How much can new housing contribute to state climate action?

Author: Zack Subin

Because solutions to the climate crisis are both urgent and unprecedented in scale, climate policy researchers routinely evaluate ambitious scenarios for climate action—well before figuring out how to make them practical. For example, over a decade ago researchers started asking how close we could get to long-term climate pollution reduction targets if we phased out gasoline vehicles, despite electric vehicles being far more expensive than today.

However, climate policy researchers have rarely considered ambitious housing policies. To help fill that gap, my co-authors and I recently published an article in Buildings and Cities.¹ In the spirit of previous climate policy studies, we estimated how much climate pollution would be reduced if each state solved its housing shortage in ten years and built this housing in compact, walkable neighborhoods.

While the paper presents data for all 50 states,²  here we focus on the findings for California. According to Up for Growth, California had an existing housing shortage of at least 880,000 homes in 2021 (6% of existing housing). To rectify this shortage and anticipate future statewide household growth by 2033,³ the state would need to build an estimated 2 million homes.

Our scenario combined this estimate with Replica’s information about neighborhood travel patterns, which Terner Center previously illustrated for the Bay Area. We then identified neighborhoods with relatively low vehicle miles traveled (VMT), defined as those near the 90th percentile most walkable (and least car-oriented) places.⁴ In the fall of 2022, people living in these neighborhoods drove 40 percent less than the state average. (Examples of cities with similarly low rates include El Cerrito and Culver City.) These neighborhoods also have room to add more housing: on average, 43 percent of the homes are single family buildings, and 75 percent are in buildings no bigger than a fourplex. These neighborhoods could be ideal places to build ADUs and other missing middle building typologies.

Our scenario assumed we could build 1.8 million of the needed 2 million new homes as infill in these relatively walkable neighborhoods, achieving 40 percent lower VMT per capita than the state average.⁵ This would result in an overall 4.5 percent reduction statewide—including the travel patterns of both residents of the new housing and current residents.⁶

This scenario would directly prevent 3.4 million tons of carbon pollution in 2033 due to less driving.⁷ Including indirect and lifecycle pollution savings—for example, avoided oil extraction—roughly 8 million tons could be eliminated.⁸  Though many states with greater anticipated growth rank above California on a per capita basis, this total represents the second highest potential carbon reduction in the scenario across all states, behind Texas.

California needs to reduce carbon pollution from all sources by 114 million tons to comply with Senate Bill 32 by 2030.⁹  At the same time, the Statewide Housing Plan calls for building 2.5 million new homes in about eight years—an even more ambitious target than the one we modeled. Our analysis shows the potential of directing the state’s housing production to infill locations. Scaling up the published scenario results to match this plan’s pace would result in a 5.6 percent reduction in VMT per capita by 2030. This would make a meaningful contribution to the very ambitious state climate plan target of reducing VMT per capita 25 percent by 2030 (below 2019 levels). Similarly, scaling up the carbon reductions contemplated in the scenario to match the Statewide Housing Plan would result in 4.3 (direct) to 10 (total) million tons avoided in 2030.

Idealized scenario investigations like the one described here can establish the scale of impact that housing reforms could have. From this study, we concluded that successful policy reforms enabling rapid construction of climate-friendly housing could meaningfully contribute to climate policy targets—at a national scale, the contribution would be comparable to strengthening electric vehicle policies.¹⁰ That said, even ambitious land use and housing production policies will not get us all the way to California’s targets for reducing transportation pollution. Additional policies will certainly be needed, such as limiting highway expansion and investing in public and active transportation instead.

Ongoing housing policy research remains necessary to achieve the scenario’s ambitious climate outcomes and ensure that new homes also advance affordability and fair housing goals. But papers such as this one lay the groundwork for understanding how removing obstacles to infill housing production could have substantial climate policy benefits.

Endnotes

¹ Subin, Z. M., et. al. (2024). “US urban land-use reform: a strategy for energy sufficiency.” Buildings and Cities, 5(1), 400–417. https://doi.org/10.5334/bc.434.

² We also included the District of Columbia, but we excluded Alaska due to data limitations.

³ This was modeled for each state based on Berrill and Hertwich (2021) population projections. See our article for details.

⁴ In other words, consider ranking every neighborhood in the state by how much people drive on a typical day. The 90th percentile neighborhood is the one where people drive less than in 90 percent of other neighborhoods. The average person in this neighborhood drove 40 percent less than the average person in the state. Note, this is distinct from the top 10 percent most walkable neighborhoods, where people would drive even less—we selected relatively walkable but not best-in-class examples to inform our scenario.

⁵ Technically, the modeling is indifferent to whether the state builds these new homes as infill in these existing car-light neighborhoods, or builds them in other locations while matching their environmental characteristics. We assumed 1.8 of the 2 million homes (90%) would be built to match the lower VMT target, as no policy can be 100 percent perfect, though this still represents a very optimistic scenario.

⁶ We assumed that residents of existing housing did not change their VMT as new housing was added.

⁷ Million metric tons CO2-equivalent.

⁸ These include emissions occurring outside the state, not typically counted to the state’s climate targets under Assembly Bill 32.

⁹ Relative to 2022 emissions. See https://ww2.arb.ca.gov/ghg-inventory-data and https://ww2.arb.ca.gov/ghg-2020-limit.

¹⁰ Nationwide, the modeled scenario would achieve 65 percent of the carbon pollution reduction as would extending California’s zero emissions vehicle policy to all states. For further details, please see our article.