In fact, the available aid is substantial. Homeowners with losses benefit from federal tax relief; they may write off non-reimbursed casualty losses greater than 10% of their adjusted gross incomes. They also receive benefits from the Federal Emergency Management Agency's disaster housing assistance program and are eligible for low-interest loans offered by the Small Business Administration.

Concerned that this relief is not sufficient, California politicians are rallying behind proposed legislation to eliminate the 10% floor in presidentially declared disaster areas. The bill, proposed by Sen. Barbara Boxer (D-Calif.) and Rep. Howard Berman (D-Panorama City), is widely supported by California Democrats and has considerable Republican support as well. Both major-party candidates for Congress in the West Valley's 24th District support it.

In my view, such subsidies to homeowners are undesirable. I am concerned that public disaster relief reduces homeowners' initiative to save and to insure against disaster.

Individuals who live on flood plains, along the beach, in fire-prone neighborhoods and in seismically active areas have elected to face unusual risks. Normally, the purchase of private insurance forces them to confront the inherent dangers. When government disaster relief is available, people will build with diminished regard to the damage potential and may not take reasonable measures to reduce risk.

When I question government disaster relief, I am commonly told, "We pay taxes!" What many people fail to realize is that government "insurance" is costly. Incentives that encourage efficiency in the private sector are absent in the public sector. Insulated from competition and managing someone else's money, government employees lack the incentives that motivate private sector efficiency in providing services. And government insurance isn't equitable. When the government acts as a universal insurer, everyone pays premiums in the form of higher taxes, even those who have chosen not to live in high-risk areas or have purchased private insurance. As things stand now, an uninsured family that has a $100,000 earthquake loss-given the current tax deductibility of casualty losses and depending on its tax bracket-may reduce its combined state and federal tax liability by more than $30,000. Congress' Joint Committee on Taxation estimates that the tax deductibility of uninsured casualty and theft losses reduces federal revenue by an average of $500 million a year. If the 10% floor is eliminated, the cost is expected to increase by more than $20 million a year.

Losses are also covered by FEMA's disaster housing assistance program, which issues checks in amounts up to $12,000 to cover uninsured quake-related losses. These checks are distributed without regard to income.

Finally, all or part of repairs related to uninsured losses may be financed with subsidized low-interest loans from the SBA. For the $100,000 loss, this amounts to a subsidy of several thousand dollars a year for the length of the loan, depending on the interest. Because these programs exist, it is not surprising that many homeowners choose not to insure their homes, choose large deductibles when they do buy insurance and fail to set aside savings that would help them cover their uninsured losses.

A pessimistic assessment is that these benefits are with us for the long term. Once the government bails people out, they come to expect it. It becomes politically impossible to change the benefits, no matter how costly they are to society. Because many remain uninsured, a major earthquake on the San Andreas Fault will burden state and federal budgets far in excess of the Northridge quake.

If I could change things, here is what I would propose. Beyond the immediate provision of basic rescue services, disaster assistance programs should require a financial means test, with all but the poorest excluded. Tax deductions for uninsured losses should be eliminated. Homeowners should assess their own willingness to bear risk and purchase the appropriate level of private insurance.