Single-Family Zoning in California Methodology Statement

Challenges

There are many daunting research challenges to studying municipal zoning. Foremost among them is the lack of consistent classifications or base formulae for zoning designations from which they may be derived. Zoning codes are generally dense, technical, and difficult to access. While there may be strong similarities in zoning taxonomies across jurisdictions with land use authorities, the particulars of each zoning code are often idiosyncratic to the jurisdiction. Certain designations may mean different things in different jurisdictions. This means that studying regional zoning patterns requires tedious and time-intensive effort to produce accurate data.

There are shortcuts researchers can use to gather data, but they lack precision. The most common and widely relied upon shortcut is the use of survey instruments (e.g. the Wharton Regulatory Land Use Index). We summarized notable zoning surveys, going back to the 1970s, and their respective strengths and limitations, in our Bay Area report.1 Although there are a number of other proposed methodological shortcuts, including restricting analysis to a sample of cities using American Community Survey samplings of housing units within cities, or drawing inferences about zoned density from property tax records, the vast majority of research conducted on this issue relies upon these survey instruments.2

While these surveys may be generally useful, there is evidence to suggest that their results contain systematic errors, partly due to the inaccurate perception of the survey respondents.3 These surveys also suffer from lack of reliability, since, in many cases, respondents fail to complete the survey, creating holes in the datasets.4 These instruments are also limited and tend to focus on the largest jurisdictions, meaning that smaller municipalities are less likely to be included. This may be another source of systematic bias in such data, since smaller municipalities may have a greater share of restrictive zoning.

Another challenge to studying zoning is that zoning ordinances and codes do not necessarily indicate actual zoning, especially for planned developments. Planned development is typically a designation that allows for more flexible or dynamic zoning and land use, where planners anticipate or intend future development. The intended use for such parcels may not be defined in the municipal zoning code, meaning that it has to be observed directly.

A related problem is that zoning codes do not always reflect the built environment, and vice versa. This is probably the single biggest source of confusion regarding our zoning maps, as evidenced from inquiries we have received about them. Our maps and analyses are of zoning, not the built environment. While the actual built environment may be the most relevant piece of information for some purposes (such as researchers or policymakers who are primarily concerned with the production of housing), our research is focused on better understanding how zoning designations themselves shape or correlate with certain community characteristics and life outcomes.

So-called “non-conforming uses,” which create a divergence between the built environment and zoning laws, exist for many reasons. Some developments, especially older ones, may have been constructed before zoning codes were adopted or prior to a down-zoning of the neighborhood to prevent new, higher density developments. Developers of more recent developments may petition for, and receive, a “spot variance,” which is permission to build at a higher density or taller than what is permitted by ordinance. Certain state laws, like density bonus laws, may similarly create override mechanisms that generate non-conforming structures. One cannot easily scan an environment and accurately infer a zoning designation.

Methodology

Because we are specifically interested in the effects and/or correlates of restrictive zoning itself, and most of the zoning codes were idiosyncratic across municipalities, for purposes of this report, we simplified zoning designations into three broad categories: Single Family Residential, Other Residential, and Non-Residential/Unknown, defined as follows: 

Single Family Residential is land designated for detached, single-family residential land use (one or two dwelling units per parcel of land) in both low and high density. This includes single-family homes and two-family detached dwellings, and usually includes single-family zones, low-density zones, agriculture zones (if single-family homes are permitted) or estate zones. The zone is not classified as single-family residential if it only permits the following: caretaker’s residence, employee housing, or live-work unit.  

• Single Family Residential is land designated for detached, single-family residential land use (one or two dwelling units per parcel of land) in both low and high density. This includes single-family homes and two-family detached dwellings, and usually includes single-family zones, low-density zones, agriculture zones (if single-family homes are permitted) or estate zones. The zone is not classified as single-family residential if it only permits the following: caretaker’s residence, employee housing, or live-work unit.  
• Other Residential includes both multi-family residential and mixed-use residential.
  • Multi-family residential is land designated to allow for multiple dwelling units per parcel of land. This includes apartment buildings, duplexes, triplex, fourplex, townhouses, condos, mobile home parks, and two-family attached dwellings.  Additionally, if a zone’s intent allows for both single-family homes and any of the multi-family housing above, it is sorted into the multi-family category. This category does not include employee housing and student housing. 
  • Mixed-Use Residential is land designated to blend multiple uses that includes residential use. This includes mixed-use commercial zones permitting residential use, zones allowing multi-family buildings with commercial spaces (e.g., on the group floor), and zones where residential use and other non-residential uses such as commercial or industrial uses are both permitted.5
• Non-Residential is land that is designated for non-residential uses. This includes parks and open space, commercial, industrial, and public land. Non-developable land was removed from the overall land area, including water, waterways, marshland, and highways or streets.

Municipal zoning code documents are the primary source to identify each zoning category for each city, describe the purpose of each, and list what is permitted to be built in each of these zones. We utilized these documents for each city to recode each zone into the typology listed above. Based on city ordinances for each unique zone in each municipality, we recorded the type of housing development permitted and conditionally permitted as follows:

P: permitted as right. We did not include building types that were permitted for accessory use only.

C: conditionally permitted; requires review; requires minor use permit

We prioritized and focused on the intent and the purpose of the zone to sort the zones into one of the three categories: non-residential/unknown, single-family, and other residential. For example, a residential zone intended for both single-family and multi-family developments would be classified in the other residential category (which includes multi-family zones), even if multi-family is only conditionally permitted. This is because the zone’s intention was for both single-family and multi-family developments.

When the intent and purpose of the zone was unclear, we classified the zone based on what is permitted in the zone. For example, consider a “general residential” zone that does not specify whether it is strictly for single-family homes in its intent. If it permits single-family homes but only conditionally permits multi-family homes, it would be classified in the single-family category. And if it permits both single-family homes and multi-family homes, it would be classified in the other residential category. Since we prioritized the zone’s stated intent and purpose in the city ordinance, we would only rely on the permitted use if the intent and purpose was unclear.

We accessed parcel-level zoning data primarily from shapefile repositories maintained by individual municipalities, in addition to the repositories maintained by The California Governor’s Office of Planning and Research (OPR).6 We applied our classification of simplified zoning categories to the zones found in these zoning shapefile for each city. 

Once we obtained the zoning shapefile, we created a list of all zoning codes available within the shapefile. We used the description of each zone in the municipal code document, and re-coded each of those to one of the three categories listed above. We then applied our simplified typology to each of the parcels in the shapefile. We mapped the data to display this typology.

We realize that our simplified taxonomy of myriad zoning designations into three categories (in order to draw out the degree to which single-family-only zoning predominates residential areas) conceals many other aspects of zoning codes that impede or restrict development or certain types of developments. For example, height limitations, setback requirements, discretionary reviews, parking requirements and the like can all inhibit density, even in multi-family residential or mixed-use zones. For that reason, in constructing this tranche of maps, we have separately created a database of additional zoning characteristics for each jurisdiction, which we will release at a future point in completing a statewide zoning atlas for California. Until then, our focus is on single-family-only residential zoning as the principal form of restrictive and exclusionary zoning.

Additional notes on methodology are below.

Quality check procedures: Human errors could impact the quality of the analysis so it was important to establish data quality checking protocols to reduce any human errors. We followed these procedures at different stages of data collection and classifying our zoning categories.

Once the zoning codes were extracted from the shapefiles and were re-coded to our three simplified zoning categories, another team member conducted a spot check to ensure the extraction and coding was done properly and with internal consistency. We cross-checked our final maps against the zoning maps provided by the city to ensure our processing was accurate.

Though attempts were made to minimize any human error while interpreting various zoning codes and recoding those to our simplified categories, it was beyond our capacity to check the input data for accuracy. We were not able to check that the zoning shapefiles used for this analysis had the data entered correctly. We solicit any inputs or suggestions from the general public or city departments to highlight any inconsistencies in identifying the correct zoning categories and/or share updated zoning data to ensure our mapping and analysis are accurate.