3.1 Introduction
Justice Cardozo wrote that when serving as a fiduciary, “thought of self [is] to be renounced, however hard the abnegation.”Footnote 1
The American health care system’s reliance on fiduciary principles accelerated with the passage of the Employee Retirement Income Security Act of 1974 (ERISA).Footnote 2 After the failure of many employer pension plans in the 1950s and 1960s, Congress passed ERISA to protect workers.Footnote 3 ERISA did not require employers to offer employee benefits. Instead, ERISA incentivized the private sector to provide these benefits. However, the statute left employers with control over not only whether and how to provide benefits but also how any such benefits plans were to be administered. Through ERISA, legislators attempted to constrain employer power with statutory fiduciary duties by, for example, requiring plan fiduciaries (employed directly or contractually by the employer) to act “solely in the interest of the participants and beneficiaries” and with “the exclusive purpose of (i) providing benefits … and (ii) defraying reasonable expenses of administering the plan.”Footnote 4
In the following years, however, the Supreme Court effectively gutted any restraints that existed by providing employers and their affiliates serving as fiduciaries with a highly deferential standard of review when participants (i.e., employees) challenged their decisions.Footnote 5 Sections 502 and 514 of ERISA have allowed employee benefit plans (which provide more than half of Americans with health insurance) to opt in to that deferential standard of review when they deny benefit claims.Footnote 6 ERISA preemption spread the illusion that fiduciary duties could protect workers – but it went beyond their finances to their health.Footnote 7 However, fiduciary protections over health are a mirage. Employers pretend they can act in the best interests of employees when maintaining health plans – even though they pay (directly or indirectly) for claims.
Physicians and other providers are also considered fiduciaries.Footnote 8 They are obligated to act in the best interests of patients, regardless of financial incentives to reduce their standard of care. Though providers have always had an incentive to see more patients, the consolidation of health care has increased the risk of systemic conflicts because of reduced competition. Physicians pretend they can act in the best interests of patients – even when hospitals and group practices pressure them to see more patients and spend less time with them, and even though physicians and their employers are incentivized to order expensive tests and procedures. Cardozo’s fiduciary ideal has become even more strained because of corporate consolidation in health care and the pressure of skyrocketing health insurance costs.
But what if, instead of ignoring these conflicts of interest, the American health care system abandoned the fiduciary fallacy? Building on recent scholarship showing that targeted legislation, regulation, and self-policing by stakeholders – rather than reliance on broad fiduciary duties – better manage these conflicts, this chapter uses game theory to analyze different behavioral scenarios involving the actors responsible for financing and providing care.
In Section 3.2, I review the literature addressing the role of fiduciary duties in health care and how game theory has been used to study strategic behavior and legal constraints. In Section 3.3, I explain the “game” as it currently exists, recognizing that the game I present is a substantially simplified version of reality – though one that highlights why fiduciary duties do not adequately constrain providers and payors. Finally, in Section 3.4, I show how abandoning fiduciary duties in this context is paradoxically more likely to result in providers and payors behaving in the best interests of patients using the same, simple game.
My goal is to advance scholarship arguing that fiduciary duties are inadequate to address conflicts in health care, while also proposing paths forward. Proposals to better resolve conflicts in health care should incorporate how rational actors are likely to behave and react to different restrictions designed to constrain their behavior. Laws restricting specific financial conflicts, in combination with financial incentives based on patient outcomes, advance the goal.
3.2 Literature Review
3.2.1 Abandoning Fiduciary Duties
Recent work by Sam Halabi argues against “fiduciary utopianism,” or the idea that fiduciary law is the best way to regulate relationships of trust.Footnote 9 Recently, fiduciary protections have expanded beyond private relationships that impact “public or quasi-public interests – e.g., attorneys and clients, directors and shareholders, guardians and wards” to juries, taxpayers, and friends. Scholars have argued that fiduciary protections can “benefit [] those locked out or marginalized by prevailing economic and political structures.”Footnote 10 However, the fiduciary framework may be largely aspirational because fiduciaries in health care are hopelessly conflicted.
Moving beyond critiques seeking to limit the application of fiduciary duties in specific cases,Footnote 11 Halabi “argues that the application of fiduciary duties to more legal relationships distracts from lawmaking and law-enforcement processes that lead to more effective solutions to the problems that fiduciary duties are asserted to solve.”Footnote 12 In other words, fiduciary duties are the wrong tool in many relationships. In lawsuits not subject to ERISA preemption, common law tort and contract claims are preferred over fiduciary claims for their clarity and judicial conservatism (e.g., compensatory damages in a malpractice action over disgorgement for a fiduciary breach).Footnote 13
Moreover, fiduciary duties do not deter bad behavior. In fact, instead of acting as a deterrent, fiduciary duties “may instead encourage [misbehavior] by weakening executives’ internal moral constraints.”Footnote 14 Fiduciary duties change the calculation of human decision-making because (1) otherwise immoral acts move within the range of justifiable action if not prohibited by fiduciary duties (including due to lack of enforcement), and (2) people engage in self-interested behavior by (roughly) weighing costs and benefits when choosing an action within that range.Footnote 15 As courts allow more self-interested behavior in spite of statutory or common law fiduciary duties through deference to fiduciaries, the range of actions that can be justified expands and more fiduciaries choose self-interested behavior.Footnote 16
Fiduciary duties may be counter to broader policies or social goals. In the corporate world, fiduciary duties are “designed to align an officer’s interests with those of the shareholders [instead of with society’s].…If the alignment is deep, it encourages executives to follow shareholder interests across ethical thresholds in pursuit of profit.”Footnote 17 Similarly, the rules governing employer-sponsored insurance are designed to align the interests of providers and administrators with the employers’ and insurers’ – not with participants’ and beneficiaries’ interests.
3.2.2 How to Play the Game
Game theory is “a powerful tool to model and understand complex interactions.”Footnote 18 In 1994, Douglas Baird, Robert Gertner, and Randal Picker published Game Theory and the Law.Footnote 19 Despite making a significant impact on other areas of the law, the application of game theory to health law has been limited.Footnote 20
In game theory, a “player” employs a “strategy” each time she acts.Footnote 21 Assumptions about the player’s motivations or the context in which the decisions are made must be clearly explained when discussing the game.
Games can be either “cooperative” or “noncooperative,” with further subclassifications within each category. Cooperative games model behavior by a group of decision-makers, while noncooperative games model behavior of individual decision-makers who cannot consult each other.Footnote 22 The Prisoner’s Dilemma is an example of a noncooperative game since the players are in different rooms and cannot cooperate. Much of game theory focuses on information asymmetries, signaling, and verification because a player’s best strategy might differ depending on whether she knows the other player’s best or likely strategy and whether she has complete or incomplete information.Footnote 23
A payoff is defined as “[t]he utility a player derives under a particular combination of strategies.…What matters is not the amount of a payoff a player receives, but the amount of that payoff relative to what the player would receive from choosing other strategies.”Footnote 24
Games can be “represented” in normal form (bimatrix), extensive form (decision tree), or beyond normal extensive form. For simplicity, I focus on normal-form games (without repeated interactions).Footnote 25
Solution concepts predict how the game will be played by focusing on which strategy a rational player will use, given their goals.Footnote 26 A dominant strategy is “a best choice for a player for every possible choice by the other player.”Footnote 27 Where a player has a dominant strategy, she will always use that strategy, and cooperation with the other player would not make a difference.
Game theory offers promise for health law because of its ability to move away from “basic assumptions: people act rationally, perfect information, zero transaction costs.”Footnote 28 While extensive-form games allow opportunities to model repeated interactions with signaling or the development of reputations and information sharing, normal-form games still illustrate discrete points about how providers and payors work in health care. These discrete games are part of a larger game in which people are repeat players and make deals over time.Footnote 29
3.3 Playing the Game without Fiduciaries
3.3.1 The Set Up
3.3.1.1 Providers
Parties in the health care game face a problem. If there is a maximum for what can be charged for health care at a given time, then any additional profits earned by one player must decrease the profits earned by another player.
Assuming that this is a zero-sum game for individual health care providers (“Providers”) and insurance companies and administrators (“Payors”) is not fully accurate, however. Health care expenses increase at a rate much faster than inflation, and employers and participants tolerate it for reasons including a lack of political will to develop a strong public option for working Americans. Pressure by employers and patients/health plan participants (“Patients”) also constrains the rate of annual increases.
This game does not rely on the assumption that Providers and Payors behave solely in their self-interest by attempting to maximize profits, which is the opposite of what is required by their fiduciary duties. Assuming Providers and Payors have some nonzero quantity of altruism, they will want to provide appropriate care for Patients and for Patients to pay less.Footnote 30 Given a range of options that meet the standard of care for Patients, however, they will prefer the choice that increases profits.Footnote 31 Assuming fiduciary duties constrain Providers at all, they should also be forced to select the strategy that prioritizes a higher payoff for Patients over a lower one – even if it lowers their own payoff. Providers need not care about the profits or costs to Payors and employers, to whom they owe no fiduciary duty or altruism, unless they are likely to be moved to an out-of-network position and lose Patients and money.
With this background, Providers should maximize their compensation by increasing reimbursement rates for patient care and keeping their costs down. Yet they know that Payors servicing health plans want to minimize, or at least reduce, Provider compensation to keep profits high, regardless of whether they pass along the savings through lower premiums.
Although employers may put pressure on Payors to reduce costs, they face issues of lack of transparency associated with increased costs and prices that are uniformly high across service providers. Employers can also pass the increased costs along to Patients through reduced wages or higher employee contributions. Finally, Patients want to reduce Provider reimbursement rates as long as they impact their cost-sharing obligations.
In this health care game, I assume the players must choose from the range of options within the standard of care. As such, there will be only two strategies: Self-Interest and Fiduciary. For Providers, this assumes that their desire to maximize profits and their desire to act in the best interests of Patients are mutually exclusive. In reality, of course, all Providers act in their own self-interest at times and put the best interests of Patients first at other times.
Nonetheless, for each decision, Providers must choose to act primarily in their own self-interest or in the best interests of a patient. The game examines that one choice and draws conclusions for the health care system.
Providers have a choice to earn higher compensation through some combination of:
increasing their reimbursement rates,
spending less time with Patients, and
ordering more visits, medications, tests, and procedures.Footnote 32
Payors and Patients have imperfect information about the necessity of each of these options that make up the Self-Interest strategy.
The second strategy for Providers is to provide higher quality care and ensure Patients pay less but results in lower Provider compensation. Providers can achieve these goals through some combination of:
decreasing their reimbursement rates,
spending more time with Patients, and
eliminating any unnecessary or questionable visits, medications, tests, and procedures.
This is the Fiduciary strategy, which models the type of behavior broad fiduciary duties are supposed to incentivize.
3.3.1.2 Payors
Like Providers, Payors have fiduciary duties, too. Therefore, even though they want to keep reimbursements to Providers low, they are obligated to prioritize the best interests of Patients. Given their imperfect information about what is medically necessary, Payors constrained by fiduciary duties should defer to Providers when they think Providers will play the Fiduciary strategy but defect (i.e., betray them and refuse to cooperate to share profits or work together in the best interests of Patients) when they think Providers will play the Self-Interest strategy. Otherwise, costs will increase for Payors with no resulting benefit for Patients.
In game theory, a player typically does not know which strategy another player will use during a particular interaction without signaling or cooperation. Instead, each player has to try to divine which strategy the other will play when making her own choice (e.g., by looking at the other player’s optimal strategy).
Payors have a choice to earn higher profits through some combination of:
decreasing their reimbursement rates to Providers;
denying approval for visits, medications, tests, and procedures;
incentivizing cheaper solutions (e.g., generic drugs); and
promoting preventive care that reduces future expenses.
The first two options put them in direct conflict with Providers and test their compliance with their fiduciary duties, and these situations are the focus of this chapter and this game.
Interestingly, the Self-Interest strategy allows Payors to claim that they are acting in the best interests of Patients because it eliminates at least some unnecessary care to Patients.
However, a Payor acting as a fiduciary would defer to the judgment of a Provider absent some sort of signal that the care requested would be harmful to a patient.
The second option for Payors is keeping their reimbursement rates high(er) so Patients receive all the care they need. They can achieve this goal through some combination of:
updating standards relating to reimbursable medications, test, and procedures when technology changes and
deferring to the judgment of Providers when it seems beneficial (the benefits outweigh the costs) to Patients.
This is the Fiduciary strategy, which models the type of behavior broad fiduciary duties are supposed to incentivize.
3.3.1.3 Patients
While Patients are the beneficiaries of the game (and all of the other parties owe them fiduciary duties), they are not actually playing the game. Although they pay an increasing share of their health care and insurance costs, the only strategies they have are to seek care or not to seek care.Footnote 33 Because of the information asymmetry between Patients and Providers, Patients cannot usually evaluate whether Providers are acting in their best interests and providing the best care (although they can seek a second opinion and play another round).
3.3.2 The Game
The game can then be represented as follows:
| Payor | |||
| Self-Interest | Fiduciary | ||
| Provider | Self-Interest | 2, 2 | 5, −5 |
| Fiduciary | −5, 5 | 0, 0 | |
The payoffs are listed in the order of the row player first and then the column player. As a reminder, the absolute value of the payoffs listed does not matter. Instead, what affects the game is the value of each payoff relative to the value of another potential payoff.Footnote 34 Here:
Each player’s payoff is the highest (5) when she plays the Self-Interest strategy and the other player plays the Fiduciary strategy because she maximizes her compensation while the other player does not try to take more of the limited health care dollars because she is focused on the best interests of the Patient.
Conversely, each player’s payoff is the lowest (−5) when she plays the Fiduciary strategy and the other player plays the Self-Interest strategy because she is not focused on maximizing compensation and the other player takes more of the pie. I use a negative payoff to reflect the fact that the Provider can lose money by expending time seeing a patient that is later not reimbursed.
When both players play the Fiduciary strategy, neither experiences a maximum or minimum payoff, but payoffs are minimal (0) because the Patient gets less care than will maximize Provider’s compensation as Provider minimizes questionable treatments and spends more time with Patient, but more care than will maximize Payor’s compensation as she defers to Provider.
When both players play the Self-Interest strategy, neither experiences a maximum or minimum payoff, but payoffs are higher than when they both play the Fiduciary Strategy (2) because Patient gets more care as Provider tries to maximize her compensation with questionable treatments and by spending less time with Patient, but less care as Payor tries to maximize her compensation by deferring less to Provider.
If Provider plays the Self-Interest strategy, then Payor’s best response is Self-Interest because it will result in a payoff of 2 instead of −5. If Provider plays the Fiduciary strategy, then Payor’s best response is still Self-Interest because it will result in a payoff of 5 instead of 0.
Without enforcement of fiduciary duties – a financial or criminal sanction for deviating – Payor will never play that strategy (absent altruism that outweighs self-interest). Because the payoffs are symmetric, the same is true for Provider; she will always play the Self-Interest strategy.
If there is coordination (i.e., interaction between the players where they agree to adopt a particular strategy), then Provider and Payor will still both play the Self-Interest strategy to minimize the risk of the worst possible payoff. Coordination then changes no one’s strategy.
Assuming no penalty from playing the Self-Interest strategy (as our system currently provides, assuming care is within a broad range of acceptable behavior), no rational actor would play the Fiduciary strategy.
3.4 Proposals
The physician–patient relationship has long been described as constrained by fiduciary duties. However, state and federal legislation has been far more important in constraining self-interested behavior by physicians than judicial enforcement of fiduciary law.Footnote 35 Among the interventions discussed by Halabi are the following federal statutes (and their state analogues): the Anti-Kickback Statute, which penalizes payments made to incentivize or reward referrals in health care paid for by federal programs; the Stark Law, which limits a physician’s ability to recommend services to a patient provided by a company that the physician has a financial interest in; and the Physician Payments Sunshine Act, which requires producers of drugs and medical devices covered by Medicare, Medicaid, or Children’s Health Insurance Program (CHIP) to report payments and other incentives to physicians and teaching hospitals.Footnote 36 The framework provided by this legislation has been more detailed and easier to enforce than fiduciary duties.
What all of these laws have in common, in terms of the game, is that they penalize self-interest by Providers and Payors to prevent the worst outcomes for Patients. In effect, they force Providers and Payors to use the Fiduciary strategy.
3.4.1 Statutes Limiting Financial Conflicts
To illustrate, I will first examine how laws that prevent Providers from referring Patients to service providers they have a financial interest in and laws that prohibit kickbacks for recommending particular services, drugs, or devices change the game. For example, the federal Anti-Kickback Statute could be applied to private health insurance instead of only federal health care programs.Footnote 37 The statute would provide that “[w]hoever knowingly and willfully solicits or receives any remuneration (including any kickback, bribe, or rebate) directly or indirectly, overtly or covertly, in cash or in kind … in return for the furnishing … of any item or service [payable under a private health insurance plan] … or … in return for purchasing, leasing, ordering … any good, facility, service, or item [payable under the plan] … shall be guilty of a felony,” imprisoned for up to ten years, and/or fined up to US$100,000.Footnote 38
The original game can now be represented as follows, with the modified payoffs in bold:
| Payor | |||
| Self-Interest | Fiduciary | ||
| Provider | Self-Interest | 1, 3 | 3, −4 |
| Fiduciary | −5, 5 | 0, 0 | |
The payoffs in bold have changed because the law constrains Provider, preventing her from maximizing her compensation in situations where it may harm Patient and is prohibited by statute. As a result, when Provider plays the Self-Interest strategy, her payoff goes from 2 to 1 where Payor also plays the Self-Interest strategy and from 5 to 3 when Payor plays the Fiduciary strategy. The payoffs for Payor also change, increasing from 2 to 3 when she plays the Self-Interest strategy and from −5 to −4 when she plays the Fiduciary strategy. This assumes that Payor keeps her prices the same when her costs from Provider decrease – increasing her profits.
The best strategy is still for Provider to play the Self-Interest strategy (and for Payor to play Self-Interest), but now the payoff has been reduced for Provider from 2 to 1 and increased for Payor from 2 to 3 because Provider cannot profit as much from acting in her own self-interest. Payor will take advantage to keep the profits that Provider would usually have gotten from her financial conflicts (assuming the market will bear it).
Coordination again changes no one’s strategy because Self-Interest is still the best strategy for each regardless of what the other player does. Even enacting similar statutory provisions to restrict financial conflicts for Payor would merely reduce her payout for playing Self-Interest but not change her optimal strategy.
3.4.2 Paying for Improved Outcomes
Increasing payoffs for Providers who achieve better patient outcomes and for Payors whose insureds achieve better outcomes at lower costs is a systemic change that could move both players from the Self-Interest strategy to the Fiduciary strategy (because the payoffs would increase). This concept is behind value-based reforms,Footnote 39 which attempted to improve the value and quality of care under Medicare when enacted through the Patient Protection and Affordable Care Act (ACA).Footnote 40 However, the consensus has been that these reforms have largely failed to achieve their goals.Footnote 41
Looking at the original game again with the new changes in bold shows why short-term value-based reforms have not been as successful as hoped.
| Payor | |||
| Self-Interest | Fiduciary | ||
| Provider | Self-Interest | 2, 2 | 5, −5 |
| Fiduciary | −5, 5 | 2, 2 | |
To get Provider and Payor to shift strategy, the incentives awarded for improved patient outcomes must be equal to the financial incentives received when Provider and Payor previously acted instead in their own self-interest. Assuming any sense of altruism or fiduciary responsibility, they may then coordinate and play the Fiduciary strategy; if there is no altruism or fiduciary responsibility, then incentives for improved outcomes will need to be greater than those previously received for acting in their own self-interest.
Why do they need to coordinate? Each player’s best strategy to avoid the worst possible outcome of a −5 payoff is to play the Self-Interest strategy (which results in a payoff of either 2 or 5). Unless a player can ensure that the other side will play Fiduciary (and she will not suffer a −5 payoff), then she will always play the Self-Interest strategy.
Even if they both play Fiduciary here, though, this results in cost-savings for the system only if the legislative solutions previously discussed are first imposed to reduce the rewards for self-interested behavior by Provider and Payor. The new incentives awarded for improved outcomes do not need to be as high to motivate a shift in strategies because of these existing laws.
From a game strategy perspective, this means that state and federal legislatures should draft specific, targeted legislation designed to limit and disclose conflicts of interest held by health fiduciaries and then implement incentive payments based on patient outcomes. In this way, patient outcomes improve while health care spending decreases. Otherwise, the system will fail to achieve the desired savings, as seen after the ACA implementation.
3.5 Conclusion
There is no evidence that health fiduciary behavior is successfully constrained at a systematic level by fiduciary duties. Fiduciaries are already behaving in a rational way when they ignore their fiduciary duties in health. The time has come to engage in a rational response.
Open access