by Mac McCullough, JP Leider, Beth Resnick, David Bishai

---

City and county governments play a major role in the funding and provision of a broad array of health, public safety, and social services in their communities. Although there is substantial federal and state spending in support of local activities, local governments are a main source of government spending in the US. Local government spending allocations reflect local political priorities and have come under scrutiny with rising COVID-19 cases and questions around policing in the wake of the murder of George Floyd and other people of color in Spring 2020.

Aside from spending for education, spending for police, corrections, and the judiciary (which we’ll call “law and order” spending) are among the largest local government expenditures in the US. Non-hospital health and social services are other major governmental expenditure areas. Newly available data on local government spending from the US Census show that median per capita spending for non-hospital health and social services combined has been in decline for the better part of a decade (Figure 1) whereas conversely, law and order spending over this same time period has increased. As of 2017, total spending on non-hospital health and social services is now surpassed by law and order spending, both overall and when viewing counties in quartiles based on wealth.

Since the Recession, the majority of US communities have trended toward more spending on law and order than spending on non-hospital health and social services.

In 2017 local governments across the US spent a median of $270 per person on non-hospital health and social services (including housing, parks, libraries, and all non-hospital health-related spending) in 2017. This represents a decrease of 4.6% since 2007.

In contrast, local governments spent a median of $308 per person on law and order (including police, corrections, and judicial activities). This represents an increase of 13.5% since 2007. These data indicate that since 2007 the majority of communities in the US have trended toward spending more on law and order than spending on non-hospital health and social services.

These concurrent local government spending trends since 2007 of increasing spending on law and order and decreasing spending on non-hospital health and social services can be seen across a variety of communities—especially poorer counties. US Census Bureau data show that the median county has seen a shift since the Great Recession of 2007-2009 in funding away from non-hospital health and social services such as public health, community health care, housing, parks, and public welfare while during the same time a shift in funding toward law and order (police, corrections, judicial) (Figure 1). In the 1970s through the mid-2000s, non-hospital health and social service spending experienced modest year-over-year growth. In most US counties during this period, non-hospital health and social services spending was greater than law and order spending. That changed around the time of the Great Recession (2007-2009).

Expressed in constant (2017) dollars, the median US county spent $283 on non-hospital health and social services and $275 per capita on law and order in 2007. By 2017, median spending for law and order had increased to $312 per capita in constant (2017) dollars while spending for social services and non-hospital health decreased to $270 per capita.

A geographic examination of where spending changes occurred found that poorer counties (ie, those with lower tax resource bases as measured by general revenue from their own sources) seemed to have experienced this spending divergence between law and order and social services and non-hospital health to a larger extent than richer counties. Figure 1 Panel 2 (poorest quartile of counties) shows a gap in spending between law and order versus social service and non-hospital health that does not appear in Figure 1 Panels 3-5 (counties in the wealthiest 3 quartiles of counties).

We found it relatively common as of 2017 for total non-hospital health and social services spending to be less than total law and order spending across a county-area. Figure 2 shows the ratio of total law and order spending versus total non-hospital health and social services spending in a county. A ratio of less than 1.0 indicates less non-hospital health and social service spending than law and order spending in that county in 2017. Total spending in a county is denoted in Figure 2 by the size of the circle.

Figure 3 shows the ratio of law and order versus non-hospital health and social services spending for counties categorized by their relative wealth as measured by per capita own revenue (Q1=lowest, Q4=highest). Figure 3 shows that poorer governments tend to more commonly spend more on law and order activities than social services and non-hospital health (top left panel). Spending ratios are slightly more heterogeneous in middle-income and richer counties. Total spending in a county is denoted in Figure 3 by the size of the circle.

When examining spending at the individual service level instead of as aggregate categories, local spending on non-hospital health and for many social services from 2012 to 2017 has remained fairly stagnant or in decline, despite the overall growth of the US economy during that period (Figure 4). The annual median per capita spending on non-hospital health declined from $63 in 2012 to $55 in 2017; per capita spending for housing and community development spending declined from $41 in 2012 to $40 in 2017. In contrast to these spending declines, the three law and order spending categories (police protection, corrections, judicial) all had increased spending between 2012 and 2017.

Looking at cumulative percentage change in spending since 2007 (ie, the period surrounding the Great Recession and its aftermath), local spending decreased for non-hospital health and social services by a total of 6%, decreased for K-12 education by a total of 8%, and increased for law and order by a total of 13% (Figure 5, panel 1). These spending divergences were observed across counties of all wealth levels (Figure 5, panels 2-5), with the exception of the poorest quartile of US counties where a 4% increase in non-hospital health and social service spending was observed.

Why Relative Spending on Law & Order Versus Health & Social Services Matters

There is increasing recognition of the importance of factors such as community health and social services on population health outcomes and health inequities. Repeated incidences of police violence against people of color have led to calls to reallocate funding away from police and toward services targeted at improving the long-term safety and health of communities.

Yet data from the US Census Bureau show that local governmental spending allocations during and after the Great Recession do not align with this thinking. In fact, findings indicate that spending in the poorest quartile of counties in the US has shifted away from non-hospital health and social services and has moved toward law and order spending. A decade ago the median county spent 2-3% more for social services and non-hospital health than for law and order. By 2012, spending for the two categories had drawn equal. As of 2017 (the most recent data year available), the median county spends some 16% more for law and order than for social services and non-hospital health. Gaps in law and order versus non-hospital health and social service spending do not appear to be closing over time nor narrowing for richer versus poorer communities. Policy interventions targeting non-hospital health and social services since 2017 may have affected spending trends, but data are not available yet. It may be appropriate to temper such expectations given that major health-related policies such as the Affordable Care Act are already captured in the time period 2012-2017, though data related to COVID spending are not yet available.

The COVID-19 pandemic and social unrest arising from systemic racism and social injustice highlight the need for a strong community health and social service infrastructure. Yet data suggest that local government spending patterns are not currently aligned to these needs nor do they give any indication of trending in that direction.

Interested in the data or methods?

We obtained and processed financial data for local governments in the US from 1972 through 2017 for the entire range of government services provided by local agencies. Data come from the US Census Bureau’s Census of Governments and contain expenditure and revenue data as reported by local governments in the US every five years. In intervening years, a sample of governments report these data. Using a similar approach as the Census Bureau, we interpolated expenditure data from governments that did not report in intercensal years. This approach yielded a dataset of county-area expenditure and revenue data for 3,136 counties and county equivalents in the US. Our unit of analysis is the county-area, meaning all local governments within a county (or county equivalent), including city, county, and other jurisdictional governments.

We report annual county-area expenditures as per capita amounts in that county-area. All spending totals are adjusted for inflation using the Bureau of Labor Statistics state and local government deflator and are shown in constant 2017 dollars. We calculated each county-area’s annual per capita spending totals for (a) non-hospital health and social services— non-hospital health (which includes public health, behavioral health, public outpatient clinics, and disability-related clinical care) and social services (which includes housing and community development, public welfare, libraries, parks and recreation) and for (b) law and order—which includes police protection, judicial, and corrections. We tracked changes in these spending areas between 2007 and 2017, allowing for full examination of changes in spending during (December 2007- June 2009) and after the Great Recession. County spending allocations are skewed both as per capita and as shares of total spending, so we summarize them as medians.

Throughout this piece, spending is shown as medians (the value at the middle of the dataset) rather than mean (the average across the entire dataset). That is because spending—and especially per capita spending—tends to have a very skewed distribution, with some counties reporting extremely high spending totals for certain categories. Medians aren’t subject to the distortionary effects of outliers and in this case are preferable because they better represent spending than a mean would.

Recommended:

• COVID-19 Resources from JPHMP

Read All Posts in this Series:

• Excess Mortality as a Canary in the Coal Mine for COVID Disparities
• Yet Another Post on Public Health Spending
• New Workforce Estimates Show Public Health Never Recovered from the Great Recession. Then Came COVID-19
• Reflecting on the Past to Divine a Better Future for Public Health
• When the Going Gets Tough, the Tough Must Be Ethical
• What’s the Deal with Public Health Funding?
• A Workforce in Transition
• Meditations on the MPH, Part 2
• Meditations on the MPH, Part 1

Many Thanks to My Guest Co-Authors:

Mac McCullough, PhD, MPH, is Associate Professor at Arizona State University College of Health Solutions. His research involves quantifying investments in public health and social services and assessing how these investments are put to use in order to improve population health. [Full bio]
 

Beth A. Resnick, DrPH, is Assistant Dean for Practice and Training and Senior Scientist at Johns Hopkins Bloomberg School of Public Health.

 
 

David M. Bishai, MD, PhD, is Professor of Population, Family and Reproductive Health at Johns Hopkins Bloomberg School of Public Health.