'Ban the Box' a Pandora's Box?

Efforts to improve job outcomes for felons released back into society have been mixed, and current legislative fixes seem to just create new problems.

By Peter Cappelli

When we see something wrong in society, we have a tendency to turn to legislation to fix it. That is particularly so with discrimination, a social problem that, over the past generation, we have created more laws and regulations than ever before to prohibit employers from contributing to, particularly when it comes to discriminating against protected groups.

The problem is that the forces that create undesirable outcomes such as discrimination are complex, which means that it's harder to find a singular point of entry to intervene with regulations. Even worse is the problem of unintended consequences, in which the regulations lead to changes in behavior that we didn't expect, and those changes then create their own problems.

Last month, I wrote about evidence that diversity training programs lead to worse diversity outcomes because they tend to irritate the intended audiences (white men), which makes them react poorly to the message. This month, the topic is similar: efforts to improve job outcomes for felons released back into society from incarceration.

"Ban the box" legislation -- which is now in place in 24 states, many cities and the federal government -- tries to improve the odds that ex-felons will get jobs by prohibiting employers from asking about their criminal record until the later stage of the hiring process. In other words, employers can't just rule out any candidate with a criminal record -- as one might by adding that criterion to an applicant-tracking system -- before reading through his or her application and, if otherwise qualified, interviewing him or her. The premise is if employers like candidates before finding out they have criminal records, they will care less about that fact and be more likely to hire them.

What's not to like about this approach? The regulations don't say you can't factor in a candidate's criminal record or even that you can't reject a candidate just because he or she has a record. They just say "Wait."

The problem is that, while regulations can prevent you from finding out about a candidate's criminal record at a particular time, they can't prevent you from thinking about it. Just thinking about it may be enough to cause a problem. The problem with thinking about it is called "statistical discrimination." Translated into plain English, that means "Can I guess who has a criminal record if I don't know for sure?"

A good guess as to who has been a felon would be a man, especially a young man. If you're looking at probabilities, you would add African-American and possibly Hispanic. lf you guess who has a criminal record, and you act on that guess, then we have another form of discrimination. If you don't want to hire ex-felons, you disproportionately rule out applicants who are young, African-American and Hispanic men. That's the unintended consequence.

Several studies years ago looked to see what happened to the hiring of young African-American and Hispanic men once it became possible to check easily online whether they had criminal records. The good news was that hiring went up: Once employers could find out, they no longer assumed that all young black and Hispanic men are possibly ex-felons because you could check and find out.  

What happens when we introduce "ban the box" legislation -- which, again, doesn't make it impossible to know but just delays knowing until later in the hiring process? Several recent studies have looked at this question, including a new National Bureau of Economic Research working paper by Jennifer L. Doleac and Benjamin Hansen titled Does Ban the Box Help or Hurt Low-Skill Workers? Statistical Discrimination and Employment Outcomes When Criminal Histories Are Hidden.

I bet you can guess the answer. The good news is that more ex-felons seem to have gotten jobs, but the hiring of young black and Hispanic men overall declined. In other words, we swapped one form of discrimination for another. The difference is that the form of discrimination that went up (against young black and Hispanic men) is less justifiable than the one that declined (against ex-felons).

It wasn't supposed to work that way. The problem is people don't behave the way the legislation anticipated. We don't wait until the law allows us to find out about criminal records. We start guessing.

Given this kind of evidence, it is hard to argue for more "ban the box" legislation. If we want to help ex-felons make the transition back into society and into jobs, and we should, some other approach makes more sense.    

Peter Cappelli is the George W. Taylor Professor of Management and director of the Center for Human Resources at The Wharton School of the University of Pennsylvania in Philadelphia. His latest book is "Will College Pay Off? A Guide to the Most Important Financial Decision You'll Ever Make."


Aug 8, 2016
Copyright 2017© LRP Publications