Real Estate in Transition: Housing, Welfare, and the Decarbonization of Cities
Abstract: Real estate accounts for the largest share of urban carbon emissions, and major cities have begun to penalize carbon-intensive buildings. In the large multifamily segment, where ownership and occupancy are generally separate, how much these penalties reduce emissions and who bears their cost depend on long-run housing market adjustments that cannot yet be observed. I build a dynamic general equilibrium model of a frictional housing market to measure both. Financially constrained landlords make lumpy, building-specific abatement investments and capture only part of the energy savings, while households choose how much to save and where to live. The model is calibrated to newly linked building-level data on emissions, energy audits, rents and operating costs, and renter sorting from New York City. I then quantify the impact of the City's Local Law 97, one of the first carbon prices on existing buildings. Because landlords abate only where investment is cheaper than the penalty, emissions from the buildings it covers fall far short of the target, and the law itself accounts for only about a third of the reduction. The penalty lowers the value of covered buildings, supply responds, and the cost of renting rises as much outside the law's reach as inside it. The incidence is regressive because poorer households are more likely to be tenants and homeownership insulates against the higher cost of renting. Of the uses of the penalty revenue, returning it only to low-income households minimizes welfare losses. A cap-and-trade program with less stringent limits achieves most of the law's cut at a lower price and a much smaller cost to households.
Asset Reallocation Under Local Decarbonization Policy with Siqi Zheng
The Costs and Benefits of Stringency: How Building Codes Shape Housing Supply, Affordability, and Resilience with Siqi Zheng
Converting Brown Offices to Green Apartments with Arpit Gupta and Stijn Van Nieuwerburgh
Abstract: The conversion of brown office buildings to green apartments can contribute towards a solution to three pressing issues: oversupply of office in a hybrid-and-remote-work world, shortage of housing, and excessive greenhouse gas emissions. We propose a set of criteria to identify commercial office properties that are are physically suitable for conversion, yielding about 9% of all office buildings across the U.S. We present a pro-forma real estate model that identifies parameters under which these conversions are financially viable. We highlight several policy levers available to federal, state, and local governments that could accelerate the conversion, and that may be necessary should policymakers desire the creation of affordable housing. We highlight the role that the Inflation Reduction Act could play.
The Hamilton Project (Policy version)
A Harmonized National Work-Type Classification of 150 Million U.S. Building Permit Records with Sean Hardesty Lewis, Yichun Fan, and Siqi Zheng
Abstract: Building permits are the most complete record of physical change to the U.S. built environment, but they cannot be compared across jurisdictions: local authorities classify work for administrative purposes, in vocabularies that differ by orders of magnitude in granularity and share no common definitions. The national statistic compiled from them, the Census Bureau’s Building Permits Survey, counts only newly authorized housing units, leaving demolition and all work on the existing stock unmeasured. PermitKey contributes at two levels. The first is a general procedure that treats jurisdiction-supplied labels as weak supervision, trains a multi-label classifier over permit descriptions against a taxonomy chosen by the analyst, and validates the result against both municipal labels and blind human review. The second is a national implementation: 150,044,555 records from 90 U.S. jurisdictions’ open data portals, classified into 25 work-type categories and released as permit counts for U.S. census tracts by year, together with the category definitions, validation tables and classification code, so that other researchers can apply the same procedure to taxonomies of their own.