I read with interest GAO report 03-702 on the subject of medical malpractice insurance rates, published in June of 2003. It is an excellent example of how government sidesteps an issue by claiming a need for more data.
Production of this report certainly cost taxpayers a significant amount of money. I am equally certain that some people eagerly awaited its release, hoping for an answer to the rising costs of medical malpractice insurance. Sadly, the report provided nothing of the kind. The report did slide along the edges of common sense, but in the end gave up, obliquely asking for more government regulation to support the collection of even more data, thus delaying the obvious answers for a few more years.
Common Sense, And A Little Midnight Oil
From Washington, DC, the world looks like a very complex place, likely because legislators are enthralled with the delicate balance between special interest groups and their own re-election efforts. They study issues and amass piles of information, but either avoid meaningful analysis or reject it outright.
Over three decades in insurance, I have seen complexity: many good companies, a few very bad ones, massive rate increases and reductions, incredible profits, and staggering losses. In the final analysis, however, I find that insurance is a simple business. Many people spend their lives trying to make it complicated, but a little midnight oil burned studying the numbers—coupled with a little common sense—can make the basics quite clear.
Fundamentally, insurers only have two decisions to make about a piece of business: whether to underwrite it—since some risks are not worth taking—and how much premium to charge if they do.
Insurance Basics: The Insurer’s Point of View
What premium should an insurer charge for a new product? The answer involves no rocket science: no one knows! You accumulate as much data as you can, and then you make what engineers refer to as a SWAG: a Sophisticated Wild-A**ed Guess. If you don’t think you can guess accurately enough, then you don’t underwrite the new product. Since many “new” products are actually new versions of existing products; old product loss data can help, to some degree, to refine the accuracy of your initial guess. Even so, past performance does not necessarily predict the future very well, and in the case of truly new products, you’re on your own.
For a new product, then, the first factor to consider in selecting a premium is simply an educated guess on the part of the underwriter. This guess also encompasses the initial policy form, because trial lawyers are going to hunt for any loophole they can find in whatever policy form you create, in order to win the claim settlement lottery (33% of litigated claim settlements go into an attorneys’ pockets). As claim follows claim, the insurer will alter the policy form until it most clearly defines the risk they wish to take.
The second factor in pricing insurance products is loss history (claims presented against the policy form). This can be a difficult number to compute, since the size of a claim is never truly known until the file is closed. Although insurance companies are pretty good at settling small claims—the costs are generally straightforward, and most parties just want to get on with their lives—closing the file on a large claim can take many years, most of them spent doing the following:
- Discovering what actually happened in the claim.
- Discovering who was really involved.
- Discovering what damages are alleged, and which actually occurred.
- Comparing these findings to the policy to determine who or what was actually covered.
- Attaching a claim cost estimate to the file, covering not only the value of the claim itself, but also the discovery and legal costs required to close it.
- Trying to settle the claim for a reasonable amount.
The size of the claim is based in large part on the amount available for settlement, or the policy limit. For example, say you knock down your neighbor’s wooden fence, at a cost of $100 to repair. You might reasonably expect your liability insurance to pay your neighbor $100 in settlement. If, however, your liability policy allows for exact replacement valued at up to $1 million, you might be surprised to discover that your neighbor’s fence—fashioned of unusually aged wood and adorned with certain attractive knots and artfully crackled layers of antique paint—is by far his most prized and valuable asset. Remember the woman who sued McDonalds over hot coffee? Heaven help the insurance company that goes before a jury!
In any case, so far as premium pricing is concerned, the point is that even well-established products may not have accumulated a long enough list of successfully closed files to take the guesswork out of the pricing process.
The third and fourth factors in insurance pricing are company expenses and investment income.
Like most organizations, insurance companies try to find places to cut costs. Yes, premiums sometimes pay for corporate jets and beautiful offices, but in general not so often as to have any real impact on average premium rates. In fact, there is a fairly clear inverse relationship between the opulence of the home office and the long term health of an insurer. Selecting a new carrier? Ask for an office tour.
Investment income certainly has an impact on premium levels. Simply put, the more money the carrier makes by investing your premium, the more they can reduce your rate. Medical Malpractice is a good example: most cases are litigated over years, and the effects show up in the premiums.
The bottom line is that, even though it is perfectly valid to draw a connection between an insurer’s corporate expenses or investment returns and its product premium pricing, the impact of such connections is minor when compared with those of underwriter pricing estimates and a product’s established loss history.
Losses and Premium Levels
We have seen that premium levels are pretty much a guess for new products. For older products—and most insurance products are very old—the premium level is determined by expected loss levels. As we discussed, though, the real costs of a loss remain unknown until the claim is settled and the file closed, which in our highly litigious society can take many years. As a result, loss levels as well are in many cases no more than SWAGs.
America is a large country, and one would think that, over time, certain trends would appear to ease loss cost projections. Unfortunately, since each state has its own laws, which directly impact loss costs—and since in many cases the courts make up the law as a claim is litigated—useful precedents are often difficult or impossible to find. In other words, that fact that the damage to your neighbor’s wooden fence settled for $100 yesterday in no way implies that it will settle for the same amount when you knock it down again tomorrow. In fact, tomorrow’s settlement may grow a hundredfold if your neighbor can tell his tale of woe to twelve jurors who hope to win a similar lottery in their own lifetimes. Do not despair, though: if you can prove that your neighbor’s fence blocked your view as you drove, perhaps the jury will anoint you the victim instead!
So, after all that, what factors go into a claim file value projection, in other words, what do we count to determine how much a risk is worth, so we can come up with a reasonable premium? Factors include the following:
- The type of claim.
- Projected costs to settle the claim.
- The policy limits available to a claimant.
- Inflation between today and when the claim is settled. In other words, when is the claim most likely to occur?
- Pain and suffering, represented by a dollar value that can range from justice to a winning lottery ticket, depending on the mood of the jury.
- The legal jurisdiction in which the claim is likely to occur. Juries in the city are very generous, as are those in California, Lower Texas and Florida.
- The probable quality of the claimant’s legal representation, and whether the presiding judge might be appointed or might have to account for his actions during an upcoming election.
After accounting for these factors—collectively known as “case reserves”—pricing actuaries add “formula reserves” for intangible factors that affect the overall portfolio of claims, such as:
- Demographic and litigation trends.
- Claim inflation. Regardless of who is right in any particular case, it is a fact that fact that US laws—and especially juries—favor claimants over insurers more each year.
- Looming societal issues like environmental exposures and medical inflation trends.
- The age of the files.
- An estimate of the losses that have been incurred but not yet reported to the insurance company.
- A factor to account for the degree to which the claim technicians have missed the actual final cost in their prior claim file reserves.
Adding all this together produces a view of what a current portfolio of claims might ultimately cost the insurance company. We said it before: setting premium levels isn’t rocket science. Given the factors listed above, it is also no picnic. Despite all the details and factors and trends and inflations and jurisdictional issues involved, however, insurers do succeed in actually setting premiums, in part because one long-term trend is very clear: medical malpractice claim costs are rising dramatically.
Litigation drives premiums up; competition drives them down. In the insurance business, as in any other, a little common sense goes a long way.
Remember That GAO Report?
Page 17 of the GAO’s Medical Malpractice report presents a chart of Inflation-Adjusted Paid and Incurred Losses for Medical Malpractice nationwide (reproduced here). In 1975 the average paid loss was about $500. In 2001 the average paid loss was about $5,300, adjusted for inflation. That should tell you something!
Insurers rarely settle a claim in the same year in which a loss occurred. As we mentioned above, large losses often litigate for decades before the file finally closes. For this reason, insurers expend a lot of effort figuring out, based on losses incurred this year, how much extra money to reserve for settling those losses in future years. In this report, the GAO spends a lot of time (and paper) trying to understand such reserving practices, finally concluding that they don’t know enough about what was happening at each of the studied insurers to figure out why rates went up. The GAO’s implication is that malpractice premiums are clearly related to insurers’ reserving practices, if only we could figure out how.
The paid losses went up! In the final analysis, this is the only significant factor affecting the steady increase in malpractice insurance rates. The effect of reserving policies is simply to smooth the volatile fluctuations of annual incurred losses by distributing their effects over a number of subsequent years. This may affect an insurer’s cash flow issues, but it does nothing to reduce the number of dollars going out the door. That number is determined almost entirely by sympathetic juries whose legislators have trained them to view the transfer of wealth through litigation as a kind of free public lottery.
In 1997, medical malpractice insurers paid losses totaling $4 billion. The next four years—the last four of the study—saw that number increase by 32.5%. Any guesses as to what 1997’s incurred losses—worth $3.5 billion in 2001 dollars—will be worth when finally paid in 2007? Is it any wonder that insurance companies are abandoning the medical malpractice business?
On page 24, after claiming they need more data, the GAO says “potential causes” for the increasing medical malpractice claims might be: “greater societal propensity to sue, a lottery mentality, older population, greater expectations for medical care because of improved technology, a breakdown in the doctor-patient relationship.”
Is more data really going to clarify these observations? Will another study finally add the critical data point proving that our society has a “greater societal propensity to sue,” that we have developed a “lottery mentality?” Everybody who can read a newspaper knows that the legal system is broken when it comes to jury awards. Sure, doctors sometimes make mistakes, and that is what insurance is supposed to cover, but can any insurance product be reasonably priced when juries routinely award settlements in the tens of millions of dollars?
An examination of real causes, rather than the kind of navel-gazing that produced the GAO’s reserve-policy argument, leads us to the inevitable conclusion that there is no future—neither for insurers, nor for the insured—in charging the kind of malpractice insurance rates that can actually pay for such judicial abuses. The only long-term solution involves fixing the nature of claimant expectations. They should expect settlement, not a one-way ticket to Millionaire Row.
The fact is that people die, and medical science can only do so much about it. Not all medical procedures are guaranteed to succeed. Society has to take responsibility for itself, and sometimes, when things go horribly wrong, the “victim” is the only one to blame. You can’t smoke cigarettes your whole life and expect the insurance company to pay millions because they could have prolonged your life one day longer if they had only done one more very expensive procedure. There should be caps on jury awards, and a claimant’s individual negligence should be taken into account in determining awards.
As I said, calculating premiums is not rocket science. It is also not the path to utopia, no matter how many billions juries award in pursuit of public morality.
The GAO has wasted your money and mine. They identified a major problem and studied it to death, but lacked either the intelligence or the guts to draw a conclusion. Instead, citing their insatiable need for more data, they punted. Politicians print the lottery tickets, and trial lawyers deliver them to the masses. The GAO depends for its existence on the patronage of such people, and it is beyond naive to expect such a house of bureaucrats to bite the hands that sign its budgets.
One would think that in Washington, as in the insurance industry, a little common sense would go a long way. So far as I’m concerned, the jury’s still out on that one.