Congressional Influence as a Determinant of Subprime Lending

Stuart Gabriel, Ryan Vaughn and I have written a new paper that uses some funky data to take a look at how a leading subprime bank priced its credit and sized its loans during the 2003 to 2006 period (the bank died in February 2007).   Everybody knows that banks seek out "good borrowers" who won't default and so we control for borrower attributes such as FICO Score and household income and the race of the borrower.  Since we know the zip code of the loan, we control for year 2000 Zip code attributes such as the socio-economic and racial attributes of the community and the state/year/month fixed effects for the loan's location.     We use County Business Patterns data to control for the count of jobs in the zip code/year.  You might expect that lenders are more willing to lend at better terms when the local economy is doing well.

Controlling for all of these factors, we explore how 7 dimensions of the borrower's Congressional Representative are associated with the loan pricing and loan amount.

The 7 dimensions are;

1.  Is the Representative a "Leader"  (Speaker of the House, Majority Leader, Minority Party Leader, Majority Whip, or Minority Whip).
2. Is the Rep on the Finance Committee
3.  The Poole-Rosenthal conservative index score (dwnominate1 from political science)
4. terms served in the congress
5.  Did the bank make direct campaign contributions to the Representative?
6. Is the Representative Black?
7. Is the Representative Hispanic?

Our major findings are as follows;


  • Borrowers in Leaders' districts get better deals measured in terms of lower interest rates and larger loans.  This is especially true if the borrower is black.
  • If the lender made campaign contributions to the congressman's district, then the bank offers lower interest rates to borrowers in that district
  • We reject the hypothesis that the representative's race matters.  Some have conjectured that subprime banks sought to curry favor with minority representatives.


So, why is this interesting?   The "political influence" literature has focused on observable campaign contributions as the main way that industry influences politicians but a more hidden form of influence is for banks to be nice to local constituents.   Using data on over 900,000 loans, we are able to detect these patterns.   The Leaders are powerful and they sought for more home ownership and a greater rate of minority home ownership in their districts.  Subprime banks knew this.  During the boom of the 2000s, you could imagine a "double bottom line" mind set that the subprime banks by lending on better terms to blacks were making money and allowing more Americans to have access to the American Dream.

We think it is interesting that New Century (the bank we study) simultaneously made campaign contributions to certain Representatives and gave better deals to their constituents.  Giving on several margins and being active in such communities is a way to earn Congressional loyalty.  While we don't know how New Century cashed in on this loyalty,  political capital is a relevant "capital stock" that major firms seek to accumulate.