Several papers have estimated that "ban-the-box" laws, restricting employers from asking about employee criminal histories, have adverse effects on Blacks because they incentivize statistical discrimination.
A new paper found that the evidence for this has evaporated post-2014.
@captgouda24
It does not appear that Ban-the-Box laws (which prohibit asking about criminal history) actually have much of an effect on the employment of Blacks. An update to strong claims that banning asking discriminates against Blacks. https://t.co/fnV5mvYmKF
The estimated effect of ban-the-box laws for Blacks turned positive and nonsignificant and the estimate for Hispanics remained as nonsignificant as ever, while the estimate for Whites became a resoundingly clear and precise null.
Hispanic results are still meaningfully ambiguous
The earlier papers finding this policy backfired and induced statistical discrimination against non-criminal Blacks can be found here:
The initial time period result for this new paper (2004-14) replicates the estimate from Doleac and Hansen (2020) exactly, but it is nonsignificant here, where it was marginally significant for them.
Marginally significant results are tenuous, so it's no surprise it was fragile.
Overall, I think I'm updating to
"This effect was unlikely to find strong support (meaningful effect/statistically strong evidence). There might be small harms to 'ban the box' policies, they're just dainty. If the goal is to help criminals, the tradeoff seems worthwhile."
Several papers have estimated that "ban-the-box" laws, restricting employers from asking about employee criminal histories, have adverse effects on Blacks because they incentivize statistical discrimination.
A new paper found that the evidence for this has evaporated post-2014.The estimated effect of ban-the-box laws for Blacks turned positive and nonsignificant and the estimate for Hispanics remained as nonsignificant as ever, while the estimate for Whites became a resoundingly clear and precise null.
Hispanic results are still meaningfully ambiguousThe earlier papers finding this policy backfired and induced statistical discrimination against non-criminal Blacks can be found here:
And here:The initial time period result for this new paper (2004-14) replicates the estimate from Doleac and Hansen (2020) exactly, but it is nonsignificant here, where it was marginally significant for them.
Marginally significant results are tenuous, so it's no surprise it was fragile.Overall, I think I'm updating to
"This effect was unlikely to find strong support (meaningful effect/statistically strong evidence). There might be small harms to 'ban the box' policies, they're just dainty. If the goal is to help criminals, the tradeoff seems worthwhile."
yes
Several papers have estimated that "ban-the-box" laws, restricting employers from asking about employee criminal histories, have adverse effects on Blacks because they incentivize statistical discrimination.
A new paper found that the evidence for this has evaporated post-2014. ... The estimated effect of ban-the-box laws for Blacks turned positive and nonsignificant and the estimate for Hispanics remained as nonsignificant as ever, while the estimate for Whites became a resoundingly clear and precise null.
Hispanic results are still meaningfully ambiguous ... The earlier papers finding this policy backfired and induced statistical discrimination against non-criminal Blacks can be found here:
And here: ... The initial time period result for this new paper (2004-14) replicates the estimate from Doleac and Hansen (2020) exactly, but it is nonsignificant here, where it was marginally significant for them.
Marginally significant results are tenuous, so it's no surprise it was fragile. ... Overall, I think I'm updating to
"This effect was unlikely to find strong support (meaningful effect/statistically strong evidence). There might be small harms to 'ban the box' policies, they're just dainty. If the goal is to help criminals, the tradeoff seems worthwhile."
Missing some Tweet in this thread? You can try to
Update
