Monday, April 26, 2010

Urban Policy Effects on Carbon Mitigation

With the UC furloughs, I have been thinking about retiring but as the year continued, I find myself mildly interested in economics again and have now written a new paper titled; "Urban Policy Effects on Carbon Mitigation".

Here is the paper's first section. I will present this at this NBER Conference next month.


Suppose that your household was choosing between living in suburban Houston or center city San Francisco. In each case, what would your household’s annual carbon footprint be? Glaeser and Kahn (2010) estimate that a standardized household would create 12.5 extra tons of carbon dioxide per year if it moved to Houston rather than moving to San Francisco. In Houston, the same household drives more, lives in a bigger home, uses more residential electricity – electricity that is generated by power plants with a higher emissions factor. Using data from 2006 for 74 major Chinese cities, Zheng, Wang, Glaeser and Kahn (2010) document that northern cities have the largest household carbon footprints due to coal burning for winter heat. This cross-sectional descriptive work creates a benchmark for comparing cities’ household carbon emissions from transportation, electricity consumption and home heating, at a point in time and tracking city trends over time. In both studies, there is clear evidence that cities differ sharply with respect to their “greenness” on this important dimension. Given that greenhouse gas emissions are a global externality, households are unlikely to internalize this cost of moving to a city like Houston when they make their locational decisions.

Why is San Francisco “greener” than Houston? San Francisco is blessed with a temperate climate. Northern California’s electric utilities emit less greenhouse gas emissions per unit of power generated than their Texas counterparts. In addition to these factors, San Francisco’s urban form and public transit system encourage more households to live a walking, compact, “new urbanist” life relative to households living in sprawling Houston.

This discussion highlights that there are a number of urban policies that can affect a metropolitan area’s per-capita carbon emissions. Rather than attempt to tease out the individual contributions of any one policy, I focus on estimating how much does a household’s carbon footprint shrink by when it lives closer to the city center. Throughout this paper, I assume that urban areas can enact a range of policies including urban transit investments, to business incentives (such as urging major employers to remain downtown) and center city quality of life improvements. Such policies increase the demand for living and working in the center city. I seek to measure the carbon mitigation externality benefits of having a more vibrant center city that attracts households to live closer to the center.

To quantify the greenhouse gas emissions reductions benefits from living at higher population density, and closer to the city center, this paper uses recent micro data from the 2009 National Household Transportation Survey and unique 2008 data from a California electric utility to measure how a household’s carbon footprint varies as a function of how close it lives to the city center, and proximity to rail transit. This paper’s main finding is that a standardized household drives less and consumes less electricity when it lives closer to the city center than if it lived in the same metropolitan area’s suburbs.

Is this a causal effect? A reasonable concern is residential self-selection; those with an unobserved taste for living a “low carbon” life cluster close to rail transit stops and close to city centers. My cross-sectional OLS estimates are likely to reflect a mixture of selection and treatment effects and thus to represent an upper bound on the policy induced benefits of moving a household chosen at random closer to the city center.