The pro-housing movement has grown rapidly over the past several years, winning legislative victories from Minneapolis to Raleigh, N.C. to Maine to Utah. The movement’s goals are to increase the production of moderately priced homes and create more diverse housing options. Policy levers to achieve these goals include zoning reforms that legalize duplexes, rowhouses, and apartments as well as procedural changes to make housing development simpler, shorter, and more transparent.
The recent legislative victories offer a unique opportunity for urban scholars across a range of disciplines to evaluate the outcomes and impacts of pro-housing policies and the political strategies that led to change. Academic researchers typically conduct backward-looking evaluations of programs that have been in effect for some time. However, the pro-housing movement is changing rapidly, with many local and state policymakers looking to replicate reforms from early adopters. This creates an urgent need for real-time data and analysis to inform current policy choices.
The overarching questions facing researchers, policymakers, and housing advocates are:
- What package of housing policy changes will be most effective at increasing the supply, diversity, and affordability of homes?
- What set of political strategies has the greatest likelihood of achieving better housing policies?
To begin answering those questions, we can break the larger agenda into smaller, discrete pieces that can be addressed with careful empirical research, as summarized in Figure 1.
Two further pieces of context are important to remember when designing empirical research on these questions. First, while the policy design questions are posed separately, policymakers adopt an entire package of housing policies, so policymakers and researchers should think carefully about the interactions between the components. For instance, legalizing apartments while capping building height at two stories is likely to make building apartments more challenging.
Second, the “right” policy design choices and political strategies are highly context-specific. A given package of housing policies or regulations will have different impacts on housing outcomes (number, size, and price of new housing built) depending on the original market conditions. Similarly, how to motivate elected officials to enact housing policy changes will depend on the partisan and ideological preferences of local/state voters, the salience of the issue, and the relative power of existing political stakeholders and constituencies. Empirical studies that do not adequately account for local and/or state housing market conditions and political contexts are likely to produce inaccurate or misleading results.
The remainder of this piece lays out a framework for empirical research to evaluate recent housing policy changes and identifies opportunities and challenges, including gaps in available data sources. It focuses primarily on the policy design side; similar analysis of political strategy questions would be helpful.
A framework for evaluating recent housing policy changes
Over the past five years, more than a dozen U.S. cities and states have passed substantial changes to their housing and land use policies, each offering an opportunity for investigation. Descriptive analyses that document key policy design issues and first-order housing outcomes are needed to provide feedback on what works and what doesn’t—hopefully in time to inform ongoing legislative action. Researchers should pay careful attention to four essential—and potentially tricky—components of research design, described below.
Clearly describe policy design and targeted housing outcomes
Specifying the outcomes of each policy change—including policies with multiple components—is critical to an accurate evaluation. Minneapolis, Oregon, and California were among the first places to pass laws that legalize specific structure types (such as ADUs, duplexes, and triplexes) across the entire city or state; measuring the impact of these laws should start by counting the production of newly legalized structures. By contrast, Massachusetts’ new “MBTA communities” law is intended to encourage multifamily buildings near transit stations in a specified set of cities and towns.
A handful of cities, including Cambridge, Mass. and Culver City, Calif., have reduced or eliminated minimum parking requirements for newly built housing. Other jurisdictions, such as Houston, have reduced minimum lot sizes. As discussed below, some outcomes will be more challenging to measure with publicly available data; enumerating which outcomes are of most interest is helpful to guide data collection.
Traditional program evaluation often tries to evaluate the impact of a single policy change, but several recent housing reforms have included multiple different policy components (either within one bill or passed in the same legislative session). For example, Minneapolis increased the allowable density around transit corridors at the same time it legalized duplexes and triplexes, although the former received less attention. Oregon passed a statewide rent regulation law in the same session as the bill that legalized duplexes and triplexes, which may interact with landlords’ incentives to expand rental housing. In one sense, these multilayered policy changes complicate researchers’ task; in another sense, what matters for policymakers is the collective impact of the full set of policies.
Lay out a timeline for policy implementation and market responses
Land use is famously sticky; even a complete overhaul of zoning codes will not replace existing buildings overnight. Researchers need to be clear about the appropriate time frame in which policies might become visible to observers. The policy implementation process itself may be lengthy: Massachusetts Governor Charlie Baker signed the MBTA communities bill in January 2021, the state agency charged with overseeing its implementation published their draft regulations in January 2022, and localities have until 2023 or 2024 to bring their zoning codes into compliance with the new state guidelines. The housing development process is likewise slow and occurs in multiple stages: the landowner or developer must apply for building permits and receive approval (often from multiple different agencies) all before shovels hit the ground.
As such, the timeline for observable impacts is likely to vary across specific policy types and market circumstances. In some cases, such as the Massachusetts example, it may take years before substantial amounts of newly legalized homes are completed; this would still constitute a “successful” policy change. Other policies may produce rapid responses through unexpected channels, as how California’s ADU legislation prompted owners of existing but informal ADUs to seek certification. And some types of policy changes could induce developers to act in advance of a new policy taking effect, such as a rush to apply for building permits before a mandatory inclusionary zoning law begins.
Consider how policy design interacts with underlying housing market conditions
What types of land use policies prohibit or impede new construction—and whether changing those policies in certain ways will effectively allow more housing—is a complicated question. Intuitively, when land is expensive and zoning rules allow it, developers will choose to use less land per housing unit, such as by using smaller lot sizes or stacking homes vertically. Material and labor costs of construction typically increase with building height (especially over six stories, which requires switching from wood to steel and concrete and adding elevators), so developers generally choose to build low-rise, lower-density structures in places where land is abundant and inexpensive. Zoning rules that cap building heights at, say, 10 stories will be a binding constraint in Manhattan or downtown San Francisco, but probably not in most suburban areas. Many post-World War II residential subdivisions allowed single-family homes on lots as small as 5,000 square feet (roughly one-tenth of an acre). But today, affluent suburbs such as Greenwich, Conn. and Atherton, Calif. require minimum lot sizes of one or two acres in some districts—substantially increasing the land costs associated with each home and limiting overall housing construction. In those communities, legalizing rowhouses or small-lot single-family detached homes would allow for substantially more housing than current rules—even before discussing multifamily buildings.
State legislation that places guardrails on local zoning to set a common baseline across all cities and counties offers an excellent opportunity for researchers to understand how policy changes interact with underlying market conditions. For instance, under Oregon’s new law, duplexes and triplexes are legal to build in nearly every community. But the additional density should be more economically valuable in areas where land values and rents are high, such as Portland’s affluent neighborhoods, compared to rural communities in eastern Oregon. Looking at differences in housing production across localities before and after statewide policy changes is one of the more promising areas for future research and may help illuminate which types of regulations were binding constraints in different types of housing markets.
One challenge in trying to compare results of ostensibly similar policy changes across states is how newly adopted policies will interact with long-standing policies. For instance, California’s legalization of ADUs and duplexes will be implemented in housing markets that are substantially distorted by the state’s environmental review (the California Environmental Quality Act) and property tax limitations (Proposition 13). This makes it especially difficult to extrapolate California’s experience to other states—while each state is unique, California has by far the most complex housing policy environment.
Collect better data to track outcomes at the right time frames and geographies
Publicly available data sources are woefully inadequate to evaluate most recent housing policy changes. For example, one of the most popular zoning reforms has been legalizing ADUs, which can take multiple forms (basement apartments, garages, or free-standing structures). But ADUs are not consistently identified by the Census Bureau’s American Community Survey, residential building permits survey, or the American Housing Survey. So how can researchers and policymakers track changes in ADU production? Researchers in California have created a website to map ADU creation, but that requires a dedicated investment of staff and resources to assemble and maintain. Most local and state legislatures that passed zoning reforms have not allocated funds for monitoring and data collection.
Public data has even larger gaps when it comes to evaluating changes in dimensional requirements. For example, to understand the impact of eliminating minimum parking requirements, researchers would need to know the number of newly built homes with and without off-street parking spaces before and after the new policy went into effect. No publicly available survey collects these metrics for individual properties; even property records held by local tax assessors do not consistently indicate the presence or type of parking spaces. For some research questions, local administrative data and proprietary data can help fill gaps, but these sources are expensive and/or labor-intensive to work with.
Tackling an ambitious research agenda requires a small army of interdisciplinary researchers
Understanding how state and local policy changes impact the availability and affordability of housing is the most important question facing the housing field today. The current burst of policy experimentation offers unique opportunities for qualitative and quantitative research across a range of academic disciplines, including economics, planning, policy, and political science. A few guidelines can help make the resulting research more usable for policymakers and advocates:
- Build on local institutional knowledge. The complex, hyperlocal nature of housing and land use policies—and the interactions between policies and market conditions—make it particularly important for researchers to have a solid understanding of local institutions and data sources. It may be more efficient to have multiple research teams with local expertise focusing in-depth on small geographic areas (cities, metro areas, or states) and sharing results with one another to draw broader lessons.
- Look at the full range of places undertaking policy reforms. The earliest or best-known policy reforms will likely draw more attention, but it is important to understand how policy changes play out across a range of housing market conditions as well as political and institutional settings. If, five years from now, there are dozens of studies of California and Minneapolis but none of Raleigh, N.C. and Montana, that will be a missed opportunity. A particularly important set of places to study are metro areas where housing has historically been abundant and relatively affordable, but are quickly becoming more expensive, such as Austin, Texas, Denver, Nashville, Tenn., and Boise, Idaho. Some early coordination across research teams to discuss geographic focus would be useful.
- Learn from both successes and failures. Pro-housing advocates and policymakers may be tempted to focus on success stories. While developing a set of potentially replicable “best practices” should be one goal of research, we can also learn important lessons from policy changes and political strategies that don’t work exactly as planned: What features of that policy design or advocacy tactics need to be tweaked or altered to be effective in boosting housing supply? Zoning reforms to encourage abundant housing are still very new; there isn’t a template for how state and local policymakers should design and implement these programs, and it’s unlikely that all reforms will get everything right on the first iteration. Being honest about what works and what doesn’t is the best way to learn.
The pro-housing movement has achieved remarkable legislative victories in the past five years, creating unique opportunities to study the impacts of housing policy changes. Careful empirical research can help policymakers and advocates design more effective policies—thus making housing more affordable and abundant for all Americans.
Thanks to Alex Baca and Henry Honorof for providing thoughtful comments.