A Little Knowledge Is a Good Thing: Empirical Evidence of the Effectiveness of Pre-Purchase Homeownership Counseling

We are deeply indebted to Jim Carey for providing us with the opportunity to do this research and for sharing his knowledge about counseling, to Oliver Zeng for his outstanding work in developing the loan matching routine, and to Jim Berkovec for his econometric insights and consultation. The views expressed in this paper are ours alone, and do not necessarily represent those of Freddie Mac or its Board of Directors. Abstract For the past three decades homeownership counseling has been an integral part of affordable lending in the United States. Counseling's popularity has been based in large part on the belief that borrowers receiving counseling are better able to handle the responsibilities of homeownership. To this point, however, there has been no convincing empirical evidence to support this view. This study uses data on almost 40 thousand mortgages originated under Freddie Mac's Affordable Gold program to pose three questions; does pre-purchase homeownership counseling demonstrably reduce 90-day delinquency rates, do the different types of pre-purchase homeownership counseling programs vary in their effectiveness at reducing delinquency rates, and are any counseling providers more or less effective in administering their programs. We find that counseling can be effective in reducing mortgage delinquency. Borrowers receiving counseling in our data have, on average, a 19 percent lower 90-day delinquency rate. We find, moreover, that different counseling programs vary in their effectiveness. In particular, borrowers receiving counseling through individual programs experience a 34 percent reduction in delinquency rates, all things equal, while borrowers receiving classroom and home study counseling obtain 26 percent and 21 percent reductions, respectively. We find no evidence that telephone counseling mitigates credit risk. Nor, after controlling for the mix of counseling programs do we find that counseling providers vary in their effectiveness in reducing delinquency rates. Finally, we also attempt to determine whether our estimated impacts capture only the effect of counseling itself, or whether they also are driven by possible endogeniety of borrower assignment/selection into the counseling programs. We find that counseling itself has a significant impact on delinquency rates, and that this impact varies across the types of counseling programs. We also confirm the specific impact of classroom counseling. However, although borrowers receiving individual or home study counseling have lower delinquency rates, we are unable reliably to confirm that this reduction comes from the counseling itself rather than the assignment/selection of borrowers into these programs.