An Indirect Approach to Estimate the Jumbo-Conforming Spread

Abstract

The Economic Stimulus Act of 2008 (ESA) was signed into law on February 13, 2008. It established increases to the two government-sponsored enterprises’ (GSEs’) conforming loan limits in high-cost areas as defined by the U.S. Department of Housing and Urban Development. In this article, the author takes the ESA as a natural experiment and uses the difference-in-differences methodology to indirectly estimate the jumbo-conforming spread. The study classifies the 360 CBSA metro areas located in the 48 contiguous states plus DC into two groups: 1) a treatment group made up of the 52 CBSA metro areas whose conforming loan limits are affected by the legislation and 2) a control group made up of the 308 CBSA metro areas whose conforming loan limits are not affected by the legislation. The basic idea is straightforward: If the jumbo-conforming spread were large, one should expect to see that the difference between the change in the mortgage demand of the treatment group and the change in the mortgage demand of the control group before and after the legislation would be statistically significant, and vice versa. The analysis finds that the difference between the change in the mortgage demand of the treatment group and the change in the mortgage demand of the control group before and after the legislation is statistically insignificant, which implies that the jumbo-conforming spread might be small.