Association for Public Policy Analysis & Management

In the summer of 2008 the U.S. government effectively nationalized the Government Sponsored Enterprises (GSEs), Fannie Mae and Freddie Mac – taking control of the secondary mortgage market.  Virtually overnight, private institutions stopped purchasing mortgages, and the government assumed responsibility for the vast majority of outstanding U.S. mortgage debt.  Eight years later, the secondary mortgage market remains structurally unchanged - the federal government still controls the GSEs, while policy makers have finally begun considering legislation to fundamentally restructure the secondary market for mortgages.  

Findings from this paper provide policy makers with insight into potential consequences of an increasingly privatized secondary mortgage market.  By examining the expansion of the private secondary market at the turn of the 21^st century, this paper documents the way the privatizing secondary market transformed conventional mortgage lending and how secondary market activity contributed to the intensification of place based inequality in the Philadelphia region.   

This paper addresses the following key questions:  

• How was secondary mortgage market activity distributed across the Philadelphia region at the turn of the 21^st century? 
• How did secondary mortgage market activity contribute to the reproduction place based inequality in the Philadelphia region at the turn of the 21^st century?

This paper relies on 12 years Home Mortgage Disclosure Act data (1996 – 2007) combined with census tract level data from three decennial censuses (1990, 2000, 2010).  Descriptive analyses examine the distribution of GSE and private institutions’ purchasing activity in the secondary market over time, and across different types of neighborhoods in the region.  Change models estimate the effect of GSE conventional market share on neighborhood level changes in homeownership from 1990 to 2010.  The results from these initial analyses are then used to estimate a set of difference in difference models to estimate the effect of GSE-informed changes in homeownership on changes along key neighborhood conditions from 1990 to 2010: poverty levels, household incomes, and minority populations.  

From 1996 to 2007, GSE conventional market share declined from over 70% to less than 50%.  During this time GSE market share was most heavily concentrated in predominately white, middle and upper income areas, while private purchasing was most active in historically underserved areas throughout the Philadelphia region.  Higher levels of GSE market share were significantly associated with net gains in homeownership, and GSE-informed changes in homeownership from 1990 to 2010 were significantly associated with net reductions in neighborhood poverty levels.  The inverse was also true: lower levels of GSE market share were significantly associated with net declines in homeownership, which were also significantly associated with increased neighborhood poverty.

These findings suggest that declining GSE market share transformed the safety and security of homeownership in the conventional market.  While the effects of subprime lending on borrowers and neighborhoods is well documented, this study points to substantive differences in the conventional market associated with different funding streams in the secondary market.  These findings provide substantive food for thought for policy makers considering the merits of an increasingly privatized secondary mortgage market.