In recent years, many U.S. cities have made their data more accessible to the public, especially through online data portals, and crime data are among the most frequently shared types of open data. In addition to making data more accessible to use for policymakers, open data portals promote transparency and accountability by offering residents insight into how their communities are being served.

But beneath the surface of these well-intentioned portals lies a pressing problem: the data are often incomplete, inconsistent, and difficult to use. Without reliable, high-quality crime data, our ability to understand crime trends or design effective policies is undermined.

In 2023, NORC set out to gather existing open crime data into a shared resource. That effort resulted in the launch of the Live Crime Tracker (LCT) in May 2024, a tool that centralizes up-to-date, incident-level open crime data for over 50 cities in eight crime categories and refreshes those data on a daily basis.

Our experience building and maintaining the Live Crime Tracker, combined with our expertise in policy research and data quality assessment, has given us a unique perspective on the opportunities for improving how crime data is collected, maintained, and made accessible across the country.

Our work began in June 2023 with a detailed review of the crime data available from the 200 largest cities in the United States. After months of reviewing and cataloging, we discovered that only a fraction of cities had any usable incident-level data, and even fewer had data in formats conducive to analysis.

We were interested in capturing the information reported in open data files, as well as how frequently data was updated. After limiting our search to cities that reported any data from the current year, we identified only 64 with open incident-level data that met our search criteria.

However, several other cities reported incident data less frequently or in a format that was not accessible. For example, 16 cities reported only aggregated numbers, such as monthly or quarterly counts for crime incidents. Another 17 cities reported crime statistics online in formats such as interactive maps, but the incident-level data could not be downloaded in any format.

Beyond the issue of data availability, the lack of standardization in data reporting across different law enforcement agencies creates challenges for data aggregation and analysis. While many agencies use platforms such as ArcGIS to share open data, adoption is far from universal. These inconsistencies necessitate complex scraping strategies to extract and standardize the data, adding another layer of difficulty to working with open crime data.

We also identified differences in how agencies define and categorize crime types. For example, one agency may adhere to following National Incident-Based Reporting System (NIBRS) offense codes while another may only include a loose description of the offense. These variations in crime categorizations led us to map each city’s reported data to a common framework. Without these standardization and harmonization processes, it would not be possible to draw accurate conclusions about regional or national crime trends from the available open data.

Another challenge is that data often lack traceability—that is, individual records change over time, records are added and removed, and entire data sources can vanish overnight when a city alters its data portal structure or changes the location of the hosted data. Changes to real-time data sources are expected at the incident level, as individual incidents are reviewed, new records are identified, or data corrections are made.

However, these changes are not often documented. We have been monitoring LCT data sources daily for more than a year now, and it is rare to have a full week pass without a change in status for at least one of our 50+ included data sources, and we see hundreds—sometimes thousands—of incident records change each day. The initial list of cities we identified in 2023 is not the same list populating the LCT today—new cities have been added as their data become more accessible, but some cities have also been removed as their data portals change or stop updating.

If we want to use open crime data to conduct robust analyses on crime patterns and inform effective policies, we need to better understand these changes. We can’t expect that changes won’t occur, but data providers can better document when, how, and why records are changed or data sources are moved or altered.

To unlock the full potential of open crime data, we must make a concerted effort to improve their quality. These improvements should include local officials: 

• Sharing up-to-date data in easily accessible formats such as APIs or downloadable data files
• Aligning with established reporting standards for data elements like incident type 
• Clearly documenting data corrections at the record level 

The onus is on law enforcement agencies to adopt standardized reporting practices and on city data offices to support uniform data formatting, commit to consistent and timely updates, and more carefully document data modifications. Where law enforcement agencies and city data officials lack the capacity or resources to dedicate to their data reporting practices, they should seek out partnerships with research organizations or universities to collaborate on improving data management and reporting.

Enhancing the accessibility, consistency, and traceability of data in open crime data portals will improve their fitness for use in monitoring patterns, completing analyses, and getting useful data into the hands of decision-makers to reduce the impacts of crime in U.S. cities. They will also directly improve how governments serve the public. With better data, residents can better understand their own public safety needs and the local government’s ability to appropriately deliver services and supports.