top of page
Jerry Paytas

University of Pittsburgh's Quantitative Economics Students Explore the Demographics of Evictions




In a blog post on May 5, we released the results of the work conducted by three outstanding students from the University of Pittsburgh’s new Master of Science in Quantitative Economics (MQE) program:


We asked Gayatri, Mason, and Natalee to utilize publicly available data from the US Census Bureau, Department of Housing and Urban Development (HUD), and the Eviction Lab at Princeton University to help us understand how underlying demographic and economic conditions might predict future changes in evictions for renters. This is an especially timely question given the economic impacts of Covid-19 — especially considering the pronounced spikes in housing insecurity that continue to affect households around the country. Now that the students have had some time to recover from their semester, we asked them to reflect on their experience. Their answers are below:


You all sought out this topic. Why did it interest you? Why do you think this is an important area for economic analysis?

  • Studying evictions and housing affordability is interesting because it will always be a relevant and ongoing topic. With respect to the pandemic, it is especially important to be able to identify factors that might contribute to eviction rates and identify how housing shortages impact different income groups.

  • Even though problems with evictions have been around for a while, studying evictions on a national scale is still in its infancy. This project presented a chance to gain a better understanding of the housing affordability problem.


What were the most exciting parts of the project?

  • One of the most exciting parts of this project was developing interactive visualizations. It was a great opportunity to learn more about data visualization through Python and Google Data Studio. Creating these visualizations allowed for a clear picture of our analysis and trends.

  • Overall, it was exciting to pull together a number of tools including Jupyter Notebook and a variety of R packages to tell a story about housing affordability in the United States.


What were the most interesting or unexpected things you learned?

  • It was surprising to learn that there are many places in the US that either do not report on evictions or do not make these records accessible for research. There is little existing literature on eviction forecasting; in addition, many evictions go unreported, making this topic especially difficult to study.

  • Strong geographic trends in housing affordability were also unexpected, with many coastal states having incredibly high-cost burden rates.


How would you like to see work like this used by others? What value could it bring to dialogue around housing policy?

  • We would like to see our work used as a tool to enable holistic affordable housing policy. We find that factors such as unemployment and population size are significant contributors to housing affordability.

  • This project, in addition to other tools, will ideally make housing policy more efficient. We hope our analysis helps spark discussion about this issue and provide a clearer overview of the affordable housing shortages.


What was the most challenging aspect of the project?

  • We had little prior knowledge about this topic; there was a learning curve regarding the history, context, and terminology.

  • Data availability was one challenge in trying to accurately represent housing markets. For example, we learned that there is no official government-based source for evictions.

  • We also ran into challenges with data quality. For example, in levels smaller than the county level, there were instances of inconsistent naming, missing information, and small population sizes.

  • Finally, we learned that finding the right type of regression is an iterative process. Overall, a lot of trial and error went into producing the final product, but every obstacle contributed to our success. With guidance from Fourth Economy, we were able to learn a lot about evaluating and visualizing a completely new area of data.

Fourth Economy would like to thank Gayatri, Mason, and Natalee for their excellent work on this project, and we are very thankful to have had the chance to work with and get to know them. For more information on this work, please reach out to us here.


bottom of page