Eviction Rate

[MathematicsCLtd]

Is your feature request related to a problem? Please describe.
Yes, Franklin County has the highest eviction rate in Ohio.

Describe the solution you'd like
Search Smart Columbus Operating System for datasets to analyze:

• United States Census Bureau
  • Poverty Level by Age and Sex at a Census Tract Level - Central Ohio 2016
• Ohio Geographically Referenced Information Program (OGRIP)
  • Need data from OGRIP for Franklin County
    • Foreclosures
    • Foreclosure Filings for Cuyahoga County

Describe alternatives you've considered
N/A

Additional context
Look in comments for link to article.

[MathematicsCLtd]

http://www.columbusalive.com/entertainment/20180314/behind-franklin-countys-high-eviction-rate

According to the Glenn College research, more than 40 percent of Franklin County’s eviction filings occur in only six zip codes, with neighborhoods like Whitehall, Northland and the Hilltop accounting for the highest numbers.

In 2016, nearly 18,000 evictions were filed in Franklin County — the most in Ohio. Chicago, on the other hand, which has nearly three times the population of Columbus, filed about 24,000 evictions.
“When you look at it by population, the numbers aren’t adding up. Something is leading to a very high eviction rate,” said Melissa Benson, a housing attorney with the Legal Aid Society of Columbus (LASC). “I don’t know if you can pinpoint it to just one thing, but it is very concerning.”

Columbus Alive

Behind Franklin County's high eviction rate

Franklin County has the highest eviction rate in Ohio. Why? And what can be done about it?

[MathematicsCLtd]

https://evictionlab.org

Eviction Lab

The Eviction Lab

We’ve built the first nationwide database of evictions. Use our map tool to see eviction rates in your community, view rankings for places across the country, and learn about the causes and effects of residential instability.

[daneshaffer88]

I can do this. I will try tonnage something next month.

[daneshaffer88]

The OGRIP website, like the smart Columbus website, only has foreclosure information for Cuyahoga and Stark Counties. I just sent an email to the Clerk of Courts for Franklin County explaining that I am a volunteer trying to get data for the smart Columbus website. I explained that I would like both foreclosure and eviction filings with addresses for a multiyear period. If I don't get an email back next week, I will follow up with a phone call. After getting these filing dates and addresses, they can be geocoded and/or related to parcels on the county auditors publicly available data.

The Clerk of Courts does provide aggregated foreclosure filing counts for the county, but this is clearly not at the level that we need: https://clerk.franklincountyohio.gov/CLCT-website/media/Docs/Foreclosures.htm.

If this doesn't prove possible, there are addresses put out by the Franklin County sheriff with scheduled sheriff sales of foreclosed properties. Note this is for foreclosures only (not evictions) and sheriff sales are foreclosures that have progressed further than the initial foreclosure filings. However, for a small area like Linden these could probably be entered and geocoded for several year period. The report can be pulled by zip code; however, there are individual addresses allowing us to be as precise as desired:

https://sheriff.franklincountyohio.gov/real-estate/search.aspx

Search

[daneshaffer88]

I can see why area poverty rate and area foreclosures would be correlated; however, I'm not sure how direct this relationship is. For starters, it's questionable how many people below the poverty line own vs rent in the first place. Nonetheless, a homeowner that drops below the poverty line due to job loss is at increased risk for foreclosure. In fact, economic theory for foreclosure suggests a two-trigger process:

Trigger 1: The value of the home compared to the value owed becomes low enough that the home cannot be sold (at least quickly enough) to cover the loan. If this were not the case, the home could be sold to prevent foreclosure.
Trigger 2: A life event occurs such as unemployment, underemployment, other drop in income, serious illness/injury, or divorce such that payments cannot be made.

A caveat must be made for 'strategic' default. In strategic default, trigger 2 does not happen. Rather, the home value falls below what the owner owes and, despite being able to make payments, the homeowner walks away for strategic reasons (why pay more for the home than it's worth?). The factors that limit strategic default are many. For instance, even if current value of the home is less than owed, the homeowner may predict future home value increases after the recession or economic shock. Further, the damage due to adverse credit reporting or other reputational factors may make foreclosure less appealing.

Nonetheless, I will look at poverty rate based on the smart Columbus uploaded dataset to judge the economic insecurity of the area and (possible) propensity for foreclosure (and later, eviction). The dataset is from the American Community Survey giving counts above and below poverty line by gender and age. The 2016 ACS 5 year would be based on data collected in the yearly 1% ACS samples from 2012-2016. It is aggregated across years at the census track level.

To relate these census tracts to Linden, I use shapefiles available from Zillow Data which delineate the boundaries for North Linden, South Linden, and East Linden. I then use 2016 Tiger Line files to find the 16 census tracts that overlap these. These census tracts I dub the 'Linden Area' and encompass a larger area than Linden proper.

I attached a copy of the SAS code to import and summarize this dataset. Poverty rate is about 14% for the Linden area. This rate is somewhat higher for females than males (15% vs 13%). The most obvious difference in poverty rates is between children (age under 18) and young adults (ages 18-24) which have a much higher poverty rate than older adults (ages 25-64) or senior adults (age 65+).

Finally, I focus on poverty for older adults (ages 25-64). These are the individuals that are most likely to have an active mortgage (although foreclosure for tax reasons is also possible). I divide the 16 Linden area tracts into quartiles based on poverty rate for this age group. From the map we can see that the highest poverty rates for this age group (quartile 1) tends to be at the southern end of the Linden area. In fact, much of this area is not in Linden proper at all but rather extends into Fair Grounds and Milo-Grogan neighborhoods. Nonetheless, it is clear that even within Linden proper poverty increases as we move southward.

sas_code.txt

[MathematicsCLtd]

@daneshaffer88 This is fantastic! What do you think about turning the work you've done here into it's own repository? A GitHub Page can be created for it so that your analysis can be forward facing. I was using this repo as a hub for all the issues I wanted to cover for Linden.

[MathematicsCLtd]

I have asked about the foreclosure data again for Franklin County. I will let you know what I hear back from my contact.