Tim Pawlenty has announced his intention to reform Medicare by applying Pavlovian theory to the practice of medicine.

During a presidential campaign stop Monday at Point of Grace Church in Waukee, Iowa, Pawlenty, the former Minnesota governor, proposed "performance pay" to improve the efficiency and quality of health care.

"We will begin to move providers from getting paid not just for the volume of procedures they crank out, but whether people are actually getting healthier and getting better," the Republican told the Des Moines Register.

The theory presupposes that the nation's hospitals and clinics are staffed by slackers who care less about their patients' well-being than they do about the bottom line. So why not tie that bottom line to patient well-being?

In classic behavioral-psychology experiments, laboratory animals learn to perform tricks in exchange for treats. And at firms like Goldman Sachs, billions in bonuses were handed out - even after the firm faltered - because, taxpayers were told, that was the only way to keep the best and most talented people.

Knowing how to push the right buttons has always been the key to success. (You knew that salivating dogs and investment bankers had something primal in common, didn't you?)

Building on this logic, the government has tried to goose educational performance by linking funding to student test scores.

Makes you wonder what else cash incentives could improve.

Hostage negotiation? They often don't go so well. What if experts were paid only when captives walked out happy?

Bomb defusing? Another iffy enterprise. Tell the guys approaching the hurt locker that if they want a raise, those IEDs had better not explode.

Parenting teenagers? Few efforts yield such pitiful results. Tell parents that their tax deductions will be based on said offspring's exemplary behavior. Money in the bank for families producing quiet, polite, studious, sober, trustworthy, and sexually abstinent adolescents who fold their laundry and return the car with a full tank.

"As a general rule, if you're going to give a reward, it's better for it to be performance-based," says Steven Reiss, emeritus psychology professor at Ohio State University and a national expert on motivational theory.

Two essential factors must be considered. The outcome has to be in the control of the person being evaluated, he says. And the goals must be clear and achievable.

It makes sense to offer raises to business managers who reduce energy use in a factory or improve customer service, but ordering doctors to make people healthy? "I don't know if that's a very good idea," Reiss says.

Control and clarity are missing.

The course of an illness is often out of a doctor's control. And the smartest and most skilled physicians are usually sent the riskiest cases, in which poor outcomes are most likely.

"Doctors can do a very good job, and the patient might not do well because of the disease," Reiss says. "Especially with Medicare, they're dealing mostly with chronic diseases, illnesses that consume a big percentage of health-care dollars, and some are incurable."

Unrealistic goals, he says, can backfire.

"If the rewards are perceived to be unfair, you encourage cheating and fudging."

Consider No Child Left Behind. Test scores went up, proving that money talks. But does it speak the truth?

Diane Ravitch, assistant secretary of education under President George H.W. Bush, doesn't think so.

"It turns out, as a result of putting so much emphasis on the test scores, there's a lot of cheating going on. There's a lot of gaming the system,"  Ravitch, author of The Death and Life of the Great American School System, told NPR last year. "Instead of raising standards, it's actually lowered standards, because many states have dumbed down their tests or changed the scoring of their tests to say that more kids are passing than actually are."

Same goes in medicine.

"If you start telling doctors and hospitals that they have to cure people," Reiss says, "they may feel that's very unfair. . . . You can't have a reward or incentive system if people feel it's unjust, or they will put their energy into how to get around it."

Before Pawlenty or any government officials propose another performance-based policy, maybe they should test the theory out on themselves.

Make taxpayer money for their salaries, health care, and office staff contingent on their ability to create bipartisan legislation.

Let's see if the country gets healthier and better.