“I used to be a drunk, but now I’m bipolar.” A guy who was applying for Social Security disability benefits once said these words, which sum up the problem with this entitlement program.

Thursday’s Wall Street Journal article about a “just say yes” Social Security administrative law judge (ALJ) in West Virginia who granted benefits to 100% of the people on his case docket during this fiscal year highlights one problem with the program, but its roots are deeper.

The Social Security Disability Insurance (SSDI) program, which was created in the 1950s, was designed to provide cash benefits to workers who became unable to work due to a physical impairment. (Supplemental Security Income, or SSI, a Nixon-era law, extended these benefits to the disabled poor.)

In the beginning, SSDI payments mostly went to people over age 50 who had limited education, engaged in low-skill work, and whose bodies were worn out or wracked with disease — factors easily verified by evidence. Over the decades, the Social Security Administration (SSA) was confronted with more cases involving mental illness. Again, some cases (e.g., schizophrenia) were pretty clear-cut. Others, involving mood and affective disorders (e.g., unable to work due to anger management problems) were more murky.

In 1984 President Reagan (yes, Reagan) signed a law that further liberalized reviews of mental disorders for SSDI and SSI benefits. Earlier in his administration (at the behest of David Stockman), SSA ramped up its audits on people receiving disability benefits — but the resulting blowback in the press was so harsh that Reagan accepted the 1984 changes as a way to end the negative publicity. Thus in 1984 about 64,000 people were awarded SSDI benefits due to “mental disorders”; by 1986 it had jumped to about 124,000. In 2009, it was over 216,000, resulting in an unsurprising cumulative effect of over 2.4 million people receiving SSDI benefits for “mental disorders other than retardation.” (SSA has oodles of SSDI stats here, but, oddly, not so much for SSI.)

Now you know why that guy was claiming to be bipolar — because being an alcoholic wouldn’t qualify him for benefits.

At this point, you might wonder why Social Security’s ALJs have so much leeway in the gray areas of mental impairments, especially when the agency has millions of data points to be gleaned from decades of decisions. Couldn’t SSA develop a checklist for the judges to use and monitor decisions to prevent judges from approving or disapproving too many cases? Indeed, SSA has tried, but the 1946 Administrative Procedures Act gives these judges “decisional independence.” In a practical sense this means the commissioner of Social Security cannot call a judge on the carpet for awarding or rejecting too many cases, or for not processing his or her share of the workload. (Rejecting more cases than the statistical norm adds to the workload, as these cases are inevitably appealed. For instance, my home state, Delaware, has an ALJ with an extremely low approval rate (16%); she’s a major chokepoint in the system.)

Currently, the only real brakes on a poorly performing judge are the threat of an audit by the inspector general or a reassignment to a hinterlands post with few amenities and lousy airline service. To make disciplining even more difficult, Social Security’s ALJs are unionized — represented by the AFL-CIO for “protection” even though ALJs are civil servants with six-figure incomes and excellent benefits.

To make matters more confusing, who qualifies for Social Security disability (in some cases) varies by federal appellate court circuit. This is a result of differing court decisions; thus, an ALJ located in one circuit may have to decide a case differently than an ALJ in another circuit facing an identical case.

Senators Coburn and Hatch are worried that “just say yes” ALJs are turning Social Security Disability into an extended unemployment compensation program. The House Ways and Means Committee says they’re investigating this situation. They all have good intentions, but I doubt it will make much difference.

As far back as the Carter administration, the executive branch has struggled unsuccessfully to find an administrative solution that was compatible with the Administrative Procedures Act. After 40+ years of futility, it’s time for Congress to enact a legislative solution.

We need statutory changes that reflect 21st-century realities, one of which is that solid statistics exist about how Social Security cases should be decided. One change could be a more narrow statutory definition of Social Security ALJ “decisional independence.” Right now, it serves as a loophole for erratic, irrational, corrupt, or just plain lazy ALJs.

Another would be tightening the rules with clearer standards for deciding “mood and affective disorder” cases to help eliminate a lot of ALJ guesswork in this area.

One radical solution would be to eliminate ALJs altogether and replace them with hearing officers who could be subject to management requirements regarding case processing procedures and workload. Some feel a Social Security federal court is needed for nationally uniform decisions, instead of the current patchwork of circuit decisions.

The challenge will be to balance the need for consistency and efficiency with the demands of the disability lobby, which wants a system that has the most opportunities for the most generous outcomes possible.

It will take legislative action to reform the disability process. Adding yet another GAO report to the stack won’t do it. Americans deserve a better Social Security disability process than the cumbersome, lengthy, and capricious one we have today.

Joanne Butler is a senior economics fellow at the Caesar Rodney Institute of Delaware. You can email her at joanne-butler@comcast.net.