How much taxpayer money is spent on police, courts, and legal services to respond to crime? This information is important to help determine municipal budgets, to measure how different agencies are allocating costs, and to help quantify the benefits of preventing crime.
RAND researchers addressed this question by using modeling to estimate costs by crime type and by state in two studies:
The following visualization presents results of the average law enforcement costs and court and legal services costs of a reported crime by crime type. See the table for the raw data on the costs of crime, organized by state.
About the estimates Download the data
Download a spreadsheet of the full data set: RAND_TLA517-1.data.zip
To download a subset of data directly from Tableau, click on the download symbol in the bottom-right corner of the table above and chose your preferred file type. The "Crosstab" option to download a .csv file will be available to select once you have made selections within the table.
How to use these estimates. These estimates are the costs (in 2010 U.S. Dollars) to respond to a reported crime, but they also can be used to quantify the benefit of preventing crime in freed-up resources.
What is included in these cost estimates? These estimates include the time spent by police officers, judges, public defenders, prosecutors, and other staff to address a reported crime. For law enforcement, this includes such activities as calling officers to a scene, controlling a crime scene, conducting investigations, filling out paperwork, conducting interrogations and arrests, and appearing in court.
The estimates also include equipment and supply costs. For law enforcement, this includes such items as fuel for vehicles.
Administrative staff and utilities also are included in these cost estimates. "Fringe benefits," such as health insurance, are included in utilities. Self-directed retirement benefits are excluded (such as contributions to a 401(k) or IRA).
What does it mean that the costs are only for reported crime? The calculations include only the cost to respond to crime, which is particularly relevant for police officers who are involved in many other activities to prevent crime and address issues that are not related to crime. For instance, police help with traffic and traffic accidents, reach out to communities, and conduct patrols. None of those activities are included in the cost estimates.
For courts and legal services, some cases do not lead to an arrest and therefore cost the courts nothing, while other cases go to trial and result in more-substantial spending. These costs implicitly consider the percentage of cases that do and do not lead to an arrest and how they are disposed.
Why do costs vary across states? The results do not necessarily reflect how much it should cost to fight crime. Higher costs in one state do not mean that police are being wasteful; it could be that they are thorough. Similarly, a low-cost state may not be underpolicing a crime, but might be using their resources more efficiently.
How were these estimates developed?
Law enforcement: We used publicly available data on (1) police expenditure from the Bureau of Justice Statistics and (2) time spent responding to crimes in urban and rural areas by role in a department (e.g., detective, patrol officer) from observational and time-diary studies. Because there is considerable uncertainty regarding various parameters underpinning these calculations, we use Monte Carlo simulation methods to incorporate the uncertainty into our estimates. For this reason, the ranges are a more-accurate representation of the cost of law enforcement response by crime type and state.
Court and legal services: We use publicly available judicial and legal services expenditure data from the Bureau of Justice Statistics, along with a variety of data sources on the percentage of criminal, felony, and misdemeanor cases by crime type; and on the time spent by court personnel on felony or misdemeanor crimes. Because there is considerable uncertainty regarding various parameters underpinning these calculations, we use Monte Carlo simulation methods to incorporate the uncertainty into our estimates. For this reason, the ranges are a more-accurate representation of the cost of court and legal services by crime type and state.
All costs are standardized across states using state-level wages for justice, public order, and public safety services jobs.
