On Tuesday, the Defense Department unveiled its "don't ask, don't tell" reform plan. The strategy is simple: slow progress toward ending a policy, and repealing a law, that doesn't work.

The idea is to modify enforcement of "don't ask, don't tell" by, among other measures, disallowing certain third-party "outings" from being used against gay troops. New rules may also require that a two-star officer approve any discharge. Depending on how they're applied, these changes could mean the beginning of officially tolerated service by known gays and lesbians. In some cases, gay troops could be honest with their peers, who could not rat them out.

But a failed policy will still be in place; 66,000 gay, lesbian and bisexual troops will continue to serve in fear of needless discharge, and the U.S. could still be forced to fire soldiers, sailors, airmen and Marines it can ill afford to lose.

The strategy for dealing with that reality? A year of further study. The problem is, the issue has been studied for half a century. "Further study" is nothing but a delaying tactic.

According to poll data, most Americans agree that "don't ask, don't tell" is unjust and should end. But many people don't grasp that the 17-year-old policy is not just unfair, it is a colossal failure that harms military "cohesion and readiness," the very thing it is supposed to protect. The insistence by opponents of reform, such as John McCain, that the policy is a "success," that it is "working well," reflects a profound detachment from the situation on the ground.

For starters, two-thirds of military members already know or suspect that there are gays in their units, so the policy has failed to achieve even its most basic goal: to protect morale and cohesion by shielding straight troops from knowledge of gay troops.

The policy has also failed to preserve desperately needed skilled personnel. Since the law's inception, roughly 13,500 gay, lesbian and bisexual service members have been discharged. According to the Government Accountability Office, nearly 800 of them had "critical skills," including more than 60 Arabic speakers. In the meantime, the military has granted an increasing number of "moral waivers" to ex-convicts and drug abusers to fill slots in a force stretched thin by two wars.

According to the military's own studies, the policy (not the presence of gays) is undermining trust and integrity in the force by mandating dishonesty, a point reiterated Tuesday by Adm. Michael G. Mullen, chairman of the Joint Chiefs of Staff.

The Pentagon's hesitant roll toward repeal makes sense only if a case can be made that, as bad as "don't ask, don't tell" is, the path out of it is best trod slowly. Is this the case?

Research on institutional change, including our own military's experience with racial integration, answers this question clearly. The two most important factors in a transition like this are decisive leadership and a single code of conduct for all personnel. A major study by the Rand Corp. in 1993 found that openly gay service could work well, but it would be important for the senior military leadership to throw their weight behind it.

The 500-page study said that a successful new policy must be "decided upon and implemented as quickly as possible" to avoid anxiety and uncertainty in the field. Finally, it said that "fast and pervasive change will signal commitment to the (new) policy," while "incremental changes would likely be viewed as experimental" and weaken compliance.

In the face of such research and experience, why is the military — and the Obama administration — trying to move slowly? Certainly political considerations and the moral opposition of many in the military community play a role, along with the slow grind of legislative realities. But the president has the authority to invoke his "stop-loss" power to bring discharges to a halt overnight, and it would be better for national security and individual troops if he would use it.

It is heartening to see movement toward ending the policy. If the Defense Department's changes are adopted, they must be implemented decisively to ensure success. Better yet, it should move decisively to end the policy once and for all.

Nathaniel Frank is the author of "Unfriendly Fire: How the Gay Ban Undermines the Military and Weakens America." He wrote this commentary for the Los Angeles Times.