A New Approach to the Housing Element Update

Paavo Monkkonen, Christopher Elmendorf, Moira O’Neill, and Eric Biber
January 2020

By law, every city and county in California must update the Housing Element (HE) of their General Plan every eight years. This requirement is not new — mandatory updates are a longstanding part of California’s statewide housing planning process. But the current planning cycle, which for most cities starts in 2021 or 2022, is different from past cycles for several reasons. Recent amendments to the state’s Housing Element Law led to an increase in the number of housing units local governments must plan for. Where in the past these targets could largely stay on paper, the new mandates are framed more like goals for actual production.1

In addition, local governments face new pressure to affirmatively further fair housing because of AB 686, a 2018 amendment to California law that pushes cities to site low-income housing in high opportunity neighborhoods. At the same time, they face new restrictions on the kinds of parcels that can be used as potential sites for low-income housing because of AB 1397, a 2017 amendment to California law that created new restrictions on which parcels local governments can use to satisfy their low-income RHNA targets. Together these new developments make the current HE process qualitatively different and arguably more challenging than in years past. Fortunately, the state Department of Housing and Community Development has new, enhanced authority over the contents of the HE.2

The four main components of the HE update are an assessment of local housing needs, an inventory of available sites for housing development (to cover the local government’s assigned housing targets), an analysis of constraints (governmental and non-governmental) to housing development, and a proposed set of programs to reduce these constraints and, if necessary, make additional sites available.

In this issue brief, we propose a revised approach to how cities and counties demonstrate their capacity to meet housing targets, expanding on the inventory of vacant and underutilized sites section of the element. This new approach offers a way to resolve a bind some local governments face in complying with new laws.

California’s Regional Housing Needs Allocation (RHNA) process assigns local governments a number of units of housing need for households of different incomes. The local governments rarely meet this number. In part, this is because governments have been able to satisfy their RHNA obligations simply by zoning for new housing, not by actually seeing it built (i.e., cities have only needed to demonstrate that the housing units could theoretically be built within their jurisdiction), and in part because cities use vacant and underused sites to count toward their zoned capacity, without considering either the probability of those sites being developed, or whether developing those sites advances the HE process’s equity goals.

The heart of our new approach, therefore, is a shift away from this traditional approach of identifying vacant and underutilized parcels and hoping for the best. We propose instead a realistic and practical approach to actually building housing, in keeping with the expectations of AB 1397 (2017) that the HE sites have “realistic and demonstrated potential for redevelopment during the planning period,” not just some theoretical possibility of development or redevelopment by some unspecified future date We propose cities and counties engage in strategic rezoning to make production of the RHNA target likely to occur during the planning period, use data to estimate the development potential of all parcels, and take a proactive approach to get affordable housing built.

The Current Approach to an Inventory of Vacant Sites: Inadequate to the Task

The RHNA process assigns local governments housing unit targets for households of different income levels. Traditionally, cities and counties meet these housing targets by inventorying their vacant and underused sites. Essentially, if a local government can identify vacant sites with enough zoned capacity3 for housing units above their RHNA numbers, then they are compliant with the law. If they cannot identify sufficient vacant sites, however, they need to assess the housing unit capacity of underutilized sites. If they can identify underutilized sites where redevelopment is “realistic” — a vacant shopping center for example — with potential for additional units above their RHNA targets, they stop the analysis and are compliant. If they cannot identify sufficient vacant or underutilized sites, then they must rezone some land that is currently built out, to allow more development.4

This approach has at least three flaws. First and foremost, it counts similarly zoned sites as accommodating the same number of units — specifically, the number of units likely to be built on the site if it is developed — even if the sites have very different probabilities of being developed during the planning period. (The sites’ development probabilities may diverge owing to variation in existing uses, location, environmental conditions, and other factors that make development or redevelopment profitable.) So long as a site is vacant or has some “realistic” potential for redevelopment, it is assumed that the site will be built out during the next eight years. To our knowledge, no local governments’ vacant sites have ever been completely built out during an eight-year planning period. Cities and counties with a zoned capacity somewhere near their RHNA number are highly unlikely to actually produce the requisite number of housing units.

Second, the current approach does nothing to explicitly advance the fair housing goals of the HE law (ie., reducing social segregation) or to advance other state policy priorities, such as mitigating the negative environmental impacts of Californians by putting higher-density housing near transit and jobs. In fact, relegating new affordable housing to whatever parcels happen to be vacant in a city likely exacerbates segregation by income. In most cases, vacant parcels — the last to be built on — are in the least affluent neighborhoods of an urban area.5

Finally, the reliance on vacant sites is a flawed approach because they are likely to be vacant for a reason. Developers build on the more desirable locations first, meaning remaining sites are more likely to be poorly located or have a patient owner holding out for a high price. This idea is reinforced by evidence that the majority of California’s recent housing development has been on parcels not listed in local governments’ HE sites inventory.6

A New Approach to Meeting Housing Production Goals

For local governments to meet their housing unit targets over the next eight years, we propose they undertake a new approach to the HE update with three parts: strategic rezoning, an evidence-based evaluation of development potential during the planning period, and a proactive approach to the production of affordable housing. In addition to assisting cities and counties in building needed housing, this combination of strategies provides a potential path to affirmatively further fair housing as required under AB 686 (2018) while remaining compliant with new restrictions on inventory sites created under AB 1397 (2017). The combination of these two new laws creates a challenge because parcels in high opportunity neighborhoods tend to have uses that are not “likely to be discontinued.” We discuss the three parts of our approach below.

Strategic Rezoning in High Opportunity Neighborhoods

We propose that cities and counties increase zoned capacity for new housing strategically — in neighborhoods near transit, near major amenities, and where the housing is more likely to actually be built (i.e., neighborhoods in high demand as reflected by high rents and prices). Similar to the allocation of regional housing needs to local governments,7 which California law mandates be based on objective criteria such as proximity to transit, jobs, and in high opportunity neighborhoods, local governments ought to evaluate which neighborhoods have high scores on an opportunity index, for example, and increase zoned capacity in them through rezoning. This can be achieved through gentle density and marginal changes to zoning — allowing three-story six-plex buildings in single-family neighborhoods, for example. It does not mean high-rise towers everywhere or even anywhere other than regional cores like downtown Los Angeles.

This approach increases the effectiveness of the HE in two ways. First, strategic rezoning in high opportunity neighborhoods will allow cities and counties to be compliant with the law under AB 686.8 AB 686 is a 2018 amendment to California law that requires that program actions for HEs due to be revised on or after Jan. 1, 2021, affirmatively further fair housing. Strategic rezoning does exactly that, by allowing low-income housing to be built in high opportunity neighborhoods.

Second, this approach will likely lead to more housing actually being built as a result of the HE update than reliance on vacant and underutilized sites. Recent research shows that zoning capacity has a much bigger housing-production payoff in high-price locations.9 Using permitting data from 2013–2017 and zoned capacity numbers from cities’ fifth cycle HEs, Monkkonen and colleagues estimate the impact of increasing zoned capacity (from the 25th to 75th percentile) for two cities, one with an average rent of $900 (the 10th percentile in the state in 2013) and one with an average rent of $2,100 (the 90th percentile). In the low-rent city, permitting is estimated to increase from 110 to 165 units over four years, but in the high-rent city, the change is from 240 to 740 units. The same logic applies to neighborhoods — developers are more likely to take up opportunities for redevelopment in places that command higher rents.

The strategic rezoning approach faces two challenges (beyond the local politics of rezoning) because of AB 1397 (2017). First, AB 1397 requires site-specific “likely to be discontinued” findings with respect to existing uses if the local government assigns more than 50 percent of low-income RHNA share to non-vacant sites. The idea being that cities and counties must demonstrate that the current uses of a site are likely to end, increasing the likelihood of redevelopment. Second, AB 1397 requires the site inventory to account for “existing leases” on non-vacant parcels. This rule is particularly challenging simply because of the impracticality of obtaining information on the leases for each parcel in large cities and counties.

Our approach to the HE update provides a potential fix for the implementation of AB 1397 and AB 686 through the use of evidence-based development probabilities described below.

Consider Development Probabilities

We propose that local governments estimate the development probability (within the eight-year planning period) of all sites with potential for housing development. In this way, they can ensure that the expected yield of housing production is equal or greater than their RHNA targets. Rather than the current approach to the inventory, which assumes all vacant sites and those underutilized sites deemed potentially developable will be built on during the eight years, local governments should actually estimate the probability that different kinds of sites will be built on using data on recent development trends. They can then assess the probability of hitting their unit targets over the eight-year period.

We illustrate this approach with an example.10 Imagine a suburban jurisdiction primarily composed of single-family homes and a few commercial corridors with some parking lots and older storefronts. If the local government rezones the commercial corridor to allow midrise, mixed-use buildings, the parking lots and run-down storefronts may be redeveloped. Staff estimate only about half of any mixed-use projects would include a residential component, the typical residential component for these projects is 50 units, and they only consider 10 of the storefronts to be underutilized. Under the traditional approach to the sites inventory, one commercial corridor with 10 parking lots and 20 storefronts would yield: (10 [number of parking-lot parcels] * 0.5 [share of projects with residential component] * 50 [number of units per project with residential component]) + (10 [number of underutilized storefront parcels] * 0.5 [share of projects with residential component] * 50 [number of units per project with residential component]) = 250 + 250 = 500 units.

This assessment of zoned capacity accounts for the fact that some development on mixed-use sites is likely to be commercial rather than residential, but not for the fact that many (most) of the sites will not be built on during the planning period. Our approach asks local government staff to estimate development probabilities based on development activity in the region and apply them to potential capacity. Under this approach, staff estimate that the 10 parking-lot sites have a 50 percent probability of redevelopment over the eight-year period and the 10 storefront sites have a 20 percent probability. In that case, the expected yield over the planning cycle from rezoning the commercial strip is 250 * 0.5 + 250 *0.2 = 175 units.

This realistic assessment of housing production over the planning period is much lower, which may seem daunting at first, especially for local governments with higher RHNA numbers than in previous years. In the past, however, local governments have focused on only a portion of their jurisdiction’s parcels. Recall that most housing development has occurred on sites not usually included as part of a sites inventory. Moreover, rezoning in high-rent neighborhoods will have a larger impact on probable development than vacant parcels in less affluent neighborhoods. In this way, housing targets become attainable when combined with efforts to affirmatively further fair housing.

The approach also addresses the implementation challenges of AB 1397 if the state Department of Housing and Community Development (HCD) accepts the reasonable discounting of sites by probability of redevelopment as the functional equivalent of site-specific “likely to be discontinued” findings. In this way, local governments could have compliant HEs by assigning RHNA numbers in high opportunity neighborhoods. Instead of interpreting “likely to be discontinued” as a greater than 50 percent chance the use will be discontinued, HCD could interpret this to mean “likely relative to the probability of redevelopment claimed for the site.” Redevelopment requires both (1) discontinuation of the existing use and (2) the proposal and approval of a housing development. Both of these events have a probability of less than one, meaning the probability of both happening (1 & 2), i.e., the probability of redevelopment, is necessarily lower than the probability of (1) alone. In all cases, therefore, a site’s existing use is “likely to be discontinued” relative to the development probability claimed for the site if the estimate of development probability is realistic.

We also propose that the evidence-based probability of development approach also be treated as sufficient to satisfy the “existing leases” requirement in AB 1397, at least if local government certifies (1) that no information about leases was used to select parcels for the inventory, and (2) that the local government has no reason to believe that distribution of leases/lease terms across parcels is different in the jurisdiction at the time of HE adoption than it was in the jurisdiction at the time that the data were generated for the model of redevelopment probability. In estimating the probability of redevelopment from data, the average “lease barrier” to redevelopment is built in, even if no information about leases is included in the regression analysis.

There are some technical challenges to this approach and it requires data analysis some local governments may not yet be equipped to carry out.11 HCD can and should provide technical assistance for cities and counties.12 HCD could calculate the probability of (re)development for categories of sites in distinct housing markets and provide these estimates directly to governments, which depends in part on reliable local-level data reporting on recent development.13

Take a Proactive Approach to Affordable Housing Production

We propose that local governments create multiple proactive programs to meet their affordable housing unit targets as mandated by RHNA. In general, most cities and counties are passive and reactionary when it comes to housing development, especially affordable housing development. History shows that by simply listing potential sites for affordable development in the HE and waiting for developers to propose projects, local governments will not meet their RHNA targets for low-income housing.

There are several simple ways local governments can increase the chance of producing the deed-restricted affordable housing targets set out by RHNA. Here is a list of four areas they can work within to promote housing production:

1. Land: Identify publicly owned land suitable for affordable housing development and sell parcels for $1 (with consideration of the Surplus Land Act as amended by AB 1486 in 2019).

2. Money: Local governments can create new sources of funds (other than fees imposed on housing development) for affordable housing production, and divert existing revenues to housing where possible. Examples are local bonds funded by parcel taxes, tax revenue from short-term rentals, and parking revenues.

3. Outreach: City and county staff and councils can meet regularly with owners of potential sites and affordable housing developers to discuss needs and constraints in the jurisdiction, as well as serving as an intermediary.

4. Super Density Bonus, By-Right Approval: Adopting a program like the City of Los Angeles Transit Oriented Communities14 program would be an effective way to create new affordable housing in many cities and counties. This program offers more concessions than the minimum required under state density bonus law,15 and a menu of options for developers with a clear path to entitlement.


California’s housing affordability crisis requires public action on multiple fronts. Local governments have a tremendous opportunity to make change through their HE update and several new requirements for the sixth cycle that force them to change their approach. It is clear that a business-as-usual approach to the HE will do nothing to address the housing needs of California residents. The new, higher housing targets facing all local governments give them the chance to make real change.

Paavo Monkkonen is the Lewis Center senior fellow for housing policy and an associate professor of urban planning and public policy at the UCLA Luskin School of Public Affairs.

Christopher Elmendorf is the Martin Luther King, Jr. Professor of Law at the UC Davis School of Law.

Moira O’Neill is a senior research fellow at the Center for Law, Energy, and the Environment at Berkeley Law and associate research scholar at Columbia University.

Eric Biber is the Edward C. Halbach Jr. Professor of Law at Berkeley Law.


  1. Notably, SB 828 (2018) and AB 1771 (2018) have changed the process of determining regional housing needs and allocating these numbers to local governments, and SB 35 (2017) has created consequences for failing to meet local housing targets. The Legislature’s intent to require local governments to assign their RHNA share to parcels that are likely to yield production of the requisite number of units during the planning period is also manifested in the text and legislative history of AB 1397 (2017). See generally Elmendorf, C., E. Biber, P. Monkkonen, & M. O’Neill. (2020). Making It Work: Legal Foundations for Administrative Reform of California’s Housing Framework, Ecology Law Quarterly, 46, available online: https://www.ssrn.com/abstract=3500139 (manuscript p. 49). For a history of the pre-2017 implementation of the Housing Element Law see Dillon, L. (2017, June 29). California lawmakers have tried for 50 years to fix the state’s housing crisis. Here’s why they’ve failed. The Los Angeles Times.
  2. Elmendorf, C., E. Biber, P. Monkkonen, & M. O’Neill. (2020). Making It Work: Legal Foundations for Administrative Reform of California’s Housing Framework, Ecology Law Quarterly, 46, available online: https://www.ssrn.com/abstract=3500139
  3. Local governments usually discount from the absolute maximum potential units on a site to account for realistic outcomes of development resulting in fewer units, e.g., the commercial share on sites zoned for mixed use, but they don’t account for the probability that the site will be developed at all.
  4. We do not know the share of local governments that rely exclusively on vacant sites.
  5. O’Neill, M., G. Gualco-Nelson, & E. Biber. (2020). Sustainable Communities or the New Urban Renewal? Ecology Law Quarterly, 46 (forthcoming).
  6. Legislative Analyst Office. (2017). Do Communities Adequately Plan for Housing? March 8, available online: https://lao.ca.gov/publications/report/3605
  7. For a brief summary of the allocation of RHNA numbers among jurisdictions in the Southern California Association of Governments, see: https://www.latimes.com/california/story/2019-11-07/housing-building-density-zoning-coastal-inland-empire-southern-california-scag
  8. For more, see: https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=201720180AB686
  9. Monkkonen, P., M. Manville, & M. Lens. (2020). Built Out Cities: How California cities restrict housing production through prohibition and process. UC Berkeley Terner Center Working Paper.
  10. Based on the example in Elmendorf, C., E. Biber, P. Monkkonen, & M. O’Neill. (2020). Making It Work: Legal Foundations for Administrative Reform of California’s Housing Framework, Ecology Law Quarterly, 46, available online: https://www.ssrn.com/abstract=3500139
  11. Gualco-Nelson, G., M. O’Neill, & E. Biber. (2019). Policy Brief: Enhancing Local Land Use Data, available online: https://www.law.berkeley.edu/wp-content/uploads/2019/08/Enhancing-Local-Land-Use-Data.pdf
  12. It is important that HCD develop this expertise, if only to prevent local governments from overclaiming site capacity on the basis of recent state legislation that authorizes two accessory dwelling units (ADUs) to be added to lots zoned for single-family homes. State law allows only ADUs which are likely to be produced during the planning period to be counted toward a local government’s RHNA share. See Elmendorf, C., E. Biber, P. Monkkonen, & M. O’Neill. (2020). Making It Work: Legal Foundations for Administrative Reform of California’s Housing Framework, Ecology Law Quarterly, 46, available online: https://www.ssrn.com/abstract=3500139 (manuscript p. 18). To enforce this requirement, HCD must estimate (or supervise the estimation of) development probabilities.
  13. O’Neill, M., G. Gualco-Nelson, & E. Biber. (2019). Comparing Perceptions to Practice: Why Better Land Use Data is Critical to Ground Truth Legal Reform, available online: http://californialanduse.org/download/O’Neill_Comparing_Perceptions.pdf
  14. For more on the City of Los Angeles TOC program, see: https://planning.lacity.org/plans-policies/transit-oriented-communities-incentive-program (last accessed 12/27/2019).
  15. The recently passed AB 1763 (2019) provides another model, and cities could apply this approach to projects with less than 100% affordable units. For more, see: https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=201920200AB1763