The U.S. "diversity visa" program -- 50,000 green cards allocated by lottery each year to applicants who need only a high school education to qualify -- has been generating embarrassing headlines lately. Shortly after notifying this year's lottery winners -- out of nearly 15 million applicants -- of their entitlement to move permanently to the United States, the government discovered a computer glitch that produced erroneous results. (Instead of a random drawing, 90 percent of the winners were from entries submitted on the first two days of the 30-day registration period.) The government told the winners, who likely were busily severing their ties to families, friends, jobs and communities in preparation for their imminent move to the United States, that their visas were rescinded.

But the visa fiasco obscures a more fundamental objection to the program: It was misconceived at its inception.

Immigration is in many ways the lifeblood, future and salvation of an aging, technology-driven America. The stakes could hardly be higher in getting our immigration policy right and bringing in those who can provide what we need: skills, entrepreneurs, close family members and investment.

A green card to the United States is one of the most valuable pieces of paper in the history of the world. So why would we want to give roughly 5 percent of them each year to people who, for all we know, have nothing more to offer America than a high school education, a winning ticket and (in many cases) an agent they paid to help them game the lottery system?

The solution is straightforward: Abolish the program and use those 50,000 visas (or more) to promote carefully defined national interests, particularly in more high-skilled immigrants who, many studies show, produce jobs, innovation and new businesses.

Indeed, this would be an opportunity to experiment with two new ways to improve visa allocation. First, the government could auction some visas to the highest bidders, just as it auctions scarce broadcast spectrum. The winners are likely to be either employers willing to pay for needed skilled workers or individuals who can finance their bids based on their ability to create value and thus earn good incomes. A second experiment would allocate some visas through a points system -- long used by Canada and elsewhere -- that awards points for English fluency, job skills, family or other ties to American society, and other predictors of successful assimilation, with visas given to those with the highest point totals.

These approaches promise to teach us more about which kinds of legal immigration reforms can best advance our national interests.

Peter H. Schuck, a Yale Law School professor, is co-editor (with James Q. Wilson) of "Understanding America: The Anatomy of an Exceptional Nation."