With the economy mired in its longest period of high unemployment since the Great Depression, the richest one percent getting near all of the gains from economic growth, and a thoroughly corrupt financial sector surviving the crisis largely intact, the thoughts of folks like Washington Post columnist Charles Lane naturally turn to the Social Security disability program. Lane is upset because the cost of the program has been rising rapidly.

The most immediate reason for this increase is the economic downturn brought about by the collapse of the housing bubble. However, there has been a longer term rise, which is mostly due to the aging of the workforce and an increase in the percentage of women who have worked long enough to qualify for benefits. But there also have been an increase due to other factors, which has Lane especially angry.

He wrote on the topic last summer, but decided to give it another stab today in the context of telling us that we should emulate Germany to get back to low unemployment. (Lane touts the Hartz Reforms, which were intended to weaken workers' bargaining power, as the main explanation for Germany's current 5.4 percent unemployment rate. However, the more obvious factor is its system of short-work, which encourages employers to keep workers on the payroll working fewer hours rather than laying them off. The German government began to promote this system at the start of the downturn. As a result, its unemployment rate fell from 7.8 percent in 2008 to its current 5.4 percent even though its growth over the last five years has been virtually identical to growth in the United States.)

Lane's big club in his attack on the disability system is a new study from the University of Michigan. He quotes from the study:

"the employment rate of new beneficiaries would have been 28 percentage points higher in the absence of benefit receipt."

He then adds:

"SSDI is one reason, in addition to recession and aging, that the U.S. ratio of employment to population declined from 62.5 percent to 58.5 percent in the past 10 years."

This sure sounds like an interesting study. If we go to the study itself, we find in the abstract:

"We find that among the estimated 23% of applicants on the margin of program entry, employment would have been 28 percentage points higher had they not received benefits."

This means that we would see an increase in employment rates of 28 percentage points among the 23 percent who are considered marginal applicants if they had not received benefiits. That translates into an increase in employment among all applicants of 6.4 percentage points. If we applied this to the entire population of 9 million workers getting disability, it would mean that employment would rise by about 580,000, or just over 0.25 percentage points of the civilian population.

It is worth noting that these estimates were based mostly on the pre-recession period when it would have been easier for people with disabilities to find jobs. It is also worth noting that the 28 percentage point figure refers to change in employment status after two years. The study found that the difference fell to 16 percentage points after 4 years (p 22). 

It is also worth noting that the study found that the difference in employment of those exceeding the Social Security Administration's $1,040 a month "substantial gainful activity" threshold was just 19 percentage points in year two. The study also found that earnings for the marginal applicants who were denied benefits averaged between 25-50 percent of their pre-application earnings.

These findings would suggest that cracking down on "abusers" may not go far towards Lane's goal of saving money on Medicare. (Disability beneficiaries are eligible for Medicare after 2 years.) A substantial portion of the people who are turned down for DI are likely to have incomes low enough to qualify for Medicaid. That would still leave the government stuck with the tab.

In addition, the actual earnings path of the people denied disability indicates that the vast majority are not earning much money even when we show them tough love and deny them disability benefits. This might suggest that many of the people denied benefits really do suffer from disabilities that make it difficult for them to work.

In short, the paper cited by Lane implies that the bonanza for the government from cracking the whip is much smaller than he claims. It also suggests that any efforts to further tighten eligibility (only 40 percent of claims are approved) is likely to lead to many more people with real disabilities being denied benefits.

This doesn't mean that the system cannot be usefully reformed. Certainly a faster application process would be beneficial. If applicants are not going to get benefits, it is helpful for them to know this as soon as possible, since most do not work while waiting. In addition, it would be a positive step if more people on disability could return to work if their health improves. (There are some experimental programs in place to test different mechanisms.) However the story of massive abuse bankrupting the government and leading to plunging employment simply does not fit the data.

Addendum: Michael Cushman points out that Germany may not be the model that Lane is looking for on this issue. According to Eurostat it spends roughly 1.7 percent of its GDP on its disability insurance program. By comparison, the United States spends less than 0.9 percent of GDP.