Comments on: Police roundup

By: Raybury

Raybury — Mon, 24 Jun 2019 20:49:21 +0000

In reply to John Fembup. According to the source report's abstract, "Boys’ race and prior engagement in delinquent behaviors did not moderate the effect" -- aha! police stops bad! -- namely more stops leading to more self-reported criminal behavior. Wait! Self-reported? Self reporting is of course the worst way to get information. Who is to say that the boys didn't perceive that they did more bad as a result of being stopped, as opposed to actually doing more bad? Well, the self-reporting of the person who is emotionally affected by having been stopped.

By: gasman

gasman — Mon, 10 Jun 2019 18:59:39 +0000

Post hoc, ergo propter hoc falacy.
Retrospective studies such as this can at best note an association. In this study, the association is between police stops and delinquent behavior.

The conclusion could just as (in)validly have been written “Delinquent behaviors in urban youth are predictable in advance by the frequency by which such youth attract the attention of law officers”.

By: John Fembup

John Fembup — Mon, 10 Jun 2019 13:28:44 +0000

Re police stops of minority teens, the linked article states “The research indicates that a pedestrian stop of an adolescent male of color slightly increased the likelihood of future delinquent behavior, regardless of the young man’s previous engagement in delinquent activities. The research also indicates that multiple pedestrian stops further increases the likelihood of future delinquent behavior. The research neither indicated that police stops had the desired deterrent effect”

The research seems to look only at subsequent behavior of teens who were actually stopped. My questions

1. Is it not at least plausible that the teens who were stopped would have engaged in more future illegal activity, regardless whether they were stopped?

2. More importantly, is it not at least plausible there is a community sentinel effect I. e., resulting from more street arrests?

The cited research does not appear to consider either possibility and perhaps there is no way to measure it. But if either or both are at least plausible, then isn’t the effectiveness of such street arrests still an open question?