The Economics of Housing Supply


chart by Smallman12q

Housing economists Edward L. Glaeser and Joseph Gyourko have posted The Economic Implications of Housing Supply to SSRN (behind a paywall but you can find a slightly older version of the paper here). The abstract reads,

In this essay, we review the basic economics of housing supply and the functioning of US housing markets to better understand the distribution of home prices, household wealth and the spatial distribution of people across markets. We employ a cost-based approach to gauge whether a housing market is delivering appropriately priced units. Specifically, we investigate whether market prices (roughly) equal the costs of producing the housing unit. If so, the market is well-functioning in the sense that it efficiently delivers housing units at their production cost. Of course, poorer households still may have very high housing cost burdens that society may wish to address via transfers. But if housing prices are above this cost in a given area, then the housing market is not functioning well – and housing is too expensive for all households in the market, not just for poorer ones. The gap between price and production cost can be understood as a regulatory tax, which might be efficiently incorporating the negative externalities of new production, but typical estimates find that the implicit tax is far higher than most reasonable estimates of those externalities.

The paper’s conclusions, while a bit technical for a lay audience, are worth highlighting:

When housing supply is highly regulated in a certain area, housing prices are higher and population growth is smaller relative to the level of demand. While most of America has experienced little growth in housing wealth over the past 30 years, the older, richer buyers in America’s most regulated areas have experienced significant increases in housing equity. The regulation of America’s most productive places seems to have led labor to locate in places where wages and prices are lower, reducing America’s overall economic output in the process.

Advocates of land use restrictions emphasize the negative externalities of building. Certainly, new construction can lead to more crowded schools and roads, and it is costly to create new infrastructure to lower congestion. Hence, the optimal tax on new building is positive, not zero. However, there is as yet no consensus about the overall welfare implications of heightened land use controls. Any model-based assessment inevitably relies on various assumptions about the different aspects of regulation and how they are valued in agents’ utility functions.

Empirical investigations of the local costs and benefits of restricting building generally conclude that the negative externalities are not nearly large enough to justify the costs of regulation. Adding the costs from substitute building in other markets generally strengthens this conclusion, as Glaeser and Kahn (2010) show that America restricts building more in places that have lower carbon emissions per household. If California’s restrictions induce more building in Texas and Arizona, then their net environmental could be negative in aggregate. If restrictions on building limit an efficient geographical reallocation of labor, then estimates based on local externalities would miss this effect, too.

If the welfare and output gains from reducing regulation of housing construction are large, then why don’t we see more policy interventions to permit more building in markets such as San Francisco? The great challenge facing attempts to loosen local housing restrictions is that existing homeowners do not want more affordable homes: they want the value of their asset to cost more, not less. They also may not like the idea that new housing will bring in more people, including those from different socio-economic groups.

There have been some attempts at the state level to soften severe local land use restrictions, but they have not been successful. Massachusetts is particularly instructive because it has used both top-down regulatory reform and incentives to encourage local building. Massachusetts Chapter 40B provides builders with a tool to bypass local rules. If developers are building enough formally-defined affordable units in unaffordable areas, they can bypass local zoning rules. Yet localities still are able to find tools to limit local construction, and the cost of providing price-controlled affordable units lowers the incentive for developers to build. It is difficult to assess the overall impact of 40B, especially since both builder and community often face incentives to avoid building “affordable” units. Standard game theoretic arguments suggest that 40B should never itself be used, but rather work primarily by changing the fallback option of the developer. Massachusetts has also tried to create stronger incentives for local building with Chapters 40R and 40S. These parts of their law allow for transfers to the localities themselves, so builders are not capturing all the benefits. Even so, the Boston market and other high cost areas in the state have not seen meaningful surges in new housing development.

This suggests that more fiscal resources will be needed to convince local residents to bear the costs arising from new development. On purely efficiency grounds, one could argue that the federal government provide sufficient resources, but the political economy of the median taxpayer in the nation effectively transferring resources to much wealthier residents of metropolitan areas like San Francisco seems challenging to say the least. However daunting the task, the potential benefits look to be large enough that economists and policymakers should keep trying to devise a workable policy intervention. (19-20)

The Economic Implications of the Housing Supply

Ed Glaeser and Joe Gyourko posted The Economic Implications of the Housing Supply which is forthcoming in The Journal of Economic Perspectives. In it, they

review the basic economics of housing supply and the functioning of U.S. housing markets to better understand the impacts on home prices, household wealth and the spatial distribution of people across markets. Section II documents the state of housing affordability in the U.S., and begins with three core facts about housing supply. First, when building is unrestricted by regulation or geography, housing supply curves seem relatively flat, meaning that we can approximate reality by referring to a single production cost. Second, both geography and regulation severely restrict the ease of building in some parts of the country. These constraints raise building costs both directly, by increasing time delays and reducing the amount of available land, and indirectly, by ensuring the homes are produced more on a one-by-one basis rather than in bulk. Third, the supply of housing is kinked and vertical downwards because housing is durable. (2, citation omitted)

These are themes that Glaeser and Gyourko have touched on before, but this essay does a service by updating them ten years after the financial crisis.

Glaeser and Gyourko have consistently hit on some important points that can garner attention at the national level , but there has been no real action on them as of yet:

  • where supply is regulated, housing costs more;
  • heavy land use regulation in places like NYC and SF reduces the nation’s overall economic output; and
  • existing homeowners tend to oppose new projects, which is consistent with their financial self-interest.

Glaeser and Gyourko do not give up hope that policymakers can craft solutions that deal with the political economy of housing construction. One first step would be to develop a toolkit of carrots and sticks that can be employed at the national and state level to incentivize local governments to take actions that are in the interest of their broader communities and the nation as a whole.

We know we need more housing in highly productive regions. We just need to figure out how to build it.

Is The Mortgage Interest Deduction Inequitable?

photo by Elana Centor

I have certainly thought so, as do many other housing policy types. Daniel Hemel and Kyle Rozema have a more nuanced view in their paper, Inequality and the Mortgage Interest Deduction, that was recently posted to SSRN. The abstract reads,

The mortgage interest deduction is often criticized for contributing to after-tax income inequality. Yet the effects of the mortgage interest deduction on income inequality are more nuanced than the conventional wisdom would suggest. We show that the mortgage interest deduction causes high-income households (i.e., those in the top 10% and top 1%) to bear a larger share of the total tax burden than they would if the deduction were repealed. We further show that the effect of the mortgage interest deduction on income inequality is highly sensitive to the alternative scenario against which the deduction is evaluated. These findings demonstrate that claims about the distributional effects of the mortgage interest deduction depend critically on the counterfactual to which the status quo is compared. We extend our analysis to the deduction for state and local taxes and the charitable contribution deduction. We conclude that the appropriate counterfactual for distributional claims is dependent upon political context — and, in particular, on the feasible set of politically acceptable reforms up for consideration.

To make this a bit more concrete, the authors offer a simple hypothetical:

to show how a provision of the tax code can provide a disproportionate share of dollar benefits to the rich while also causing the rich to bear a larger share of total tax liabilities. Imagine a society with two households—a rich household with pre-tax income of $100, and a poor household with pre-tax income of $50. Further imagine that the rich household pays $12 in mortgage interest and the poor household pays $9 in mortgage interest. Say that the tax system is structured such that the tax rate on the first $50 of income is 20% and the tax rate on all income above the $50 threshold is 40%.

If the tax system does not allow a deduction for mortgage interest, the rich household would pay a tax of $30 ($10 on the first $50 and $20 on the next $50), and the poor household would pay a tax of $10. Thus the government would collect a total of $40 in revenue; the rich household would bear 75% of the total tax burden ($30 divided by $40); and the poor household would bear the remaining 25%. If the tax system allows each household to deduct mortgage interest, the rich household would receive a benefit from the deduction of $4.80 ($12 times 40%), and the poor household would receive a benefit from the deduction of $1.80 ($9 times 20%). The benefit of the MID in dollar terms is clearly greater for the rich household than for the poor household. In percentage terms, the rich household received 72.7% of total MID benefits, while the poor household received 27.3% of total MID benefits. And yet the rich household now bears 75.45% of the total tax burden ($25.20 divided by $33.40), as compared to 75.0% before, while the poor household now bears 24.55% of the total tax burden ($8.20 divided by $33.40), as compared to 25.0% before. (Government revenue decreases from $40 without the MID to $33.40 with the MID.) (8, footnote omitted)

The authors conclude that their analysis “simultaneously confirms and challenges widespread beliefs regarding the effect of tax expenditures on inequality. The mortgage interest deduction does indeed appear to be inequality-increasing relative to a counterfactual in which the deduction is repealed and revenues are reallocated to all households on a equal basis; when the mortgage interest deduction is evaluated against other counterfactuals, however, the distributional effects are more nuanced.” (40)

Where does this leave us? It reminds us that we should be careful about promoting simple policy proposals without taking into account the likely context in which they might be implemented.