Ed Glaeser and Joseph Gyourko's e-book (pdf) on federal housing policy really is worth a read.
Glaeser and Gyourko argue that federal housing policy wrongly treats every jurisdiction alike. The country really has three different housing markets.
First, there are markets with high demand but elastic supply -- e.g., Houston and Atlanta. Housing in these markets closely tracks the cost of construction. Because developers cannot supply housing more cheaply than this over the long run, it makes no sense for the federal government to intervene directly in these markets -- there simply is no market failure. (For the poor, Glaeser and Gyourko suggest outright transfers so they can buy more housing, or housing vouchers as a second-best solution).
The second kind of market is the low-demand market -- Detroit, for instance --where housing can be had for less than the cost of construction. It is silly to subsidize the construction of new housing in these places.
The third kind of market is the high-demand, inelastically-supplied market -- the West Coast, the Amtrak corridor in the northeast, and a few other markets in the country's interior. These markets exhibit government-created market failures. Tight zoning regulations in these markets combined with high demand have raised the cost of housing well above the cost of construction, sometimes by hundreds of thousands of dollars. New construction permits amount to a tiny fraction of the housing stock.
The market failures in these jurisdictions endure because local governments have no incentive to increase the supply of new housing: the benefits from new housing go mostly to non-residents, while the costs from polution and congestion -- not to mention the effect of new competition on housing prices -- fall on existing residents.
Federal housing policy -- particularly the mortgage interest deduction -- has different effects depending on the market. In elastically-supplied markets like Houston or low-demand markets like Detroit, the mortgage interest deduction increases consumer demand and stimulates the construction of more housing.
In high-demand, inelastically-supplied markets, though, the benefits of the mortgage interest deduction flow to existing homeowners. The mortgage interest deduction paradoxically makes housing in these markets less affordable.
Glaeser and Gyourko argue that federal housing policy ought to target the broken markets. The goal, they suggest, should be to incentivize the high-demand, inelastically-supplied markets to allow more housing.
They propose capping the home mortgage interest deduction at $300,000 for households in these markets. (They would phase out the deduction for current existing mortgages for equitable reasons.) This cap would generate additional tax revenue. The federal government would then refund that revenue to the local governments (via the states) provided the local governments made sufficient progress toward increasing their housing supply. Glaeser and Gyourko essentially propose bribing local jurisdictions with their homeowners' tax dollars.
Needless to say, I think this is a great idea. Rich housing markets are becoming increasingly stratified by income. The lower and middle classes are being run out of these markets by high home prices. Their allegedly progressive citizenry, embarrassed by the increasing inequality, flail about for solutions. But they won't allow the one thing that would improve affordability over the long-term.
Incentivizing new development in these cities would benefit the rest of us, too. The high-demand cities have high demand because they make their citizens more productive. Restricting access to them creates a net welfare loss for the country as a whole.
But Glaeser and Gyourko's proposal will be dead on arrival. The costs of their proposals would fall on a small, clearly identifiable group -- wealthy homeowners in high-priced markets. These homeowners would suffer a double-whammy, through the loss of their tax deductions and the loss of home value due to increased supply. Everyone else would benefit, of course, but those benefits would be diffuse. Local governments would benefit from the extra revenue, but local governments don't vote for senators or congressmen. If this proposal were enacted, Barbara Boxer would hear the howls of Marin County homeowners from her seat on the Senate floor.
