In the two summers that I worked for government in Washington during the mid-1960s, I reached two conclusions about government. First, most government workers are dedicated and believe they are providing for the common good. Second, most production of goods and services by government was hopelessly contained by customs, rules, and managers who were avoiding problems rather than solving them.

I realize these are blanket statements, but I cannot forget when my office colleagues would join me for two hour lunches and brainstorm how we could improve the operations of the Post Office. We would then finish our assigned tasks and write memos about how to improve postal operations.

Many times, this would require working until well into the night and on weekends. But we thought we were doing something useful in managing resources for the country. (Indeed, a memo proposing the creation of a public corporation for the Post Office came from those efforts).

Then a new office manager was assigned to oversee office activities. He quickly informed us that he did not care how many hours we worked as long as 8 of them were between 8:45 a.m. and 5:15 p.m. That ended the brainstorming, the memos and the enthusiasm for improving the government's use of the people's resources.

This inability by government management to be creative, reward initiative, and fire incompetence is the major reason why I do not agree with the Democrats in Congress that the security inspectors at airports should be federal employees.

Indeed, the arguments for federalizing that function have not been compelling. Clearly, the alternative of allowing the airlines to provide that function has not worked. Airlines are not in the security business.

Also, the finances of air transportation are so fragile that airlines probably cannot provide adequate security services. Instead, the airport authorities that operate airports should be responsible for the security. A ticket surtax or higher landing fees would then be used to finance the security services.

Government would determine what level of security is desirable and would provide regulators (investigators) to insure that those service levels are achieved. Government provides oversight very well but tends to provide production very poorly.

The problem is not that security currently goes to the lowest bidder. We build very good buildings using the lowest bid of those contractors deemed competent to do the job. We also inspect the construction work.

A similar method of establishing guidelines and testing the inspectors would provide at least as good a level of safety as a federalized force would do. Indeed, we are evolving into incentivizing private providers to meet safety objectives. When errors are made, the workers involved are suspended or fired. Those hiring these workers are fined.

Who do we fine if errors are made by a federalized force? Also, will we pay security people the same whether they work in Starkville, Mississippi or San Francisco?

Once again I remember my post office experience. Mail carriers were being paid the same in Mississippi and San Francisco. In one place, they were highly desired jobs. In the other, vacancies could not be filled.

To provide what is needed for airport security, we need to have incentives, flexibility, and accountability. Federalizing the airport security function provides none of that.

Now lets get on with providing a security system that is efficient and effective. We are achieving effectiveness (those knives and stun guns were uncovered before they were available to potential hijackers on planes) but we have abandoned efficiency.

I believe we can improve both (though my own longest time to clear security in the ten plane trips I have taken since September 11 is 45 minutes). However, I do not see that happening with a federalized force.