In a recent post citing our data on metro cap-and-trade costs, Free Exchange, the blog of the Economist, feared that people are moving from “clean” metropolitan areas with low carbon emissions to “dirtier” ones with higher emissions. As evidence, the author points to a recent paper by Ed Glaeser and Matthew Kahn, who argue that cleaner cities have set up the most onerous barriers to population growth.
There are two empirical problems here, but before getting into those, it is worth affirming the link between reform efforts related to exclusionary zoning and the environmental movement. Denser areas, with less suburban sprawl, which is created in large part by zoning, not only provide more economic opportunities for minorities and the poor, but they also reduce emissions by shortening commutes and facilitating public transportation. Reform efforts should seek the elimination of anti-density barriers, as part of an effort to reduce carbon emissions.
Now for the problems: Are people really flocking to “dirty” metros? Not any more than they are moving to clean metros. There is an insignificant positive correlation between emissions and housing supply growth at the metropolitan level. Using growth data from 1990 to 2000 for the 50 largest metros, I find that the relationship disappears completely if you adjust for initial population density, which is negatively correlated with both growth and emissions. A graph in Glaeser and Kahn’s paper also demonstrates the absence of a significant relationship between emissions and growth using data from 2000 to 2006.
Second, both Free Exchange and the two economists claim that stringent anti-growth zoning regulations are associated with low carbon emissions (though, to their credit, neither argues that anti-growth zoning causes low emissions). The trouble is that this conclusion depends on the measure of zoning, and the one used by Glaeser and Kahn does not adequately represent anti-growth regulations. They rely on an index from a survey conducted by Joseph Gyourko and his colleagues at Penn’s Wharton School. The Wharton Index has the noble but flawed ambition to be all-encompassing. As I’ve discussed elsewhere, this index is not correlated with housing supply, seemingly because it includes too much information, such as local political pressure, approval authority, the degree of local democracy, permit delays, development fees, open space criteria, state regulations, density regulations, and supply restrictions. Not all of these things are relevant to the basic question that a developer of a multi-family housing unit wants to know: Can I build or not? And at what density? Permit delays, for example, may speak to staffing capacity more so than an anti-growth disposition.
Fortunately, Gyourko and his team also have a density regulation index that can be isolated from the all-encompassing index used by Glaeser and Kahn. Adjusting for the effect of all other regulations (i.e. the Wharton Index), the correlation between the anti-density index and carbon emissions (from residential and transportation sources) is significant and positive in the top 50 metros. This is depicted below, and it is the opposite of Glaeser and Kahn’s claim. In fact, the relationship remains significant using just the two regulatory variables or including controls for pro-environmental politics (measured as the percentage of House members in each state voting for Waxman-Markey’s Cap-and-Trade bill), population density, July temperature, January temperature, and the presence of a coastal seaport (which should mitigate extreme temperatures).
Although Gyourko’s density index works, I prefer an alternative measure of density zoning from Rolf Pendall of Cornell. His survey technique made it more representative at the metropolitan level for the 50 most populated metros. The figure below shows that metros that allow dense multi-family housing in their municipalities have less residential- and transportation-related carbon pollution. Again, the relationship holds regardless of environmental politics, weather, coastal proximity, and overall population density.
In short, equitable housing policy is also good environmental policy not only because it allows people to move to clean places. Equitable housing policies actually make places cleaner. More work should be done, but there is compelling evidence that zoning policies do have a causal effect on emissions (the results obtained from using an “instrumental variable”--an econometric trick to assess causality--are consistent with this). The cap-and-trade bills being debated in Congress have some measures to promote better regional planning, but do not go far enough to discourage exclusionary zoning. Although the politics of local land use are intense, they could be overcome by allocating some share of allowance revenue to exclusionary jurisdictions who agree to abolish their anti-density laws.