Richard K. Green: The United States has idiosyncratic institutions whose purpose is to provide capital to mortgage markets while not originating loans. Two of these institutions, Fannie Mae and Freddie Mac, are among the largest financial intermediaries in the world, with assets of about $800 billion each.1 Each company guarantees well over $1 trillion of off-balance-sheet mortgages.2
Beyond being large, both of these companies are highly profitable, with typical book returns on equity of 25 percent.3 Critics of the firms argue that, on a risk-adjusted basis, they are too profitable.4 Specifically, they argue that shareholders whose debts are implicitly guaranteed by the U.S. government should not earn such large returns.
The size and profitability of the companies are likely the reason that they are required to meet affordable housing goals. The original charters of the companies were silent on the issue of affordability. Rather, they emphasized stability, liquidity, and ubiquity.5 It was not until 1992, with passage of the Federal Housing Enterprises Financial Safety and Soundness Act, that Fannie Mae [End Page 241] and Freddie Mac faced a regulatory requirement to target mortgage funding to "low- and moderate-income" borrowers, to "underserved" census tracts, and to "very low-income" borrowers or low-income borrowers in "low-income" areas. It is not a stretch to think that Congress felt that, in light of the companies' special status and profitability, they had a special obligation to help those at the margins of the housing market. These targets became known as the "affordable housing goals."
Click for larger view
|Table 10 |
GSE Performance on the Low- and Moderate-Income Housing Goals, 1996–2003
The paper by An, Bostic, Deng, and Gabriel asks a very simple question: Did the regulatory requirements put in place in 1992 work? More specifically, did the goals actually increase the flow of mortgage credit available to low- to moderate-income borrowers and underserved neighborhoods? The paper is typical of the authors in that it is thorough, is careful, and reveals an impressive mastery of data. Yet my interpretation of their findings is different from theirs.
Before getting into the authors' methods and results, however, data from the 2004 rulemaking for the affordable housing goals shows that they did seem to have a powerful effect on the companies' behavior.6 Table 10 presents the performance of Fannie Mae and Freddie Mac for the low- and moderate-income housing goal between 1996 and 2003.7 A unit counted toward meeting the low- to moderate-income goal if its household had income at or below the median family income for the tract in which it lived. "Official" performance [End Page 242] is the measure of whether the companies complied or not with the goal. Official performance was not a straightforward count of units that complied with the goal. I discuss this further below. In any event, one cannot help but notice that the official performance of both companies, and especially of Freddie Mac, changed dramatically between 1996 and 2001, the first year in which the companies were required to meet the 50 percent threshold for the low- to moderate-income goal. Freddie increased its performance by 12 percentage points, or 29 percent, between 1996 and 2001, a remarkable change consistent with the hypothesis that the goal influenced its behavior.
The story gets even more interesting when we consider the difference between the "official" and the "baseline" performance of Freddie Mac. Before 2004, Fannie Mae and Freddie Mac received extra credit for the affordable housing goals when they made loans to small rental properties.8 Freddie Mac also received a temporary adjustment factor (TAF) that further rewarded it for making multifamily loans.9 These bonuses explain the difference between the companies' baseline performance and their official performance. Sure enough, between 1996 and 2001, Freddie Mac's purchases of multifamily loans exploded, increasing from a little over $2 billion in 1996 to more than $9.5 billion...