University Honors Program News

  • RSS
  • Print

Getting to the Core of Bankruptcy

Mary Hansen

(Photo: Jeff Watts)

To understand the importance of economics professor Mary Hansen’s efforts to preserve and digitize bankruptcy court records, a good place to start is to understand what we don’t know about bankruptcy, including its relationship to business cycles.

“We don’t really have any clear idea of how financial crises and bankruptcy and credit all fit together,” says Hansen. “We don’t have any idea whether using the bankruptcy law is at all related to long-run growth. We have a pretty good idea that having a bankruptcy law is essential to long-run growth at the national level,. . . but we don’t really know beyond that whether there are aspects of bankruptcy or using the bankruptcy law that can enhance growth.”

Moreover, data sources for information on the finances of bankruptcy filers—whether they’re credit card data, publicly available federal court databases, or others—are incomplete.

And the amount of wealth being transferred from debtor to creditor via declaring bankruptcy is huge, about as much money as is transferred from taxpayers to poor people and uninsured people through Medicaid, Hansen says.

Research into bankruptcy, Hansen argues, has been held back by a lack of ground-level data. The best source for such data is U.S. District Courts’ bankruptcy case files. But the sheer mass of the records is daunting, and the expense of maintaining 2 million cubic feet of archives is becoming prohibitive.

So Hansen and her collaborators initially helped the Federal Records Centers and the National Archives to design a sampling scheme to preserve 2.5 percent of the records. In pilot studies they tried to digitize all records, but that effort proved too expensive. The current goal is to get a 1 percent sample of cases filed since 1898. Records from every court and from every year will be photographed, digitized, and made available for scholars from any discipline to use.

To show the potential merit of the data, Hansen and her co-investigator, Michelle McKinnon Miller of the Rutgers Business School, have started two pilot programs. Those pilots are looking at a sample of the bankruptcy case file data for Mississippi during the Great Depression and a sample from the District of Maryland and Eastern District of Virginia from 1940 to 2000.

The first pilot project studies a district that received bank bailouts during a financial crisis versus one that didn’t. The second project tries to figure out how changes in federal garnishment law may have dramatically raised bankruptcy cases in Maryland.

“They’re hard to access, they’re scattered across country, but the bulk of them are in Kansas City because that’s where the biggest storage facilities are,” Hansen says of the records. “If you don’t have a legal history background, you open these boxes and say, ‘What is this stuff?’”

So it’s a labor intensive, hands-on process.

And that’s only the beginning. Once the data are digitized they must be put into a format useful for researchers. Hansen hopes to develop procedures to use optical scanners for printed information, but older, handwritten records require the human eye to discern the information.

The projects have received five separate grants, including faculty grants from American University and Rutgers University and funds from the Institute for New Economic Thinking, the Alfred P. Sloan Foundation, and the National Conference of Bankruptcy Judges Endowment for Education. Funding from non-AU sources amounts to about $333,000.

But Hansen’s sure the project is worth the trouble. Beyond their obvious academic benefits, she notes at least two potential policy uses:

  • Since we’re in the dark about fundamental questions involving bankruptcy, we can’t know the effects of proposed changes in credit law. How will the Dodd-Frank Act, especially as a result of the Bureau of Consumer Financial Protection, impact bankruptcy cases? We have no idea.
  • Do minorities and women suffer inequality in bankruptcy proceedings? We don’t know.

This new data could help answer such basic questions.