier-induced demand, the developing literature on reimbursement mechanisms such as in Ellis and McGuire (1986) and McGuire (2000)—which provides a summary of the approach up to that point—and then more recent contributions such as Barham and Milliken (2015). In regard to empirical investigation, the rapid development of methods for establishing causal relationships using observational data from natural experiments (Dunning, 2012) has led to a substantial expansion of the literature in recent years. For examples, see Dumont, Fortin, Jacquemet, and Shearer (2008), Kantarevic and Kralj (2013), Rudoler et al. (2015), van Dijk et al. (2014), Li, Hurley, DeCicca, and Buckley (2014), and Erickson, Winkelmayer, Chertow, and Bhattacharya (2014). Recent developments in the use of experimental methods in economics have also led to controlled experimental investigations of these issues as well (Godager & Wiesen, 2013). One robust finding from these disparate empirical studies is that FFS is associated with increased treatment. A second is that differences in treatment induced by payment systems alone are often quite modest and difficult to disentangle from the effects of HCPs selecting between payment systems and differences in their case-mix of patients.
Economic Approach to Incentives
While there has been substantial progress in establishing the potential and actual linkage between different payment systems and the performance of HCPs, the approach has been piecemeal. Specific examples of payment mechanisms are modeled and compared, predictions are made, and evidence is sought. Economics has contributed to the development of thinking about incentives and performance for HCPs in a more fundamental way by pursuing a different research agenda. Rather than consider how different payment mechanisms might impact on HCP performance this approach sets out to ask questions regarding what sorts of payment mechanism are appropriate for dealing with the fundamental problems involved in delivering healthcare. This approach of necessity requires a high level of abstraction, but its benefits are substantial.
The starting point for understanding this alternative approach is to view healthcare delivery as a problem of delegation. The patient or the third-party payer wants certain things done but must rely on the HCP to do those things. What is in the interests of the HCP to do depends not only on their preferences and objectives but also upon how they are paid—the incentives that are built into the payment system. Can the third-party payer ensure that the HCP does what it wants? What payment mechanism will ensure that? These are the questions that are addressed by the branches of economics concerned with incentives (Laffont & Tirole, 1993), mechanism design (Mookherjee, 2008), and contracts (Bolton & Dewatripont, 2005).
The standard paradigm utilizes a principal-agent framework (Laffont & Martimort, 2001) and casts the HCP as an agent and the third-party payer as the principal. Depending on the setting the principal may be a public funder or a private insurance company and may therefore, to differing degrees, reflect the preferences of a patient. This allows for the possibility that the HCP is altruistic in the sense of being concerned with patient well-being and yet may still not act in the principal’s interests. For a given specification of the HCP’s objectives and a given structure of payment the choices of the HCP—their performance—can typically be derived as the solution to a constrained optimization problem. The third-party payer can then be conceived as choosing the payment system and its associated parameters so as to best achieve its objectives. By convention the principal’s choice of payment system in this setting is referred to as a contract, because it becomes important to constrain these arrangements in a way that they could subsequently be enforced, typically by a court.
A fundamental question in this framework is whether the third-party payer can specify a payment system that achieves the same result as if it did not have to rely on the HCP. This is by convention termed the first-best. Because by design, and reflecting the practical obstacles that surround the delivery of healthcare when HCPs have better information than third-party payers, it is often necessary to fall short of the ideal, the focus then becomes describing what second-best (Lipsey & Lancaster, 1956) performance entails and how payment can be arranged to deliver it.
Within this principal-agent/contract/mechanism (PACM) paradigm two sources of failure to achieve first-best outcomes are a particular focus. Where the agent’s performance is not observed, their choices are hidden from the principal. Such hidden actions are one potential source of failure. Actions occur once any agreement between the principal and agent is being implemented, but there may be asymmetric information even before the execution of any contract. Hidden information is thus a second source of failure. Many variations on this theme are possible, and more recent theories have evolved to consider the simultaneous presence of hidden action and hidden information. These terms and their interpretation have particular resonance when considering interactions with, and the performance of, HCPs. The care taken during procedures when the patient is incapacitated or unconscious is a hidden action. The exact medical condition of a patient is something only the HCP might be able to determine—it is hidden information, and so on.
The PACM paradigm is a very powerful and flexible tool for understanding the nexus between incentives and performance for HCPs. Like any such tool its effective use requires skill and experience. Rather than illustrate those uses by examples from the hands of experts, a simple illustration of this paradigm in action is presented. It applies it to consider how and when a third-party payer might wish to augment FFS with a payment that reflects an imperfect assessment of the hidden action or hidden information of an HCP.
Illustration of Designing Incentives
A Hidden-Action Problem
This example follows in the tradition of Holmstrom (1979) in considering how conditioning payment on imperfect information can potentially improve outcomes. In this simplified model it is assumed that the HCP can choose between two treatments labeled g or b. Good treatment g represents a careful and diligent examination of the patient, whereas bad treatment b, which appears superficially the same, is an inferior and careless treatment. Assume that the HCP is purely selfishly motivated and derives utility where p is the payment they receive and is the disutility of carrying out the treatment. It is reasonable to suppose that .
In this setting FFS implies choosing the payment p for treatment and in the spirit of making the issue at stake as stark as possible it follows that whatever payment is chosen, the HCP will choose the treatment provided that they are better off treating the patient than not. In this case and assuming that the HCP has no alternative (i.e., they will participate in treatment provided only ) then the best the payer can do is to set . Paying any more will simply increase the cost of care without improving the quality of treatment, while paying any less will involve the patient going untreated, because the HCP will not agree to the contract and will walk away.
Now suppose that the payer could ask patients about the quality of their treatment and get an assessment of that. Bearing in mind that patients are often poorly informed about medical practice we should not expect that assessment to be 100% correct. To capture this in a simple way suppose there are three types of patients. Type 1 is discerning and always knows the sort of treatment they have received. Type 2 is optimistic and always regards treatment as good, and type 3 is pessimistic and always reports treatment as bad. For simplicity assume that these types of patients occur with equal frequency and that neither the HCP nor the payer can determine a patient’s type.
The important point is that by asking patients for an assessment of the treatment they have received the payer gets a signal of the quality of that treatment and by basing payment on that signal they can induce the HCP to provide good treatment. Here is how it could work in practice. If the HCP provides good treatment, then on the assumptions made so far on 2/3 of occasions they will receive a favorable patient report. If they provide bad treatment then on 2/3 of occasions they will receive an unfavorable report. Now suppose the payer makes the payment depend on that report and sets a FFS of conditional on observing a favorable report and conditional on an unfavorable report. If the HCP chooses good treatment their utility is while for bad treatment it is .
The remainder of the payment design process is simply to ensure that the incentive is sufficient, and then to achieve delivery at the lowest cost. The incentive will be sufficient if the utility derived from choosing the good treatment exceeds that from choosing the bad treatment: . This in the language of the paradigm is called the incentive-compatibility constraint. Delivery will be as cheap as possible if , which is called the individual rationality constraint, because it reflects the need to pay enough such that a rational HCP will agree to the contract.
This model establishes that even with noisy information the third-party payer can achieve the good treatment that it wants. So it acts as a proof of concept of the effectiveness of incentives in this setting. A little further investigation can establish that the payer has to pay more than it would if it could actually observe the quality of treatment given (the first-best). Further analysis can establish what that extra payment depends upon and how it might be mitigated. In other words the model provides a guide to the design of an optimal incentive payment.
The Hidden-Information Case
Changing notation and interpretation of the model can illustrate a hidden-information incentive problem. Re-label the two treatments as “simple” s or “intensive” i and again assume these cannot be distinguished from each other either by the payer or patients, but now consider there to be just two types of patient, with type 1 being amenable to simple treatment while type 2 is amenable only to intensive treatment. The HCP can determine a patient’s type, but the payer cannot. Here the fundamental problem of ensuring the appropriate quality of care is different from that previously discussed. The payer might reasonably want type 1 patients to receive treatment s but type 2 patients to receive treatment i, but if it pays a fixed fee for treatment the HCP, who again is assumed to be purely financially motivated so as to make the problem stark, that HCP will deliver low-cost, simple treatment to all patients.
Again, suppose that patients can assess the outcome of treatment and report that, but their reports are imperfect. In this case suppose that patients report the improvement they experience following treatment and that these reports are made with the following probabilities:
Table 1. Probabilities of Different Reports According to the Treatment Received
Type 1 Patient
Type 2 Patient
Who receives s treatment
Who receives i treatment
Who receives s treatment
Who receives i treatment
reports favorably with probability
reports unfavorably with probability
Once again the HCP can be given an incentive to treat patients in the way that the payer desires by conditioning payment on the patients’ reports. Indeed the algebra is the same as before; it is only the description and interpretation that have changed. As long as the HCP will prefer to treat type 2 patients with treatment i. It is both cheaper and more likely to result in a favorable report to treat type 1 patients according to s. So this condition defines incentive compatibility while setting prices so that will ensure that the HCP finds the arrangement worthwhile.
In this setting by appropriately conditioning payment on patient reports the payer has induced the HCP to act appropriately upon their hidden information. The payer has ensured that different types of patients receive the appropriate treatment even though it cannot establish what type a patient is.
Both cases illustrate that careful conditioning of payment can create incentives to improve performance even in circumstances where HCPs are at a natural advantage relative to payers in regard to what they know. Generalizations of this approach have generated a wealth of insights into both the possibilities for enhancing performance and some of the risks that might arise through taking a broad approach of “pay for what you observe.”
Risks and Unintended Consequences of Incentives
There is a certain allure in the idea that if the payer can observe some aspect of performance, even imperfectly, then payment should be made conditional on that so as to enhance performance. The PACM paradigm, however, provides a reason to be cautious. In a hidden-action setting there may be many aspects of performance that cannot be observed at all. While it is intuitively appealing to condition payment on those that can be observed, the problem is that the agent’s performance on the remaining dimensions of performance might deteriorate. This issue had been formalized in what is termed the multitasking problem (Holmstrom & Milgrom, 1991; Eggleston, 2009).
In essence the multitasking problem is endemic because even quality of healthcare is likely to have many dimensions, not all of which will be capable of being monitored. In the prior model, for example, suppose that the HCP can influence reports by offering some cheap but unwarranted treatment. Then the existence of a payment based on patient reports has created an incentive to produce the wrong treatment. It has recently been suggested that the expansion of prescription drugs, especially opioid painkillers, might be accounted for by doctors providing medication in order to please their patients. Whether multitasking is a reason to abandon incentive payments depends on what those payments are made for, and whether the performance that is being remunerated occurs at the expense of, or in cooperation with, other aspects of performance (Kaarboe & Siciliani, 2011).
More generally incentive payments give rise to the risk that HCPs will seek to manipulate the performance measures upon which they are rewarded. This manipulation is frequently referred to as gaming and is discussed extensively in the literature based on the PACM paradigm (Baker, 1992). Even the illustrative example of an incentive scheme could be susceptible to such manipulation. Suppose that the HCP can influence patients’ reports by choosing which patients report. She may then be able to generate favorable reports without dispensing the appropriate treatment. One response to gaming is to engage in more careful monitoring of performance measures—perhaps a sample of patients could be interviewed regarding their treatment or whether the HCP had tried to influence their report, for example. Monitoring is, however, costly, and the relative ease of gaming might mean that the cost is not worth incurring given the best performance gain that can be achieved through incentives. Gaming opportunities arise from precisely the kinds of information asymmetry that motivate the use of performance incentives, and so it is possible to build such risks into the design of the payment system. The presence of gaming opportunities suggests that incentives might be restricted and is often argued to militate against the use of high-powered incentives in which agents receive a large proportion of their reward from conditional payments. For HCPs this suggests that P4P might be only a small component of overall remuneration (Gaynor, Rebitzer, & Taylor, 2004).
If gaming is viewed as an artifact of residual information asymmetry in respect to hidden actions, its corollary in respect to hidden information is selection. Here the concern is that when faced with incentive payments HCPs might focus their activity choosing who to treat in ways that increase payment but are not intended or wanted by the payer. Positively selecting low-cost and easy-to-treat patients has been termed cream-skimming Ellis, 1998), while shunning high-cost patients is referred to as dumping (Ma, 1994; Ellis, 1998). One response to such issues is to refine payment so as to distinguish between different types of patients, and that is one rationale for seeking better proxies for the costs of treatment in the form of risk adjusters. Once again this solution may entail costs, both in terms of complexity of payment design and in supporting the relevant information systems, such that the payer is better off not trying to make incentive payments.
These kinds of issues arise naturally within the PACM framework. There are also, however, concerns that financial incentives that are fully rational in that context and that properly account for multitasking, gaming, and selection may, nevertheless, result in poor outcomes.
Financial and Intrinsic Incentives
The idea that HCPs have a number of moderating influences on otherwise selfish behavior was discussed in the Incentives in Healthcare. At one level such influences are easily incorporated into the PACM paradigm in that they suggest different forms of objective function for HCPs. The most common way of incorporating this is by assuming that HCPs have a degree of altruism (Jack, 2005; Godager & Wiesen, 2013).
This approach broadly suggests that the degree of altruism will influence the design but not the underlying rationale for incentive payments, as can be seen even in the previous simple model—adding some element of patient benefit into the utility function of the HCP will change the report-contingent fees that the payer will offer, but it does not suggest that a contingency payment mechanism will be abandoned.
In the domain of education, however, it has long been argued that explicit rewards for attainment may undermine what has been termed intrinsic motivation (Jordan, 1952). In this view financial incentives, which are referred to as extrinsic motivation, may compete with or crowd out intrinsic motivation. There had been considerable support from both behavioral science and psychology for this view (Deci, Koestner, & Ryan, 1999).
Standard economic models of incentive such as those within the PACM framework treat preferences and motivation as exogenous, which makes the incorporation of these effects problematic. Bénabou and Tirole (2003) offer a potential reconciliation of competing intrinsic and extrinsic motivation within an extension of a PACM model, to consider under what circumstances crowding out might occur, but there has not yet been any translation of these ideas to apply them to the case of HCPs.
If one takes a step back from the detailed analysis of incentives and performance undertaken by economists, how does the landscape of payment mechanisms for HCPs appear?
In spite of the very longstanding criticism and concern that it may result in excess treatment and cost, FFS prevails in many healthcare systems and for many professionals (Schroeder & Frist, 2013; Haelle, 2018). Much policy discussion of capitation and salary payments makes reference to incentives, but these are often aspirations rather than practical implementations of P4P and where those schemes have been introduced the evidence regarding their effectiveness is mixed. As an example, the National Health Service in England introduced an extensive system of performance payments—the Quality and Outcomes Framework (QOF)—that has served as a natural experiment and whose analysis continues. It is noted as much for its unintended consequences such as gaming and selection as it is for its impact on performance (Gravelle, Sutton, & Ma, 2010). Similarly mixed experiences are reported across different jurisdictions and for different schemes (Conrad & Perry, 2009; Glazier, Klein-Geltink, Kopp, & Sibley, 2009; Kantarevic & Kralj, 2013; Nassiri & Rochaix, 2006; Rogowski, 2013, Scott, Schurer, Jensen, & Sivey, 2009; Van de Poel, Flores, Ir, & O’Donnell, 2016; Greene, 2013; Gosden et al., 2000).
This contrasts starkly with the relatively successful role that economic incentive arguments have played in the process of reforming payments to healthcare organizations and in particular the adoption of prospective payment mechanisms for acute hospital care and the associated setting of fixed prices for complete episodes of care, in place of reimbursement for treatment costs. Following the inception of Diagnostic Related Groups (DRGs) as the basis of payment for U.S. Medicare in the early 1980s (Baker, 2002) similar systems have been adopted across many healthcare systems (Busse et al., 2011) and have generally been found to increase efficiency, through reduced duration of treatment, without adverse impact on quality of care (Farrar et al., 2009).
Economic concepts have certainly been instrumental in focusing attention on the potential importance of incentives and performance for HCPs. The basic economic approach of considering choice, subject to constraints, suggests that there are many aspects of performance, that HCPs have considerable, discretion being relatively unconstrained by conventional market mechanisms, and that evidence supports the view that this discretion is sometimes associated with undesirable consequences—such as cost inflation.
In addition to providing a framework for identifying specific problems and guiding the interpretation of evidence, economics has a rich framework in the PACM paradigm for suggesting remedies and indicating the risks and benefits of pursuing those. This is indeed the process behind the refinement and adoption of DRG systems, but it seems to have been less successful in respect to incentives for HCPs where take-up has been slow, patchy, and not-convincingly successful.
Economics provides some reasons for that lack of take up. Intuitive approaches to adopting incentive payments are fraught with risks; paying for one aspect of performance may compromise other aspects; payment schemes are subject to manipulation, and there may be unintended and undesirable consequences in regard to cream-skimming and dumping. So one reason for the relative lack of successful incentive payments for HCPs might simply be that it is very difficult to design such schemes and we have not yet found the right formula. It may also be the case that there is no right formula, or that the formula is so idiosyncratic and particular to the circumstances (the motivation of HCPs, the institutional structures they work within, the healthcare financing environment, and so on) that we cannot expect to see general adoption of simple P4P mechanisms. In that case FFS, capitation, and salary arrangements will be subject to small but important additions in terms of incentive payments in the future.
Economic models within the PACM paradigm are not intended to be a blueprint for practical policy implementation. The role of these models is to explore and highlight trade-offs, and to explore the potential limits of economic mechanisms for delivering performance. Nevertheless there is often a large gap between what economics suggests is optimal and what can be practically achieved. This is especially true in regard to publicly financed healthcare. For example, economic arguments for at least some degree of patient cost sharing are very strong but remain an anathema and politically toxic in many healthcare systems, especially in the United Kingdom (see, e.g., Campbell, 2014). The same may be true of incentive payments for HCPs who remain largely trusted and where the implementation of such payments may undermine public trust. Furthermore, even relatively straightforward features of incentive payments might be hard to introduce in practice. The simple model outlined here is not controversial from the perspective of economic analysis. It logically suggests that an HCP could receive a payment that varies with the reports of their patients and is built on the premise that such reports are imperfect. It thus entails sometimes paying the HCP a low price even though they have delivered high quality care, or treated a difficult patient. That randomness of payment is optimal, but it may not be acceptable.
Rather more fundamentally the possibility that extrinsic incentives crowd out intrinsic motivation might be a permanent obstacle to the widespread adoption of incentive payments for HCPs. The state of knowledge regarding what motivates professionals is still evolving, and the translation from motivation to action may be much less mechanistic than the rational optimization of an exogenous objective function that forms the foundation of the economist’s toolkit in this regard. Along with other areas of economic science, health economics and the study of incentives for HCPs are likely to be influenced more in the future by interaction with behaviorists, psychologists, and experimentalists, in an effort to better understand the evolution and determinants of motivation and the trade-off between intrinsic and extrinsic incentives.
Research in the area of incentives and performance is constantly evolving and developing. From the theoretical perspective, future work is likely to draw on the rapidly expanding fields of contract theory and mechanism design and apply emerging analysis to the particular circumstances influencing the performance of HCPs. There may be a need in particular to extend the PACM approach to better integrate its approach with the insights of behavioral economics and psychology. Beyond that, for the reasons set out it would seem important to bridge the gap between theory and application and perhaps integrate an understanding of implementation into the political economy landscape of healthcare. Finally, and as with other areas of health economics, empirical research that establishes causal linkages between incentives and performance utilizing both natural and laboratory experiments is likely to be an important part of the research agenda in incentives and performance of HCPs.
Andersen, L. B. (2009). What determines the behaviour and performance of health professionals? Public service motivation, professional norms and/or economic incentives. International Review of Administrative Sciences, 75, 79–97.Find this resource:
Andersen, L. B., & Pedersen, L. H. (2012). Public service motivation and professionalism. International Journal of Public Administration, 35, 46–57.Find this resource:
Arrow, K. J. (1963). Uncertainty and the welfare economics of medical care. American Economic Review, 53, 941–973.Find this resource:
Ash, E., & MacLeod, W. B. (2015). Intrinsic motivation in public service: Theory and evidence from state supreme courts. Journal of Law and Economics, 58, 863–913.Find this resource:
Baker, G. P. (1992). Incentive contracts and performance measurement. Journal of Political Economy, 100, 598–614.Find this resource:
Baker, J. J. (2002). Medicare payment system for hospital inpatients: Diagnosis-related groups. Journal of Health Care Finance, 28, 1–13.Find this resource:
Barham, V., & Milliken, O. (2015). Payment mechanisms and the composition of physician practices: Balancing cost-containment, access, and quality of care. Health Economics, 24, 895–906.Find this resource:
Bénabou, R., & Tirole, J. (2003). Intrinsic and extrinsic motivation. Review of Economic Studies, 70, 489–520.Find this resource:
Bolton, P., & Dewatripont, M. (2005). Contract theory. Cambridge, MA: MIT Press.Find this resource:
Busse, R., Geissler, A., Quentin, W., & Wiley, M. (2011). Diagnosis-related groups in Europe: Moving towards transparency, efficiency and quality in hospitals. Maidenhead, UK: Open University Press.Find this resource:
Callison, K. (2016). Medicare managed care spillovers and treatment intensity. Health Economics, 25, 873–887.Find this resource:
Campbell, D. (2014, March 29). Britons overwhelmingly against GP charges to help NHS balance books. Guardian.Find this resource:
Choné, P., Coudin, A., & Pla, A. (2014). Are physician fees responsive to competition? Health, Econometrics and Data Group (HEDG) Working Papers 14/20. HEDG, Department of Economics, University of York.Find this resource:
Conrad, D. A., & Perry, L. (2009). Quality-based financial incentives in health care: Can we improve quality by paying for it?. Annual Review of Public Health, 30, 357–371.Find this resource:
de Pouvourville, G. (2013). Paying doctors for performance. European Journal of Health Economics, 14, 1–4.Find this resource:
Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125(6), 627–668.Find this resource:
Devlin, R. A., & Sarma, S. (2008). Do physician remuneration schemes matter? The case of Canadian family physicians. Journal of Health Economics, 27, 1168–1181.Find this resource:
Dumont, E., Fortin, B., Jacquemet, N., & Shearer, B. (2008). Physicians’ multitasking and incentives: Empirical evidence from a natural experiment. Journal of Health Economics, 27, 1436–1450.Find this resource:
Dunning, T. (2012). Natural experiments in the social sciences: A design-based approach. Cambridge, UK: Cambridge University Press.Find this resource:
Eggleston, K. (2009). Provider payment incentives: International comparisons. International Journal of Health Care Finance and Economics, 9, 113–115.Find this resource:
Eijkenaar, F. (2013). Key issues in the design of pay for performance programs. European Journal of Health Economics, 14, 117–131.Find this resource:
Ellis, R. P. (1998). Creaming, skimping and dumping: Provider competition on the intensive and extensive margins. Journal of Health Economics, 17, 537–555.Find this resource:
Ellis, R. P., & McGuire, T. G. (1986). Provider behavior under prospective reimbursement: Cost sharing and supply. Journal of Health Economics, 5, 129–151.Find this resource:
Ellis, R. P., & McGuire, T. G. (1988). Insurance principles and the design of prospective payment systems. Journal of Health Economics, 7, 215–237.Find this resource:
Epstein, A. M. (1990). The outcomes movement—will it get us where we want to go? New England Journal of Medicine, 323, 266–270.Find this resource:
Erickson, K. F., Winkelmayer, W. C., Chertow, G. M., & Bhattacharya, J. (2014). Medicare reimbursement reform for provider visits and health outcomes in patients on hemodialysis. Forum for Health Economics and Policy, 17, 53–77.Find this resource:
Evans, R. G. (1974). Supplier-induced demand: Some empirical evidence and implications. In M. Perlman (Ed.), The economics of health and medical care: Proceedings of a Conference held by the International Economic Association at Tokyo. International Economic Association Series. London: Palgrave Macmillan.Find this resource:
Farley, P. J. (1986). Theories of the price and quantity of physician services: A synthesis and critique. Journal of Health Economics, 5, 315–333.Find this resource:
Farrar, S., Yi, D., Sutton, M., Chalkley, M., Sussex, J., & Scott, A. (2009). Has payment by results affected the way that English hospitals provide care? Difference-in-differences analysis. British Medical Journal, 339, b3047.Find this resource:
Gaynor, M., Rebitzer, J. B., & Taylor, L. J. (2004). Physician incentives in health maintenance organizations. Journal of Political Economy, 112, 915–931.Find this resource:
Gibbons, R., & Roberts, J. (2013). Economic theories of incentives in organizations. In R. Gibbons & J. Roberts (Eds.), The handbook of organizational economics (pp. 56–99). Princeton, NJ: Princeton University Press.Find this resource:
Glazier, R. H., Klein-Geltink, J., Kopp, A., & Sibley, L. M. (2009). Capitation and enhanced fee-for-service models for primary care reform: A population-based evaluation. Canadian Medical Association Journal, 180, E72–E81.Find this resource:
Glied, S. (2000). Managed Care. In A. J. Culyer & J. P. Newhouse (Eds.), Handbook of health economics (pp. 707–753). Amsterdam, The Netherlands: Elsevier.Find this resource:
Glied, S., & Zivin, J. G. (2002). How do doctors behave when some (but not all) of their patients are in managed care? Journal of Health Economics, 21, 337–353.Find this resource:
Godager, G., & Wiesen, D. (2013). Profit or patients’ health benefit? Exploring the heterogeneity in physician altruism. Journal of Health Economics, 32, 1105–1116.Find this resource:
Gosden, T., Forland, F., Kristiansen, I., Sutton, M., Leese, B., Giuffrida, A., . . . Pedersen, L. (2000). Capitation, salary, fee-for-service and mixed systems of payment: Effects on the behaviour of primary care physicians. Cochrane Database of Systematic Reviews, 2000(3), CD002215.Find this resource:
Gravelle, H., Sutton, M., & Ma, A. (2010). Doctor behaviour under a pay for performance contract: Treating, cheating and case finding? Economic Journal, 120, F129–F156.Find this resource:
Greenberg, W. (2002). Competition in the health care sector: Past, present, and future: Proceedings of a conference sponsored by the Bureau of Economics, Federal Trade Commission, March 1978. Washington, DC: Beard Books.Find this resource:
Greene, J. (2013). An examination of pay-for-performance in general practice in Australia. Health Services Research, 48, 1415–1432.Find this resource:
Grytten, J., & Sørensen, R. (2007). Primary physician services—list size and primary physicians’ service production. Journal of Health Economics, 26, 721–741.Find this resource:
Haelle, T. (2018). Fee-for-service still dominates in United States. Medscape.Find this resource:
Hillman, A. L. (1987). Financial incentives for physicians in HMOs. New England Journal of Medicine, 317, 1743–1748.Find this resource:
Holmstrom, B. (1979). Moral hazard and observability. Bell Journal of Economics, 10, 74–91.Find this resource:
Holmstrom, B., & Milgrom, P. (1991). Multitask principal-agent analyses: Incentive contracts, asset ownership, and job design. Journal of Law, Economics, & Organization, 7, 24–52.Find this resource:
Iversen, T., & Luras, H. (2000). The effect of capitation on GPs’ referral decisions. Health Economics, 9, 199–210.Find this resource:
Jack, W. (2005). Purchasing health care services from providers with unknown altruism. Journal of Health Economics, 24, 73–93.Find this resource:
Jian, W., & Guo, Y. (2009). Does per-diem reimbursement necessarily increase length of stay? The case of a public psychiatric hospital Health Economics, 18, S97–S106.Find this resource:
Johansson, V., Axtelius, B., Söderfeldt, B., Sampogna, F., Lannerud, M., & Sondell, K. (2007). Financial systems’ impact on dental care: A review of fee-for-service and capitation systems. Community Dental Health, 24, 12–20.Find this resource:
Jones, A. M. (2006). The Elgar companion to health economics. Cheltenham, UK: Edward Elgar.Find this resource:
Jones, A. M. (2012). The Elgar companion to health economics. Cheltenham, UK: Edward Elgar.Find this resource:
Jordan, A. M. (1952). Intrinsic versus extrinsic motivation in learning. High School Journal, 36, 76–83.Find this resource:
Kaarboe, O., & Siciliani, L. (2011). Multi-tasking, quality and pay for performance. Health Economics, 20, 225–238.Find this resource:
Kantarevic, J., & Kralj, B. (2013). Link between pay for performance incentives and physician payment mechanisms: Evidence from the Diabetes Management Incentive in Ontario. Health Economics, 22, 1417–1439.Find this resource:
Kantarevic, J., & Kralj, B. (2016). Physician payment contracts in the presence of moral hazard and adverse selection: The theory and its application in Ontario. Health Economics, 25, 1326–1340.Find this resource:
Khalil, F., & Lawarrée, J. (1995). Input versus output monitoring: Who is the residual claimant? Journal of Economic Theory, 66, 139–157.Find this resource:
Kristensen, S. R. (2017). Financial penalties for performance in health care. Health Economics, 26, 143–148.Find this resource:
Labelle, R., Stoddart, G., & Rice, T. (1994). A re-examination of the meaning and importance of supplier-induced demand. Journal of Health Economics, 13, 347–368.Find this resource:
Laffont, J. J., & Martimort, D. (2001). The theory of incentives: The principal-agent model. Princeton, NJ: Princeton University Press.Find this resource:
Laffont, J. J., & Tirole, J. (1993). A theory of incentives in procurement and regulation. Cambridge, MA: MIT Press.Find this resource:
Leonard, K. L., & Zivin, J. G. (2005). Outcome versus service based payments in health care: Lessons from African traditional healers. Health Economics, 14, 575–593.Find this resource:
Li, J., Hurley, J., DeCicca, P., & Buckley, G. (2014). Physician response to pay-for-performance: Evidence from a natural experiment. Health Economics, 23, 962–978.Find this resource:
Lipsey, R. G., & Lancaster, K. (1956). The general theory of second best. Review of Economic Studies, 24, 11–32.Find this resource:
Ma, C. A. (1994). Health care payment systems: Cost and quality incentives. Journal of Economics and Management Strategy, 3, 93–112.Find this resource:
McClure, W. (1984). On the research status of risk-adjusted capitation rates. Inquiry, 21(3), 205–213.Find this resource:
McGuire, T. G. (2000). Physician agency. In A. J. Culyer & J. P. Newhouse (Eds.), Handbook of health economics (pp. 461–536). Amsterdam, The Netherlands: Elsevier.Find this resource:
Meacock, R., Kristensen, S. R., & Sutton, M. (2014). The cost-effectiveness of using financial incentives to improve provider quality: A framework and application. Health Economics, 23, 1–13.Find this resource:
Melichar, L. (2009). The effect of reimbursement on medical decision making: Do physicians alter treatment in response to a managed care incentive? Journal of Health Economics, 28, 902–907.Find this resource:
Mishan, E. J. (1957). A re-appraisal of the principles of resource allocation. Economica, 24, 324–342.Find this resource:
Mookherjee, D. (2008). The 2007 Nobel Memorial Prize in mechanism design theory. Scandinavian Journal of Economics, 110, 237–260.Find this resource:
Morris, S., Goudie, R., Sutton, M., Gravelle, H., Elliott, R., Hole, A. R., . . . Skåtun, D. (2011). Determinants of general practitioners’ wages in England. Health Economics, 20, 147–160.Find this resource:
Nassiri, A., & Rochaix, L. (2006). Revisiting physicians’ financial incentives in Quebec: A panel system approach. Health Economics, 15, 49–64.Find this resource:
Nelson, P. (1970). Information and consumer behavior. Journal of Political Economy, 78, 311–329.Find this resource:
Newhouse, J. P. (1998). Risk adjustment: Where are we now? Inquiry, 35, 122–131.Find this resource:
Olsen, K. R. (2012). Patient complexity and GPS’ income under mixed remuneration. Health Economics, 21, 619–632.Find this resource:
Oyer, P., & Schaefer, S. (2010). Personnel economics: Hiring and incentives. Working Paper No. 15977. National Bureau of Economic Research.
Quast, T., Sappington, D. E. M., & Shenkman, E. (2008). Does the quality of care in Medicaid MCOs vary with the form of physician compensation? Health Economics, 17, 545–550.Find this resource:
Reeves, D., Doran, T., Valderas, J. M., Kontopantelis, E., Trueman, P., Sutton, M., . . . Lester, H. (2010). How to identify when a performance indicator has run its course. BMJ, 340(2010), c1717.Find this resource:
Roberts, J. (2010). Designing incentives in organizations. Journal of Institutional Economics, 6, 125–132.Find this resource:
Rochaix, L. (1993). Financial incentives for physicians: The Quebec experience. Health Economics, 2, 163–176.Find this resource:
Rogowski, W. (2013). An economic theory of the fourth hurdle. Health Economics, 22, 600–610.Find this resource:
Rudoler, D., Deber, R., Barnsley, J., Glazier, R. H., Dass, A. R., & Laporte, A. (2015). Paying for primary care: The factors associated with physician self-selection into payment models. Health Economics, 24, 1229–1242.Find this resource:
Schroeder, S. A., & Frist, W. (2013). Phasing out fee-for-service payment. New England Journal of Medicine, 368, 2029–2032.Find this resource:
Scott, A., Schurer, S., Jensen, P. H., & Sivey, P. (2009). The effects of an incentive program on quality of care in diabetes management. Health Economics, 18, 1091–1108.Find this resource:
Shafrin, J. (2010). Operating on commission: Analyzing how physician financial incentives affect surgery rates. Health Economics, 19, 562–580.Find this resource:
Sherry, T. B. (2016). A note on the comparative statics of pay-for-performance in health care. Health Economics, 25, 637–644.Find this resource:
Van de Poel, E., Flores, G., Ir, P., & O’Donnell, O. (2016). Impact of performance-based financing in a low-resource setting: A decade of experience in Cambodia. Health Economics, 25, 688–705.Find this resource:
van Dijk, C. E., Verheij, R. A., te Brake, H., Spreeuwenberg, P., Groenewegen, P. P., & de Bakker, D. H. (2014). Changes in the remuneration system for general practitioners: Effects on contact type and consultation length. European Journal of Health Economics, 15, 83–91.Find this resource:
Woodward, R. S., & Warren-Boulton, F. (1984). Considering the effects of financial incentives and professional ethics on “appropriate” medical care. Journal of Health Economics, 3, 223–237.Find this resource:
Wu, Y., Chen, Y., & Li, S. (2018). Optimal compensation rule under provider adverse selection and moral hazard. Health Economics, 27(3), 509–524.Find this resource:
Zweifel, P., Breyer, F., & Kifmann, M. (2009). Physicians as suppliers of medical services. In P. Zweifel, F. Breyer, & M. Kifmann, Health economics (pp. 293–309). Berlin: Springer.Find this resource: