At the end of June, the House of Representatives voted to extend the $8,000 homebuyers' tax credit, by an extraordinary margin of 409-5. The Senate approved the measure on a voice vote. At a polarized political moment, this near unanimity was noteworthy in itself. Conservative Republicans and liberal Democrats, from cities and suburbs and small towns across the country, joined together to shower a bit more taxpayer largesse on one of America's favorite industries: real estate. But there's a problem with this bipartisan idyll.
Though the homebuyers' credit was sold as a stimulus measure, we have no reason to believe that it is anything other than another wealth transfer to a large and powerful industry, one with allies conveniently situated in every congressional district. Casey Mulligan of the University of Chicago has suggested that the credit had almost no economic impact. As Harvard economist Edward Glaeser observed, it did little more than create an incentive for "mindless house swapping." It didn't even have a meaningful impact on the behavior of first-time homebuyers -- people already planning to make purchases simply moved them forward a few months. Yet this is where we find a consensus in policymaking: We can't agree on balancing the budget or reforming entitlements or the tax code, but we can agree to churn the housing market so that a handful of real-estate agents can make a buck on commissions while the economy crumbles.
Across the world, governments have spent vast sums on doomed industrial policies. We often hear about the occasional success of efforts in East Asia to nurture shipbuilding and automobile manufacture and electronics. But we don't hear about the countless failures, in which cronies of the party in power receive endless subsidies and concessions, getting richer at the expense of the economy as a whole.
In the United States, our industrial policy for most of the last century has been centered on housing. Tax subsidies and the government-sponsored enterprises (GSEs) Fannie Mae and Freddie Mac have helped channel hundreds of billions of dollars into housing. There was a certain logic, however flawed, behind this policy. As opportunities for less-skilled workers declined, construction jobs provided a much-needed income boost to many working- and middle-class households. But like any arrangement built on government favoritism, this one was bound to fall apart. Long-term unemployment has skyrocketed in no small part because of the evaporation of construction jobs that date from the overbuilding that occurred during the bubble years.
What we need now is to turn away from this disastrous policy and find new, sustainable sources of jobs and economic growth. That will require a series of painful steps -- among them, phasing out the mortgage-interest deduction and eliminating the GSEs -- that will minimize the privileges housing enjoys relative to investments in other industries. By shifting resources from housing to more productive sectors, we will have higher and more sustainable growth. With trillions of dollars and the health of the economy at stake, the question isn't whether we must do it, but whether we will do it now or wait until our economy is in even worse shape.
The basic argument for housing subsidies is that homeownership allows Americans of modest means to accumulate wealth. From the New Deal on, the federal government has played a decisive role in the mortgage marketplace. As journalist Alyssa Katz recounts in her 2009 study of the housing industry, Our Lot: How Real Estate Came to Own Us, homeownership was far less common in the United States of the 1920s than it is today. Borrowers had to make down payments that approached half the purchase price of a house to secure a three- to five-year mortgage. For families without enough savings, there was a market in second mortgages to finance down payments, at startlingly high interest rates. As housing prices collapsed with the onset of the Great Depression, millions found themselves underwater, and this created intense pressure for what we might reasonably call a government takeover of the mortgage industry.
The political case for federal intervention was strong. Americans had come to believe that homeownership was essential to economic security and that it made for better citizens. Research had found that housing was a particularly important component of total wealth accumulation for lower-income households and suggested that it led to improved educational outcomes. The portion of the monthly mortgage payment that pays down the principal constituted a source of savings for households that were unlikely to have other significant savings or investments.
The high down payments and short-term mortgages meant that households all over the country held a significant amount of equity in their homes just a few years after buying them. In some cases, the value of this equity grew as the value of the home appreciated. These capital gains, in conjunction with the forced savings of mortgage payments, meant that millions of families had assets they could pass on to future generations. The New Dealers believed this was the path industrial workers could take to reach the independence once associated with prosperous ranchers and farmers in the American West.
The formula, however, changed dramatically at the end of the 20th century. From 1994 to 2005, the homeownership rate reached record highs, thanks largely to innovations in the mortgage-finance market that reduced down payments and minimized equity. This shifted the basic wealth-building proposition of homeownership away from savings to an almost exclusive focus on capital gains. Average down payments fell, reducing the savings required to "get in the door." More significant was the rise of mortgages that involved no forced savings: the interest-only loan, in which no equity is built because the principal is never paid down, and the "negative amortization" loan, in which payments are so low that they do not even keep up with the interest, leaving homeowners more indebted, rather than less, each month. By 2006, more than one-third of subprime mortgages had amortization schedules longer than 30 years, nearly half of Alt-A mortgages were interest-only, and more than one-fourth were negative-amortization loans.
One effect was to reduce the social benefits of homeownership, because the benefits are a product of equity and not of the mere fact that a contract has been signed and a mortgage taken out. The relationship between homeownership and social goods had been misunderstood: The traits that enabled households to build up the savings necessary for significant down payments -- hard work and the deferral of gratification -- were misattributed to homeownership itself. Paying a mortgage did nothing to improve children's educational outcomes; instead, the factors that gave rise to homeownership also led parents to raise children in a manner that led to greater educational attainment.
Without substantial down payments and conservative amortization schedules, the entire proposition of homeownership as a social good is turned on its head. Think of a homeowner with a zero-down, negative-amortization mortgage: The balance would equal at least 100 percent of the value of the house at origination and would steadily grow, putting him ever deeper in debt unless the market value of the house grew at an even faster rate. Rather than being a source of wealth, the mortgage would actually reduce the net worth of this homeowner below what it would have been had he rented.
Rather than providing a social benefit, then, homeownership without equity imposes costs. Andrew Oswald of the University of Warwick has argued that such homeownership can exacerbate unemployment by making workers less likely to move from one labor market to another. Labor mobility is badly undermined when homeowners in a depressed market can't sell their property for anything approaching the principal balance of the mortgage they originally took out to buy it.
The macroeconomic consequences of this shift toward low-equity homeownership are visible in research from the Federal Reserve that examines the assets and liabilities of U.S. households. In the first quarter of 2001, U.S. households' home equity stood at $7.7 trillion, or 61 percent of the value of all residential real estate. By the third quarter of 2008, it had declined to $7.6 trillion, even as outstanding mortgage debt increased by $5.6 trillion over the same period. By the first quarter of 2009, home equity was $1.35 trillion lower than it had been in 2001. Put another way: Despite the housing boom, the portion of residential real estate actually owned by households declined. This means that the increase in homeownership rates (and the subsequent rise in housing prices) was entirely debt-financed.
These developments provide important lessons for policymakers. First, subsidies designed to turn renters into homeowners likely did harm to many households, given that home equity declined over the 2001-09 period. Second, there was a reduction in real mortgage rates, thanks to the subsidies provided by the GSEs, the Federal Housing Administration (FHA), and the tax code. By increasing households' purchasing power, such measures drive up the prices of homes -- over the period in question, by as much as 25 percent -- without doing anything to encourage real affordability. This is why homeownership rates in Canada and in European countries that do not offer a mortgage-interest deduction are roughly the same as in the United States. While ending these subsidies would probably not alter homeownership rates, it would likely shift capital away from artificially bid-up residential real estate to more productive uses.
Admittedly, ending the subsidies would probably depress housing prices overall. Since most homebuyers base their purchase decisions on the monthly after-tax cost of housing, reducing the deduction for mortgage interest would mean that the same monthly payment would buy "less house." For example, a 25 percent deduction for mortgage interest allows buyers with a 6 percent mortgage to spend an extra $30,000 on a house without seeing any increase in their monthly payments. Similarly, an increase in down-payment requirements from the current 3.5 percent to 20 percent would mean that $20,000 of savings could be used to buy only a $100,000 house, rather than one priced at $570,000.
A general decline in housing prices would constitute a one-time wealth transfer from current homeowners to future ones -- but this would be well worth it if phased in over a period of years. In 2007 (the last year of the bubble), households' primary residences accounted for only 31.8 percent of total family assets. While primary residences make up a larger share of the assets of lower-income than of higher-income households, housing subsidies are less significant for the former because their tax rates are lower, which makes the value of deductions smaller. Because the value of subsidies provided by the FHA and the GSEs accrues to the borrower on a per-dollar-of-debt basis, their reduction is unlikely to be felt as strongly by lower-income households. The well-off take out bigger mortgages, pay more interest, and have bigger income-tax bills against which to apply a deduction: The median house value for households in the 40th through the 59th income percentiles is just $150,000, compared with $500,000 for households in the top income decile.
According to the Office of Management and Budget (OMB), the mortgage-interest deduction is expected to cost $637 billion over the five years ending in 2015. The exclusion of capital gains on primary residences is expected to cost another $215 billion over the same five years, with the deductibility of state and local property taxes on owner-occupied homes adding $151 billion. In total, these subsidies will reduce federal revenue by well over $1 trillion over a decade during which the federal government is expected to run a $9 trillion deficit. A gradual phase-out of these subsidies is therefore not only smart economics, but a fiscal necessity.
Over the years, tax experts have also zeroed in on how some of these subsidies are distributed. Under the status quo, 80 percent of the benefits from the mortgage-interest deduction go to the top 20 percent of households in terms of income. The deduction helps only those taxpayers who itemize deductions on their tax returns, which is much more common among high earners, and the value of the subsidy rises as one moves up the tax brackets. Further, as Joseph Gyourko and Todd Sinai of the University of Pennsylvania have documented, the subsidies are unevenly concentrated, with net benefits going to only 20 percent of states and 10 percent of metropolitan areas. Not surprisingly, over 75 percent of these benefits go to three high-cost metropolitan areas: New York City-Northern New Jersey, Los Angeles-Riverside-Orange County, and San Francisco-Oakland-San Jose.
A better approach would be to provide a flat tax credit to all homebuyers. This would preserve an incentive for people to buy a home but would not provide a larger incentive for people who buy bigger homes or take on outsized debts. The size of the credit could be reduced over time. Under this sort of policy, the federal government could aid middle- and working-class homebuyers at a small fraction of the cost of the current mortgage-interest deduction.
Dismantling the GSEs is a more difficult proposition. Taxpayers have already committed roughly $150 billion to the bailout of Fannie and Freddie. The Congressional Budget Office projects that losses could balloon to $400 billion over time, while other analysts suggest the taxpayer hit could be closer to $1 trillion if default and foreclosure rates stay high. The reason these estimates vary so much is that taxpayers can expect three different kinds of losses from the GSEs: those linked to the $5 trillion of mortgage-backed securities and loan guarantees that they are responsible for; those that will continue to occur as a result of regular, ongoing operations in a declining housing market; and those that may result from their being used as de facto government agencies, subsidizing foreclosure-prevention efforts.
Fannie and Freddie function today as off-balance-sheet conduits for taxpayer spending on housing, and there is no mechanism in place to end this practice. What's particularly disappointing is that Congress is on the verge of sending the president a sweeping financial-reform bill that doesn't account for Fannie and Freddie, the most expensive part of the bailouts.
A lot of thoughtful proposals for reforming Fannie and Freddie have been issued over the past year. In late May, Donald Marron and Phillip Swagel of Georgetown University put forth one of the more balanced and straightforward plans. The crux of it is to make the GSE guarantees explicit rather than implicit, and to charge an appropriate fee for them. Marron-Swagel would turn Fannie and Freddie into private companies and force them to compete with other firms. These new businesses would have a narrow mission: to buy conforming mortgages and bundle them into securities that are eligible for government backing. The key is that the federal guarantee would be transparent, and offered only in exchange for the firms' paying the government an actuarially fair price for what would amount to insurance.
An explicit government backstop might seem an unwarranted interference in housing markets, but recent experience suggests that it is unrealistic to believe that the government will stand aside next time. Some government backstop will always be implicit; better to make it explicit and price it. Once a price is established under the Marron-Swagel plan, the government would have the option of raising it, thereby reducing its support for the market, slowly and over time. The government could also reduce its footprint in the housing market by putting a ceiling on the size of the mortgages eligible to be packaged into government-backed securities. If the loan limit were capped in nominal terms, then future inflation and house-price increases would, over the course of several years, work to reduce the government's presence in the marketplace.
Likewise, other subsidies, such as the mortgage-interest deduction, can and should be gradually eliminated.
Reforming the housing sector won't miraculously restore robust economic growth. It will, however, help stanch the bleeding of productive resources into a sector that has been distorted for decades by misguided government subsidies. And over time, that will give workers and entrepreneurs the tools they need to build a stronger and more sustainable economy.