Quantitative Data Sources

I am proposing to research urban displacement because I want to find out why patterns of eviction in cities become spatially concentrated among certain areas in order to help my reader understand the structures responsible for producing the housing (in)stability and precarity of contemporary cities.

While this research problem statement already lends itself to a large-n analysis, further specifying the problem statement to a more specific case: What explains variation in the eviction rates of census block groups in Washington DC.

There are two databases that I am primarily interested in. The first is the list of scheduled evictions released every couple weeks by the Office of the Tenant Advocate which provides the location and scheduled date of evictions throughout Washington DC.[1] The second is the American Community Survey’s 1-year estimates for 2019 demographic data, which provides the most updated demographic characteristics for census tracts and block groups in Washington DC.[2] Much of that demographic data – ie average income, median rent, race – will be used as my control variables. For my independent variable, I am planning on creating a gentrification index that will be determined with that demographic data (though how exactly I am doing that I am still figuring out).

The scheduled eviction report provides the data that I will use for my dependent variable of eviction rate. Unfortunately, right now, that report does not show actual rates, but rather the list of the hundreds of evictions that were scheduled between August and November. To turn them into eviction rates I will have to geocode the addresses in the spreadsheet into a coordinate system that can then be input into GIS. Then I will be able to tell the number of evictions from that dataset that fall within each block group, allowing me to determine an eviction rate for each block group (% of renter-occupied homes that received an eviction in a block group over a set period of time, typically one year).

Bibliography

“2018 ACS 1-year Data Profiles,” United States Census Bureau. (Washington, D.C.: GPO: August 29, 2019) https://data.census.gov/cedsci/table?d=&table=DP02&tid=ACSDP1Y2018.DP02&lastDisplayedRow=22&hidePreview=true&q= (Accessed: October 11, 2019).

“Scheduled Evictions,” Office of the Tenant Advocate. (Washington, D.C.: GPO, n.d.), https://ota.dc.gov/page/scheduled-evictions (Accessed: September 29, 2019).

 

 

[1] “Scheduled Evictions,” Office of the Tenant Advocate. (Washington, D.C.: GPO, n.d.), https://ota.dc.gov/page/scheduled-evictions (Accessed: September 29, 2019).

[2] “2018 ACS 1-year Data Profiles,” United States Census Bureau. (Washington, D.C.: GPO: August 29, 2019) https://data.census.gov/cedsci/table?d=&table=DP02&tid=ACSDP1Y2018.DP02&lastDisplayedRow=22&hidePreview=true&q= (Accessed: October 11, 2019).

Evan Margiotta

Evan Margiotta

View more posts from this author
2 thoughts on “Quantitative Data Sources
  1. Avatar Carly Holencik

    Evan, your project sounds very interesting! I like how it is centered around DC; it makes the issue feel even closer to home, and makes me more aware of my surroundings and local problems. Would your dependent variable be the location of the evictions or eviction rates? There are so many options for independent variables as well, like demographics of residents, tax rates, and more. What independent variables are you considering? Income, employment status, race, changes in taxes and housing costs, etc? This issue is also interesting in relation to the gentrification that seems to come up often when talking about certain neighborhoods in DC, like Georgetown and Shaw. It raises important questions like where are these displaced people going, how is it related to changes in taxes and cost of living in certain areas, and even how does implicit gentrification and displacement happen. Your system for analyzing your data seems interesting but very complex and time-consuming. Best of luck with your project!

     
    Reply
  2. Avatar Dr. Boesenecker

    Overall this is a very good start for conceptualizing your project in this methodology, Evan. As we’ve discussed, you don’t need to construct the entire database just yet, but knowing what data is there (and how you would have to work with it to assemble a dataset at this stage) is very important as you think about the direction that you might take with your research. I gather from this post that you’ve settled on census block groups as the unit of analysis (at least for now)? What tradeoffs come with this choice?

     
    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *