24 November 2021

Today, the Ohio Department of Public Safety made the use of force database available to the public. The data goes back several years. With this data, we can finally begin to put to rest assumptions about police and use of force in Ohio. The topic of current interest is how police use of force varies depending on race, gender and ethnicity, and whether the use of force is appropriate to the crime and the encounter.

Updated 7/2/2022: Fatal police shootings in the United States account for 1042 deaths as of June 29, 2022, according to the Washington Post. Add at least one death for the shooting in Akron, Ohio where police fired 90 bullets at Jayland Walker as he ran away from his car. In comparison, police across the country have shown that they are capable of handcuffing white people with assault rifles who have admittedly killed people (or even ignoring them), but not Black people who have committed minor traffic violations, those who have legally carried a gun, or have been stopped for "suspicion" that does not carry over to non-Black populations ( 1, 2, 3, 4).

From the organization, Mapping Police Violence, the data shows that Black people are 2.9 times more likely to be killed by police than white people in the U.S. Yet, the Black population is only 13.6%, where as 75.8% of the population identifies as white.

Ohio Police Data Analysis

With the CRISP-DM for data analytics process in mind, we downloaded data from 2020 to do a bit of data exploration. After quite a bit of data clean-up (including disaggregating incidents with >1 officer involved, and relabeling for consistency), the data shows that the majority of police officers involved in use of force incidents were white (81.2% of incidents) and male (89.08% of incidents). But keep in mind that the majority of officers employed in Ohio fit that demographic so it is inevitable that they will have more use of force incidents. That is key to consider before making any conclusions with that data alone.

Looking at this data I have additional questions, some of which can be answered by performing statistical analyses on the cleaned data:

1. What was the total count of police officers in Ohio in 2020? With new hires, transfers and departures, some additional statistical data would be useful for me to perform additional analyses.
2. What is the racial/ethnic and gender distribution of the total Ohio police force, and how has it changed over the past several years since this data has been collected?
3. Is there a difference in the rate of use of force by officers who are male or female, Black, white, Hispanic or other?
4. There were 1698 reported use of force incidents. The gender counts and race/ethnicity counts are greater 1698 because there were incidents that involved multiple police officers. The question is then, how many of the officers are repeat offenders?
5. What are the races and genders of those classified as "unknown/not reported" and "pending further investigation"? They represent a large percentage of missing data.

Since the data I need is not 100% available in the provided data set, I will have to look for additional data sets to integrate and help finish telling the story about use of force in Ohio. More importantly, I have to trust that any data I use is data is accurate, complete and reliable. Since no data is 100% accurate, complete and reliable, I estimate the error.

As a nation, in order to correct the problem that police are more likely to use force when there are Black people involved. the police must use data-driven decision making to remove, reassign or replace repeat offenders among their ranks. Like any problem, the one of excessive use of force must be resolved from within the organization. Victims of police brutality cannot fix police brutality.

Creating a police force that reflects the communities is another decision that is driven by data. Given efforts in Ohio to recruit a more diverse police force, and screening out and firing individuals that show bias in their interactions with the public, we are optimistic about the opportunities for change.