The case for unwinding Fannie and Freddie
If lawmakers could design a housing policy from scratch what might it look like?
Today Fannie Mae and Freddie Mac--two government-sponsored enterprises originally designed to increase the availability of loans and thereby raise levels of home ownership--dominate the US mortgage lending market. Fannie Mae, which was established in 1938 as part of Franklin Delano Roosevelt's New Deal, provides local banks with federal money to finance home mortgages. Freddie Mac, created in 1970, underwrites mortgages that fall below a certain size threshold with the intention of helping homeowners get access the housing market. These mortgages are cheaper since they implicitly--and after 2008 explicitly--benefited from a government guarantee.
In 2008, after incurring significant losses on their portfolios, Fannie and Freddie were taken over by the government; still, they finance the majority of mortgages in the US.
According to my latest research,* this arrangement is problematic. My research shows that houses financed with loans through Fannie and Freddie have higher prices than comparable homes bought with unsubsidized (jumbo) mortgages. In other words: the credit that Fannie and Freddie provide ostensibly to help homebuyers get a foot on the property ladder, has the unintended consequence of increasing house prices. That means a fraction of the lower cost of credit is passed on to the sellers of homes.
My colleagues, Manuel Adelino at Dartmouth College and Felipe Severino a PhD candidate at Sloan, and I looked at deed records from ten big US cities, including New York and Boston, for the ten-year span of 1995-2005. We compared the sale prices of houses that were eligible for financing through Fannie and Freddie with houses that were sold for prices just above the conforming loan limit (CLL). Fannie and Freddie underwrite home loans that fall beneath the CLL, an amount set by Congress each year. Mortgages for amounts greater than the CLL are considered non-conforming loans or jumbo loans.
We find that houses that were eligible for Fannie and Freddie loans cost $1.10 more per square foot than houses eligible for jumbo loans. Considering that the average home size is 1,800 sf, this represents a disparity of $1,980 or a .5% difference in price per square foot from year to year. This credit helps buyers afford a home, to be sure, but a big fraction of that subsidy goes to home sellers in the form of higher prices.
My concern is that the costs of Fannie and Freddie might vastly outweigh their benefits. On one hand, Fannie and Freddie provide a slight reduction in borrowing costs to homebuyers. On the other hand, the cost imposed on taxpayers through the bailout of Fannie and Freddie, and the lobbying efforts of these entities in the period leading up to the crisis, have proven to be humongous.
In February, the White House announced plans to reduce the government's outsized role in mortgage funding and wind-down Fannie and Freddie. This is a very welcome goal. But the findings from our study also caution that the winding down of government support for the mortgage market has to be gradual, since we would surely see a reduction in the average price of houses. In the short run a drop in asset prices could have negative multiplier effects, which is certainly not what we want in the current economic environment.
Unwinding Fannie and Freddie over a period of time seems the best way to go. This could happen through restructuring the overall loan support that is provided or by lowering the CLL every year by a preset amount. The influence on house prices would be smoother, and more incremental, as opposed to a big shock that might take place if the government were to, say, close down Fannie and Freddie tomorrow. Of course, the big question is whether the political process allows for such a smooth transition or if these efforts would be diverted over time by opposing political interests.
Bio: Antoinette Schoar is the Michael Koerner '49 Professor of Entrepreneurial Finance at MIT Sloan School of Management. The opinions expressed in this commentary are solely those of the author.