A map created by researchers at the University of Cincinnati can predict with startling accuracy how the racial makeup of neighborhoods will change.
Tomasz Stepinski, a professor of geography at the UC College of Arts and Sciences, has created a machine learning algorithm to predict in detail how neighborhoods will become more or less segregated over the next 10 years.
Stepinski, who works at UC’s Space Research Institute for Discovery and Exploration, analyzed data collected by the US Census Bureau each decade. They mapped the data by racial composition in high-resolution 300-meter squares called cells.
The algorithm had to be “trained” to interpret data from two census years spaced 10 years apart. The algorithm also looked at individual cells against those around them.
“The name ‘machine learning’ suggests there’s something magical about it, but it’s just more powerful statistics,” Stepinski said.
Stepinski validated his algorithm by comparing his predictions to actual data from the 2010 and 2020 censuses and found that it was up to 86% accurate.
“Our assumption that you can predict the class of a cell 10 years from now based on the previous two classes and the surrounding classes was correct,” he said. “It’s not perfect, but you can see it’s pretty good.”
Stepinski and co-author Anna Dmowska, an assistant professor in the Department of Geoinformation at Adam Mickiewicz University in Poland, applied their algorithm to Cook County in Chicago, Illinois, once considered one of the most racially segregated from America.
The UC map showed that many neighborhoods dominated by white and black populations will become less segregated by 2030 with less noticeable changes in neighborhoods dominated by Hispanic and Asian American populations.
Stepinski said researchers in Chicago conducted pioneering sociological research on race, ethnicity and gentrification using Cook County as a model. The sprawling, heavily populated county also provides a good model for studying the algorithm because it still has many segregated neighborhoods despite a trend toward greater racial and ethnic diversity over the past 50 years, he said.
Stepinski has also applied the algorithm in Houston, Texas, Los Angeles, and San Francisco, California with similar success.
“The ability to predict demographic changes is essential from a scientific point of view and for policymakers, city development, etc.,” said co-author Dmowska.
“As the document shows, the predictive maps are quite accurate and show what the region could look like in the next 10 years,” she said.
Stepinski said the predictive maps could be used to help schools or governments plan for more services like classrooms or Spanish-speaking interpreters. It could also help sociologists understand the driving forces of neighborhood demographic change.
“My interest is not sociological. My specialty is calculus,” Stepinski said. “I leave the why to someone else. But I can imagine what is happening.
Stepinski said younger generations are often likely to stay in nearby neighborhoods if they stay in one area. And if a particular racial population declines in an area, others usually fill the void.
“It’s broadcast,” Stepinski said. “So in Cook County you have a Hispanic population that is growing faster than the white population. They will move nearby. They are not going to move away from home.
Michael Chavarria, executive director of the nonprofit HOPE Fair Housing Center in Illinois, said he was surprised by the predictive accuracy of the map.
“One incredibly powerful thing it does is affirm the narrative that people don’t have as much control over where they live as decision makers think they do,” he said.
“People think that segregation is the result of individual choices. And some people might choose to live in places where everyone is like them,” Chavarria said. “But the idea that an algorithm can predict where people live shows that there are other factors that control those decisions. It tells me that the choice of where we live is not rooted in free arbitrator.
In the case of Cook County, Chavarria said the snapshot view offered by the predictive map could suggest certain neighborhoods in Cook County are becoming less segregated. But it could also simply capture the transition from one type of neighborhood that is segregated to another that is equally segregated.
“Seeing these neighborhoods change might just be a precursor to more segregation,” Chavarria said.
Olivia Cobbins, an investigator with the Cook County Commission on Human Rights, said it was important to look at racial segregation in the context of issues such as housing discrimination or public services.
“Chicago has a history of racial discrimination, redlining and inequality towards African Americans,” she said.
Cobbins said policies and laws must keep pace with changing demographics to ensure equal opportunities for all.
“Public servants and society have a moral responsibility and obligation to ensure justice, equality and fairness and to guarantee public assistance and decent and affordable housing to those in need.”
Machine learning with apps
The title of the article
Machine learning models for spatially explicit prediction of future racial segregation in US cities
Publication date of articles
Sep 15, 2022
Conflict of Interest Statement
None to report.
Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of press releases posted on EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.