New data gives hope for informed reform
A foundation of data is essential in conversations from the squad room to the halls of legislative bodies
This article originally appeared in the February 2021 Police1 Leadership Briefing. To read the full briefing, visit Data-driven reform | Violence reduction | Tech to fight crime, and add the Leadership Briefing to your subscriptions.
COVID-19 and police reform are two themes continuing through the news. Numbers are a constant in the analysis of the pandemic guiding medical professionals and researchers in the response to the virus. In contrast, police reform is often being guided by speculation, assumptions and politics instead of data.
“Policing by the Numbers” is a recently published report from the Council on Criminal Justice (CCJ) that may provide useful data for policymakers. The CCJ describes itself as “a nonpartisan membership organization and think tank created to advance understanding of the criminal justice policy challenges facing the nation and build consensus for solutions based on facts, evidence, and fundamental principles of justice.”
As part of the CCJ’s report, several areas of statistical significance are graphically represented for readers.
Police agency size and composition
The increase in the number of police officers has not kept up with population growth across the U.S. Measured from 1987 to 2016, the number of full-time, sworn police officers peaked in 2013 with 724,690 but declined by 3% by 2016. When compared to the number of serious offenses (FBI UCR Part 1 crimes), the ratio of officers to offenses showed more officers per offense over the period of the downward trend in crime in the 1990s.
The number of female police officers increased from 7.6% of the law enforcement workforce in 1987 to 12.3% in 2016, with most of the increase occurring in the 1990s and leveling off in subsequent years. By race, Black officers during the same time period increased by 60% while Hispanic representation quadrupled. The statistics are aggregate and do not reflect percentages of ethnicity relative to populations served by individual agencies.
Spending on law enforcement as a percentage of government budgets has remained at under 4% for the last four decades even though actual dollars spent shows triple the inflation-adjusted dollars from 1977 to 2018. For the percent of expenditures to remain constant while actual expenditures increase is an indicator of the increased costs of other government services over time. The distinction between percentage of government budgets compared to actual dollars, especially given the reduction in actual numbers of sworn personnel, is an important number in the debate on defunding police.
Police contacts with the public
This category was sourced largely from a Federal Bureau of Justice Statistics report from 2015. Data differentiate, where possible, between police-initiated contacts and citizen-initiated contacts. The source is interviews of citizens rather than data provided by law enforcement. Some other estimates of police and citizen interactions count only the total interactions without regard to citizen requested contact, a practice that can leave the reader with the assumption that all contacts are assertively police generated.
Traffic stops, not counting crash investigations, and counting passenger contacts separately from driver contacts, constitute the greatest number of police-initiated interactions. In total numbers, Blacks are the subject of contacts more than whites as a total of the population although, once again, the numbers are a national aggregate and not percentages based on populations by jurisdiction.
The CCJ’s graphs represent nearly 30 million police-initiated contacts and 60 million citizen-initiated contacts, indicating that citizens seek out the police more often than the reverse. One potential indicator or cause of mistrust of police in minority communities is that Blacks are somewhat over-represented in police-initiated contacts and under-represented in citizen-initiated contacts.
The actual number of contacts fell by nearly 20% in the period from 2011 to 2015.
Crime rate trends, victimization reporting rates, and arrest and clearance rates
Crime rates, though measures are flawed in many respects, are one of the most consistently monitored statistics in the criminal justice arena. As already widely reported, both violent and property crimes dropped dramatically after a late 1980s peak, while the homicide rate has very recently spiked dramatically in certain urban areas.
The percentage of crimes unreported to the police is determined by comparing the reported crime numbers to surveys of the population asking if they have been a victim of a crime that was unreported. A decline in the percentage of crimes reported to police by victims was recorded from 2010 to 2019 after a slight rise in accurate reporting in years prior.
Clearance rates on reported crime have remained fairly steady over time with the exception of a poorer clearance rate for homicides between 1987 and 2018. Arrest rates have declined in recent years, mostly from public order offenses. Racial disparities have not disappeared but have lowered in all categories of crime in comparing arrests of blacks to whites.
Officer-involved fatalities, officers killed or assaulted
The most controversial issue in law enforcement is the number of persons killed while in contact with the police. This is also the set of data least accurately reported, as is the number of dangerous assaults and threats perpetrated against officers. This report’s aggregation of five different sources of officer-involved fatalities does show that the number of such deaths has remained steady near 1,000 annually. The report notes that “no clear aggregate trend” is evidenced over the past five years, which likely shows that deaths related to police intervention are singular anomalies in the milieu of millions of interactions with police.
An interesting trend is an increase in police-involved deaths in rural jurisdictions by 34% from 2013 to 2019, and suburbs by 8%, while urban areas experienced a 27% decrease. Deaths of whites rose 34%, marking the largest increase by race in contrast to a 24% decrease of Blacks in urban areas from police interaction. Most deaths were related to firearms and vehicles.
Deaths of officers have always fluctuated and defied conclusions about causation. Assaults on officers and officer injuries are among the least accurately reported, but a significant uptick in reported assaults began being reported in 2014.
Public opinion of police
This final category of the report reflects the public attitude as measured after the in-custody death of George Floyd.
Trust in the police had remained steady prior to Floyd’s death but dropped among Blacks by 16% compared to 1994 levels. White respondents’ opinions remained steady. Still, the majority of Black respondents reported that they would expect to be treated with respect during a police contact.
Only 14% of persons surveyed wanted less police presence in their neighborhoods and only a quarter believed that funding for police should be reduced. While indicating some weakness in public support of law enforcement, the numbers are not as dire as some would believe.
Policy implications of the data
Whether the data outlined in this report has a direct impact on policy formulation remains to be seen. As in all sociological studies, determining precise cause and effect relationships is difficult and defining problems and solutions just as challenging.
A foundation of data is essential in conversations from the squad room to the halls of legislative bodies. There are many positives in the report that police leaders can emphasize to their public to leverage improvements for their agencies. Law enforcement staffing, budget trends and improvements in reducing racial inequities may counter some persistent negative stereotypes about law enforcement. Many agencies will be able to boast that they are ahead of national trends or, conversely, need support to reach national averages.