We examine behavior changes associated with the use of insurance and how they drive health care spending even higher.
In parts 1 and 2 of this series, we discussed why paying for primary care directly is less costly than paying with insurance. Now we turn our attention to behavior changes associated with the use of insurance, which drive health care spending even higher. And in part 4 we will examine how the incentives in a direct doctor-patient relationship lead to beneficial results in addition to cost savings.
How patients and physicians behave differently when insurance is present
It has long been known that the presence of insurance affects behavior; this is true for all types of insurance, and insurance companies use various policy provisions to limit their losses due to these types of behavior change. In a nutshell, our focus here is on ways in which people can change their behavior to benefit themselves financially when they are insured. The effects of these behavior changes are larger when the person has more control over what’s being insured. Returning to our example of homeowners’ insurance that protects against loss due to fire (from part 2), the homeowner has a great deal of control over setting their house on fire in order to collect on the insurance policy. This is moral hazard, people changing their behavior due to the presence of insurance. Homeowners’ insurance policies exclude payment when arson is involved, thus protecting them against this type of moral hazard.
In healthcare, moral hazard takes a few different forms. People may be less likely to invest in their health or engage in healthy behaviors when they are insured. They may consume more health care at a given level of health if they are insured compared to if they are not insured. And physicians may recommend more and more expensive care when they know the patient is insured.
Another type of behavior change in the presence of insurance involves how people choose the insurance coverage they want when they have options. Adverse selection occurs when the insurance company is unaware of a potentially expensive condition of a patient and the patient selects an insurance plan with richer benefits because of the likelihood they’ll need expensive care. The insurance company is unable to price for the higher risk of the patient since they are unaware of it. As a result, everyone will have to pay a little more since the premium charged for that person is likely to be inadequate.
Of course, to the extent that large amounts of claims are generated by moral hazard and adverse selection, the insurance company has to charge everyone higher premiums. The best protection against these costs is to avoid using insurance to pay for predictable, routine, likely events over which the customer has a great deal of control. Insurance, then, should focus exclusively on events that are largely random or unavoidable, are unlikely to happen to a particular person, and are high cost. If at all possible, insurance should only be used for insurable events. Long ago actuaries figured out that insurance works best, is most efficient with stable premiums over time if it only is used for insurable events.
While all of the above is well-established, we don’t necessarily have accurate estimates of how much these behavior changes add to the cost of healthcare, but a few studies shed light on this question. In the RAND health insurance experiment patients were randomized into insurance plans that varied by levels of cost sharing required. Patients responsible for 95% of costs had 40% lower health care spending than patients responsible for none of the cost, and additional services had little value to patients, although there was variation. (Some sicker patients benefited from the additional care while some healthier patients were harmed.) (1)
Before Medicare was enacted in 1965, three-fourths of Americans over the age of 65 (who comprised about 10% of the population) were uninsured. Increasing the ranks of the insured in America by 7.5% overnight had significant impact on the healthcare system. Total hospital spending (not just spending on the elderly) increased 40% in the first 5 years after Medicaid was implemented. One expert estimates that the spread of insurance between 1950 and 1990 explains half of the six-fold increase in per capita healthcare spending over that time (2). All of this suggests that both patients and providers change their behavior considerably when insurance is present.
To the extent that large amounts of claims are generated by moral hazard and adverse selection, the insurance company has to charge everyone higher premiums.The best protection against these costs is to avoid using insurance to pay for predictable, routine, likely events over which the customer has a great deal of control.
However, it’s not known how much of the additional spending associated with these behavior changes are related to primary care vs other levels of care (specialty, emergency, hospital, etc). In fact, spending more on primary care is generally associated with spending less on downstream care. Specialists tend to use more resources (medications, procedures, ancillary services and longer hospital stays) than primary care physicians; as a result, patients who receive more of their care from specialists tend to have higher expenditures, all else equal (3-5). We must be cautious about estimating how much these behavior changes add to primary care spending. It seems safe to say that when extra spending associated with these behavior changes is added to the extra spending that results in the primary care physician’s office and the insurance company doing extra administrative tasks, we conclude that paying for primary care with insurance adds at least 50% to the cost, compared to paying directly.
Please note that in this blog post we’re looking at ways in which using insurance to pay for primary care leads to behavior changes that increase costs. For a more thorough discussion of other structural & technical issues created by improper use of insurance and other dysfunctions of our health care system, please consult the paper “The Role of Insurance Principles in the Sustainability of Risk Transfer Systems” which can be found under Digging Deeper at www.concernedactuaries.org.
The spread of insurance between 1950 and 1990 (due largely to the implementation of Medicare and Medicaid) explains about half of the six-fold increase in per capita healthcare spending over that time.
In part 4, we’ll examine two common approaches to delivering primary care directly, and we’ll consider how the physician and patient incentives inherent in direct approaches differ from the incentives inherent in paying for primary care with insurance. Stay tuned!
1. Finkelstein A. Moral hazard in health insurance. New York: Columbia University Press; 2015.
2. Finkelstein A. The aggregate effects of health insurance: Evidence from the introduction of medicare. Q J Econ. 2007;122(1):1-37.
3. Donaldson M, Yordy, K., Lohr, K., Vanselow, N. Primary Care: America's Health in a New Era. Washington DC: National Academy Press; 1996.
4. Harrold LR, Field TS, Gurwitz JH. Knowledge, patterns of care, and outcomes of care for generalists and specialists. J Gen Intern Med. 1999;14(8):499-511.
5. Franks P, Fiscella K. Primary care physicians and specialists as personal physicians. Health care expenditures and mortality experience. J Fam Pract. 1998;47(2):105-109.