(Edwin Torres/ProPublica)

How We Analyzed Racial Disparity in Debt Collection Lawsuits

An explanation of how we analyzed whether debt collection lawsuits disproportionately impact black communities.

The primary goal of our analysis of debt collection lawsuits was to explore whether the number of court judgments against blacks was disproportionate when compared to whites of a similar income level.

Because civil court records do not typically identify the race of a defendant, we based our analysis on where the defendants lived. This allowed us to compare the number of judgments lodged against residents of majority black communities to those against residents of majority white communities.

Who is Filing the Most Lawsuits in Black Neighborhoods? A Count of Court Judgments

A variety of companies filed collections lawsuits against debtors. Below is a breakdown of the which of these groups obtained the most judgments against debtors in each of the three cities we studied. We only included groups with more than 1,000 judgments.

St. Louis area

Plaintiff Types Judgments
Debt Buyers13,679
Utility Companies10,700
Major Banks*8,554
Medical Providers5,799
Auto Lenders4,402
High-Cost Lenders**4,267

(The total population of mostly black neighborhoods in the St. Louis area is about 348,000)

Newark area

Plaintiff Types Judgments
Debt Buyers37,557
Major Banks*11,797
Medical Providers3,150
Bail Bond Companies2,399

(The total population of mostly black neighborhoods in the Newark area is about 293,000)

Chicago area

Plaintiff Types Judgments
Debt Buyers37,221
Major Banks*27,066
Auto Lenders6,651
Miscellaneous Lenders***6,275
High-Cost Lenders**3,535
Government Agencies1,175

(The total population of mostly black neighborhoods in the Chicago area is about 1,156,000)

* Only a few large banks account for almost all filings in this category: Capital One, Discover, Citigroup, Bank of America, and JPMorgan Chase.

** A lender that offers credit at rates exceeding 36 percent APR

*** Lenders that didn’t fit into any of the other categories (mostly nonbank lenders).

Notes: St. Louis includes St. Louis City and St. Louis County. Chicago includes all of Cook County, Illinois. Newark includes Essex County, New Jersey. For St. Louis and Chicago, the judgments stemmed from suits filed from 2008 through 2012. For Newark, the time window is slightly different: July 1, 2007 through June 30, 2012.

Source: State court data, ProPublica analysis

To conduct the analysis, we examined court judgments arising from debt collection suits in three large metropolitan areas: Chicago (Cook County, Illinois), St. Louis (St. Louis City and County, Missouri), and Newark (Essex County, New Jersey). Our study selected a five-year window of cases from 2008, the beginning of the financial crisis, through 2012. (Our window for Essex County was slightly different: July 1, 2007 through June 30, 2012.) For our analysis, we only looked at suits that resulted in judgments against defendants, because judgments allow plaintiffs to garnish defendants’ wages or bank accounts.

We started with line-by-line court case data from each area. Each line provided basic information for each case (who filed the suit, when it was filed, who the defendant was and where he or she lived, etc.).

Using the defendant’s address, we grouped the suits by census tracts. The tracts are defined by the U.S. Census Bureau, and each tract typically has a population of a few thousand people. To this count of suits, we added census data like race and median household income.

From our initial analysis of the numbers, we found that majority black communities had higher per-capita judgment rates than majority white communities. However, because race and income level are often correlated (blacks typically have less income than whites), we needed to control for income in order to isolate the role of race in the judgment rates.

To perform that analysis, we employed more advanced methods that allowed us to hold income constant. We explain this process in detail in our white paper, but it allowed us to determine the comparative risk of a judgment for black communities as opposed to white ones. The risk of judgment, we found, was about twice as high in majority black census tracts as majority white census tracts, holding income constant.

Because our case data also had details about which company filed each suit, we also analyzed which companies had obtained judgments against residents of mostly black neighborhoods at higher rates than residents of mostly white neighborhoods. We found that in St. Louis and Chicago, high-cost and subprime auto lenders had very large racial disparities. (In Newark, state interest rate caps limit such lending.) Debt buyers also showed relatively large disparities in all three of our jurisdictions, especially when compared with large banks.

When we looked for the causes of the disparity, we found one likely contributor: the gap in assets between black and white households. National survey data shows that black households have fewer assets (such as home equity, investments and savings). Because black households have less of a safety net than white households, they could be more vulnerable to falling far behind on debts, resulting in a larger number of collection lawsuits.

In the white paper, we present two pieces of evidence to support this hypothesis. First, there is local data showing that black families in St. Louis and Chicago do indeed have fewer assets than whites. And second, blacks tend to be sued over smaller amounts, which may happen because white consumers are better able to resolve smaller debts before being sued than black consumers.

For a longer, more technical discussion of our methodology, please see our white paper.


author photo

Paul Kiel covers business and the economy for ProPublica, reporting on the foreclosure crisis, consumer debt and other financial issues.

author photo

Annie Waldman is a senior reporting fellow. She recently graduated with honors from the dual masters program at Columbia’s School of International and Public Affairs and the School of Journalism.

Comments powered by Disqus