Read our investigation: How felony murder locks up Black Michiganders

Michigan does not track who county prosecutors charge with felony murder. Nor does it keep statistics on how those cases resolve through plea deals and convictions. To understand how prosecutors use the charge, Michigan Public and the Investigative Reporting Lab at Yale compiled data from correctional records, court documents, and charging reports.

To obtain criminal conviction data, Michigan Public used the Python Requests library to extract records from the Michigan Department of Corrections Offender Tracking Information System (OTIS). The resultant dataset captured all people in OTIS during the scraping period, May 4-7, 2026. OTIS includes people currently incarcerated, on parole, or who exited the system within the past three years.

Michigan Public narrowed our analysis to offenses committed since 2000 to focus on recent prosecutorial decisions and avoid inaccuracies in older cases. In total, that dataset included over 62,000 people convicted on more than 180,000 counts.

OTIS labeled 2,089 people as convicted under some subsection of the first-degree murder statute. Within that group, we identified and resolved two major inconsistencies in subsection labelling.

First, 212 convictions were labeled as “open murder.” That charge allows prosecutors to pursue second-degree, premeditated, or felony murder at trial. The jury reaches a specific verdict on each charge pursued. In some open murder cases, OTIS reflected the final verdict, but for these 212, the label simply read “open murder.”

Prosecutors across the state described different policies for charging open murder. Some told Michigan Public they charge every homicide as open murder; others almost never use the charge.

Michigan Public identified and removed two second-degree murder convictions labeled as “open murder.” For the remaining convictions, because we could not identify or predict the subjection of first-degree murder, Michigan Public designated them as “other” — a category distinct from both pre-meditated and felony murder in our analysis.

Second, 129 people had first-degree murder convictions that were not labeled with a specific subsection. For those convictions, Michigan Public manually reviewed court records, charging reports, and, as a last resort, news articles, to determine the final conviction type. The remaining six convictions Michigan Public could not definitively identify were included in the “other” category.

“Other” also includes two people convicted of First-Degree Murder of a Peace Officer or Corrections Officer.

Recent court rulings have given people under 21, who were sentenced to mandatory life without parole, the chance at release. Michigan Public identified only two cases of youthful offenders convicted post-2000 who had exited the system. We added those profiles to the dataset.

Individuals with both felony murder and premeditated murder convictions were classified in our dataset as “premeditated.” Resultantly, those in the felony murder category were convicted solely under that charge.

In total, we identified 685 people convicted of felony murder for offenses committed since 2000. That number represents a minimum; others may be represented in the “other” category or may have exited OTIS through death or transfer. Our analysis also excludes the hundreds of people convicted of felony murder before 2000 who are still serving life sentences in Michigan prisons.

Racial and geographic analysis was based on OTIS' record of the individual’s race and the county in which they were prosecuted.

Michigan Public compared the racial composition of people imprisoned for felony murder to that of people imprisoned for a “serious felony,” defined as one of the crimes enumerated in the felony murder statute as a possible predicate to a felony murder conviction. That group included people who had been convicted of an enumerated felony or of felony murder, since all felony murder convictions included an underlying felony.

This group served as a proxy for those eligible for a felony murder conviction, which we believe is a better point of comparison than, for example, the general population.

Still, it was an imperfect proxy. Most felonies don’t result in death, and those that do aren’t randomly distributed amongst this group. Further, because these felonies carry a range of sentences — as opposed to life without parole — people who have completed their parole and exited the system are not recorded. We don’t believe conviction date significantly impacted racial disparity: We found a similar disparity within recent convictions, where the concern of unrecorded cases is minimal, and we also found disparity when comparing to annual datasets like the FBI’s NBIRS.

Michigan Public also obtained charging records to spot check our dataset from OTIS, as well as to analyze how felony murder charges resolved into plea deals.

To obtain charging records, Michigan Public sent Freedom of Information Act requests to Michigan’s largest counties for a “charge inquiry report,” a digital form maintained in the office’s case management system. The report contained all charges filed by prosecutors, regardless of how those cases were resolved.

Wayne County was the only office among a dozen to which Michigan Public sent FOIA requests that refused to release their charging records. The office argued that doing so would require them “to create a compilation, summary, or report of information,” which is beyond their obligation under the law.

For plea-deal analysis, Michigan Public looked at cases from 2015 to 2025. That date range was selected to minimize two concerns: first, some offices warned that older cases were less reliably transferred from previous case-management systems; second, because people released from prison are eventually removed from the MDOC system and therefore from our data, we could only identify plea deals amongst those currently or recently incarcerated.

We matched records by name and county of prosecution, manually correcting records where we were able to confirm an identity even when spelling varied slightly. Of 200 total charges, we linked 127 to MDOC records. About half appeared to have pled to a lesser crime. We manually verified docket numbers to check that plea deals arose from the same crime as the felony murder charge. After review, we confirmed 57 matches, including 35 plea deals for second-degree murder.

Due to the challenges of matching cases, Michigan Public’s analysis represents a minimum number of plea deals which arose from felony murder charges. There are likely more.

The data collection and analysis was completed by reporter Etai Smotrich-Barr. Michigan Public Data Reporter Adam Yahya Rayes reviewed and contributed to the analysis in this story.

Thousands of people rallied at the National Mall on Friday, August 28, the 63rd anniversary of Martin Luther King Jr.’s 1963 March on Washington. But a group of over 50 Michiganders who had planned to attend the event instead spent Thursday evening in a church parking lot in Adrian, waiting for a bus that never came.

The canceled bus trip was one of several reported across the Midwest, all on buses scheduled by the National Action Network, a civil-rights organization founded by the Rev. Al Sharpton.

The rally in Washington, cohosted by Sharpton and Martin Luther King III, focused on protecting voting rights amid recent changes to election laws in several states, and a Supreme Court ruling that undermined a key provision in the 1965 Voting Rights Act.

Denise Benjamin, a coleader of Lenawee Indivisible, said she coordinated with the National Action Network to request the bus and organized the group of attendees. She said she spoke to a division leader at the network around 8 p.m. Thursday, an hour before the bus was scheduled to depart.

“And they just said, we're sorry, but all the buses in Michigan, all the buses from Chicago and the buses from Indiana have been canceled,” Benjamin recalled. “That’s all they would tell me.”

Media outlets in Philadelphia, Chicago, and Cleveland reported similar last-minute cancellations of Action Network-scheduled buses.

The National Action Network has not publicly explained the cancellations or identified specific bus companies that had been contracted for the transit. The organization did not respond to a request for comment.

In a Facebook video, Bishop Clyde Posley Jr., a leader with the Indiana chapter of the network, claimed that although the organization had scheduled and paid for buses, “the bus companies in those areas somehow were somehow gotten to, and for whatever reasons … they backed out.”

Benjamin called the experience “disheartening and very disappointing.” She said her group had looked forward to joining the event and getting a chance to voice their concerns about potential voter suppression in the upcoming midterm elections.

“But we're going to continue the work,” Benjamin said. “We're going to continue to protect the vote as best we can, in any way we can.”