Did Affordable Housing Tank the Economy?

What were people thinking? That's the question the housing bubble and resulting financial crisis immediately bring to mind. Lenders offered loans that borrowers plainly could not afford, and borrowers went right along with it. Easy credit drove prices up. Securities based on the bad mortgages spread the risk. And as soon as house prices dropped--and borrowers with little equity could no longer sell to get themselves out of trouble--the entire thing fell apart.

Was the problem private-sector greed and a lack of regulation? That's certainly the thinking behind the Dodd-Frank law that Democrats enacted in 2010. But Peter Wallison, a fellow of the conservative American Enterprise Institute and a member of the federal government's Financial Crisis Inquiry Commission (FCIC), has long held a very different position: He says that the government, trying to promote affordable housing for the low- and moderate-income, led the charge to lower lending standards, dragging the rest of the market--and eventually the economy--down with it.

His latest enunciation of this view is Hidden in Plain Sight: What Really Caused the World's Worst Financial Crisis and Why It Could Happen Again. In it, he proves beyond any doubt that the government did incredibly ill-advised things in its drive to make mortgages easier to get. But his more aggressive argument--that affordable-housing policies were "the element without which there would not have been a widespread financial breakdown in 2008"--is a dicier proposition.

The story begins in 1992, when the Department of Housing and Urban Development (HUD) decided to impose affordable-housing goals on "government-sponsored enterprises" (GSEs) Freddie Mac and Fannie Mae, institutions that buy mortgages and securitize them to provide liquidity to the market. Historically, Fannie and Freddie had been rather conservative about the loans they acquired, insisting on sizable down payments, good credit histories, and reasonable debt-to-income ratios.

At first the change wasn't a big deal. In fact, the initial goal--that 30 percent of the GSEs' loans be to borrowers below the median income of the area they lived in--was below what the GSEs were already doing. There's nothing inherently wrong with loaning to these "LMI" (low- or moderate-income) borrowers, so long as they demonstrate an ability to pay and a history of meeting their credit obligations.

But the goals, which covered several other categories besides LMI, quickly ramped up. By 2001--at which point the bubble had been expanding for about four years, judging by the Case-Shiller index of home prices--the LMI goal reached 50 percent, and in 2008 it was 56 percent. The goals certainly seemed to work: The GSEs' loans to the targeted categories closely tracked the percentages they were required to meet.

Wallison does a terrific job of documenting how much of a struggle it was to find enough qualified borrowers. A 2003 Fannie presentation noted that, in the scramble to meet the previous year's goal, the GSEs "did deals at risks and prices we would not have otherwise done." Two years later, in another presentation, Fannie complained about "having to compromise credit standards," deals that were "producing negative cash flow," and exotic products that "encourage[] continuation of risky lending."

And all this indeed corresponded to an increase in risky loans on the GSEs' books. In 1990, for instance, about 1 in 100 loans they purchased had down payments of 3 percent or less; this rose to 1 in 7 in the early 2000s and 1 in 2.5 when the bubble popped.

The government had other ways of promoting affordable housing as well. One was the Federal Housing Administration, which insures loans to people who cannot meet conventional standards. Wallison cites a Quicken ad from 2000 bragging that, with FHA insurance, "borrowers need as little as 3% of the 'total funds' required," "credit requirements are flexible," and borrowers can have a "higher debt-to-income ratio than most insurers typically allow." Another was the Community Reinvestment Act, through which the government pressured banks to lend to "underserved" borrowers.

Whether you're a conservative or a liberal, this makes little sense. If we want to help the poor, we can give them money--and if we'd like to make sure that the money is spent on basic needs, we can instead give vouchers for things like housing. But that level of specificity wasn't good enough for the government. Instead, it decided to facilitate homeownership in particular, which, whatever its virtues, imposes serious financial responsibilities. To achieve this, the government worked to undermine the various screens that lenders use to make sure borrowers are up to the challenge--down payments, credit scores, debt-to-income ratios--mainly by taking on the risks associated with the loans. And it did this while a housing bubble was inflating toward a spectacular explosion.

Proving that the government made horrible decisions in the lead-up to the crisis is not the same as proving the government caused the crisis, however, and Wallison's attempt to go this extra mile is where things get messy. Wallison is, after all, a minority of a minority: Of the ten members of the FCIC, six Democrats wrote a majority report calling for regulation, three Republicans wrote a minority report arguing that a variety of disparate global factors caused the crisis and regulation wasn't the answer, and Wallison filed a report by himself. Wallison argues that the government was directly responsible for a huge share of the nation's worst mortgages, and that its lowered underwriting standards went "viral"--private actors in the mortgage market were forced to lower their standards to compete, higher-income homebuyers availed themselves of the lower requirements so they could get bigger houses, and the housing bubble hid the problems with this situation as it spiraled out of control.

The most striking claim Wallison makes, drawing on research by his American Enterprise Institute colleague Edward Pinto, is that there were 31 million nontraditional (subprime or "Alt-A") mortgages in the financial system in 2008. This constituted 57 percent of all mortgages outstanding, and the government could be blamed for 76 percent of them, with an unpaid principal balance of $3.4 trillion. The GSEs held the credit risk for 16.5 million loans; the FHA and other federal agencies, 5.1 million; and private institutions covered by affordable-housing policies, 2.2 million.

Wallison says this was especially problematic because the GSEs had led everyone to believe the numbers were much lower. In their reports, for example, the GSEs repeatedly claimed that they had negligible exposure to subprime loans, even though they had many loans that had been made to borrowers with FICO credit scores below 660 and thus were, by Wallison's preferred definition, subprime. In 2008 the Federal Reserve estimated that there were only 6.7 million subprime loans in existence, with a total value of $1.2 trillion. It did not count an additional 9 million loans to borrowers with scores below 660.

Wallison is correct that Fannie and Freddie understated their exposure to risky loans: The Securities and Exchange Commission charged former Fannie and Freddie executives with fraud for that very reason, saying they failed to report about $300 billion in de facto subprime loans and also understated their exposure to Alt-A. But even the SEC allegations didn't define all loans with sub-660 FICO scores as subprime -- and as Wallison concedes, starting in 2003, Fannie filed reports to the SEC that included credit-score data and no one seemed alarmed. Obviously, lower credit scores mean more risk, but is 660 really a magic line separating prime from subprime?

A 2001 ruling by federal bank regulators did include FICO scores below 660 as one element that defines a subprime loan. But when Wallison quotes the definition, he cuts it off before this: "This list is illustrative rather than exhaustive and is not meant to define specific parameters for all subprime borrowers. Additionally, this definition may not match all market or institution specific subprime definitions, but should be viewed as a starting point from which the [bank regulators] will expand examination efforts." As the FCIC noted, the definition makes an "imprecise distinction"--and critics say there's a big difference between real subprime loans and the sub-660 loans the GSEs didn't count in that category.

David Min of the liberal Center for American Progress, for example, has noted that sub-660 Freddie loans not classified as subprime had a serious-delinquency rate of about 11 percent in 2010--compared with 7 percent for conforming loans, a 9 percent national average, and 28 percent for loans actually classified as subprime. The FCIC majority made a similar point. Wallison dismisses the FCIC's argument as merely pointing out that "other loans were worse" than the ones he accuses the GSEs of misclassifying, and he has gone back and forth with Min. But the bottom line is that, while these disputed loans were moderately riskier than the loans we all agree are "prime," they were much less risky than the loans we all agree are "subprime." This raises red flags given how dramatically Wallison's definition hikes the number of subprime loans.

At the very least, the direct aspect of Wallison's case--that most mortgages were risky and the government was involved with three-quarters of risky mortgages--isn't as conclusive as it appears. And when we dig into the question of where the financial crisis hit hardest, we see that government-backed loans to the low-income aren't the answer. This puts more pressure on Wallison's attempt to blame the government for the private sector's behavior--the theory that the government's lending practices went "viral"--than it can probably bear.

Min has noted, for example, that while the federal government owned or guaranteed 67 percent of all mortgages, it was responsible for just 32 percent of serious delinquencies--while private-label securities generated 13 percent of loans and 42 percent of serious delinquencies. Similarly, at a recent event for Wallison's book, Moody's analyst Mark Zandi pointed out that, relative to debt outstanding, realized losses on residential mortgages at Fannie and Freddie were just 3 to 4.5 percent, compared with 6 percent for banks and 23 percent for private-label securities. Jason Thomas of the Carlyle Group has written that, while Fannie and Freddie imploded and needed to be placed into conservatorship, had they "simply been required to hold equity capital in roughly the same proportion that banks are, shareholders would have absorbed all of the losses."

Further, an intriguing new working paper suggests that delinquency, if measured in dollar amounts, actually increased most among middle- and higher-income borrowers when the bubble popped. "Since these borrowers have much larger mortgages, a small increase in their default rate has [a] large impact on the amount [of] dollars in delinquency," the authors note. Another interesting finding is that lower-income borrowers' share of mortgage originations changed little between 2002 and 2006--the authors say credit expanded for all consumers, not just the poor.

The final obstacle for Wallison's thesis is what happened in markets that the affordable-housing goals didn't cover: Many countries had housing bubbles around the time that we did, despite not having similar policies. As Wallison says, these countries' bubbles weren't as bad as ours. But as Zandi pointed out at the book event, the countries that had lively markets for private-label mortgage securities tended to fare particularly poorly. Others have noted that commercial real estate in the U.S. experienced a bubble similar to the one that afflicted residential real estate.

It's possible to weave a narrative that's consistent with all of these facts and still lays the blame at the feet of the government's affordable-housing push, and Wallison gives it a good shot. To him, federal housing policy was like a lit cigarette tossed aside in a dry forest: It still started the fire even if the flames burned brightest far from where it landed. But fair-minded readers will remain skeptical. That the government stupidly encouraged risky lending to promote homeownership is undeniable; that the financial crisis wouldn't have happened otherwise is harder to believe, if ultimately nondisprovable.