"Ban-the-Box" versus Alternative Screens

The “ban the box” (BTB) policies limiting an employer's ability to learn about job applicants’ criminal records are intended to improve employment outcomes for those with criminal records, with a secondary goal of reducing racial disparities in employment. However, since employees with a criminal record tend to have more workplace problems, it is valuable to employers to develop other screens to weed out these applicants. One possible screen is to reducing hiring among the demographic groups that include more ex-offenders.

In "The Unintended Consequences of “Ban the Box”: Statistical Discrimination and Employment Outcomes When Criminal Histories Are Hidden," Doleac and Hansen recently investigated this issue by examining the outcomes in about one hundred jurisdictions enacting these policies over the 2004-2014 period.

We find that BTB policies decrease the probability of employment by 3.4 percentage points (5.1%) for young, low-skilled black men.

The term for this is "statistical discrimination" because discrimination is statistically related to the characteristic you want to screen out. The irony is tragic, because as the policy harms the group it was intended to help.

Removing a noisy signal improves decision making: Army removes photos from promotion sheets

 From The Army Times:

The Army will no longer include official photos for officer selection boards, beginning in August, to help eliminate unconscious biases in the promotion process...

To their credit, the Army first ran an experiment to determine the consequence of the change and found that when the photo was removed:

...there was less variance between voters’ scoring, meaning voters ranked candidates more similarly across the board. After removing the photo, voters also took less time to make decisions on each individual file, and the outcomes for minorities and women improved.

A similar change, Ban the Box, a ban on employers asking about criminal background, saw some employers turn to race as a proxy for criminal background.  The difference seems to be in the value of the signal:  removing a noisy signal results in better decisions; but removing a valuable one does not.  

One student uses questions about criminal background to screen for honesty, which results in better hiring decisions at his firm.  Unsurprisingly, honesty seems like a valuable signal.  

HT:  Evan W.

Adverse Selection, the ACA, and Criminal Records

The Economist Magazine is devoting space to explaining some fundamental economic ideas.  They begin with information asymmetry, which leads to adverse selection, the topic of chapter 19.

Information asymmetry remains a tricky problem for policymakers. Adverse selection is plaguing America’s Affordable Care Act, better known as “Obamacare”. Fewer healthy people than expected have signed up to the government-sponsored insurance exchanges, which limit how much premiums can vary with risk. Insurers are making losses; as a result, they are raising prices substantially (or pulling out altogether). Critics say those price rises will drive away more healthy customers, leading to a “death spiral”. Information economics should also give pause to the “ban the box” campaign, which seeks to forbid employers from asking about job-applicants’ criminal records prior to interview. Having no criminal record is a positive signal; removing that information makes information asymmetry worse. Recent research suggests that banning the box causes American firms to discriminate by race, such that employment of low-skilled black and Hispanic men falls. Adverse selection indeed.

Most significant economic research of 2018

Good article summarizing results from 10 papers. Here is the one that speaks to Chapter 19, Adverse Selection:

Ban the Box, Criminal Records, and Racial Discrimination: A Field Experiment by Amanda Agan and Sonja Starr

Main finding: Restricting employers from asking about their criminal history can lead to more discrimination based on race.