In June 2016, voters approved Proposition C, a measure placed on the ballot by Supervisor Aaron Peskin & Jane Kim, then a candidate for State Senate. Prop. C was sold to voters as a way to double the amount of “affordable” housing produced by market-rate developers. (Affordable is in quotes here because the below-market-rate housing produced by this program is still pretty expensive.)
The San Francisco public were sold a bill of goods when it came to Prop. C. Voters were told that developers were going to continue to produce thousands of units of housing, and instead of getting a measly 1,200 BMR apartments per 10,000 apartments proposed, the city could have 2,500 BMR apartments.
What actually happened? Overall housing proposals tanked — they are 20% of what they could have been based on the pre-Prop. C trend. Instead of getting 2,500 BMR units over the last three quarters as promised, or even keeping the 1,200 that the trend suggests would have been proposed, developers proposed barely 500 new BMR units in the last three quarters.
Industry analysts and actual affordability advocates (not slow-growth advocates in disguise) warned that Prop. C wouldn’t work the way proponents promised and would instead make San Francisco’s housing situation worse.
Real anti-displacement activists knew Prop. C wouldn’t work because the theory of why it should work doesn’t make any sense. Proponents of Prop. C claimed it would reduce developers’ profit margin — instead of pocketing the (evidently ill-gotten) gains gleaned from building housing, developers would be forced to put those profits back into the project by offering reduced rents to 25% of the tenants.
But developers are like any other kind of business person — if a job isn’t worth their while, they just won’t do it. Indeed, San Francisco city planners told me that some developers who were going to do housing projects before Prop. C switched their proposals to hotels after Prop. C passed.
Imagine you’re a wedding photographer. You can usually earn $1,000 for a wedding in San Francisco, $975 for a wedding in Oakland, or $1,200 for a wedding in Marin. Suddenly San Francisco passes a new law that means you can only earn $700 to photograph a wedding in the city. Suddenly jobs in Oakland take precedence over ones in San Francisco, and when you work in San Francisco you photograph at other kinds of parties — anniversaries, graduations, etc.
Before June 2016, San Francisco’s inclusionary zoning program required developers of market-rate projects to set aside 12% of their project to be affordable to lower income people. After June 2016, when Prop. C passed, the new requirement was for 25% of the project to be below-market-rate.
Prop. C opponents warned voters that the measure wouldn’t work — instead of making developers produce more below-market-rate housing, Prop. C would discourage developers from proposing new housing developments at all. This would reduce the amount of housing proposed overall, including the amount of below-market-rate housing, and this is exactly what wound up happening:
The San Francisco city economist projected a modest decline in proposed housing units. The actual decline has been much worse:
Before Prop. C passed, proposed housing units were trending up:
After Prop. C, hold a funeral for housing:
It’s reasonable to ask whether housing proposals might have gone down anyway after the second quarter of 2016. We don’t have an alternate universe San Francisco that didn’t pass Prop. C in June to look at. What we do have is data on projects smaller than 25 units.
Projects smaller than 25 units weren’t affected by Prop. C. Projects smaller than 10 units have no inclusionary requirement at all, and projects 10 to 25 units had to set aside 12% BMR both before and after Prop. C:
No, small project applications don’t show a slowdown. Large project applications have been declining every quarter since Prop C. Small projects are zig-zagging uninterrupted.
It’s also reasonable to ask why I’m asking Google Sheets to draw trend lines on data sets with only six or nine points, or using lines of best fit to extrapolate average values outside of the range of my independent variable. Good questions.
The Planning Department has data going back further than January 2015, but doesn’t make it available online. This report is the best I can do with what I have access to: http://sf-planning.org/pipeline-report.