Introducing The Real-Time Crime Index

A new way of evaluating national crime trends.

One of the greatest challenges to understanding crime trends is that data is frequently old and stale by the time it is received. Imagine only learning about the Texas Rangers winning the 2023 World Series winner 9 months after the last out. That’s essentially the world of crime trends.

The solution to date for crime data has been to collect data from dozens or hundreds of cities that make it publicly available in order to guesstimate what’s happening nationally. That’s what our YTD murder dashboard attempts to do and it’s also how groups like the Council on Criminal Justice and Major Cities Chiefs Association approach the problem.

The advantages of this methodology are clear: it's fast, it uses publicly available data, and it's reasonably predictive of the national trend.

The disadvantages are equally clear: the data that individual agencies publish is not standardized, there's a decent sized margin of error — especially when the sample is smaller, you’re stuck doing YTD comparisons which aren’t particularly useful for a good chunk of the year, and going beyond murder boosts the difficulty quite a bit.

The Real-Time Crime Index

All of those problems will hopefully be a thing of the past as we start to build a new project called the Real-Time Crime Index (RTCI). The RTCI is being undertaken thanks to a grant from Arnold Ventures with the objective of creating a repository of crime data from 500 to 1,000 cities that is updated monthly. The goal is to more or less mimic what BLS does with employment data to create an understanding of national crime trends in as close to real-time as possible.

This project is being done with a goal of eventually transitioning stewardship of the RTCI to the Federal government by the end of the grant.

Why 500 to 1,000 agencies?

Good question! This number of agencies was chosen as our goal because it is small enough to be attainable while being large enough to do a good job of predicting the national trend. There are a little over 800 police departments serving populations of 50,000 or more nationwide, so that group of agencies serves as a good target. If you get 500 or more agencies reporting monthly data then that should be predictive within 1 percent of the national trend (if crime rose 4 percent in a sample of that size then the national change should typically be between 3 and 5 percent).

What data are you collecting?

Another good question! The plan is to start with collecting aggregated UCR Part I index crime data. We’re starting with this because agencies regularly collect and report this exact data to their state UCR programs on a monthly basis, so the ask should be easy to accomplish for individual agencies. The hope is that eventually we will be able to add clearance rate, staffing, NIBRS, and even shooting data, but that is a longer term goal.

What is the cost to joining?

None! The RTCI is not a commercial endeavor and it is fully supported by Arnold Ventures. Agencies will never be asked to provide funding for participating in the project, and the RTCI is designed to require minimal effort from agencies while still providing essential data to the public, academics, law enforcement, and policymakers. Understanding national crime trends is in the national interest and the RTCI should be a public good.

What if I want to participate?

Please get in touch if you work at an agency or city that you think might want to participate! Or maybe you have a friend or colleague that works at an agency that might want to participate. Or maybe your organization can connect us with agencies that might want to participate. If so then you can send us a note at RTCI@AHDatalytics.com or drop a comment and I will be in touch.