In 2018, Larry Krasner was elected on a platform of bringing transformational change to the Philadelphia County District Attorney’s Office.
He’s kept those promises in the five years since by restricting prosecution and the use of cash bail for certain low-level charges, reducing mass supervision for people on probation or parole, increasing the review of potential cases of wrongful conviction, and holding law enforcement officers accountable for criminal conduct.
But in the middle of those reforms, the COVID-19 pandemic struck the nation, followed by a nationwide spike in homicides that was also felt in Philadelphia.
While homicides appear to be trending downward in 2022, Krasner became a scapegoat. Despite easily winning re-election to a second four-year term, he was impeached at the state level. Politicians pursuing his impeachment have based their allegations on a study on Philadelphia’s crime rate that researchers say is inaccurate.
That paper was first published in July 2022 by Thomas Hogan, a former federal prosecutor and Republican district attorney in Pennsylvania, and purported to show that “de-prosecution” since 2015 was responsible for the increase in homicides in Philadelphia County (note: Krasner took office in January 2018). Hogan’s claim was more than the usual partisan allegation or pundit accusation — though he’s made his share of those, including referring to Krasner as a Manchurian Candidate for the Ku Klux Klan in a blog post. Hogan has also called Krasner one of the five worst prosecutors in America.
While Hogan’s paper was published in a peer-reviewed journal — the standard for rigor and scientific analysis — other researchers are claiming it is riddled with conceptual and analytical flaws that bring its conclusion into doubt. They’re also calling for the journal to issue a correction.
Among the paper’s numerous flaws, researchers highlight Hogan’s use — or, in some instances, misuse — of data, including blatant data omissions to support his arguments. They also point out Hogan’s use of homicide counts instead of homicide rates to delineate outcomes and his use of outdated modeling methods that provided inaccurate results.
“These flaws are fatal to the author’s findings, and therefore the study should not be used to inform criminal justice policy,” said Jacob Kaplan, a researcher at the Princeton School of Public and International Affairs, in response to Hogan’s study.
Kaplan was part of a group of researchers that reanalyzed Hogan’s paper and found that when those flaws were corrected there was “no effect of de-prosecution on homicide.”
These flaws are fatal to the author’s findings and therefore the study should not be used to inform criminal justice policy.Jacob Kaplan researcher at the Princeton School of Public and International Affairs
Oren Gur is Krasner’s director of research and policy advisor, and directs the District Attorney’s Transparency Analytics (DATA) Lab. (The Philadelphia District Attorney’s Office (DAO) is an Arnold Ventures grantee.) Gur works with a team at the DAO and with external partners to research and understand on a granular level the impacts both current and former policies have on the people of Philadelphia County.
When he learned of Hogan’s first academic paper, he was shocked by what to him was “a clear effort to cloak personal disdain in academic regalia,” but not surprised.
“Hogan has shown a particular animus for Krasner for a long time,” Gur said.
Gur saw the study as a different tactic to attack Krasner and reform-minded prosecutors. As he dug into the paper, he found “it doesn’t make any sense theoretically, and there are no specific prosecution policies studied or causal mechanisms.”
If there are fewer arrests, there will likely be fewer cases coming into the system, and fewer cases resolved, too. The paper ignores this reality.Oren Gur director of research and policy advisor for the the District Attorney’s Transparency Analytics (DATA) Lab
Gur said that Hogan’s paper finds a reduction in the number of cases prosecuted and sentenced, and attributes this to what Hogan terms “de-prosecution,” when “every student learns in their introduction to criminal justice class that police are the gatekeepers to the criminal justice system.
“If there are fewer arrests, there will likely be fewer cases coming into the system, and fewer cases resolved, too. The paper ignores this reality.”
In fact, arrests have been trending down in Philadelphia for some time, from over 88,000 in 2010 to over 45,000 in 2018, according to the FBI’s Crime Data Explorer.
Economist John Pfaff has also noted there were reductions in clearance rates in homicides in Philadelphia during this same time period. There were also broad systemic efforts to reduce the jail population and improve fairness through the city of Philadelphia’s engagement with the MacArthur Safety and Justice Challenge.
Gur lamented that none of these factors are accounted for in Hogan’s work.
“He’s a former district attorney, so he knows how important police are to whether a case is successfully prosecuted, but ignores that in addition to the tremendous role of arrest. He ignores the discretion of individual officers, and the autonomy of the police department more broadly.”
Gur said he’s concerned with Hogan’s previous blog posts in which he has “an axe to grind” against Krasner and how those views shaped and inject bias into his paper.
“I just wonder how reviewers may have scrutinized his manuscript differently if they knew about and read these blog posts,” Gur said.
Meanwhile, Kaplan and other researchers have called on the journal to retract Hogan’s study or, at minimum, provide addendums to his work. While the journal has so far rejected those calls, Kaplan continues to push for better practices for peer-reviewed papers moving forward.
“Considering [Hogan’s] unwillingness to share their data and code, we call for a greater dedication to open science and reproduction/replication in criminology,” Kaplan writes.
As far as Krasner’s fate is concerned, the impeachment proceedings in the Senate are at a standstill after a judge ruled on Dec. 30 the articles filed against him don’t meet the threshold of “misbehavior in office.”
Krasner’s Senate trial is expected to begin Jan. 18, though Senate Republicans have said they’re reviewing the judge’s ruling and drafting an “appropriate response.”