Violent crime -- is it up ?
The FBI has a way with data
I was first introduced to crime and police statistics and analysis by faculty at Old Dominion University where I received my Master’s degree in Public Administration. Many of the faculty in Public Administration had made the study of police and crime their life work, and I learned a lot about research and methodology there that has served me well throughout my career. I remember a light bulb moment about police research when we discussed that it would be intuitive that crime would go down, when the number of police officers were increased, but in fact, it often increased! This was because the increase in police officers resulted in more arrests, and so the crime would appear to be “increasing”.
My skepticism about these kinds of intuitive conclusions about police and crime was triggered by the recent FBI release of their 2023 report in October with their one major conclusion — violent crime in 2022 was down from the previous year, 2021 by 1.7%.1
The Data Collection
The FBI tracks crime statistics from law enforcement agencies and so we have a great picture of trends in crime in American. The FBI has been collecting data for the Uniform Crime Report since the 1930s2 and has evolved to electronic reporting since 1995, and then most recently in 2019, a major shift in how crimes were reported transitioned to the “more comprehensive” NIBRS or National Incident-Based Reporting System.3 The Uniform Crime Report, provides data for about 20 different types of crimes from homicide and rape to property crimes.
Violent crime in the FBI category includes murder, manslaughter, excluding negligent manslaughter, rape, robbery, aggravated assault— crimes against persons. Each state has their own definition of murder and manslaughter, but in general they are distinguished by the state of mind of the perpetrator, with forethought being attributed to murder, and “crimes of passion” generally attributed to manslaughter.
They also use a term, “clearance” meaning an incident has been addressed by an arrest or a lack of evidence that causes them to drop the case.
So if we want to see trends in crime, this would be the place we could expect to find it. They do not make it easy to find, in part because around 40% of reporting agencies failed to report their data in 2021 during transitioning to a new reporting system.
The FBI reports that violent crime is down for America, but they do not tell us if they mean incidents? arrests? clearances? They do not tell us if arrests are up or down, are numbers of police officers affecting the numbers of arrests due to the ill-conceived “defund the police” movement or is their report based on clearance rates for violent crime?
The analysis. Arrests are down.
Looking at the years 2022, 2021 and 2020, the years of the pandemic, arrests were down in 19 and up in 32 states and the Dist. of Columbia. Looking at ten year trends of arrests, 2011 through 2020, the population increased by 12 million, but the arrests decreased by 35%! 4
The FBI’s claim
The FBI claimed in their press release in October 2023, that national violent crime decreased 1.7% in 2022 compared to 2021.5 Without providing any explanation, we are to conclude that because there is less violent crime, life is better in the U.S. this year than it was last year.
Economic and Unemployment explanation
Intuitively, we know that the worsening economy, high unemployment in 2021 (8.06%) probably contributed to crime,6 in general, but this is also found in scientific studies. The drop in the rate of unemployment in 2022 (3.65%)7 could certainly tend to reduce criminal activity.
Number of police officers explanation
With a significant increase in arrests for violent crimes from 2021 to 2022, could that be due to more police officers? From my opening light bulb moment, it is important to see if the increase in arrests are due to more police officers rather than a behavioral change in criminals or policy change in jurisdictions or something else. Below, is a graph showing the number of full-time law enforcement officers over time (defined as full time, paid by the federal government, and generally carry a gun and have full arrest powers).
From our period of 2021 to 2022, nearly 50,000 police officers were added of the 708,001 in 2022. This is a likely explanation for the increase in all arrests from 2021 to 2022, and for the 37.1% increase in arrests for violent crime in that same period, 2021 to 2022.
Violent crime incidents vs. violent crime arrests—-explanation?
We have focused on arrests but let us examine the number of actual violent crime incidents compared to the number of arrests for our years, 2019, 2020, 2021 and 2022. There is a public perception that arrests are down due to local government policies that refuse to make arrests for property crimes, but this may not be the case for violent crimes.
Violent crime incidents in 2020 was 565,822 and violent crime arrests in 2020 for the U.S. was a total of 338,443 violent crime arrests, 147.9 arrests per 100,000 population. Violent crime arrests in 2021 were 219,981 arrests. For 2022, a total of 349,283 violent crime arrests, 129.1 arrests per 100,000 population. In 2022, there were 809,381 total violent crime incidents,8 but only 349, 283 arrests. In 2021, there were 769,182 total violent crime incidents, but only 219,981 arrests.9 Here is this data in table form:
Violent crime
Year Incidents Arrests % arrests:incidents Rate/100K10 # p. officers
2019 454,218 297,195 65.4% 363.9 697,195
2020 565,822 338,443 59.8% 387.8 (+5.6%) 696,644
2021 769,182 219,981 28.5% 377.6 (-2.7%) 660,288
2022 809,381 349,283 43.1% 369.8 (-2.1%) 708,001
From this comparison we can conclude from the FBI’s own data, there is an increase of 5% in violent crime incidents from 2021 to 2022. However, the rate of violent crime in relation to the population per 100,000 people has dropped by 2.1% from 2021 to 2022, but still above the pre-2020 rate. There are inconsistent rates reported by the FBI report and the DOJ/BJS for the year 2019. The DOJ/BJS has widely different data than the victimization report also from DOJ, which has eroded confidence among serious analysts of crime.11 While the federal statistics bureaus have scientific standards and methodologies, the FBI is not a scientific agency and how they interpret the data from those statistics bureaus is not necessarily based on scientific methodologies.
The 2.1% drop is good news story, but not the one the FBI chose to highlight in their most recent report of a 1.7% drop in something to do with violent crime.
The “playing with the reporting period” explanation
In 2020, there was a rise in the number of violent crimes in the U.S. for the first time in four years — a 5.6% increase over 2019. The other statistic that the FBI touts in its press release is the reduction in murder and nonnegligent manslaughter offenses in 2022. In 2020, there was a whopping increase of 29.4% of murder and nonnegligent manslaughter offenses, the biggest increase in at last six decades!12 So by using the reporting period of 2021-2022, one can avoid a focus on the 29.4% increase in murders and nonnegligent manslaughter; but it also avoids the ultimate conclusion that one would expect that the murder and nonnegligent manslaughter offenses would improve in the following years, after a 29.4% spike which set a six decade record.
By using the period of 2021 to 2022, the analysts avoid drawing attention to the 2020 year spike of 29.4%. The 2020 year is most likely the explanation for a drop in 2021 to 2022, and not necessarily a good news story.
A conclusion.
The DOJ/Bureau of Justice Statistics data in the table above, shows a steady increase in violent crime incidents from 2019 to 2022 — 5% increase in violent crime, not a decrease of 1.7%. So what data is the FBI using to claim that violent crime has dropped 1.7% as their good news story? It is not based on the violent crime incident data, the violent crime arrest data or the rate of violent crime.
However, using the violent crime rate there is a drop in the rate of violent crime from 2021 to 2022 of 2.1%, coming closer to their claim of a 1.7% drop. Any drop can be likely attributed to the dramatic decrease in the rate of unemployment from 8.06% to 3.65% which would be a better story that could be replicated from the data.
It is not comforting to know you cannot replicate the statistical claims of the FBI from their own data.
The FBI has a way with data.
https://www.fbi.gov/news/press-releases/fbi-releases-2022-crime-in-the-nation-statistics
https://www.fbi.gov/how-we-can-help-you/more-fbi-services-and-information/ucr
https://www.fbi.gov/news/press-releases/fbi-releases-2022-crime-in-the-nation-statistics
https://cde.ucr.cjis.gov/LATEST/webapp/# Table 33. Ten Year Arrests Trends by Sex.
https://www.fbi.gov/news/press-releases/fbi-releases-2022-crime-in-the-nation-statistics
Axios agrees with me. https://www.axios.com/2021/09/27/us-2020-violent-crime-rate-rose-fbi-data
https://www.statista.com/statistics/269959/employment-in-the-united-states/
https://cde.ucr.cjis.gov/LATEST/webapp/#/pages/explorer/crime/crime-trend
https://cde.ucr.cjis.gov/LATEST/webapp/#/pages/explorer/crime/crime-trend
https://www.statista.com/statistics/191219/reported-violent-crime-rate-in-the-usa-since-1990
https://www.themarshallproject.org/2023/11/03/violent-crime-property-data-nibrs-ucr-fbi-2022
https://www.axios.com/2021/09/27/us-2020-violent-crime-rate-rose-fbi-data