Behind the story: How CBS News analyzed homicide solve rates across the country
Since the start of the pandemic, homicide rates in the U.S. have jumped at alarming rates. In nearly every major city, the number of homicides after the start of 2020 are far higher than they were in previous years, reversing decades of lower homicide rates that began in the early 2000s.
The investigative team at CBS News wanted to understand how police were addressing this nearly unprecedented spike in murders — and why many were going unsolved. The team, along with the CBS Innovation Lab, analyzed data from major cities across the country and found declining homicide clearance rates as well as big disparities depending on the race of the victim.
What is a "clearance rate?"
Clearance rates are a tool used by law enforcement and researchers to measure how well a police agency can solve cases. They're calculated by dividing the number of incidents that were "cleared" or solved each year by the number of murders that occurred in that same year. It doesn't matter if the murders cleared were committed that same year, or 10 years prior — they're counted in the year that they're cleared.
Why not just use arrest rates?
Clearance rates are different from arrest rates. A case can be cleared if a person is arrested — this is the most common reason a case is cleared — but can also be cleared under other circumstances.
According to FBI guidelines, cases can be cleared under something called "exceptional" circumstances. These so-called "exceptional clearances" include cases closed because the offender is dead, serving a sentence on another case, if the victim refuses to cooperate, or in a handful of other less common circumstances such as if the offender is in a jurisdiction that refuses extradition, according to Thomas Hargrove, who runs the Murder Accountability Project, which tracks unsolved murders nationwide.
The FBI data, however, doesn't distinguish between cases cleared by arrest or by exceptional means, and it's the only way to get consistent data for all law enforcement in the country.
Why only use 2020?
The FBI is slow to release data, and, at the time this analysis was conducted, the most recent data release was for 2020.
The data
CBS News used two sets of FBI data for this project: the "Return A" files from its Uniform Crime Reporting (UCR) program, and Supplemental Homicide Report data. Both hold data related to homicides reported to the FBI by thousands of police agencies nationwide.
The data is published by the FBI in an archaic format originally used in the 1960s, so it first had to be extracted and transformed by translating data manuals the FBI publishes into computer code.
CBS News analyzed that data, then sent that analysis to 14 of its stations across the country. The journalists in those local newsrooms used that data to inform their own reporting, uncovering individual stories of unsolved murders and families left searching for justice, while the national investigative team did its own investigative work to tell the story of a country whose law enforcement were struggling to keep up with the rise in homicides — and whose communities were paying the cost of unsolved murders.
Want to learn more about CBS News' methodology? We've published our code here for those interested.
The data files used to calculate clearance rates didn't contain race data, however, and CBS News wanted to investigate how the cases of victims of different races and ethnicities were being solved. CBS News used the FBI's Supplemental Homicide Report data, which contains detailed information on homicide victims and offenders, to answer those questions.
That data doesn't directly indicate whether a case was cleared, however, so CBS News had to use the presence of offender demographic data to determine whether a case was solved. If there was demographic data present for the offender for a given incident, CBS News considered the case cleared, and if not, we considered it open.
This approach has some downsides: namely that some agencies, according to research by the Murder Accountability Project, submit offender demographic data to the FBI even when a person hasn't been arrested and the case isn't cleared, such as when a witness described the offender, but police haven't made an arrest. This approach can lead to clearance rates that are higher than what the UCR data suggests but is the only way to measure clearance by demographics without requesting data from each individual agency in the country.
