January 2, 2015 | By Moussa Diop | Back to blog

The dramatic expansion in subprime mortgage credit fueled a remarkable boom and bust in the U.S. housing market and spawned a global financial crisis in 2007 and 2008. While research has examined the housing and mortgage markets of the previous decade, how the expansion in mortgage credit affected the rental market has remained unclear, although over 30 percent of all U.S. households reside in the rental market.

Moussa Diop
Moussa Diop, assistant professor in Real Estate and Urban Land Economics at the UW-Madison Wisconsin School of Business.

Brent W. Ambrose, at The Pennsylvania State University, and I sought to fill this gap by showing how the residential rental market was adversely affected by the development of subprime lending long before the advent of the 2007-2008 subprime-induced financial crisis.

While the benefits of homeownership are widely acknowledged, the costs associated with policies designed to promote the housing market and homeownership can be substantial. We expanded on studies of homeownership externalities by focusing on the impact that the credit-fueled growth in homeownership during the recent housing boom had on the risk of the rental sector. In particular, we examined how subprime lending created ripple effects across the residential rental market.

Our formal analysis rests on the fundamental decision of whether to rent or own. Since most households typically borrow the bulk of the purchase price of their homes, the availability of mortgage financing influences these decisions. Thus, the sustained growth in mortgage lending from 2001 to 2006, attributed in part to the interaction of looser underwriting standards and the development of innovative mortgage products targeted at under-served populations, enabled numerous households previously excluded from the mortgage market to buy homes. As a result, the national average homeownership rate grew 2.4 percent from 67.5 percent in 2000 to 68.9 percent in 2006. This phenomenon was more pronounced in urban areas, where average homeownership rates in metropolitan areas and major cities rose by 2.9 percent and 5.6 percent, respectively.

However, while the homeownership rate was increasing, the risk profile for the population of renters also changed; the median renter household income as a percentage of all household median income declined from 67.5 percent in 2001 to 62.7 percent by 2005, indicating a significant shift in the income level of the renter population. This increase in the risk profile of the rental population is consistent with the notion that expansion in mortgage credit through subprime lending altered the underlying risk distribution of the rental population.

Given the remarkable expansion of mortgage credit in the previous decade, we wondered to what extent the growth in homeownership adversely affected the residential rental market. We addressed this question by examining the performance of residential leases, using a national database of multifamily rental data. We analyzed the probability of lease payment defaults during the period of explosive growth in subprime lending and found a significant (both economically and statistically) positive relation between subprime lending and the likelihood of lease defaults. Our results indicate that a 1 percent increase in an area’s subprime activity corresponds to a 1.9 percent increase in the area’s lease default index. We also show that the increase in lease defaults resulted from the migration of low risk renters into homeownership. As a result of this shift in the underlying risk profile of the renter population, which led to higher rental default rates and losses, we provide evidence suggesting that areas that experienced higher rental default rates subsequently experienced higher rent rates.

Furthermore, we document a simultaneous deterioration in the performance of multifamily properties; for example, our analysis indicates that a 1 percent increase in rental defaults results in a 0.16 percent decrease in the average annual income component of the property return. Finally, we also document a positive and significant connection between rental default rates and multifamily property capitalization rates, verifying that an increase in overall rental contract defaults results in a decline in multifamily property values. This confirms the fundamental spillover mechanism whereby subprime origination activity affected multifamily asset values by allowing lower risk renters to migrate into homeownership, leaving behind a riskier renter population.

Our results provide an important cautionary note when evaluating the benefits associated with public policies designed to promote a particular social outcome; there can be unintended consequences for another segment of society.