Finding the Hidden Data on Systemic Racism: From COVID-19 to George Floyd

Changemaker Catalyst Award recipient Kwasi Agyeman participated in a data visualization course in August 2020 to investigate how data impacts policies related to major current issues (covid-19 and police brutality). Kwasi is a law student with an interest in business and technology.

Turn on the news and it seems that our society is in a constant upheaval. From national lockdowns to local protests, it can all feel overwhelming yet still distant. As Americans, there is a deep sense that something is wrong yet we struggle to agree on a real solution. As a data scientist and law student, I wanted to learn the skillsets on how policy and data can help our communities. How do we develop frameworks to understand the impact of COVID-19? Which policies impact the numbers behind police interaction with Black and Brown communities?

My first step on my data policy journey started when I was selected as a Changemaker Catalyst winner. My original plan was to take a general data visualization course and then research the connection between large datasets and local policies. However, the pandemic effectively blindsided the instructor of the general data course and the class was put on a long-term hold. While I researched other data visualization courses, our health crisis continued to spike and then George Floyd’s death occurred. At that moment, I realized I had to take a course that would help me understand the trend lines and data points behind the systemic health and racial issues in our country.

As a member of the Gulf Coast Evaluation Network, I was introduced to a course titled “DataViz for Anti-Racism” and instantly knew it was the right fit for our moment in time.

The course is setup into four modules that each showcase a scenario, sample datasets and data exercises. The one module that caught my attention was the analysis of local city/county arrest patterns. The instructor showcased a case scenario in Raleigh, North Carolina which has a white population of 53% and a Black population of 29% and yet the police engagement with Black people was dramatically higher than other populations. The course used a standard distributed metric (possession of marijuana) which has been proven to have equal use between both groups yet the arrest numbers still disproportionately impacted Black people. In Raleigh, Blacks are 3 times as likely to be arrested for marijuana possession as compared to whites (1,169 vs. 380).

Police interaction in Raleigh, NC / Wake County

Why is the arrest number so much higher? I wanted to better understand this issue and began to delve into the policy framework. I discovered that most policing reforms such as 8 Can’t Wait miss one key component – average people have the extraordinary responsibility of policing other average people, therefore the arrest results carry internal biases. The data in Raleigh specifically showcased that it was far more common for other groups to receive a verbal warning or small fine then be arrested.

Current policy status at Raleigh Police Department

The reality is that data points cannot control prejudiced officers or limit the likelihood that an officer acts improperly. However, data can help us understand that there are deep underpinning issues that need real solutions.

One of the modules focused on childhood education and it made me think of how COVID-19 has impacted classrooms around the country. It was an unexpected learning experience because I limited my COVID-19 research primarily to reported case numbers. The education module highlighted the challenges disabled students routinely faced. Now that a significant portion of students are studying online. How are disabled students going to gain the resources they need to succeed?

The data visualization course taught me the basics of data modeling from Excel tables to infographs. Often the raw numbers can be lost in translation and the ability to transform a spreadsheet into an accessible product for long-term action is important. As a changemaker, I gained the skills to build informative data visualizations. This skillset is important because the gap between data scientists and the public is usually determined by whether a dataset is accessible and understandable. The data visualization course has helped me to further close the gap. I finished the course with a renewed sense that the power of data visualization should be rooted in raising awareness for broad societal change.