Friday, January 29, 2016

The Madness of Our Health Insurance Scheme

I discussed this topic before in a number of essays1, but it seems important enough to address once more. Thus, allow me to take a few minutes to explain why, viewed objectively, our approach to health insurance, both before and after ObamaCare, makes no sense. Well, to be accurate, it makes sense for consumers to take advantage of the system as it currently stands, as to do otherwise would be to forfeit a considerable source of income, but viewed as a whole, the system itself is nonsensical, existing solely because of historical accident and government intervention.

Before I begin, perhaps it would be best to examine what insurance is, or least what the traditional role of insurance has been. It is important to do so, as pseudo-insurance such as our current medical system, along with other schemes such as Social Security2, have allowed us to forget how insurance really works, and why the insurance model is such a bad fit for medicine, at least as currently employed3.

There are a number of ways to view insurance, as a wager, as risk pooling, as the use of collective payment to obtain better investment returns, and so on, but let us not bother trying to decide which term fits best, and simply define insurance by what it does, or how it works. Basically, insurance works quite simply for the insured. The insured anticipates that some unwelcome event may befall him, costing him a considerable sum to make right -- anything from a fire, to a flood, to disability leaving him unable to work, to death. Since he does not know if the event will befall him, or if it is certain (eg death), precisely when it will occur, he agrees to pay a certain regular fee to the insurer, and in exchange the insurer agrees to pay either a set sum, or some or all of the costs incurred, should the event come to pass. The crucial elements being that the event is unwelcome, and that its occurrence is uncertain. Without these two aspects, traditional insurance would not work, as we shall discuss in a moment.

On the side of the insurer, the process is a bit more complex. The insurer knows he will pay out, at least to some of those he insures. That being the case, he has to be careful both in who he insures and the fees he charges. Working from actuarial tables based on past experience, the insurer knows with a degree of certainty that within a given pool of customers having certain characteristics, he will end up paying a certain amount. Using this, and his experience in investment telling him the likely return on investment over the same period, he can calculate a fee which will not only pay for claims on the policies, but also provide a reasonable return to the shareholders. In the case of the insurer, the essential feature is that he must collect more in a combination of fees and investment than he pays out, as well as having the ability to adjust those fees to suit the pool of those insured, or else the ability to accept or reject applications to make the pool of insured match his schedule of fees. (Or some combination of the two.)

Allow me to explain the essential features I mentioned above.

First, those for the insured. Most essential is that the event be unexpected, happening irregularly, or sometimes not at all. It is this feature alone which allows the insurer to pool various individuals, offsetting the risk of one against the other, making the event relatively predictable to the insurer while still unpredictable tot he insured. If it were otherwise, if the event were regular, or just fully predictable, insurance would not work. The insured would want to pay a stream of payments with a net present value(NPV)4, less than or equal to the expected expense, as if it were equal or greater, then he could simply invest the same money and offset the costs himself. However, if the NPV is less than the expected cost, since NPV is largely based on market rate of return, the insurer would likely not earn enough in investment to pay the expected costs, much less any profit for investors.

The second essential feature is that the event be unwelcome, such as a house fire, flooding, theft, serious injury or death. It must be an event which the individual will not wish to bring about, as otherwise he could cause the triggering event himself and collect the full value of insurance after paying only a trivial amount5. Only if the event is something unwelcome can the insurer be certain the insured will do all in his power to avoid the event, that is exercise diligence in protecting himself,  allowing insurers to rely on actuarial charts to predict the probability of having to pay out to the insured.

The statements above should help explain the essential feature of the insurer, the ability to adjust either schedule of fees, membership in the pool or some combination of the two. Since the insurer combines individuals based upon risk to make the payout amount, and to a degree period, predictable, and since he wants to earn enough to both pay out against claims and pay a return to investors, it is essential that he be able to either control who is insured, or else adjust the fees for each individual to reflect his anticipated risk. If he lacks either ability, then it becomes likely he will end up paying out more than he will take in, resulting in eventual bankruptcy.

Given this working definition of insurance, it should be obvious why I argue that, for example, Social Security is not insurance. And the same for unemployment insurance. In both cases, there is no ability to control the risk pool, as both pools are made up of all eligible workers. Similarly, the fee is not adjustable, as it is set by law as a fixed percentage of the worker's income. As a result, these supposed insurance plans often run the risk of bankruptcy. And the reason is easy to see, it is because, unlike true insurance, there is no risk pooling, or adjustable fees. There are other problems, such as the fact that many workers do not fear hastening unemployment, or in some cases unemployment -- such as seasonal unemployment in many trades -- is predictable, as is retirement, and thus, since workers would be unwilling to pay in more than they will collect, premiums will almost inevitably fail to cover the costs6. In short, these plans feature none of the aspects of insurance.

But I intended to look at our current health insurance plan, so I will ignore these side issues and turn my attention to the way we choose to pay for medical care.

Perhaps the biggest problem with our current approach to medical insurance is that it covers everything. That is, rather than covering unexpected injury, unplanned illnesses and the like, events which are both unpredictable and unwelcome, it covers absolutely all manner of medical expenses, both the unexpected and the predictable. Rather than simply paying for unplanned hospitalization, it pays for ordinary prescriptions, routine office visits, some forms of preventative care7 and the like.

I have explained the problem of predictable expenses, as well as insuring welcome, or just neutral, events, but perhaps an example, or two, would help.

Let us suppose, for the sake of example, a company decided to offer "auto insurance". In this case, not just insurance against collision damage, or unanticipated breakdown or damage, but insurance to pay for any costs related to owning an auto. Instead of having to pay for new tires, car washes, routine maintenance, even the cost of filling the tank, you simply pay the insurer, and he takes care of everything for you. If it is a cost related to your care, it is paid by the insurer. All you need to do is pay the monthly cost (and maybe a co-pay for some services), and you are in business.

Or, to embrace a more extreme example of insurance covering predictable, welcome events, let us imagine an insurer who agreed to pay your food and drink bill for you in exchange for a monthly premium. No longer would you have to pay at the grocer, or at a restaurant, all you do is pay your insurance premium, and the insurer will take care of all your bills.

Hopefully we are not yet so economically illiterate that anyone imagines either of these is a good idea. Taken at face value, both offer two possibilities, either bankruptcy for the insurer, or crippling premiums for the policy holder. (The possibility of excessively restrictive rules could also make these work to a degree, but we shall discuss that later.) And hopefully the reason that these are the only two possible outcomes is equally obvious. But, as I have learned that many times what I think is obvious is not clear to everyone, let us take a moment to examine the situation in detail.

First, and most important, we need to recall that, whenever someone buys something, someone must pay for it. I will grant, if an insurer buys in sufficient quantity, he may be able to negotiate a discount, but in general such discounts are relatively negligible8, especially in areas of strong consumer competition, such as petroleum, car repair and food. In these and similar fields where consumers can shop around for a good deal, profit margins tend to be narrow, and so the maximum possible discount is likewise quite small, at least so long as sellers want to avoid bankruptcy.  Thus, in most cases, insurers will end up paying close to retail, close enough as to make very little difference.

Why do I say that? Because, when dealing with insurance, we must consider not just the costs paid by the insurer, but also the overhead involved in managing policy membership, collecting premiums, accepting claims, paying out on those claims and so on. In the end, these overhead costs alone would almost surely consume any discount and more. And that does not even take into account the need to pay out some sort of return to the owners of the company, be they private owners or shareholders. Adding in that requirement, it should be clear that the costs of paying the bills, administering the plan and paying a profit to the owners would require more cash than would be needed if the plan holders simply paid the bills themselves.

Thus, it is inevitable that the premium required to support such an insurance plan would cost the policy holder more than would simply paying the bills himself. Granted, this may not be the case for every policy holder, some who consume a lot of resources may end up paying less in premiums than they would for the goods themselves, but they would doubtless be a minority, and the greater their benefit, the smaller the minority. Overall, on average, each individual would pay out more in premiums than the same goods and services would cost him in the market, and so, for each dollar saved by a big consumer, that dollar must be added to the cost of everyone else.

But that is not the only problem with such a plan. As we said above, it is not only essential the event insured against must be unpredictable (as these examples are not), but the event must also be unwelcome. And, again, these examples provide triggering events which, for the most part, are not unwelcome. Eating a meal, washing your car or filling the tank is not something which unduly troubles the average individual. That being the case, by divorcing such actions from any sort of obvious cost, insurance would encourage policy holders to consume more than they would otherwise. With gasoline being effectively free -- as the premium is the same regardless of consumption -- they would be inclined to drive more. Likewise, with car washes effectively free, there is every incentive for policy holders to become overly fastidious about their cars. And the list goes on and on. Admittedly, some of the events insured against, such as collisions or breakdowns, are still unwelcome and the related coverage would not likely be overused, the remaining coverage is for events policy holders have no reason to avoid, and thus would tend to increase consumption.

That being the case, insurers would have two possible responses. First, they could adjust their premiums, charging a rate such that the payments would compensate for the potential consumption of the most ardent consumer, but such a plan would make even more obvious the fact that such insurance is a bad deal for most buyers. With their rates tied to the habits of the most profligate, for most consumers the premium would be clearly far greater than the cost of paying their own bills9.

Since raising premiums to such a level is unlikely to be popular, insurers would try one of two alternate approaches.

The first is to make the events insured against less appealing to the consumer. Since it is not possible to change their nature, the only way to do so is to impose some sort of cost, and this is accomplished by assigning co-pays. The idea is simple, since consumers keep consuming more because it costs them nothing to increase consumption, let us impose some sort of per-use cost, so they will not keep consuming more with abandon. And it works, to a degree. With a co-pay the amount consumed does affect cost, and thus it does act to discourage unrestrained consumption. On the other hand, there are two obvious drawbacks. First, because the cost is still only a fraction of the real cost, while it discourages consumption, it does in a relatively attenuated way, and thus consumption will still be significantly higher than normal. Second, by making the policy holder pay out of pocket, it makes the insurance seem ever less useful. After all, if he needs to pay a co-pay on top of a premium, he sees the premium as more of a burden. So co-pays, while at least partially effective, also tend to reduce the appeal of insurance.

Which brings us to the second alternative, the imposition of rules. As mentioned above, various policies can be used to make consumption less appealing, or at least to place some limits upon how much of a bill the policy holder can run up. In the most basic form, these policies could simply refuse to pay for certain goods or services. In our examples, the car policy could pay for regular gasoline but not premium, or food policies would pay for ground beef but not prime rib. While this would obviously cut out more pricey items, it does nothing to check the total amount of consumption, and so it would pose only a small control on total consumption. Thus, in addition to limiting what will be covered, such policies would also limit the total number of times a service could be used. In our example, let us say, limiting the holder to one car wash a week, or three meals a day. And, if that proved too weak a control, there is always the possibility of requiring approval from the insurer before using certain services, thus farther limiting possible consumption. And the list goes on. Just imagine every restriction you ever read in the description of an HMO plan and you will get the idea. While unpopular, for the most part, such restrictions are still a very effective way to curb consumption.

But even with such restrictions, and even with every cost cutting approach imaginable, we still run up against the one big deal breaker, the problem with any such scheme. Quite simply, if we insure against routine and regular happenings, in other words if we substitute insurance for simple payment for services rendered, in the end, either the premiums will exceed the out of pocket costs for the same things, or the insurer will end up running at a serious loss. And since insurers don't stay in business long operating at a loss, in the end this means inevitably that the premiums for such insurance will cost more than it would to pay out of pocket.

And that is the situation we have with much of our medical insurance. Routine checkups, prescription medication, vaccinations and the like, which form a considerable part of our health insurance business,  are predictable, regular events, more akin to the food or gas payments I described above than the traditional concerns of insurance. And so, as a consequence, we end up paying more in premiums than we would out of pocket for these things.

Now, to a degree this is concealed, but not all that well. Fist, it is hidden by the fact that doctors have a tendency to pad out bills, then "discount" them for people paying out of pocket. It is not mentioned often, and seems a bit unethical, but as insurers -- especially government insurance -- tends to pay only a fraction of the amount charged, they routinely set fees well above what they expect to get in reimbursement. And thus, looking at the "sticker price" for medical services, and comparing to our insurance premiums, we may believe we are getting a good deal. But spend some time paying out of pocket, and it becomes obvious that the sticker price for medical care is even farther divorced from reality than the MSRP for cars10.

And then there is the other part of medical insurance, the part that truly is insurance, that is the amount charged to offset the unpredictable payment of hospital bills for sudden injury or illness and the like. As this true insurance is confounded with the pseudo-insurance that simply passes money through the insurer to pay routine bills, it is hard to tell exactly how much we are losing by using insurance as a way to pay ordinary costs. But, if nothing else, logic tells that, since the insurer has to pay the same amount we do, or perhaps a slightly discounted amount, and must add to that operating costs and profits for investors, there is no way we could pay him less than we would pay out of pocket.

Finally, there is the fact that part or all of our health insurance premium is paid by our employer, and that our portion, if any, is paid in pre-tax dollars, which takes some of the bite out of the premium inflation. However, if we think about it logically, if our employer were not paying for our insurance, he could redirect that money to our salary, since it obviously is figured into the cost of employing us, and so, though it helps to distract us from the true cost, the cost is still there.

So, if our current way of paying bills is such a bad idea, why do we do things this way? The answer is a mix of historical accident and government meddling.

During World War II there were any number of government intrusions into the economy, and one of the most foolish was the effort to freeze wages. As in any state that employs such a scheme, this produced two problems. First, it makes it impossible for employers to compete for employees. Since they cannot pay more to get the best workers, workers end up randomly distributed, with the most efficient companies, or the most strongly felt needs, as likely to go short of labor as the least efficient and least urgent demands. Second, because wage freezes usually also imply low differential in wages between jobs, it also becomes difficult to fill many jobs. After all, the most demanding jobs usually pay more to begin with, and,on top of that, often offer the best prospects for wage increases, as that is what is needed to attract labor. If there is no possibility of a pay raise, and the easiest job pays a considerable fraction of the most demanding, there is little incentive to take on more challenging work, and it becomes impossible to fill any number of upper echelon positions11.

To remedy this, government -- officially and unofficially -- allowed a number of substitutes for wage competition, with companies replacing wages with a number of non-wage forms of competition, such as retirement plans. Among this assortment of wage surrogates was employer paid health insurance. And, since it was being used to replace wages, not only was it traditional insurance, covering the unforeseen, but it was also another way to smuggle in additional compensation, paying for routine medical services traditional insurance would not normally have covered.

That is the historical accident, after which began the government meddling. By making such plans payable in pretax dollars, offering various incentives for employers to provide such plans and otherwise encouraging the continuation of such schemes, the government made it unthinkable that we would ever return to a situation where insurance was truly insurance again. We had turned our insurance into a scheme for funneling our payments to doctors through insurers, and we were not going to go back. And now, with the advent of "universal health care", laws mandating we all possess such plans, and with the government underwriting the purchase of these plans for those who do not receive them from employers or family members, it seems impossible to imagine a time when we would revert to true insurance once again.

However, that does not change one simple fact, that every time we involve another party in making a payment, it increases costs. In some cases (eg paying a grocer rather than negotiating with farmers) the convenience is well worth the cost, but in this case, the truth is, we are not getting anything for our expense. If anything, with the proliferation of rules about what services we can receive, and the the increasing acceptance of insurer gate keepers limiting access, we are actually getting less while paying more. And yet, it seems, whenever someone points out this madness, the public treats him as if he were insane, and insists we not only persist in this madness, but expand it, until it involves everyone.

I just do not understand.


1. See "Redefining Insurance... To Actually BE Insurance", "Medical Reform, An Overview", "The Absurdity of Mandatory Insurance", "Clarification of my Argument for a Free Market in Medicine", "High Cost of Medical Care", "The Devil is in the Definitions (And Assumptions)", "Three Ideas That Never Work" and "Preexisting Conditions".

2. See "Social Security is Not Insurance".

3. It is conceivable a true insurance model could be developed, paying the costs of emergency medical care, the costs of catastrophic injury, or other irregular, unexpected medical costs. It is not medicine itself that is a poor fit for insurance, but rather the attempt to fit routine, regular costs into an insurance model. But both of these topics will be covered in the essay proper at a later point.

4. For those unfamiliar with the term, the net present value (NPV) is the value of one or more cash amounts, adjusted by the present interest rate (or some other expected "rate of return"), allowing financiers to compare payments received at different times, or calculate the purchase price of, for example, a perpetual annuity. It is calculated by taking each payment and dividing by one plus the interest rate per period, raised to the power of the number of periods (C/((1+i)^n)), and adding together the sum of all such adjusted payments. There are special calculations for situations such as perpetual payments, but in general they are based upon this formula. In theory, if the rate of interest and compounding period are accurate, the NPV of a future value should be the amount you would need to invest to get that amount at the given date.

5. This is why insurers will not insure property for more than its value or replacement cost. If I own something worth $100 and insure it for $1000, it would be very tempting to "accidentally" destroy it, collecting the full $1000 after paying only a few premium payments, replacing the item for $100 and essentially collecting a $900 profit, less the handful of payments made. It is also why insurers are reluctant to take policies for unusual situations, such as life insurance on strangers, as the death of the insured would likely be meaningless to the policy holder, making it tempting to hasten his demise.

6. It is more tricky with unemployment insurance than Social Security, as technically unemployment is not supposed to be predictable, so workers do not often complain of paying in more than they receive. And, in the case of unemployment, as many workers never collect, they help subsidize those who do collect. On the other hand, since Social Security is paid for a predictable period, workers would complain if they were paid less than they paid in -- and the public is aware enough of the employer-employee split that even that subterfuge would not serve to hide a lesser payout --so Social Security must inevitably pay out something more than it collects. But, since most of those paying in do eventually collect, there is no large pool of non-recipients to subsidize those who do collect, as there is with unemployment insurance. This is why Social Security is chronically at risk of bankruptcy, saved only by inflationary periods, where loss of purchasing power makes it possible to pay "larger" amounts using devalued post-inflationary dollars. (See "The Rubber Yardstick", "What is Money?", "What is a Dollar?", "Why Gold?", "The Free Market Solution", "When Help Hurts II", "Inflation and Uncertainty", "Monetary Issues Made Simple Part I", "Monetary Issues Made Simple Part II", "Bad Economics Part 7", "Bad Economics Part 8", "Bad Economics Part 19",  "The Gold Question, Not "Why?" But "When?"", "Fiscal Discipline", "Putting the Bull in Bull Market" and "Misunderstanding Money".)

7. Preventative care is even more troubling than simple routine care. See "The Problem with Common Sense Solutions".

8. Health care is a bit different, as licensing, insurance requirements for referrals and a number of other features reduce competition significantly, and allow for considerable price inflation. Thus, in these fields it is routine for insurers to anticipate a considerable discount. This sort of price inflation is not possible in competitive fields, and thus would not be an issue in my examples. (Nor, for that matter, would it exist in a less regulated health care industry. As with student loans and university tuition, our defective health insurance scheme inflated prices badly. See  "Denying Reality", "When Help Hurts", "The State Versus Universities", "Help and Harm", "Medical Regulations" and "Medical Regulation II".)

9. I actually saw a real world example of such a policy. At one time I picked up a brochure for pet health insurance, and was shocked at how high the premium payments were. Then I thought about it, and realized that there are a considerable number of pet owners who are overly cautious with the health of their pets, who would happily increase the number of preventative procedures and routine checkups. As a result, the policy was established with the habits of such individuals in mind, and thus was priced absurdly high for anyone who had a more modest approach to pet health care.

10. Actually, this is not just the case for those paying out of pocket. Many times my insurer will cover only part of the amount charged, and more often than not my doctor will "forgive" the remaining amount, or all but a trivial sum. It is not universal, but it does seem to occur quite often.

11. See "The Basics", "Employment A to Z", "The Cart Before the Horse, or, Some Thoughts on the Iron Law of Wages", "The Sado-Masochist Society, or, Would Primitive Communism Work?" and "Socialism, Communism, Democracy, Authoritarianism and Freedom - Is It Possible to Have a Non-Authoritarian Socialism?".

No comments:

Post a Comment