The Health Insurance Shell Game

The insurance industry had a rocky start a century ago. It was clear that there were untoward events that could befall any of us with catastrophic results, from the incineration of a home to the loss of the ability to maintain gainful employment from injury or death.

Insurance offers a mechanism to share this risk. The stumbling block was the possibility that the insured might burn down their home to collect. Once it was realized that “moral hazard” could be held at bay by investigating for fraud, there was little to hinder the growth of an industry designed to serve our risk adverse proclivities. Almost every adult has some experience valuing the expense of sharing risk for a variety of hazards. After all, automobile insurance is generally compulsory and most of us are familiar with notions of deductibles and riders when it comes to homeowners’ policies. The possibilities are not an abstraction; we can envision the house or its contents damaged, destroyed, or stolen leaving us bereft. What would reducing that prospect be worth to us? As is true for many value-based decisions, the answer brings a mix of reason and intuition (1)that can produce surprising outcomes (2).

Health insurance is even more complex, and has always been so. The industrial revolution saw the development of “Friendly Societies” in Britain and the Prussian “Krankenkassen”. These were trade-based institutions that allowed advantaged workers to purchase insurance to provide “sick pay” but there was little else. The sea change was the Prussian “welfare monarchy” (3), an extensive insurance scheme that encompassed universal health care and a complex approach to disability insurance (4). Modifications of the Prussian scheme spread across the industrial world. It made landfall in the United States in time for the presidential election of 1912. Only one component took root in America: Workers’ Compensation Insurance but not as a national insurance scheme. It fell to the each state to regulate an insurance scheme to compensate injured workers for lost income and medical expenses.

This set the stage for state-based regulation of employer-sponsored private health insurance schemes going forward. But forward momentum appears anything but swift or linear in a country that trusted physicians to charge “commensurate with the services rendered and the patient’s ability to pay” (AMA Code of Medical Ethics, 1957.) Health Insurance as both an industry and a product has become a frustrating web of inefficiency and confusion.

No one needs to be reminded of the escalating costliness of this approach to sharing the risk in healthcare or discordance between the costliness in the United States and elsewhere. No one needs to be reminded of the bizarre machinations already in place and now unfolding to reign in this costliness. What are less discussed are the processes by which these machinations distort peoples’ values, all of us, for whom purchasing health insurance is to provide security against the untoward consequences of clinical misadventures.

The standard approach is to ask employees to decide how much insurance should be purchased and for what coverage. If only it were that simple, but it is not. The average employee has to decide on an insurance plan that hosts a dizzying array of fees and coverage: How much will you pay out-of-pocket before insurance kicks in (deductible)? What’s the maximum coverage (cap) if I get really sick? Is a million dollars enough? Two million? What about co-insurance?  What is co-insurance? Do co-pays count towards the deductible? What about tests?

Answers to these and similar questions dramatically alter the monthly premium, which is not a trivial consideration for the average American employee. In homeowner’s insurance such questions can test one’s mettle when valuing the loss of a family heirloom. How are they even ponderable when valuing health care for one’s self or one’s family? Research suggests they are not, at least not in an optimal way (5).

Today’s American employees register their preference for health insurance at the start of employment or during an annual confusing and worrisome time called “open enrollment”. The process is familiar to most employed Americans who must consider and value variations in premiums, deductibles, co-pays, and the like. It is an annual rite because the menu of options often varies year-to-year reflecting employer prerogatives.

The workers’ input to the design of healthplans and their component fees is after the fact; their preferences are constrained by the options offered. They are faced with a complicated choice that pits financial savvy and preconceptions as to the value of health care against their tolerance of risk. This is the model that has been adopted by the Affordable Care Act (“Obamacare”) in the formulation of the Exchanges which are about to make their way deeper into the American landscape.

One might predict that few of us are a match for weighing the array of benefits in light of the costs we might predict we would face. That assumption did not prove true when we put it to the test. We invited members of the North Carolina State Employee Association and faculty/staff of Duke University to volunteer for a web based survey. The survey assessed demographic features of 400 volunteers and presented them with a series of panels, each of which offered up 3 options in health plan design with variations in the description of doctor visit fee, annual deductible, proscription fee, lifetime coverage, choice of doctors, and monthly premium. The Figure (below)  is a display of the features of the plans that influenced the preferences of the State Employees.

With further analysis we could demonstrate that the State and Duke employees would be willing to spend several out-of-pocket dollars on co-pays rather than a dollar more in premium. Furthermore, this sensitivity to premium is greatest for younger employees insuring only themselves; non-white families are more sensitive to annual deductibles (6). Clearly these volunteers were weighing the value to them of greater expense assuming that the health policy had intrinsic value worthy of as much as they could afford. The more their resources were limited, the more they were willing to run the risk of incurring expense modulated by their assessment of their risk for sickness.

aRange, $100, $300, $500. Results show a preference for the lowest deduction amount. bRange, $1 000 000, $2 000 000, unlimited. Results show a preference for unlimited coverage. cRange, $0, $1 000, $2 000. Results show a preference for the lowest deductible amount. dRange, $0, $20, $40. Results show a preference for the lowest prescription copay amount. eRange, 100, 200, 400 doctor network. Results show a preference for more choices. fRange,  $0, $25, $50. Results show a preference for the lowest copay.

Utility for Plan Attributes in the State Employee Sample (click to enlarge). aRange, $100, $300, $500. Results show a preference for the lowest deduction amount. bRange, $1 000 000, $2 000 000, unlimited. Results show a preference for unlimited coverage. cRange, $0, $1 000, $2 000. Results show a preference for the lowest deductible amount. dRange, $0, $20, $40. Results show a preference for the lowest prescription copay amount. eRange, 100, 200, 400 doctor network. Results show a preference for more choices. fRange, $0, $25, $50. Results show a preference for the lowest copay.

Generations of Americans have now been lulled into thinking all of this is sensible. “Gaming” the system is the way it is and the way it will be. Health insurance may seem like a logical variation on homeowners’ insurance. But there is nothing sensible about it. Choosing health coverage is a totally different exercise from choosing to insure a family home or heirloom. For one, the home or heirloom can be appraised. Health can’t be appraised; in fact, it is difficult to define. How valuable is the care, which elements of care are less valuable and which might one rationally do without? No one offers a policy designed to such particulars, and no person can be reasonably expected to fully know what they will want or need as a sick, scared, patient during the savvy consumer-minded extravaganza of open-enrollment. For health insurance, the “moral hazard” does not pertain to the insured; it pertains to the other stakeholders participating in the system.

Medicine in the 21st century should be an exercise in informed medical decision making. For each option in diagnosis and intervention, the patient must be encouraged to ask, “Based on the available science, what is the best I can expect?”  And then actively and with comprehension, listen to the answer. For some options, there is no informative science. For many, the science may be robust but demonstrations of efficacy have proved elusive, inconsistent or marginal. Examples include interventions for occlusive atherosclerotic disease, elective procedures on backs, shoulders and knees, pharmaceutical management of cognitive impairment, situational affective disorders, type 2 diabetes and essential hypertension, along with various screening protocols that afford minimal if any demonstrable benefit. The therapeutic decision hinges on how the patient values the remote possibility of benefit and the probability of harm.

Shouldn’t health insurance in the 21st century live by the same razor? How valuable is the care, which elements of care are less valuable and which can one rationally do without? Rather than tolerate increases in co-pay and deductible, shouldn’t we be able to pay less because we do not value particular options. Better yet, shouldn’t the options relate to the likelihood of benefit? Those of us who consider interventions with unlikely or small benefits of little value should not be asked to burden the cost of providing such for those who value such. We should be offered a “high efficacy option” at lower cost than an “any efficacy option” and no one should be offered an option that indemnifies for interventions that have been studied and cannot be shown to offer a clinically meaningful benefit. In fact, if the “any efficacy option” was transparent in terms of limitations in efficacy and risk of toxicity, would anyone want to share in the expense of indemnification or balk at a policy that did not cover interventions that lacked substantive evidence for meaningful effectiveness (7)?

Focusing primarily on cost, costliness, and administrative priorities takes the health of the “health care system” as the primary goal. It is a focus that provides no measurable advantage in caring for people. Persisting in this approach, and even expanding it, provides a temporary diversion, but not a solution. Neither the practice of medicine nor its infrastructure is the reason for medicine to exist. Furthermore, becoming a savvy consumer of “health care” is not what is meant by informed medical decision making. Medicine’s primary calling is to the personal, unique, idiosyncratic needs and values of each person who chooses to be (or must become) a patient. And in that calling is the solution to the crisis of costliness (7).


1. Kahneman D. Thinking Fast and Slow. New York, NY: Farrar, Straus and Giroux; 2011.

2. Ariely D. Predictably Irrational. The hidden forces that shape our decisions. New York, NY: Harper Collins; 2009.

3. Beck H. The Origins of the Authoritarian Welfare State in Prussia. Ann Arbor, MI: University of Michigan Press; 1995.

4. Hadler NM. Stabbed in the Back. Confronting back pain in an overtreated society. Chapel Hill, NC: University of North Carolina Press; 2009, 93-138.

5. McGraw AP, Schwartz JA, Tetlock P. From the Commercial to the Communal: Reframing Taboo Trade-offs in Religious and Pharmaceutical Marketing. Journal of Consumer Research 2012; 39: 157-73.

6. Schwartz J, Hadler NM, Ariely D, Huber JC, Emerick T. Choosing among employer-sponsored health plans. What drives employee choices? J Occupational and Environmental Medicine 2013; 55: 305-9.

7. Hadler NM. The Citizen Patient. Reforming health care for the sake of the patient, not the system. Chapel Hill, NC: University of North Carolina Press; 2013.

Nortin Hadler, MD, is a professor of Medicine and Microbiology at the University of North Carolina – Chapel Hill.  Much of his scholarship has focused on issues in employee health and safety. Hadler is the author of numerous articles and essays and a series of a popular books on the state of medicine today. His most recent work, Citizen Patient, was published earlier this year.

Janet Schwartz, PhD is an Assistant Professor of Marketing in the A.B. Freeman School of Business, Tulane University, New Orleans, Louisiana. Her current research is at the intersection of marketing and public policy where she uses insights from psychology and behavioral economics to investigate how consumers navigate the healthcare marketplace.

This post originally appeared in The Scientific American Blog Network.

2 replies »

  1. I work in the insurance industry and there are so many things wrong with your description of insurance, I don’t have time today to write it all. The biggest problem is the way you conflate risk insurance with social insurance. I will stick to just two things: 1. The difference between risk insurance and social insurance. 2. And how insurance industry deals with moral hazard.

    Risk insurance is a contract where, for a consideration, the risk of financial loss, from a peril, is transferred to another party. Risk insurance has nothing to do with “sharing risks” or “sharing costs”. The risk is merely transferred. It can be transferred to one person or a group of people. Fundamentally, it is nothing more than a bet. The transferrees are betting the transferrer will not experience a loss. In order for risk insurance to remain profitable it must be non-coercive for both the transferrer and transferree parties. The property coverage section of a homeowners policy is an example of risk insurance.

    Risk insurance has to deal with the element of moral hazard. The way the insurance carrier deals with moral hazard is by insuring the risk for no greater than its actual value and by requiring the insured to retain the risk of the first part of the loss, otherwise known as the deductible. Thus, even with a total loss, the insured is never made completely whole. Because the insured has incurred the premium and he has also incurred the first part of the loss via the deductible. Moral hazard is reduced in this manner. Investigating for fraud doesn’t reduce moral hazard, it only uncovers fraud. Not the same thing.

    Social insurance is about sharing the costs of a set of benefits which a group of people want to partake. Examples of social insurance are social security and medicare. Social insurance is necessarily coercive. All parties entitled to a benefit must participate by paying or they do not receive the benefit. The benefits cannot be properly funded if all parties aren’t forced to pay. There is no moral hazard in social insurance. There can be fraud when people attempt to claim more benefit than they are entitled to.

    Health insurance can be pure risk insurance as long as the insured retains the first part of the risk and the carrier is limited in the ultimate amount of risk he takes. Health insurance can also be social insurance as it is with medicare. Health insurance, in the United States, is neither risk insurance nor social insurance but a amalgam of both.