Housing Market Interventions and Residential Mobility in the San Francisco Bay Area

The San Francisco Bay Area is an extreme case of a constrained housing market, with job growth outpacing new housing production and resulting in supply shortages and price spikes that date back at least 30 years. The Bay Area’s structural shortage of housing that is affordable at all income levels affects the regional economy by increasing commuting and housing costs, which creates barriers to full economic participation, especially for lower income workers. An array of solutions have been considered, including subsidized housing production, affordable housing preservation, and tenant protection programs. However, there is little evaluation research available to inform which housing solutions will be most effective in stabilizing communities so that those who wish to stay are able to, even in the midst of an influx of newcomers.

This study seeks to fill this gap by examining the impacts of market-rate development, subsidized development, and tenant protections, including rent stabilization and just cause for evictions protections, on movers. Specifically, this study builds two unique and cross-validated datasets on mobility and links them to a bespoke block-level housing construction database. We use granular data on individual and household mobility to assess how specific housing interventions impact both direct and indirect displacement by looking at moves both out of and into neighborhoods with different characteristics in the nine-county San Francisco Bay Area.

Our research reveals that new market-rate construction in a neighborhood results in a slight increase in people of all income levels moving in and moving out, i.e., churn. The increase in rates of displacement (involuntary moves) for very low- to moderate-socio-economic groups is not as high as commonly feared, at 0.5% to 2% above normal rates. However, the highest socio-economic group disproportionately benefits from new market-rate housing production—they are the least likely to move out and the most likely to move into neighborhoods with new construction. We also find that rent stabilization and just cause eviction protections help residents of the lowest socio-economic status remain in their neighborhoods. At the same time, these protections may have exclusionary impacts as we find that fewer low-income people move into neighborhoods with tenant protections. Together, these findings suggest that equitable solutions to the housing crisis will require more than just upzoning and tenant protections—these are complementary solutions, but not enough. Preserving unsubsidized affordable housing and substantially expanding social housing would help mitigate displacement and exclusion while addressing the housing affordability crisis through market rate housing production and tenant protections. Social housing is the provision of rental or homeownership units affordable at a moderate income or below, and is run by a public or nonprofit entity. To work, it would need to be widely implemented, requiring government investment at levels that match the urgency of the housing crisis.

What follows is a short summary of the main findings of the study. Practitioner-oriented findings on the implications of new production, subsidized development, and tenant protections will be presented in a series of forthcoming policy briefs by the authors.

A Note on Methods

We use individual and household mobility and the type of neighborhood moved to (similar or downward) as proxies for displacement, or forced moves, and assess exclusionary displacement by examining who moves into neighborhoods following specific interventions. To measure displacement, we track the movements of individual households by income and financial stability levels in and out of neighborhoods, using two different proprietary datasets on individual and household characteristics: Infogroup and the Federal Reserve Bank of New York Consumer Credit Panel/Equifax data. We categorize residents into five socio-economic status (SES) groups: extremely low, very low-low, moderate-middle, middle-high, and high.

Impacts of Market-Rate Production

About 20% of Bay Area renters live in a block group with new housing. We find that when new market-rate housing is built, there is a slight increase in both people moving out of the neighborhood and people moving in (churn) across most socio-economic groups.

Opposite Effects on Lower SES Outmigration and High-SES Moving Out Rates

In particular, we find that new market-rate housing production slightly increases outmigration from the neighborhood for people of lower SES, and slightly decreases moving out for high-SES people.

For SES groups ranging from very low through moderate, 100 new market-rate housing units are associated with marginal increases in outmigration, ranging from 0.5 to 2% above normal rates in the first year after construction, generally declining from the second to fourth year. In other words, in a typical year, about 13 of every 100 moderate-SES households moves out, but with 100 new market-rate units in the neighborhood, about 15 move out.

The highest socio-economic status residents are slightly less likely to move out when new housing is built. Where 15 high socio-economic status households would have moved out without the new construction, 14 move out when 100 new units are built (a decrease in moving out rate of less than 1%).

Residents of extremely low and middle socio-economic status experience little change in moving out of their neighborhood.

Our model suggests that if 1,000 new units are built instead of 100, the impacts on inmigration and outmigration for all income groups would increase only very slightly; very low- to moderate-income groups would still experience increases in outmigration and inmigration of 1-2% in each subsequent year for four years when new market-rate construction occurs in their block group.

Increase in Inmigration

We find that that new market-rate housing production increases migration into neighborhoods throughout the Bay Area, for all socio-economic groups, though higher socioeconomic groups are more likely to move in compared to others.

For extremely low-SES households, inmigration is approximately one percentage point higher with the new construction of 100 market-rate units, so that 13 households would move in for every 12 that would move in without the construction.

Varying Effects on Gentrifying Neighborhoods

In the Bay Area’s gentrifying neighborhoods (neighborhoods with fast-rising incomes and an influx of high-income or highly-educated residents), new market-rate housing construction neither worsens nor eases rates of moving out, according to our study. It increases rates of people moving in across all socio-economic groups, particularly high-socio-economic residents.

Rates of people moving out remain the same in gentrifying areas for four years after construction of 100 units, with the exception of the highest socio-economic status residents; they are much more likely to move out (increasing from 22% to 31%). Middle socio-economic status residents are slightly less likely to move out.

Rates of people moving into a gentrifying area after new construction increases at first for all socio-economic groups, but by four years later goes back to normal for all groups except high-socio-economic status folks, who continue to move in at higher rates.

The slight impacts occur despite controlling for previous churn patterns, i.e., controlling for outmigration and inmigration rates in previous years.

Slight Increase in Constrained Moves

We find that market-rate housing construction is associated with a slightly higher chance of making a downward move—a move to a lower-opportunity neighborhood—for all socio-economic groups.

For every 100 new market-rate units built, residents who move out are slightly more likely (0.2 to 0.62 percentage points across SES groups) to move to neighborhoods with a lower median income or higher poverty rate (i.e., a “constrained move”).

Impact of Subsidized Development

There is too little subsidized housing construction to identify substantial impacts on displacement. Subsidized housing is effective at encouraging low-socio-economic status residents to move into neighborhoods, but effects do not last long after units are built.

Our findings on the insignificant impacts of subsidized housing construction on displacement are likely due to the lack of affordable units available for analysis. In addition, new subsidized housing units may be reserved for residents from outside the neighborhood, thus failing to mitigate local displacement effects.

Impacts of Tenant Protections

Rent Stabilization

Our analysis suggests that rent stabilization helps some—the lowest socio-economic status residents—to stay in a neighborhood. At the same time, it may be exclusionary as it discourages moving in for all socio-economic groups except moderate-middle socio-economic status.

• Rent stabilization decreases the probability of moving out for the lowest socio-economic status residents in our sample by about 1 percentage point.
• However, in gentrifying areas, rent stabilization has no significant effects on keeping very low socio-economic status residents in place.
• With an increase in units covered by rent stabilization in a neighborhood, all residents except those of moderate-middle socio-economic status experience declines in moving in.

These exclusionary impacts of rent stabilization are likely due to the lower numbers of available units, as tenant protections effectively reduce move-out rates by allowing renters to stay.

Just Cause Protections

Just cause protections help to keep the lowest socioeconomic status residents in place in gentrifying neighborhoods, where displacement pressures may be especially strong, according to our findings. However, these policies of course are limited to protecting existing residents and do not encourage residents to move in.

As more units are covered by tenant protections, we find that lower socio-economic status residents are not much more or less likely to move into neighborhoods, and high socio-economic status residents are less likely to move in.

We find that renters make fewer downward moves from tracts where more units are covered by just cause protections—suggesting people are able to make planned moves.

When there are more units covered by tenant protections (just cause for evictions and/or rent stabilization) in a census tract, renters in all income groups who move away from that tract are more likely to stay within the same city, less likely to move elsewhere in the Bay Area, and more likely to move outside of California. In other words, tenant protections shift migration patterns in two opposing ways, facilitating moves either nearby to their original homes or out of the state entirely.

Overall Impact of Interventions on Residential Mobility

Overall, new market-rate housing units are associated with a slight increase in both outmigration and inmigration from neighborhoods, i.e., more churn, for different income groups. Over the short-term, the net impact is minimal, suggesting a level of impact that is mitigable through policies to promote housing affordability. Tenant protections will also help to mitigate displacement but may increase exclusion (by decreasing inmigration to neighborhoods) for some low-socio-economic groups without being coupled with other tools.

Acquiring multi-unit rental properties that are at risk of becoming unaffordable via a program like San Francisco’s Small Sites Acquisition and Rehab Program may be an effective strategy. Other potential approaches include tenant or community opportunity to purchase policies such as San Francisco’s, transfer tax breaks for building owners when selling to a nonprofit or community land trust condominium conversion restrictions, and community land trusts. Considerations to address the housing affordability crisis and mitigate displacement and exclusion include the preservation of unsubsidized affordable housing, as well as initiatives that substantially expand social housing.

Disclaimer: The views expressed in this report are those of the authors and do not necessarily reflect the views of the Federal Reserve Bank of San Francisco or the Federal Reserve System.