Yep, it's a proxy but it's the best we have available. Obviously some decent share of their criminal activity won't be detected by police/result in an arrest and some share of the rearrests won't result in charges or a conviction.
All things being equal we're a bit more likely to detect crimes by those on pretrial release than the general population because of GPS monitoring data but it's far from guaranteed.
Thank you for a wonderfully nuanced article on the issues. But I still can’t reconcile something: Many of your posts state that people arrested for violent crimes often have multiple prior arrests. But you seem to be saying that pre-trial detention has a marginal impact on reduction of violent crime. Only 1.2% were re-arrested for violent crime? What am I missing? How does this mesh with your prior claims that many people arrested for violent crime have multiple prior arrests?
Thanks! The difference goes to how PSA slices the data:
- 1.2% were re-arrested for a crime of violence while on pretrial release for a previous crime of violence (so it's a very specific stat)...meaning that as soon as the case is over if another arrest happens it doesn't "count" towards PSA's stats because the defendant is no longer their concern. When we look back at how many prior arrests a homicide suspect has for instance many of those alleged offenses take place outside of the timeframe of when they are under pretrial release and therefore aren't counted by PSA.
- The 1.2% rearrest rate given by PSA only covered if the 2nd (or more) crime was for a crime of violence. So if someone is accused of say ADW, released then rearrested for carrying an illegal gun it doesn't count towards that 1.2% stat. PSA didn't provide the overall rearrest rate for the "crime of violence releasees" unfortunately but the overall PSA population was 92% arrest free while on release.
I also think that the simple act of being charged and released (usually with conditions) causes SOME reduction in criminality relative to not charging them at all. This is incredibly hard to prove because we don't know the true "base rate" of criminality for the cohort that is ending up under PSA supervision. But given how many crimes they rack up when NOT on pretrial release vs. the relatively low rate of rearrests during release it seems like what PSA is doing helps on the margin. It at least makes sense that "some supervision" leads to fewer crimes than "nothing."
Lastly it is worth highlighting (and I didn't explicitly say in the piece) that rearrest rate is just our best proxy for criminal activity while on release. Obviously some activity may not be detected by police and in many cases an arrest won't lead to charges or a conviction.
Doesn’t using the re-arrest rate assume that those released pre-trial who commit new crimes are actually caught?
Yep, it's a proxy but it's the best we have available. Obviously some decent share of their criminal activity won't be detected by police/result in an arrest and some share of the rearrests won't result in charges or a conviction.
All things being equal we're a bit more likely to detect crimes by those on pretrial release than the general population because of GPS monitoring data but it's far from guaranteed.
Thank you for a wonderfully nuanced article on the issues. But I still can’t reconcile something: Many of your posts state that people arrested for violent crimes often have multiple prior arrests. But you seem to be saying that pre-trial detention has a marginal impact on reduction of violent crime. Only 1.2% were re-arrested for violent crime? What am I missing? How does this mesh with your prior claims that many people arrested for violent crime have multiple prior arrests?
Thanks! The difference goes to how PSA slices the data:
- 1.2% were re-arrested for a crime of violence while on pretrial release for a previous crime of violence (so it's a very specific stat)...meaning that as soon as the case is over if another arrest happens it doesn't "count" towards PSA's stats because the defendant is no longer their concern. When we look back at how many prior arrests a homicide suspect has for instance many of those alleged offenses take place outside of the timeframe of when they are under pretrial release and therefore aren't counted by PSA.
- The 1.2% rearrest rate given by PSA only covered if the 2nd (or more) crime was for a crime of violence. So if someone is accused of say ADW, released then rearrested for carrying an illegal gun it doesn't count towards that 1.2% stat. PSA didn't provide the overall rearrest rate for the "crime of violence releasees" unfortunately but the overall PSA population was 92% arrest free while on release.
I also think that the simple act of being charged and released (usually with conditions) causes SOME reduction in criminality relative to not charging them at all. This is incredibly hard to prove because we don't know the true "base rate" of criminality for the cohort that is ending up under PSA supervision. But given how many crimes they rack up when NOT on pretrial release vs. the relatively low rate of rearrests during release it seems like what PSA is doing helps on the margin. It at least makes sense that "some supervision" leads to fewer crimes than "nothing."
Lastly it is worth highlighting (and I didn't explicitly say in the piece) that rearrest rate is just our best proxy for criminal activity while on release. Obviously some activity may not be detected by police and in many cases an arrest won't lead to charges or a conviction.