Navigate to:
National Municipal Budget Analysis
Carceral Resource Index (CRI)
Identifying Budgetary Priorities
We are in the midst of a nationwide debate rethinking how to achieve public safety and also promote public health, equity, and justice. If “money talks,” then budgets tell a story about a community’s priorities.To document how spending on carceral systems relates to investments in services that prevent or diminish the need for incarceration, we have completed a systematic analysis of local, city-level budgeting across American cities to document the investment disparity between carceral systems and non-punitive, supportive services.
The result is the Carceral Resource Index (CRI), a novel composite measure comparing municipal expenditures on health and social support services versus carceral systems. The CRI is intended to be a tool to help inform this long overdue nationwide debate.The CRI offers values ranging from -1 to 1. A CRI coefficient of -1 represents a jurisdiction’s fiscal prioritization of carceral systems to the exclusion of health and support, while 1 represents fiscal prioritization of health and support systems to the exclusion of carceral expenditures. Actual composition of budget outlays in each CRI category will vary by local government design. Expenditures that do not fall squarely into these three categories are excluded. ​
Carceral Systems Versus Supports
Carceral systems -- encompassing all institutions of law enforcement, prosecution, courts, incarceration, and supervision -- have a declared aim of achieving “prevention through deterrence.” In the belief that public safety is predicated on punishment and incarceration, enforcement has become more aggressive, prison sentences are longer, and criminal laws are harsher, all at enormous expense to the taxpayer. The United States now has the highest incarceration rate in the world (see Figure 1).
Figure 1
Adapted from the US Bureau of Justice Statistics & The Sentencing Project
Unfortunately, increasing the severity of punishment has not been found to deter crime, and rates of incarceration have no effect on rates of substance use in a state (see Figure 2). On the contrary, aggressive drug law enforcement has been shown to increase the incidence of violent crime, HIV, overdose and other health harms, while prisons themselves may be criminogenic as well.
Figure 2
Courtesy of PEW Charitable Trusts
On the other hand, supportive policy measures such as employment and housing assistance programs, access to healthcare and public space, substance use treatment, and community engagement have been shown to reduce the probability of coming into contact with the criminal legal system.
A Budget is Worth a Thousand Words
At a time of renewed interest in budget priorities, there is a need for a holistic approach that moves beyond a singular focus on policing.Our research aims to document how spending on carceral systems relates to investments in services that prevent or diminish the need for incarceration. We have completed a systematic analysis of local, city-level budgeting across American cities to document the investment disparity between carceral systems and non-punitive, supportive services.
​
During our analysis, each city’s budget was obtained through the city’s budget office website. The budget document was analyzed by two researchers who independently organized department and sub-department expenditures consistent with the established exclusion and inclusion criteria.
​
Our Tool: The Carceral Resource Index (CRI)
The CRI is a composite measure of the contrast between municipal expenditures on health and social support versus carceral systems. It answers the question, “What does the jurisdiction prioritize in their budget: support services (known to prevent the need for criminal legal system intervention) – or surveillance and punishment?”
Tableau graphic created in 2021.
Figure 3
The higher the CRI value, the more supportive; the lower the CRI value, the more carceral.
Note: not all cities in our sample are present on the above graph (n=28)
See Full Tableau Public Link Here: CRI by City (2017-2020)
What is a CRI Value?
The CRI value ranges between -1 and 1. (see Figure 3) The higher the number, the more prioritization of health and support systems with a lower share of carceral expenditures.
​
A value of -1 would represent a jurisdiction’s fiscal prioritization of carceral systems to the exclusion of any health and support services. Meanwhile, a value of 1 would indicate total investment in health and support services to the exclusion of carceral systems. All jurisdictions fall somewhere in between these extremes.
​
Each city’s budget was obtained through the city’s budget office website. The budget document was analyzed by two researchers who independently organized department and sub-department expenditures consistent with the established exclusion and inclusion criteria (see Additional Resources for further clarification regarding these criteria).
​
The current CRI calculation is unadjusted or crude, meaning that it does not consider:
​
-
External financial resources from the federal, state, or county levels (e.g. federal block grants)
-
Complex accounting techniques
-
Actual spending, but rather, adopted budget values
-
Departments that span both supportive and carceral functions (ex. education and child welfare agencies)
-
Municipal bonds (e.g. police brutality bonds) or litigation payments (e.g. wrongful death settlements)
-
Pension obligations
​
The crude CRI accordingly does not present a complete picture of budgetary outlays in any given category in any given municipality. Indeed, for many of these jurisdictions, carceral spending and functions beyond municipal police exist entirely at the county and state levels. Future phases of the project will address this complexity. Even so, the CRI can be used to compare jurisdictions, to track the expenditures of a single jurisdiction over time, or as a novel variable in the epidemiological modeling of the structural determinants of health.​
How is the CRI Calculated?
CRI is calculated using data extracted from the adopted, general fund budgets on either the city website or the Census of Governments. The line items in these budgets are coded or categorized as either carceral, support, or health spending, based on the following inclusion and exclusion criteria.
The end result is an index value ranging from -1 to +1, with -1 representing extreme carceral spending, and +1 representing extreme health/support spending.
Since the input for CRI is city budgets, one limitation is that CRI does not account for county or state-level spillover spending. It is important to identify spending outside of the city budget (e.g. county or state-level funding) that may affect the city’s carceral, health, or support systems, as well as how this spending may vary for cities in the same county or state. For example, in a given county or state, one city may receive a majority of the funding because it is the largest or most populated. CRI is limited in that it does not recognize these externalities (e.g. county or state-level funding) and disparities (i.e. disproportionate external funding) between cities.
​
For more information on budget items excluded from these calculations and the literature supporting the exclusion criteria, refer to this resource.
​
As of August 2023, the Action Lab has calculated the following sets of CRI scores.
Here is what these breakdowns look like in Los Angeles, CA & San Francisco, CA:
San Francisco
Los Angeles
Massachusetts 2021 CRI Score Analysis
Methods:
​
From September 2022-June 2023, we extracted municipal general fund budgets for FY 2021 for all 59 designated cities in Massachusetts and coded these budgets per our aforementioned CRI calculations.
The data source for crude and refined/manual scoring is different. For crude scoring, departmental breakdowns of budgets for multiple years and cities can be found on the Census of Government and other state or federal databases. For refined/manual scoring, complete budgets can be found on individual city websites. For example, for Massachusetts cities, crude CRI scores can be calculated based on departmental expenditures recorded by the MA Department of Revenue.
​
Findings:
​
The median city CRI score in Massachusetts was -0.31 (SD = 0.32). This is higher than the national CRI city average = -0.65.
How Do Cities Line Up?
There is enormous variation in the CRI values across our cities, but patterns do emerge. The average CRI value is squarely negative. On the whole, these cities invest substantially more in systems of punishment and control than they do on health and social supports, as evidenced by a median CRI value of -0.56. The majority of cities in our dataset have a CRI value between -0.45 and -0.65.
Tableau graphic created in 2021.
Figure 5
With these scatter plots (see Figure 6), you can see the relationship between a city’s CRI value and the percentage of the population that identifies as BIPOC. Opaque, red dots indicate higher percentages of BIPOC.
Tableau graphic created in 2021.
Figure 6
We have also visualized the relationship between each city’s population size and its CRI value (see Figure 7). Note that the Y-axis is populated in thousands and the X-axis covers the range of CRI values. There are signals in the data that larger cities have CRIs of greater magnitudes, although the strength of this relationship is statistically insignificant and largely driven by outliers. Furthermore, even those higher CRIs tend to remain negative.
Tableau graphic created in 2021.
Figure 7
Note that NYC was excluded from this figure due to its large population size
Tableau graphic created in 2021.
Figure 8
The 4-year average distribution of CRI totals.
See Full Tableau Public Links Here: CRI & Proportion of the Overall BIPOC Population , Average CRI & City Population, Average CRI Distribution
Previous Research
Our initial analysis of the crude budget data for FY 2020 yielded a median CRI of -0.27, suggesting that the cities in our sample spent significantly more on carceral systems than on health and social support systems combined.
Try Out Your Skills
Additional Resources
The Tax Burden of Incarceration
Healthcare Access
Housing and Employment
Substance Use Treatment
Learn More
Leo Beletsky, Britt Skaathun, et al
PLOS One, December 15, 2022