You've heard a lot about the "savings and loan bailout" - but that is inaccurate language. It's not the thrifts who have been bailed out with taxpayer money. They have, in fact, failed by the hundreds. Rather it's their depositors who have been rescued, thanks to insurance the government wrote and is now paying.

Fuzzy language impedes sound analysis. It's true that weak laws, lax regulation and political mischief bear major responsibility for the thrift debacle. However, a flawed insurance operation contributed as well.If massive bailouts of depositors are to be avoided in the future, we have to start thinking as private insurers might and indeed bring them into the picture.

In particular, the government's insurance program needs to be overhauled to incorporate the first law of insurance underwriting: rates must be based on risk.

To date we've allowed the incompetent, the crook and the optimist to exploit federal insurance at rates identical to those paid by prudent and able managers. If we continue that practice, we're asking for more troubles.

But we should also understand that government is not well-equipped to judge the risk of insuring one institution versus another. What is needed is a system that combines the ability of private insurers to evaluate risk with the ability of government to bear it. Co-insurance arrangements, varying by size of bank, would appear to be the direction to go.

For example, our largest banks and thrifts - say, those with $10 billion or more of deposits - might be required to purchase pro-rata insurance covering 3 percent of their deposits from private insurers, with the remaining 97 percent to come from the FDIC.

The government would be paid the same rate per dollar of insurance as the rate charged by the private insurers. If losses occurred, they would be shared in that same proportion: 3 percent for the private insurer, 97 percent for the FDIC.

The arrangements for smaller banks would need to take into account the possible reluctance of private insurers to put time and energy into relatively small transactions. Therefore, banks with deposits of, say, $500 million to $1 billion might be required to buy 10 percent of their coverage from private insurers, with the rest to come from the FDIC.

At our smallest banks and thrifts - say, those with less than $500 million of deposits - the FDIC would be the sole insurer, charging a fixed rate of perhaps 0.2 percent of deposits, which is approximately the rate the FDIC has proposed that all banks pay next year.

There would remain the problem of larger institutions that could not, because of their perceived weaknesses, obtain private insurance. They would be placed in an "assigned risk" plan and pay the FDIC a high rate - say, 0.4 percent initially, rising gradually to 0.6 percent within a few years.

That rate would be burdensome but not confiscatory: Well-run banks currently earn 1.5 percent to 2 percent pre-tax on deposits while paying 0.12 percent for FDIC insurance. In any case, a 0.6 percent rate would give its victims a powerful incentive to clean up their act so that they could meet the standards of private insurers and obtain lower rates.

A private insurer can quickly adjust rates to new circumstances. Insurers dealing with higher-risk institutions might write policies covering only six months. Renewal rates would act as a scorecard on performance.

For this system to work, private insurers themselves would have to meet some tough standards: A minimum of $500 million of net worth, say, and a limit on the liabilities they could pile on each dollar.

Weaker banks will howl over this proposal. Though they regard it as sound practice to require that weak borrowers pay higher interest rates than strong borrowers, they will protest that differentiated insurance rates will make their already tough economics even tougher. Precisely. If an insurance structure doesn't produce pain, it won't produce change.

Financial folly will always be with us, but a well-designed insurance system can significantly reduce the costs it inflicts upon society.

(Warren E. Buffett is chairman of the board of Berkshire Hathaway Inc., a diversified company with insurance operations.)