Another Natural Experiment: Removing SSI Payments

Here's another natural experiment: In 1996, due to Clinton's "welfare reforms," 18-year-olds were knocked off the Supplemental Security Income rolls when they became adults, a change from how they had been treated in the past. About 40% of the newly minted adults were removed from payments after their birthdays, particularly young people with mental and behavioral conditions like ADHD. (As opposed to young people with intellectual disabilities, who were much more likely to be allowed to remain on the rolls.)

What happened with those young people vs. young people who had been born earlier and got a softer transition out of their teen years?

A new study gives us the answer in aggregate:

Using data from the Social Security Administration and the Criminal Justice Administrative Records System, researchers estimated the effect of losing Supplemental Security Income benefits at age 18 on criminal justice and employment outcomes over the next two decades....

They found that terminating the cash welfare benefits of these young adults increased the number of criminal charges by 20% over the next two decades. The increase was concentrated in what the authors call “income-generating crimes,” like theft, burglary, fraud/forgery, and prostitution. As a result of the increase in criminal charges, the annual likelihood of incarceration increased by 60%.

The effect was still being seen after two decades. Overall,

In response to losing benefits, youth were twice as likely to be charged with an illicit income-generating offense than they were to maintain steady employment.

Okay, so the human cost to the kids was a net negative. What about the literal financial cost to society?

While each person removed from the program in 1996 saved the government some spending on SSI and Medicaid over the next two decades, each removal also created additional police, court, and incarceration costs. Based on the authors’ calculations, the administrative costs of crime alone almost eliminated the cost savings of removing young adults from the program. (emphasis added)

"Administrative costs" is the authors' umbrella term for police enforcement, court costs, and incarceration — in other words, the direct cost to taxpayers. It does not include what the study's authors call "costs to victims," which the authors priced at more than twice the direct cost to taxpayers. 

And of course, there's no mention of the lost opportunity cost of each person who might have become a noncriminal member of society if they had been supported a bit longer into adulthood, as people with a similar profile had been in previous decades. Which in turn has effects on their children and other family members.

__

The full paper, published in the Quarterly Journal of Economics, is here.

__

Here are a couple of my earlier posts about natural experiments from 2017 and from 2015.