Some Thoughts on Self Insurance

Retention Levels Should Not be Insurance-Driven

A recent RIMS poll asked, when evaluating the self-insurance option, how important is advice from a broker. About three-quarters of risk manager respondents said "vital" or "important. Without more information on what type of self-insurance was in the respondents' minds, not much can be inferred from this except that insurance is still dominant in a risk manager's strategic thinking.

Why should a broker advise on a management decision? Only if it relates more to insurance than management. If this is so, insurance is being treated as a commodity, not as a strategic tool.

Risk Management theory says insurance is the last step in the risk management process. In practice, it's still the first. It shouldn't be, but it is. Decisions on whether or not to self-insure are, in common practice, based on considering insurance premiums at different retention levels. Such decisions are insurance-driven; expedient for the short-term, but not for most effective long-term risk management.

Risk management should follow these steps:

  1. Develop corporate policy on risk.
  2. Choose retention level.
  3. Assign responsibilities for all risk functions.
  4. Buy insurance only over the retention level.

Insurance should be a long-range program of the broad coverage desired, chosen irrespective of deductible credits, services offered, or other non-risk-transfer considerations. Only if this is done will the risk manager have the freedom to conduct his orchestra of virtuosi in a way to produce the most beautiful music.


Is Self-Insurance "Insurance"?

This can be an important question in several situations. A company self-insured for liability, for example, might be asked to contribute to a claim payment as "other insurance. Or a state might tax self-insured plans as they do insured plans.

The latter situation occurred in New York when the state imposed a 13% tax (they call it a surcharge) on hospital bills of employees covered by employers who self-insure health care benefits. According to page 1 of the August 21, 1995 Business Insurance, the appeals court ruling validating that tax is likely to lead other states to do the same. The court said there is no reason to treat self-insured plans differently than insured arrangements.

Retired, but eagle-eyed, risk manager Douglas Barlow caught that and wondered: "Is this a penalizing by nemesis of the use of the term 'self-insurance' to describe a corporation's internal accounting in respect of losses, where there is no provision for indemnification?" He asks "whether the taxing authorities would have thought of taxing the hospital bills...had the term 'self-insurance' not been in common use.

A thoughtful question, to which I might suggest one answer. There is a difference between self-insurance and losses deliberately assumed without insurance. Self-insurance implies that certain risks are covered and some are not. In the case of employers who self-insure health care, benefits paid to employees are sharply defined, just as with an insured plan. Self-insurance of workers compensation, also, has clearly delineated coverage, so is truly self-insurance rather than non-insurance.

On the other hand, a program of in-house-funding has no coverage criteria, no exclusions, and no conditions. It is simply expensing of losses as they occur, just as maintenance of machinery is expensed as it occurs. No one would say that machinery is repaired under a "maintenance self-insurance" plan.

Unfortunately, in-house-funding programs are often popularly termed "self-insurance", which could lead to semantic problems as suggested by Barlow. It did in the following 1983 case.

A truck driver employed by Trans-Con was injured while loading his truck at a Nabisco plant. Nabisco did not insure its lower-level liability claims. Nabisco claimed additional insured status under Trans-Con's liability policy, which had a clause saying it was excess of any "other insurance or self-insurance. Nabisco said its plan was not self-insurance; they simply had no insurance. Further, they felt the term was ambiguous.

The court disagreed, saying the term "self-insurance" is common and accepted today. Well, yes, it's common and accepted, but do we agree on what it means? Obviously not.

This is a problem that is only going to get worse as both self-insurance and in-house-funding programs grow. I know of no elegant solutions, but you should keep the problem in mind as you negotiate contracts and come into contact with situations that involve these concepts. Negotiate special wording to clarify exactly what is intended.


The Need for Greater Risk Retention

Decades ago, almost full insurance was the norm; deductibles in property were low and in liability, almost non-existent. Then a few companies tried assuming some risk. "Hey, it works", was the cry, and the amounts of retention gradually rose. Most companies go through a predictable sequence: they assume a little risk, then a little more, then more. Today, high retentions are common, but in my view, the evolution still has far to go. Let's look back just a bit.

In the 1960s, the Factory Insurance Association (now IRI) wouldn't even accept large deductibles. Its management wanted to (the Factory Mutuals were way ahead of them in this respect) but FIA was too dominated by its big stock company owners. Insurance was their game.

To illustrate the temper of the times, in 1966 Ford Motor Co. sent a few of its top financial executives around the country to deliver a seminar on finance. They passed out a financial report, which showed assets way up in the billions. I looked at it and asked their treasurer what property deductible they carried. The answer was $5,000. Relatively speaking, that was like my carrying about a 7¢ deductible on personal auto insurance. I asked him why it was so low and he responded candidly that they just hadn't thought the subject through.

We have progressed a long way since then, but still have far to go. The principal reason is that commercial insurance is too full of holes. Coverage is too limited. Instead of trying to expand umbrella coverage or all-risk property coverage and charging a carefully calculated premium for broad coverage, underwriters choose to delete coverages selectively: pollution, employee practices, earthquake, flood, lead, asbestos, trampolines, and so on far into the night. Risk managers have to buy separate policies for these risks or go bare. Too often, they go bare and cross their fingers. Unfortunately, fingers don't cover all the vital spots.

In a sense, the insurance industry has shot itself in the groin by limiting coverage. Even though their loss database is greater than any other, they have not analyzed it in sufficient detail to develop accurate rating procedures for all risks, geographic locations, and other variables. If they did, they would probably find that broad coverages could be priced to be acceptable. That statement goes counter to the combined wisdom of insurance executives who know more about the subject than I, but what I do know is that informed buyers put a higher value on protection than on price. If the product is fairly priced, it can be sold. Underwriters' actions may bring them short-term profits but long-term extinction.

As for risk managers, they are still too timid about retaining loss. Consider the recent Daiwa Bank debacle where an unrestrained trader of T-bonds frittered away $1.1 billion in 11 years. This is about $400,000 lost every working day of that 11 years. Here we have a company with $180 billion in assets tossing away $400,000 every day, and nobody even knew about it for 11 years. $400,000 - that's a fairly respectable annual retention, yet Daiwa assumed that much every day for 11 years! It hurt, but according to the Wall Street Journal, Daiwa "was able to absorb the loss easily.

Risk managers sometimes justify moderate retentions by saying that insurance rate credits are not sufficient. That's too limited a view. The goal is less minimum premium than financial stability. With very large retentions, contracts become possible where many otherwise uninsurable risks become transferred. It takes some doing, but this is why finite risk and other more sophisticated financial products are becoming popular.

Copyright © 1997 by David Warren

[Return to Warren Report]