Do Large-Scale Refinancing Programs Reduce Mortgage Defaults?

from the Congressional Budget Office

In March 2012, the Federal Housing Administration (FHA) announced that it would reduce the premiums it charges to participate in its streamline refinance (SLR) program. The reduction in premiums was substantial: A borrower with a $200,000 mortgage and a loan-to-value (LTV) ratio greater than 95 percent could expect to save $3,480 in up-front premiums and $1,200 per year in annual premiums. However, only borrowers whose mortgages FHA endorsed by May 31, 2009, were eligible for the reduced premiums.

Therefore, borrowers with endorsement dates on opposite sides of the cutoff date faced different financial incentives to participate in the SLR program. This retroactive eligibility rule creates a natural regression discontinuity (RD) design with which to measure how reduced mortgage payments affect borrower behavior. The results suggest that reduced mortgage payments lower default rates substantially and that the reduced fees more than doubled the number of SLRs between July 2012 and December 2013. We estimate that the policy will prevent more than 35,000 defaults of FHA-insured mortgages.

Policymakers may wish to reduce mortgage defaults for several reasons. First, defaults cause losses to taxpayers through the mortgage guarantee programs of FHA, Fannie Mae, and Freddie Mac.2 Reducing default losses was one of FHA’s stated goals in announcing reduced fees on SLRs (Department of Housing and Urban Development [HUD], 2012). Second, defaults and subsequent foreclosures not only have adverse spillover effects on nearby properties and neighborhoods but also entail significant deadweight losses. For instance, Anenberg and Kung (2014) estimate that listing a foreclosed property for sale reduces home prices within a third of a mile by 1.5 percent, although Gerardi and others (2012) find that such effects last no more than one year. Ellen, Lacoe, and Sharygin (2013) estimate that an additional foreclosure leads to a 1 percent increase in crime in the block of the foreclosed property. Considering many of these effects and the administrative costs associated with a foreclosure, HUD in 2010 estimated the deadweight loss associated with a foreclosure to be approximately $50,000. Third, defaults and foreclosures may reduce real economic activity by reducing the value of households’ assets, in turn reducing consumer demand. For instance, Mian, Sufi, and Trebbi (2014) estimate that between 2007 and 2009, foreclosures accounted for one-third of the decline in house prices and one-fifth of the decline in residential investment and auto sales.

Source: https://www.cbo.gov/sites/default/files/114th-congress-2015-2016/workingpaper/50871-FHA.pdf