In the world of Washington, incentives—carrots and sticks—seem to be the answer for everything, including how to get people on disability back to work. But a new study suggests the problem is the same one the Americans With Disabilities Act identified 25 years ago: discrimination.

I took a certain amount of impolite criticism for my last post, in which I decried the “veritable national jihad” against disability fraudsters. The amount of abuse in Social Security’s Disability Insurance (DI) system isn’t anything like the monster it’s made out to be, and it’s unfortunate, to say the least, that the just-passed Bipartisan Budget Act throws more resources at ferreting out fraud and creating stiffer penalties for false benefit claims. For calling the vast overreaction to disability fraud a jihad, I was accused of insulting conservatives and indulging in hyperbole.

I’m not sorry to have used strong language to describe a years-long campaign against DI that’s out of all proportion to the size of the problem (would that as much effort was going into exposing and cracking down on the Pentagon’s cozy relationships with its contractors) by people who seem to have no idea of the challenges facing working people with disabilities. (The Washington Post was at it again today.)

But if the goal is to get people back to work who have something to contribute and would be better off for it, the real issue is where “incentives” ought to be applied: to disability recipients? or to their potential employers? Washington’s working assumption for a long time—borne out in the new budget act—is that it’s the beneficiaries who need tough love.

A story in today’s New York Times reports on a new study by researchers at Syracuse and Rutgers universities that suggests Washington’s got it backwards. According to the researchers, Lisa Schur of Rutgers and Mason Ameri of Syracuse, discrimination against the disabled is alive and well, even for individuals who apply for jobs in fields where they have strong qualifications. In fact, the more experienced you are, the worse your chances of landing a job if you have a disability. And the nature of the disability doesn’t seem to matter too much—whether it’s a spinal cord injury or Asperger’s Syndrome, employers are just as likely to shy away.

Earlier studies had suggested that better qualifications might help disabled candidates overcome employment discrimination, but the researchers found the opposite. Employers were about 34 percent less likely to show interest in an experienced disabled candidate, but only about 15 percent less likely to express interest in a disabled candidate just starting out his or her career….

“We created people who were truly experts in that profession,” said Ameri. “We thought the employer would want to at least speak to this person, shoot an email, send a phone call, see if I could put a face to a name.” For the gap between disabled and nondisabled to be larger among experienced candidates than among novice candidates, he said, came as a surprise.

The researchers sent cover letters and fabricated resumés to employers seeking to fill accounting jobs; the bosses expressed interest in candidates who mentioned having a disability about 26% less often than for other candidates. One possible answer: disabled workers often require special accommodations, such as wheelchair ramps. But then how to explain the fact that Aspergers, which isn’t a physical impairment, was just as likely to scare off employers as a spinal injury? That points to “a general bias against people with disabilities” as the culprit.

Today, only 34% of working people with disabilities are employed, compared with 74% of others—this despite the fact that the Americans With Disabilities Act of 1990 banned discrimination. Does that mean the law didn’t do any good? No, Ameri and Schur found. Businesses with less than 15 employees—which are not covered by the Act—were least likely to express interest in the inquiries the researchers sent. Publicly traded companies, and companies that receive federal contracts, were more likely to show interest. This is to be expected, based on what we know about prejudice and discrimination in plenty of other areas: that it will continue as long as there are no penalties to deter it and no daylight to expose it to view.

This is a real problem: Disability benefits are paltry, and despite the Reaganesque campaign to vilify them, the vast majority of disabled would prefer to work—if they could. What the new study implies, however, is that Washington has been wrong for a long time about how to get the disabled back to work. “Nudging” them back into the job market, either by making it harder to get benefits, shortening the benefits period for some disabled persons, stiffening the enforcement of anti-fraud laws, or even offering them a carrot by increasing the amount they are allowed to earn by working while still receiving benefits, is not where real action is. The problem is not the disabled themselves, but employers who are deeply biased against them. The government should be focusing on changing their behavior, not working people’s.