In my last blog, I argued that the benefits of extending health insurance coverage are probably overstated and are not likely to reduce health care costs or improve health outcomes. In this blog I will argue that we do not need more health insurance, rather, we need less.
Here is the point. Healthcare is not a service that should be funded by insurance companies. An individual should ensure against expensive and undesirable calamities: tornadoes, fires, auto accidents. These need to be insurable risks, or insurance will not be made available. This means the events need to be reasonably random and relatively rare, with calculable probabilities that do not change much over time. As discussed in a previous blog, we need to make sure that the existence of insurance does not increase the probability of insured losses: (http://neweconomicperspectives.blogspot.com/2009/09/selling-death-wall-streets-newest.html) This is why we do not let you insure your neighbor’s house. Insurance works by using the premiums paid in by all of the insured to cover the losses that infrequently visit a small subset of them. Of course, insurance always turns out to be a bad deal for almost all of the insured—the return is hugely negative because most of the insured never collect benefits, and the insurance company has to cover all costs and earn profits on its business. Its operating costs and profits are more or less equal to the net losses suffered by its policy holders.
Ideally, insurance premiums ought to be linked to individual risks; if this actually changes behavior so that risk falls, so much the better. That reduces the costs to the other policy holders who do not experience insured events, and it also increases profitability of the insurance companies. Competition among insurers will then reduce the premiums for those whose behavior modifications have reduced risks.
The insured try to get into the low risk, low premium classes; the insurers try to sort people by risk and try to narrow risk classes. To be sure, insurers do not want to avoid all risks—given a risk/return trade-off, higher risk individuals will be charged higher premiums. Problems for the insurer arise if high risk individuals are placed in low risk classes, thus, enjoy inappropriately low premiums. The problem for many individuals is that appropriately priced premiums will be unaffordable. At the extreme, if the probability of an insurable event approaches certainty, the premium that must be charged equals the expected loss plus insurance company operating costs and profits. However, it is likely that high risk individuals would refuse insurance long before premiums reached that level.
Of course insured risks change over time—which means premiums charged might not cover the new risks. Cars become safer. More people wear seat belts. Fire resistant materials become standard and fire fighting technologies improve. Global warming produces more frequent and perhaps more severe hurricanes and tornadoes. But these changes are generally sufficiently slow that premiums and underwriting standards can be adjusted. Obviously, big and abrupt changes to risks would make it difficult to properly price premiums.
In any event, once insurance is written, the insurer does its best to deny claims. It will look at the fine print, try to find exclusions, and uncover pre-existing conditions (say, faulty wiring) that invalidate the claim. All of that is good business practice. Regulators are needed to protect the insured from overly aggressive denials of claims, a responsibility mostly of state government.
Let us examine the goal of universal health insurance from this perspective. It should now be obvious that using health “insurance” as the primary payment mechanism for health care is terribly inappropriate.
From the day of our births, each of us is a little bundle of pre-existing conditions—congenital abnormalities and genetic predispositions to disease or perhaps to risky behavior. Many of these conditions will only be discovered much later, probably in a doctor’s office. The health insurer will likely remain in the dark until a bill is submitted for payment. It then must seek a way to deny the claim. The insurer will check the fine print and patient records for exclusions and pre-existing conditions. Often, insurers automatically issue a denial, forcing patients to file an appeal. This burdens the insured and their care-givers with mountains of paperwork. Again, that is just good business practice—exactly what one would expect from an insurer.
And, again, it would be best to match individual premiums to risk, but usually people are placed into groups, often (for historical reasons) into employee groups. Insurers prefer youngish, urban, well educated, professionals—those jogging yuppies with good habits and enough income to join expensive gyms with personal trainers. Naturally, the insurer wants to charge premiums higher than what the risks would justify, and to exclude from coverage the most expensive procedures.
Many individuals are not really insurable, due to pre-existing conditions or risky behavior. However, many of these will be covered by negotiated group insurance due to their employment status. The idea is that the risks are spread and the healthier members of the group will subsidize the least healthy. This allows the insurer to escape the abnormally high risks of insuring high risk individuals. It is, of course, a bum deal for the healthy employees and their employers.
This is not the place for a detailed examination of the wisdom of tying health insurance to one’s employer. It is very difficult to believe that any justification can be made for it, so no one tries to justify it as far as I can tell. It is simply accepted as a horrible historical accident. It adds to the marginal cost of producing output since employers usually pick up a share of the premiums. It depresses the number of employees while forcing more overtime work (since health care costs are fixed per employee, not based on hours worked) as well as more part-time work (since insurance coverage usually requires a minimum number of hours worked). And it burdens “legacy firms” that offer life-time work as well as healthcare for retirees. Finally, and fairly obviously, it leaves huge segments of the population uncovered because they are not employed, because they are self-employed, or because they work in small firms. In short, one probably could not design a worse way of grouping individuals for the purposes of insurance provision. Would anyone reasonably propose that the primary means of delivering drivers to auto insurers would be through their employers? Or that auto insurance premiums ought to be set by the insurable loss experience of one’s co-workers? That is too ridiculous to contemplate—and so we do not–but it is what we do with health insurance.
Extending coverage to a diabetic against the risk of coming down with diabetes is like insuring a burning house. An individual with diabetes does not need insurance—he needs quality health care and good advice that is followed in order to increase the quality of life while reducing health care costs. Accompanying this health care with an insurance premium is not likely to have much effect on the health care outcome because it won’t change behavior beyond what could be accomplished through effective counseling. Indeed, charging higher premiums to those with diabetes is only likely to postpone diagnosis among those whose condition has not yet been identified. Getting people with diabetes into an insured pool increases costs for the other members of the pool. Both the insurer as well as the other insured members have an interest in keeping high risk individuals out of the pool. Experience shows that health care costs follow an 80/20 pattern: 80% of health care costs are incurred due to treatment of 20% of patients. (Steffie Woolhandler http://www.prospect.org/cs/articles?article=more_than_a_prayer_for_single_payer) If only a fraction of those high costs individuals can be excluded, costs to the insurer can be cut dramatically.
We have nearly 50 million individuals without health insurance, and the number grows every day. Most health “reform” proposals would somehow insure many or most of these people—mostly by forcing them to buy insurance. All of them have pre-existing conditions, many of which are precisely the type that if known would make them uninsurable if insurance companies could exclude them. While it is likely that only a fraction of the currently uninsured have been explicitly excluded from insurance because of existing conditions (many more are excluded because they cannot afford premiums)—but every one of them has numerous existing conditions and one of the main goals of “reform” is to make it more difficult for insurers to exclude people with existing conditions. In other words, “reform” will require people who do not want to buy insurance to buy it, and will require insurers who do not want to extend insurance to them to provide it. That is not a happy situation even in the best of circumstances.
So here is what the outcome will look like. Individuals will be forced to buy insurance against their will, often with premiums set unaffordably high. Government will provide a subsidy so that insurance can be provided. Insurance companies will impose high co-payments as well as deductibles that the insured cannot possibly afford. In this way, they will minimize claims and routine use of health care services by the nominally insured. When disaster strikes—putting a poorly covered individual into that 20/80 high cost class of patients–the insurer will find a way to dismiss claims. The “insured” individual will then be faced with bankrupting uncovered costs.
That is not far fetched. Currently, two-thirds of household bankruptcies are due to health care costs. Surprisingly, most of those who are forced into bankruptcy had health insurance—but lost it after treatment began, or simply could not afford the out-of-pocket expenses that the insurer refused to cover. As Woolhandler says, in 2007 an individual in her 50s would pay an insurance premium of $4200 per year, with a $2000 deductible. Many of those currently without insurance would not be able to pay the deductible, meaning that the health insurance would not provide any coverage for routine care. Only an emergency or development of a chronic condition would drive such a patient into the health care system; with exclusions and limitations on coverage, the patient could find that even after meetinging the $2000 deductible plus extra spending on co-payments, bankruptcy would be the only way to deal with all of the uncovered expenses. Of course, that leaves care providers with the bill—which is more-or-less what happens now without the universal insurance mandate.
In truth, insurance is a particularly bad way to provide payments for health care. Insurance is best suited to covering unexpected losses that result from acts of god, accidents, and other unavoidable calamities. But except in the case of teenagers and young adult males, accidents are not a major source of health care costs. In other words the costs to the insurer are not the equivalent of a tornado that randomly sets its sights on a trailer park. Rather, chronic illnesses, sometimes severe, and often those that lead to death, are more important. Selling insurance to a patient with a chronic and ultimately fatal illness would be like selling home insurance on a house that is slowly but certainly sliding down a cliff into the sea. Neither of these is really an insurable risk—rather each represents a certain cost with an actuarially sound premium that must exceed the loss (to cover operating costs and profits for the insurer). So if the policy were properly priced, no one would have an economic incentive to purchase it.
Another significant health care cost results from provision of what could be seen as public health services—vaccinations, mother and infant care, and so on. And a large part of that has nothing to do with calamity but rather with normal life processes: pregnancy, birth, well child care, school physicals, and certification of death at the other end of life. Treating a pregnancy as an insurable loss seems silly—even if it is unplanned. It does not make much sense to finance the health care costs associated with pregnancy and birth in the same way that we finance the costs of repairing an auto after a wreck—that is, through an insurance claim. Many of these expenditures have public goods aspects; while there are private benefits, if the health care cannot be covered through private insurance or out-of-pocket the consequences can lead to huge public sector costs. For this reason, it does not make sense to try to fund all private benefits of such care by charges to the individuals who may—or may not—be able and willing to pay for them. Nor does it make sense to raise premiums on one’s co-workers to cover expected pregnancies as young women join a firm.
Health care is not similar to protecting a homeowner against losses due to natural disasters. The risks to the health insurer are greatly affected by the behavior of the covered individuals, as well as by social policy. Discovering cures and new treatments can greatly increase, or reduce, costs. To a large extent that is outside the control of the insurer or the insured—if a new treatment becomes standard care, there will be pressures on insurers to cover it. Death might be the most cost effective way to deal with heart attacks, but standard practice does not present that as a standard treatment—nor would public policy want it to do so. In other words, social policy dictates to a large degree the losses that insurers must cover; acts of congress are not equivalent in their origins to acts of god—although their impacts on insurers are similar.
We currently pay most health care expenses through health insurance. But people need health care services on a routine basis—and not simply for unexpected calamities. We have become so accustomed to health insurance that we cannot understand how absurd it is to finance health care services in this manner. Our automobiles need routine maintenance, including oil changes. Imagine if we expected our auto insurer to cover such expected costs. We are, of course, all familiar with various “extended warranty” plans sold on practically all consumer items—from toasters to flat screen TVs. But we recognize that these are little more than scams—a way to increase the purchase price so that the retailer gets more revenue. We tolerate the scams because we can “just say no”—caveat emptor and all that. But health care “reform” proposes to force us to turn over a larger portion of our income to insurance companies—who will then do their best to ensure that any health care services we need will not be covered by the plan we are forced to buy. Unlike a broken toaster that can just be thrown out when the warranty fails to cover repairs, we do not, and do not want to, throw out people whose insurance coverage proves to be inadequate.
It is worthwhile to step back to look at the costs of providing health care payments through insurers. According to Woolhandler, 20 cents of every health care dollar goes to insurance companies. Another 11 cents goes to administrative overhead and profit of the health care providers. Much of that is due to all the paperwork required to try to get the insurance companies to pay claims (there are 1300 private insurers, with nearly as many different forms that health care providers must fill out to file claims); it is estimated that $350 billion a year could be saved on paperwork if the US adopted a single payer system. (Matt Taibi, “Sick and Wrong”, Rolling Stone, September 3, 2009). Hence, it is plausible that a full quarter of all health care spending in the US results from the peculiar way that we finance our health care system—relying on insurance companies for a fundamentally uninsurable service. Getting insurance companies out of the loop would almost certainly “pay for” provision of health care services to all of those who currently have inadequate access—including the under-insured.
In sum, using insurers to provide funding is a complex, costly, and distorting method of financing health care. Imagine sending your weekly grocery bill to an insurance clerk for review, and having the grocer reimbursed by the insurer to whom you have been paying “food insurance” premiums—with some of your purchases excluded from coverage at the whim of the insurer. Is there any plausible reason for putting an insurance agent between you and your grocer? Why do we put an insurer between you and your health care provider?
Next time: How to build a better mousetrap.