One of the most hotly debated measures during Oklahoma's 2008 legislative session was Senate Bill 1537-colloquially referred to as "Nick's Law"-which would have required insurance companies doing business in Oklahoma to cover autism. Though the bill made it through the Senate, it did not make it out of committee in the House.

Autism is a brain-development disorder that affects approximately 6,000 Oklahoma children. It manifests itself as pattern of symptoms of which the primary characteristics are problems with communication, limited interests, and repetitive behavior. While autism is not curable, early intervention and treatment of the type that would be mandated under this bill have been shown to be important to the long-term life prospects of autistic children.

For proponents of an autism mandate, the existence of these clear benefits to autistic children makes the case for an autism mandate self-evident. "There's going to be several hundred kids that are going to be basically locked into autism because they were unable to receive services and treatment because of this delay," Wayne Rohde told The Oklahoman. (Mr. Rohde's son, Nick, is the namesake of Nick's Law.) This perspective is both understandable and important because it not only conveys the hardship borne by families with autistic children, but also how far our current "health insurance" is from being insurance.

Properly understood, insurance is about mitigating risk, an inherent part of life. As Richard Zeckhauser points out in the Fortune Encyclopedia of Economics, risk can be dealt with in two ways. First, people can take steps to reduce their exposure to a risk. An example of this type of risk mitigation would be a smoker who gives up smoking in order to reduce the likelihood of cancer in their future. The second way to minimize risk is through the purchase of insurance, which spreads risk across similarly situated individuals.

The first method is useless in the case of autism, a disorder whose causes are not well understood. This only increases the importance of health insurance, yet many families find that their insurance does not cover autism-related health care expenses. The inability to purchase in the marketplace an insurance policy that protects them against the high costs of autism-related treatments is the source of the political pressure for this mandate.

Yet no one is asking why insurance companies do not offer this coverage. It cannot be because autism is "too costly" of a condition to insure against. First of all, saying something is too costly is meaningless without reference to its value. One actuarial study mentioned in an Associated Press report puts the cost of autism coverage under the proposed autism mandate at $1.66 per month. Given the financial repercussions of an autism diagnosis, I certainly would be willing to spend $20 a year for coverage and I suspect many others would too-certainly anyone who knows someone with autism.

So why is this coverage not offered? While I cannot know for sure, I suspect the reason is that very few of us purchase our insurance individually. Instead, we get it through a third party, such as our employer. When combined with the fact that modern health insurance is not just insurance against catastrophic health care expenses but also prepayment and cost-shifting of routine expenses, the end result is that many conditions do not get insured.

The trade-offs employers face when evaluating different insurance plans are different from those faced by individuals. Companies care about how their employees value the health plan as a benefit, not just as insurance. In my experience, people get angry if their health plans do not cover routine, foreseeable expenses, such as annual check-ups.

In addition, autism coverage would likely only be a concern for households with children younger than five. Therefore, the average employee probably does not care about autism coverage. As a result, employers are likely acting rationally by maximizing coverage of regular conditions instead of insuring against rare ones.

If this is true, then the ideal public policy would be reducing the incentive for employers to offer health care benefits in lieu of wages by eliminating its favored tax treatment. That is unlikely to happen, however, so what can Oklahoma policymakers do to help improve upon the current situation?

A potential downside to health care mandates (such as an autism mandate) is that they drive up insurance costs and end up pricing people out of the market. The costs of autism coverage have to be borne by someone, and it is folly to think that the full incidence of any mandate will fall entirely on insurance company profits.

Some of the cost increase will be passed along to consumers in the form of higher premiums. This is not the real cost of the mandate, however, as this is just a transfer of resources. The real cost of the mandate will come from those individuals who are priced out of the market as a result of the higher premiums. Thus the size and impact of any proposed autism mandate is an important issue and a primary reason for careful study.

Another, more innovative, solution comes to mind after reading a May 30 story in The Oklahoman headlined "What Promise Does Ohio Hold for Autistic Boy From Oklahoma?" It seems that an Oklahoma family decided to respond to the failed passage of "Nick's Law" by selling their home and moving to Ohio.

Why would they undertake such a drastic move? Because Ohio has a school voucher program for autistic children. In Ohio, parents of autistic children can choose to have their children educated through their local school district or through private special education programs. For those who choose an alternative school, up to $20,000 of taxpayer money follows the child.

While an autism scholarship program for Oklahoma would not address all the financial issues surrounding autism treatment, it would have several appealing qualities.

First, unlike a mandate, it does not harm others by crowding out private insurance.

Second, a tax-credit scholarship program would focus attention on the issue of early diagnosis and treatment. (Thanks to a provision in Oklahoma's constitution known as the Blaine Amendment, such a program in Oklahoma would likely need to be funded through a charitable tax credit rather than a voucher.)

Third, a scholarship program provides needed flexibility for parents to seek and craft the best treatment solution for their child. Given the fact that autism is a bundle of symptoms of varying severity, the flexibility that choice allows seems especially important.

Fourth, any tax revenue lost because of charitable donations to an autism scholarship program would be, at least in part, offset by the fact that students using the scholarships would no longer be enrolled in the public schools.

Economics teaches us there are very few "free lunches" when it comes to public policy, which often involves trading off the benefits received by one group against the costs imposed on another. Nick's Law, for example, raised concern and uncertainty among policymakers about who might be harmed by rising insurance costs.

A tax credit for donations to scholarship programs geared to autistic children, however, might be close to a free lunch for the reasons described. Though it's no panacea, it is a simple and elegant solution that would help a large number of families.

Joshua Hall (Ph.D., West Virginia University) is an assistant professor of economics at Beloit College.