A recent study published in the Journal of Political Economy puts a hard number on something free-market advocates have been saying for decades: building more housing lowers rents. Specifically, a 1% increase in new housing supply pushes average rents down by about 0.19%.
Let that sink in for a second. Then let's do the math for San Francisco.
If you want rents to drop 10% — a meaningful reduction that might actually let a nurse or a teacher live in the city where they work — you'd need roughly a 52% increase in new housing supply. San Francisco currently has around 400,000 housing units. That means we'd need to build approximately 208,000 new units to hit that target.
For context, SF permitted about 4,000 units last year. At that pace, we'd get there in... 52 years. Hope you're not in a rush.
Now, this isn't an argument against building. It's the opposite. The data confirms what basic economics has always told us: supply matters. Every unit built puts downward pressure on rents. The problem isn't the theory — it's that San Francisco has spent decades making it nearly impossible to build anything, anywhere, at any scale.
As one local on Reddit pointed out, there's an important nuance here: "It's NOT 1% of housing units increase equals 0.19% lower rents. It IS 1% of new supply increase equals 0.19% lower rents." In other words, the baseline matters. If you're already building a lot, marginal increases have marginal effects. If you're building almost nothing — hello, San Francisco — even modest increases in construction represent a large percentage jump in new supply.
The takeaway isn't that building is futile. It's that half-measures are. Approving a 50-unit project here and a 12-unit project there after years of review and appeals isn't a housing strategy. It's a jobs program for land-use attorneys.
If the city is serious about affordability, it needs to stop treating every new building like a constitutional crisis. Streamline approvals. Cut fees. Upzone aggressively. Let the private sector do what it does best: respond to demand.
The math is simple. The politics are what's expensive.
