Deep in the Heart of Texas Crime Data

Texas as bellwether of the national crime trend.

I've spent most of January so far getting over my Longhorns losing a heartbreaker in the Sugar Bowl and figured there was no better way to do that than by writing about crime data in Texas. Texas may not have been able to “hook em” against Washington but the state's data can help us “hook” the national crime trend (I'm sorry for that, I understand if you want to stop reading now).

Texas's state crime data is an interesting set of data to write about for a number of reasons: Texas has good public data and serves as an early barometer of the national trend, but the Texas data is also not complete in ways that are sometimes predictable and sometimes interesting — interesting to a nerd like me, not to a normal person.

All told, the not-totally complete-data from the Texas Department of Public Safety (DPS) points to a large scale decline in murder as well as declines in violent crime and property crime outside of auto theft statewide last year. If that sounds familiar it's because it mimics, and hopefully increases confidence in, what the FBI quarterly data found.

Texas has long been one of the leaders in the FBI’s National Incident-Based Reporting System (NIBRS). Texas became NIBRS compliant in 1998 and had 98 percent of the state’s population covered by a NIBRS reporting agency in 2022 — well ahead of the national average.

The Texas DPS also has an excellent website that allows users to generate reports on crime in Texas law enforcement agencies. This means that Texas is one of the first states in the nation providing somewhat complete data on crime in 2023 compared to 2022.

I say somewhat complete because most agencies have only reported 11 months of data so far this year. The DPS spells out the advantages and cautions in using its data pretty well:

“The availability of this data is dependent on local agency timely and accurate submissions, which can be impacted by local agency resource constraints, system updates, and technical issues. As data is submitted, routine data validations are applied to ensure compliance with FBI and state-level reporting guidelines. The UCR data is a ‘live’ collection; meaning agencies can continue to update their incident data per their investigation findings, when arrests occur, for any corrections needed, and in response to data quality checks. As such, this report is a reflection of all the data currently contained within the TXDPS UCR System at the time of inquiry for the timeframe specified.”

In other words, the data we have is almost certainly an undercount as agencies are still reporting data, but this undercount becomes a much less big deal if you can figure out when most agencies have reported data for and count from there. Fortunately, one can sort the data by month of offense, so if we use a cutoff through October it provides a few extra months of buffer against late reporting while being late enough in the year to allow for a strong understanding of the statewide trend in 2023.

There are just over 1,000 agencies covering about 29.3 million of the state’s 29.5 million people that reported data through October 2023. Below is the change in crime through October in Texas using the state’s NIBRS Monthly Crime Summary Report. These figures will undoubtedly change slightly with two more months of data, but it is late enough in the year that the change will likely not be meaningful enough to effect the bottom line. To paraphrase Voltaire, don’t let precise be the enemy of accurate enough.

Change in Texas Crime Reported Through October, 2022 vs 2023 (Source: Texas Department of Public Safety)

One can audit the data in the above table by comparing it to publicly available data published by agencies themselves to add confidence in the above finding. There are a handful of Texas agencies that publish monthly crime data and for ease of use I took the 6 with 10 or more murders so far this year. Comparing data in those agencies to the DPS data suggests the above trends are accurate rather than an artifact of agencies failing to report data yet.

Comparing Murder Counts in Texas Cities With Available Data (Source: Texas DPS and Agencies)

Dallas is the obvious outlier when comparing DPS data to the agency's website data, but a closer look suggests the issue is some weirdness in what Dallas is reporting publicly rather than what they're sending to DPS. Dallas reported 222 murders in 2023 through November on their website compared to 186 on the DPS site. BUT Dallas reported 201 murders in all of 2022 on their website compared to 150 on the DPS site (the FBI reported 157 murders in Dallas in 2022).

It's the latter figure that got reported to the FBI suggesting there’s a methodological difference in what the FBI requests and what is being presented on the website (just a guess!). The percent increase presented by Dallas police is smaller than the one shown in the DPS data, so the impact on the overall trend is minimal (and actually hurts the above DPS percent change).

The DPS data is important because it gives a very strong barometer of the crime trend in Texas in 2023, and the crime trend in Texas tends to give a good barometer of the national trend — especially as it relates to murder. See below the graph showing the percent change in murder in Texas for every year between 1980 and 2020 (x-axis) versus the percent change in murder nationally (y-axis).

More and more data will help to confirm (or refute) the assessment from December that murder fell at a historic rate in 2023 while reported violent and property crimes fell at a steady clip. It has only been a month since that assessment was published but so far it is holding up better than the Texas offensive line on January 1st (and now I’m sad again).