Solving the SF Housing Crises: Loop 1
San Francisco may not have an app that can solve its astronomical housing prices, or console the 3,374 families who have lost their homes as a result of the Ellis Act, but Silicon Valley has produced a guide book for solving complex problems just like this – it’s called The Lean Startup. In it, author Eric Ries prescribes a methodology of rapid prototyping, experimentation and learning that can work for both new tech products and government programs. At Civinomics we’ve decided to put this innovation methodology to use solving the housing crises. The methodology outlines a Build-Measure-Learn feedback loop. The goal of a new startup, or government skunkworks in our case, is to get through this loop as quickly as possible, learning and improving the solution each time.
As you can see, the loop starts with an idea, so here’s one as good as any: We can solve the shortfall of housing in San Francisco by rezoning large parts of the western neighborhoods (Richmond, Inner Richmond, Sunset, Lake Merced) for higher density. Anything along Geary, where there is already great public transit would be a good start.
Take one look at the City’s zoning map and the same idea will probably occur to you as well. Nearly 70% of San Francisco’s existing land area is zoned for low density residential – light yellow on the map. So rezone and problem solved, right? Well, not exactly. At SF Public Press’s recent event, “Hack the Housing Crises” I asked Kearstin Dischinger of the San Francisco Planning Department just that, “Why haven’t you rezoned western neighborhoods?” Her response was that a) they actually did do a lot of rezoning of the Civic Center and SOMA areas during the 2000 tech boom and b) that it is currently politically unfeasible to rezone other neighborhoods even though they would like to.
So now the question becomes how do we make it politically feasible? Making sure people get to stay in their homes/neighborhoods would be a good start. Revised idea:
Create a policy that gives existing residents housing in new developments that happen in their neighborhoods, these residents will then support increased zoning densities.
Under this model, everyone wins: 1-2 story residential is replaced with 4-6 story mixed use development, more housing stock is created and existing SF residents now live in more modern housing while remaining in the neighborhood communities they’ve lived in for years.
Now for an initial product. Ultimately our product will be a way to easily approve new housing developments, allowing the City to grow and accommodate all of its happy residents, new and old. For now, lets see if we can get just one new theoretical development approved by talking to people directly in a promising neighborhood. Our initial product is thus a survey. Code for America start-up LocalData has a nifty survey app that lets you tie response data to a parcel map – allowing us to literally map public approval of increased zoning densities. Armed with this app, we identified a few promising blocks in the Inner-Richmond that meet all the conditions for transit oriented development, and hit the street. Here is the data we collected after initial exposure to customers:
Residents were initially 60% opposed to increased zoning density as shown in the first graph. When residents were guaranteed an apartment in the new development opinion moved to 50% opposed, 40% support 10% not sure. Both extremes were reduced and some respondents were simply not sure what to make of the proposal. In other words, we learned that the policy proposal needs more specificity. When posing the second question we faced a lot of incredulity. As one respondent put it, “Yeah I would support that, but I don’t believe it could happen.” We also gained a human understanding of our customer archetype (i.e. the people our product is trying to serve). A high percentage of respondents only spoke Chinese and we need Chinese speaking interviewers to engage them. Half of all respondents had lived in their homes for 11 years or more, possibly contributing to their resistance to change.
This first trip through the Build-Measure-Learn feedback loop brings us back to ideas for the next round. First and foremost, we need to design a more believable re-housing policy. We could accomplish this by talking to real-estate developers and understanding what a feasible project would look like. Exactly what accommodations could they make to existing residents? What might the resulting contracts look like? This would allow us to explain the specifics in more detail when confronted with questions at the door. Second, now that we know more about our customer, we know we need a policy or program which addresses their real challenges. Some residents are simply too old to feel they can make a significant life change like moving. A successful program may need to facilitate several aspects of this change for them. Others may be determined to maintain an investment property. Perhaps developers would be willing to structure a deal with equity in the new building for current homeowners?
Stay tuned, we’ll be moving through the feedback loop again shortly.