Formula Retail: How did we get here?

Into the blue again, after the money's gone

El Farolito, a favorite taqueria of San Franciscans, grew over the years to have multiple locations throughout the Bay Area. As is typical of San Francisco government, we're now punishing that success.

El Farolito wants to open a new store in North Beach, but San Francisco's "formula retail" ban is preventing them from opening their 12th family-owned taqueria. It is hard to find a city in America more hostile to small businesses than San Francisco, so on the one hand this is unsurprising. What should surprise you is that a law ostensibly created to keep out McDonalds and Starbucks is being used to hurt this hometown success story.

What is "formula retail"

The definition of formula retail is, basically, any store that has 11 or more physical establishments. The definition spans multiple categories: a coffee shop like Starbucks, fast food like McDonalds, a gym like Planet Fitness, or even a grocery store like Safeway. Here's the definition of formula retail, as found in San Francisco Planning Code Section 303.1(b):

A Formula Retail use is hereby defined as a type of retail sales or service activity or retail sales or service establishment that has eleven or more other retail sales establishments in operation, or with local land use or permit entitlements already approved, located anywhere in the world. In addition to the eleven establishments either in operation or with local land use or permit entitlements approved for operation, the business maintains two or more of the following features: a standardized array of merchandise, a standardized facade, a standardized decor and color scheme, uniform apparel, standardized signage, a trademark or a servicemark.

Combined with 303.1(c), which defines what retail uses are included, we understand the full breadth of this law:

[A] retail sales or service activity or retail sales or service establishment shall include the following uses[...]:

- Bar
- Drive-up Facility
- Eating and Drinking Use
- Liquor Store
- Sales and Service, Other Retail and Retail Sales and Service, General;
- Restaurant
- Limited-Restaurant
- Sales and Service, Retail
- Service, Financial
- Movie Theater
- Amusement Game Arcade
- Service, Limited Financial, except single automated teller machines at the street front that meet the Commission’s adopted Performance-Based Design Guidelines and automated teller machines located within another use that are not visible from the street
- Service, Fringe Financial
- Tobacco Paraphernalia Establishment
- Massage Establishment
- Service, Personal
- Service, Instructional
- Gym
- General Grocery
- Specialty Grocery
- Pharmacy
- Jewelry Store
- Tourist Oriented Gift Store
- Non-Auto Vehicle Sales or Rental
- Cannabis Retail

The problem isn't necessarily that we're extracting wealth from big chain stores by forcing them into a year-long conditional use approval process (though I'm no fan of that), rather that this definition of formula retail is far too strict and it's killing small businesses while exacerbating retail vacancies across the city.

Take North Beach, where El Farolito wants to open. The North Beach Neighborhood Commercial District has some of the strictest land-use rules of any neighborhood and a commercial vacancy of over 10% in 2018 (which has only worsened during Covid) — among the worst in the city.

Changing an old restaurant into a coffee shop, or a clothing store into a bar requires 10-14 months navigating San Francisco's kafkaesque bureaucracy to get permission to change the "use" of the space. Note that the business or individual seeking this change must be paying rent while jumping through these hoops. That can mean paying as much as $200,000 in rent for a vacant store while you wait for City Hall and the Planning Commission to allow you to pursue your dream.

If it's banned, why are there so many Starbucks?

Well, formula retail is only outright banned in three neighborhoods: North Beach Neighborhood Commercial District, Hayes-Gough Neighborhood Commercial Transit District, and Chinatown Visitor Retail District. A number of other commercial districts only ban certain kinds of commercial activity. See SF Planning Code Section 303.1(f) for the full list.

All other neighborhoods require formula retail to apply for a "Conditional Use Authorization" (CUA). The CUA process works like any other discretionary permit: approval is not guaranteed, the applicant must spend months doing "community outreach" and meeting with local residents and business owners and merchants associations, as well as attend a handful of mandatory public meetings to plead their case.

A small business like El Farolito doesn't have the resources to invest in this process and may opt to just give up, instead. But very large businesses like Starbucks are willing to sink $100,000 into getting their permit because their stores are so profitable and in such high demand.

Coffee shops may be fundamentally more suited to large chains. Independent cafes have wildly different quality (compare the fantastic Four Barrel to the absolutely awful Muddy Waters down the street), but consumers know exactly what to expect at Starbucks. It's not good coffee, but it's also not a gamble. Some business types are just more likely to both become successful chains and to better serve consumers as chains.

The prevalence of formula retail varies significantly by business type and size. For example, 49 percent of San Francisco’s coffee shops are formula retail, compared to 11 percent of all restaurants. The vast majority of pharmacies over 3,000 square feet and supermarkets over 10,000 square feet are formula retailers, while smaller establishments are much more likely to be independent retailers. More than 80 percent of all banks are formula retail.

- San Francisco Formula Retail Economic Analysis report, p. 4

Economic effects

Many advocates for the formula retail controls claim that this cumbersome process makes it easier for small businesses to find an affordable lease, which should lead to a boom of small businesses in neighborhoods that ban big businesses. A 2014 study by the SF Office of Economic Analysis disagrees (emphasis added):

In [neighborhoods at risk of commercial decline due to blight conditions], policies that discourage formula retail tenants could have negative consequences on the surrounding neighborhood and the city's economy.

- Expanding Formula Retail Controls Economic Impact Report, Office of Economic Analysis, p. 17

And another 2014 study by Economic Strategies, commissioned by the SF Planning Department, found a similar negative impact:

Expanding the application of formula retail controls to other types of land uses could affect a significant number of businesses considering new locations in San Francisco, and make it more challenging to fill vacant storefronts in some neighborhood commercial districts.

- San Francisco Formula Retail Economic Analysis report, p. 9

Indeed, there even seems to be a positive impact of formula retail on neighborhoods struggling with commercial decline and vacancy:

[A] formula retailer that serves as an anchor and draws new customers to a revitalizing neighborhood commercial district can have a positive effect on other retailers in the district, and potentially lead to increased sales and rents. [...] In the Ocean Avenue Neighborhood Commercial Transit District, for example, a new Whole Foods has attracted new customers and contributed to efforts to revitalize the area.

- San Francisco Formula Retail Economic Analysis report, p. 6

Additionally, formula retail not only employs more people, but also charges lower prices. This causes consumers to have more money in their pocket, which they can either save or spend at more businesses.

[T]he OEA's research suggests that prices at non-formula retailers are 17% higher than they are at formula retailers.

This price difference means that, even though policies that effectively divert spending to non-formula retailers do lead to higher levels of spending on local factors of production such as business suppliers, consumers that shift their purchases to non-formula retailers will have less to spend at other businesses.

As the table on the next page illustrates, the economic cost of higher prices on local consumers outweighs the potential benefit of greater local spending by non-formula retailers, and the net local spending impact is somewhat negative.

- Expanding Formula Retail Controls: Economic Impact Report, Office of Economic Analysis, p. 25

How did we get here

Finding the full story of how San Francisco came to ban large businesses reveals the small-c conservative nature of our local progressive government. Local politicians all sought to have more control over who got to set up shop in their little kingdom, and so-called small business advocates saw the chance to give themselves an unfair advantage and keep out competitors.

The SF Planning Department provides a helpful timeline view of how formula retail rules tightened like a one-way ratchet. In short: the first bans were enacted in 2004, expanded in 2007's Proposition G, and dramatically expanded and made less flexible in 2013 and 2014.

In 2011, as San Francisco was in rapid recovery from the economic crisis, there was rabblerousing about small businesses being hurt by big chain stores. There wasn't actually evidence of this happening, just a bunch of anecdotes and a strong feeling that gatekeeping the neighborhood was the Right Thing To Do.

This 2013 memo from the Planning Department was created "to review and analyze planning controls for formula retail uses in San Francisco due to the numerous pending proposals to change these controls." The stated goal of the report "to develop a series of controls that are clear, concise, and easy to implement that will protect neighborhood character and provide necessary goods and services" notably says nothing about fostering new business or growing the economy — instead the true small-c conservative nature of progressivism shows that it's more interested in protecting aesthetics. Here's a short excerpt from that report summarizing proposed and existing controls on formula retail:

2. Supervisor Breed would create the Fillmore [BF 120814] and Divisadero [BF 120796] NCDs which, among other controls, originally sought to prohibit new formula retail uses. Her new proposal would seek to weigh the community voice over other considerations (including staff recommendation); generally weigh the hearing towards disapproval; legislate a requirement for pre-application meeting; and codify our current formula retail policy for Fillmore and Divisadero. [...]

3. Supervisor Breed would also amend the definition of formula retail but only in the Hayes Gough NCT [BF 130468] [...]  to modify the definition of formula retail to include formula retail that is a type of retail sales activity or retail sales establishment and has eleven or more other retail sales establishments located anywhere in the world (emphasis added). [...]

4. Supervisor Kim introduced interim controls [BF 130712] at the July 9th, 2013 Board of Supervisors' hearing that would impose interim zoning controls requiring conditional use authorization [...] on Market Street, from 6th Street to Van Ness Avenue, subject to specified exceptions for grocery stores, for 18 months.


6. Mobile Food Facilities. Supervisor Wiener's recently approved ordinance amended the Department of Public Work's code [BF 120193] to restrict food trucks that are associated with formula retail establishments in the public right of way.[...]

You'll notice some familiar names there. Much to my disappointment, both then-Supervisors London Breed and Scott Wiener were proponents of these restrictions. Their support is illustrative of the misaligned incentives at the very local level vs higher office.

Supervisors, who represent roughly 75,000 people, respond to parochial interests and tend towards NIMBYism. But at higher office, where Breed and Wiener now represent 850,000 and 930,000 people, respectively, they're our best advocates for growth and YIMBYism. Understanding Supervisors as the lords of their little kingdoms explains a lot about the dysfunction of SF politics.

A later report from the Planning Department in 2014 did seek to balance economic concerns against the preservative nature of the formula retail controls. 

The ensuing policy recommendations are representative of the Planning Departments desire to put forth a balanced policy proposal designed to regulate formula retail in a manner that encourages economic development and job creation while maintaining the unique and distinctive nature of San Francisco’s neighborhoods.

Of note, this report recommended that the threshold of 11 stores be raised to 19:

  • Of all existing FR in SF, only 5% have fewer than 20 locations

  • Small businesses like Blue Bottle and Philz Coffee with 14 locations currently are reviewed under the same process as Subway and Starbucks with over 20,000 locations.

  • Raising the threshold [to 19] would allow for greater small business development in SF

The economic analysis report agreed that 11 was too few:

Changing the definition of formula retail to businesses with at least 20 or 50 other establishments (rather than the current 11) would exempt some fast-growing start-ups, while still capturing the vast majority of large, established chains. […] only 5 percent of formula retailers in San Francisco are associated with businesses with fewer than 20 total branches or subsidiaries. Another 4 percent have between 20 and 50 locations. The remaining formula retailers are either franchises (about 17 percent) or have more than 50 locations (nearly 75 percent).

- San Francisco Formula Retail Economic Analysis report, p. 10

But like most recommendations based on economic vibrancy over aesthetic preferences, this recommendation was rejected and a limit of 11 stores was written into law. 

Do we even need these limits?

In short: no.

A recent research paper titled "Measuring McCities: Landscapes of chain and independent restaurants in the United States" sought to understand how chain stores spread through America. They found that cities like San Francisco are exactly the kinds of places you would expect local independently-owned stores to thrive:

Coastal metropolitan areas such as New York (lowest in percent chain) and San Francisco (lowest in replicability) tend to have low chainness. In contrast, cities in the South (e.g., Nashville), Midwest (e.g., Indianapolis), or Southwest (e.g., Phoenix) are more subject to chains.
Areas with very low or very high Walk Score, population density, and median household income tend to have medium or low chainness, and thus maintain a locally run food scene. These areas may correspond to rural tracts or wealthy downtowns.
In particular, areas with high Walk Score, such as pedestrian-friendly main streets or tourist areas, have the lowest predicted chainness across all sociodemographic features.

In our stupor to regulate The Big Bad Thing out of our city, we failed to realize that chain stores already are extremely unlikely to proliferate in San Francisco.

The biggest predictors of chain stores are car dependence and Trump voters:

Planning regulations such as zoning and parking requirements also privilege chains that tend to cluster and serve customers traveling by car.
Medium density census tracts (i.e. 1000–4500 people per square mile) that are car-dependent (Walk Score 15 to 40) and have a medium household income ($50,000) are most likely to be McCities.
The percentage of Trump voters is still the most indicative factor in defining a McCity, i.e. an area with high chainness, across spatial scales (census tracts, UA, or CBSA), even after controlling for race, age, income, and education.

- Measuring McCities

Not only do the formula retail controls slow down business and job creation, erect huge financial and bureaucratic barriers in front of growing small businesses, and end up being mostly useless at boosting local businesses, they also make San Franciscans poorer along the way.

Where do we go from here?

If I can put on my realpolitik hat for a minute: the best path forward is to propose a rule change that raises the limit from 11 to 50 stores. It will liberate actual small businesses from the bureaucratic nightmares we currently impose on successful businesses while applying the same tight controls on actually large businesses. But most importantly, people will vote for this change.

It's easy to understand why El Farolito getting screwed is bad. It's easy to understand that a business with 20 stores is different from one with 2000. It's easy to understand that recovering from the Covid recession will require loosening the noose around local businesses. And we have economic impact reports from several years ago all warning us that 11 was too low.

All we need to do is run a ballot measure.