Thoughtful Reform of U.S. Housing Policy

The following remarks were addressed to attendees of the Conference on Restructuring U.S. Residential Mortgage Markets hosted by the Brookings Institution on February 11, 2011.

The meltdown we have experienced in housing finance has been very painful and taught us lessons new and old. Millions of people have lost their homes, savings, businesses, and jobs. These enormous costs should not be minimized or forgotten; we should learn from them. The meltdown has taught us a lot about the potential instability of our whole financial system--especially housing finance--that should help us avoid another such disaster.

Some lessons had been learned many times before and should have been obvious:

  • Lesson 1: Lending to people who probably can't pay back is a bad idea for both borrower and lender-no matter how fancy the hedges against risk.
  • Lesson 2: People with an opportunity to make large amounts of money quickly will probably take it.
  • Lesson 3: Herd behavior frequently trumps rationality and risk aversion.

In addition, several questions were definitively answered that were unclear before the crisis:

  • Can we have a national or even international housing bubble? Yes, indeed, we can.
  • Did the implicit government guarantee of the Government Sponsored Enterprises (GSEs) have a real basis? Yes, and it proved expensive.
  • Were Fannie and Freddie too big to fail? Yes. (So were all the big banks).
  • Can a company with private shareholders and public responsibilities navigate thorough a boom-bust of this magnitude? No-or, at least, not without strong price and profit regulations that would turn the company into a public utility.

We now have the opportunity to set the framework for something better. Housing finance is still in disarray and taxpayer-owned Fannie Mae and Freddie Mac are currently financing an overwhelming proportion of new mortgages. Fundamental changes in housing finance, especially in the role of the GSEs cannot be executed quickly. However, now is the time to ask what role we want the federal government to play in housing finance going forward and what kind of role should the GSEs play, if any.

However, before we even get to housing finance we should raise the fundamental question of how important housing investment should be in the economy of the future. Over the years we have created a complicated set of housing policies-including tax provisions, spending programs, and financial institutions-dedicated to increasing the flow of investment into housing, especially owner occupied housing. Some (notably the mortgage interest deduction) favor higher income groups and others favor low and moderate income groups, but all favor housing over other types of investment.

Americans have a taste for size and quality in housing, but we have overdone it. At this moment in our economic history, we ought to shift emphasis away from housing and other forms of consumption toward investment designed to increase productivity growth. We are facing a declining ratio of workers to retirees in a highly competitive world. Unless we increase the productivity of those workers, the American standard of living cannot continue to grow. We must increase investment in skills, research, plants and equipment, public and private infrastructure, and non-residential construction-almost everything but housing. Moreover, it is time to stop encouraging Americans to put their limited savings into upgrading and expanding their houses and shift instead to forms of saving that support productivity growth.

An equally important goal of economic policy should be reversing the trend in income and wealth inequality, while encouraging a broader sharing of the fruits of productivity growth. But here again policies to subsidize housing for low-income people may be less effective in increasing their opportunities than income or wage supplements, incentives to build financial assets, and investment in their skills, health care, child care, and access to jobs. Pressure from the housing industry has led us to over-emphasize housing as a tool for redressing income disparities. The argument on the other side, of course, is that helping low-income groups get ahead does not have broad support, so making common cause with the housing industry, even if it does not result in the most effective policies, is better than nothing at all.

The mortgage interest deduction is a particularly counter-productive policy in light of current economic priorities. As part of a broader tax reform, we should phase out the mortgage interest deduction over time in favor of a limited mortgage interest credit, at the lowest bracket rate, available without respect to itemization. This shift would reduce the relative attractiveness of investment in high-end housing and focus a smaller tax subsidy to lower and moderate income home-owners.

With respect to housing finance policies and the GSEs, the housing bubble and the crisis it precipitated have given us some lessons in what not to do and strong incentives to seize the opportunity to transition, albeit not too hastily, to a more robust set of policies.

To determine the desirable future role of the GSEs it is not necessary to answer the question of exactly how much they contributed to the bubble and the crisis.

Clearly, GSE purchases of subprime mortgages added to the demand for those mortgages, especially in the latter stages of the boom-a demand already flowing from private sources. Pressure from share holders to get into this lucrative market, along with lax regulation of GSE capital standards and the mistaken assumption that subprime lending contributed to affordable housing goals, led to the excessive risk taking that made the federal takeover necessary.

While it may no longer be desirable for policy to favor housing over more productivity-increasing investment, we certainly want policy to ensure an adequate and reasonably stable flow of credit into housing and to make housing credit available to everyone who can demonstrate a reasonable probability of making the required payments. The question before this conference is whether transformed GSEs are essential to these goals or whether we would be better off without them. The Treasury White Paper released today poses the question clearly, but does not answer it and instead offers three options for debate.

The extraordinary decline in lending standards in the last decade fed the housing boom and the financial house of cards built on top of it. While GSEs contributed to that decline, they certainly didn't start it. If we want to prevent another similar disaster, the place to begin is at the originator level. We need to set conservative national mortgage lending standards that are strictly enforced: significant down payments, documentation of income, and no teaser rates or balloon payments. There will always be some credit risk in lending, but housing is not a good place for creative finance. The second line of defense is originator incentives: requiring originators to retain a significant portion of the credit risk or even all of it, as in Denmark. The third line of defense is rules for GSEs-as long as we have GSEs-not allowing GSEs to buy subprime loans at all or only allowing them to buy a small visible percent.

But once the housing markets return to normal do we need GSEs at all? The papers presented at this conference offer a variety of well-reasoned answers to this question, as does the Treasury White Paper. Any transition will have to be slow. We have time to downsize the GSEs over the next few years and decide their ultimate fate later. Because GSEs inevitably involve an implicit guarantee, which will likely make them too big to fail, the burden of proof is on those who think we should retain GSEs. However, we probably cannot avoid future boom and bust cycles in housing, so it may useful to keep smaller well-regulated GSEs in case we need backstop financing in housing after a crash. Alternatively, such a lender of last resort might well be the Federal Reserve. I look forward to the discussion of these issues during the rest of the conference