The Latest Social Insurance Freak Out

The enemies of social insurance are many and persistent. They pump out trumped up story after trumped up story, hoping to gain traction in a perpetual quest to take from the poor and vulnerable in order to give to the rich and powerful.

A couple of years ago the story was Social Security bankrupting, which was ridiculous then as it is now and seems to have gone away. Medicare was a big deal to the haters, but its cost problems were all about health care inflation, and now hardly a peep is said about it. In December of last year, Nicholas Kristof set off a storm by reporting on an imaginary underclass of Kentuckian disability insurance cheats he heard about second-hand from some local woman. This fevered nonsense was shortly followed by a widely debunked NPR story also arguing that there is widespread disability insurance fraud. And then it was the Weekly Standard last month in horror that food stamp “fraud” (read: not actually fraud) rose from 1.0% to 1.3%. Out of control!

And now the latest in this tired saga is, once again, disability insurance. The Government Accountability Office put a report about potential disability insurance overpayments, and the response from the hyena media was swift and reckless, as is typical for social insurance stories. The Associated Press ran the story with the headline “Investigation: 36,000 game disability system." Other articles ran with headlines mentioning the dollar figure, like this one saying “Social Security Makes $1.3 Billion in Overpayments."

As a rule, big numbers that are not presented as fractions of larger numbers or otherwise put into meaningful context are hackish propaganda. And that’s precisely what we see here. The 36,000 figure seems high, I guess, but as the report explains, it actually represents 0.4% of disability insurance recipients. The $1.3 billion also seems big, I guess, but it represents somewhere near 1% of disability insurance benefit payments. Honest headlines would include those figures, but those figures are so unremarkable that, were they to be honestly reported on, nobody would bother to read them.

The problems don’t stop there either. The AP headline claims that the recipients that are the subject of this report “game” the disability system. But in fact, the report says nothing about anyone gaming anything. It suggests that there may be failures by the Social Security Administration in running the program. To game a program requires that you intentionally try to get around its requirements, not that the program administrator applied them in mistaken ways. But to put it on the SSA would not serve the favored narrative of there being some massive cheating ring going on, the narrative the social insurance haters love so much.

As far as the report goes, it’s important to note that GAO does not even make a hugely confident case for the minor overpayment estimates it presents. The GAO emphasizes that these are potential overpayments that their method, by itself, cannot really confirm. Moreover, the estimates have pretty serious margins of error. The number of potential overpaid recipients has a margin of error of 19.4% and the dollar figure has a margin of error of 27.3%. So we have a fairly low confidence assertion of potential over payment on a tiny fraction of disability insurance recipients with high margins of error. And that’s just on the face of the report. At this point the appropriate response is: who cares?

But you can go further still. Any time one of these stories comes out, those in the know shoot off an email to Rebecca Vallas or Harrold Pollock to see what’s up. In an email exchange, Vallas indicated to me that the GAO report likely counted income that it should not have counted. The way the GAO generated the report was to match DI recipients to a national database of new hires, comparing the income reported in the database to the income requirements of the DI program. But not all income that may show up in the database of new hires is counted towards the income requirements of the DI program. In particular, impairment-related work expenses and earnings from unsuccessful work attempts are excluded from consideration for the purposes of the DI program. So even the GAO’s minuscule estimate that 0.4% of DI recipients are receiving over payments likely overshoots the real amount, if indeed there are any.

The lesson here is, as always, don’t believe the goofy hype any time one of these social insurance stories catches fire. It is the same reheated nonsense again and again driven, not by any real concern for social programs, but by a rather transparent political agenda.