Date of Award

Spring 4-1-2021

Document Type


Degree Name

Doctor of Philosophy (PhD)


Forestry and Environmental Studies

First Advisor

Gillingham, Kenneth


Buildings play a key role in determining global demand for energy and materials, and therefore have a major influence on anthropogenic greenhouse gas (GHG) emissions. This dissertation examines recent trends and future trajectories of energy demand and GHG emissions from residential buildings in the United States. Around 10% of residential GHG emissions in the United States come from material production and construction activities, a share that will increase as energy supply decarbonizes. Decarbonization of electricity generation has been the primary source of residential sector emission reductions since 2005. Demand side energy efficiency improvements have helped to reduce emissions, but compared to transformations in energy supply, their contribution has been minor. In recent decades, reductions in residential energy demand and emissions from energy efficiency and decarbonization were offset by population growth and increases in floor area per person. The types of structures that provide housing services are an important determinant of energy demand and emissions. Both federal and local policies limit supply of multifamily and small-lot single-family structures. An analysis of changes in federal housing policies and their effects on housing construction of different types found that several changes in federal housing policies in the 1970s and 1980s reduced new construction of multifamily housing by 14 million units up to 2015. A separate analysis found that single-family houses consume substantially more energy than multifamily, and that older single family houses consume much more energy for space heating and cooling, while newer single-family houses consume more energy for non-thermal end-uses such as lighting and appliances. Combining estimates of policy effects on housing construction by type, and the effects of house type on residential energy use, overall influences of policy changes on residential energy and GHG emissions in 2015 were estimated. Without the federal policy changes, total urban residential energy demand in 2015 would have been 4.6-8.3% lower, or 27-47% per affected household. Removing policy barriers and disincentives to supply of multifamily housing has large potential for reducing future energy and emissions. Informed by these findings, a highly spatially disaggregated housing stock model was developed and applied to all counties in the United States to analyze scenarios with varying rates of stock turnover, rates of home energy renovation, shares of multifamily housing, and floor area distributions of new housing. An important novelty in this model is the identification of a natural vacancy rate specific to house type and region, and incorporation of a local flexible vacancy rate in housing stock projections. Representation of flexible vacancy rates is particularly important for projecting housing stock growth in regions with low or negative population growth. The housing stock scenarios compare GHG reductions from increased renovation vs increased rates of housing stock turnover, and find that increasing the turnover rate would increase overall residential sector emissions. Energy consumption per floor area is much lower in new buildings due to higher energy efficiency codes and standards, but new single-family houses are generally much larger, which limits the energy reductions from stock turnover. This outcome is very sensitive to the floor area characteristics of housing built in future decades, which in most scenarios are assumed to be unchanged from housing built in the 2010s. Removing houses larger that 3,000 ft2 (279 m2) from new housing supply results in a 25% reduction in the average size of new single-family houses, and produces a 4% reduction in cumulative 2020-2060 emissions. Increasing the share of multifamily housing in new construction, or increasing the rate and depth of energy renovations to existing housing, would produce emission reductions of similar magnitude. Different decarbonization strategies will be optimal in different regions. Regions with older housing are less likely to experience strong future growth, and will rely to a greater extent to renovations of existing housing. Regions with high projected stock growth have a much greater potential to reduce future sectoral emissions by building smaller and more multifamily housing. The single factor with the biggest influence on future residential emissions is the rate at which electricity decarbonizes. A more ambitious combination of efficiency and sufficiency measures in housing supply and renovation, combined with greater decarbonization of electricity supply and material production, is required to achieve sectoral targets consistent with the Paris Agreement or limiting climate change to 1.5-2°C.