Study: Not prosecuting low-level crimes can improve public safety

New, first-of-its-kind research looking at Suffolk County criminal cases has found that declining to prosecute some low-level offenses can actually lead to less crime.

Researchers from three universities analyzed 67,553 misdemeanor cases in Boston, Winthrop, Revere, and Chelsea from 2004 to 2018, which didn’t include the tenure of current District Attorney Rachael Rollins. She took office in 2019.

People arrested but not prosecuted on low level, non-violent misdemeanors — like shoplifting, drug possession, or motor vehicle offenses — were 58 percent less likely to commit another crime in Suffolk County in the following two years, according to the study.

“Our results imply that a prosecutor’s decision to not charge a defendant with a non-violent misdemeanor significantly reduces their probability of future criminal legal contact,” Rutgers University professor Amanda Agan, one of the researchers, said. “Or put the other direction: Prosecuting these defendants actually decreases public safety.”

Most non-violent misdemeanor cases, even if they are prosecuted, don’t end in a conviction. Three out of four end without a conviction, but will show up on a person’s criminal record, affecting their job prospects and ability to secure housing.

When she took office, Rollins drew criticism from police unions, state public safety officials, and business owners over her directive to assistant district attorneys to decline prosecution for certain non-violent misdemeanor crimes, like disorderly conduct, driving with a suspended license or trespassing.

A more limited analysis of cases since Rollins took office finds significantly fewer non-violent misdemeanor crimes were prosecuted, according to researchers. There’s been no change in prosecutions of violent crimes or more serious felonies.

And the researchers’ review of Boston crime data from January 2017 to February 2020 found significant reductions in reports of property damage, theft and fraud after Rollins took office, and no increase in disorder or drug crime reports.

“We see no evidence that her inauguration and this expansion of presumptive non prosecution decreases public safety,” Agan said. “If anything, it increases it.”

Agan said the researchers weren’t able to look deeply at the effect of race, ethnicity, or gender on whether cases are prosecuted or not, because often that information was missing from the data. (In the Massachusetts court system, information is written on paper forms and then entered into a database by hand.)

But from what they could glean, the likelihood of someone reoffending is the same across different races and genders.

This article was published by WBUR 90.9FM on March 29. The Reporter and WBUR share content through a media partnership.