Analyst notes: Why do murder rates differ between federal sources?
Go behind the scenes with our team as we find and make sense of the numbers.
Some of USAFacts’ most visited articles, year after year, answer variations of the same questions: What are murder rates? How do they vary across the United States? Are they increasing or decreasing?
The interest makes sense. Americans can use the numbers to understand safety, shape policy, and track broader social trends. And on a personal level, the data can inform decisions about where to live, for example.
So, I’m here to give the people what they want: more information to understand murder trends across the country.
If you’ve looked closely across different sources (including USAFacts articles), you may have noticed something confusing: the number of nationwide homicides across sources doesn’t always match. This is because there is more than one dataset tracking homicides in the US. The two sources we most commonly use are the Centers for Disease Control and Prevention’s (CDC) National Vital Statistics System and the FBI’s Uniform Crime Reporting Program.
So what gives? Which one is right?
The short answer: both. They just measure different things.
Why do homicide counts and rates differ between the CDC and FBI data?
CDC data counts deaths from homicide based on death certificates. FBI data counts homicides reported by law enforcement agencies. This ends up being a pretty important distinction.
Although murder and homicide are often used interchangeably, they aren’t actually the same thing. Homicide is a neutral term for one person killing another through intentional force — it does not imply judgment about the act’s criminality and could include self-defense and other commonly justifiable acts. The CDC uses this definition when classifying deaths as homicides.
Murder, though, is a subcategory of homicide that captures intentional and unlawful homicides. The FBI’s homicide statistics counts two criminal acts as “homicide”: murder and something called non-negligent manslaughter. These are killings that are intentional, but aren’t premeditated (think crimes of passion).
TL;DR: the CDC has a broader definition of homicide (all homicides) than the FBI (criminal homicides only).
Because the CDC and FBI rely on different definitions and reporting systems, their homicide totals and rates can differ even though they describe the same underlying issue. Check out this table from the Bureau of Justice Statistics for more details on how:
Comparing homicide data from the CDC and FBI
Aside from definitional differences, there are a couple other distinctions between FBI and CDC data that further explain the disparate numbers. First, the CDC data is mandatory while the FBI data is voluntary. This means state health agencies must report their data to the CDC while law enforcement agencies don’t have the same mandate to report to the FBI. Second, the FBI data does not include federal law enforcement agencies. This means homicides reported only to these agencies are not counted in FBI’s crime data at all. For example, it excludes homicides occurring in federal prisons, on military bases, and on tribal lands.
TL;DR: the CDC has a broader definition of homicide (all homicides) than the FBI (criminal homicides only).
How do these differences show up in the data?
These differences mean that CDC provides a more thorough accounting of the number of people who die by homicide and a higher reported homicide rate (the number of homicides that occur per 100,000 people) compared to the one reported by the FBI at the national level, anyway. (But more on that in a minute, or, more accurately *counts* 102 words).
Let’s look at how these differences have manifested in homicide data over time by looking at the homicide rate for both. (I usually prefer to look at rates rather than counts when considering trends over time because it reduces changes that occur as result of population growth.) Here’s the chart:
CDC consistently reports higher homicide rates than the FBI.
Homicide rate per 100,000 people
The CDC homicide rate has exceeded the FBI-reported rate every year for the last 24 years (excluding 2021 because comparable FBI data isn’t available for that year due to a change in reporting systems). Data from the two sources show similar trends across the period, largely moving up and down in tandem.
(Time to get to the “more on that in a minute” stuff.)
The data differs more, and in slightly unpredictable ways at the state level. For example, unlike at the national level where CDC consistently reports a greater number of homicides and a higher homicide rate than FBI, CDC showed fewer homicides in five states.
The CDC reported nearly 2.5 times as many homicides in Mississippi than the FBI in 2024.
The percent difference in the number of homicides between CDC and FBI data, by state
In states where the CDC reported more homicides, the order of magnitude exceeded what we saw at the national level. For example, the FBI reported 217 homicides in Mississippi while the CDC reported 538 — more than twice as many.
So how do you pick a dataset?
Picking whether to work with FBI homicide data or CDC data depends on your use case. Are you most interested in the total number of people who died from homicide, and perhaps how that compares to other causes of death? CDC data is your best bet. Want to understand homicides in the context of crime rates, or do you want details about the incidents (like the relationship between the offender and victim, or demographic characteristics of the offenders)? Stick with the FBI.
Or maybe you’re like me. Chronically interested in data, constantly diving deep, making data vizzes galore, and obsessed with methods. If so, then why not check both out and see how they differ where you live?
The bottom line
CDC and FBI data on homicide measure overlapping but ultimately different things: homicide deaths versus criminal homicides. Differences in definitions, data collection, and coverage lead to different totals. Looking at both sources together can help provide a clearer understanding of lethal violence in the United States.
P.S. A lot of localities, or individual states, also publish homicide data on crime dashboards. It can be head-spinning! But that’s an analyst notes story for another time (and hopefully a different analyst).
