San Francisco Rents Are Surging Again. The Bigger Story Is the Housing We Aren’t Building.

For several years, San Francisco’s housing market seemed to be telling a very different story. The pandemic emptied offices, weakened apartment demand, and briefly gave renters something they had rarely enjoyed in the city: leverage. That period is now decisively over. According to a recent San Francisco Business Times article citing Zumper’s July rental report, the median rent for a two-bedroom apartment in San Francisco has crossed $6,000 for the first time since Zumper began tracking the market more than a decade ago. The median two-bedroom reached $6,020 in July, up 25.9 percent from a year earlier, while the median one-bedroom climbed to $4,180, a 22.9 percent increase.

Those numbers are striking, but what is happening underneath them is even more important. The Business Times describes a market being squeezed from both directions. San Francisco’s growing artificial intelligence sector is bringing highly paid workers into the city at the same time that the supply of available apartments is shrinking. Zumper reports that active San Francisco rental listings are down about 30 percent year over year. With fewer apartments available and more renters competing for them, the city has moved rapidly from the soft rental environment of the early pandemic years to one of the tightest markets in the country.

The scale of that rebound stands out against the rest of the country. The San Francisco Standard reported in July that no other U.S. city comes close to San Francisco’s roughly 22 percent year-over-year rent growth. The next-fastest large market, St. Louis, was rising at about half that rate, and its median rent was roughly a quarter of San Francisco’s. That gap is the clearest sign that this is not a nationwide rent cycle but a local supply-and-demand story playing out at an extreme.

This Is a Demand Story, but It Is Also a Supply Story

It would be easy to describe this simply as another consequence of the AI boom. That would only tell half the story.

Strong employment growth normally creates demand for housing. In a market capable of responding, rising rents and improving apartment economics should eventually encourage additional construction. New units come onto the market, absorbing some of the additional demand and moderating future rent increases. San Francisco currently has a much more difficult problem: demand is recovering at precisely the moment when new housing production has slowed dramatically.

The article points specifically to what Zumper describes as a “near-empty construction pipeline.” Even projects that have already received approvals are not necessarily moving forward. Hayes Point, the prominent site at Market Street and Van Ness Avenue, was approved for a 47-story residential and office tower, but the project has been put on hold and the property remains vacant. The Business Times also notes that higher interest rates are making multifamily developments even more difficult to pencil.

Independent market data confirms just how thin that pipeline has become. Grant Montgomery, national director of multifamily analytics at CoStar, told the San Francisco Standard that fewer than 1,000 market-rate apartment units broke ground in the San Francisco metro area over the prior twelve months, a pipeline he described as thin by historical standards, with fewer projects moving forward because construction and financing costs remain elevated. For a region of this size and with this much demand, that is a remarkably small number of units entering construction.

That distinction matters. San Francisco does not necessarily suffer from a lack of approvals to build housing. It has entitled thousands of units and has spent years changing zoning, adopting housing programs, and identifying places where additional development should occur. But an entitlement is not an apartment. If construction costs, financing conditions, land values, insurance expenses, fees, and expected rents do not add up, an approved project can remain an empty lot indefinitely.

The Market Is Sending a Very Loud Signal

One of the most remarkable comparisons is with New York. San Francisco’s median two-bedroom apartment now rents for $570 more than a comparable unit in New York City, according to Zumper. New York still holds the top spot for one-bedroom rents at $4,560, but San Francisco is now considerably more expensive for two-bedroom apartments. By comparison, the national median is just $1,520 for a one-bedroom and $1,906 for a two-bedroom.

That is not just an interesting statistic, it is a market signal.

The pressure is even more dramatic at the neighborhood level. The San Francisco Standard found that in the neighborhoods closest to downtown job centers, rents have moved far faster than the citywide average. SoMa, where AI firms have concentrated their office footprint, saw one-bedroom rents rise nearly 42 percent in a single year, while Mission Bay became the city’s most expensive one-bedroom market at close to $5,500 a month. On the north side, two-bedroom rents in Nob Hill and Cow Hollow ran more than 57 percent above the prior year, and the Marina reached roughly $7,660 for a two-bedroom. As one leasing agent told the Standard, a renovated property with parking, outdoor space, and a walkable location can now command almost any price. These small-sample neighborhood figures move sharply month to month, but they show where the demand is concentrating: within walking distance of the new AI workplaces.

Rents climbing by more than 20 percent in a single year indicate that the housing market is having difficulty adjusting to renewed demand. The problem can also become self-reinforcing. As rents rise, existing tenants become increasingly reluctant to move because leaving an older or below-market apartment may expose them to dramatically higher current rents. That reduces turnover, removes additional units from the available inventory, and makes competition for the remaining apartments even more intense. Zumper specifically identifies this dynamic as one reason San Francisco’s inventory is tightening.

But there is a somewhat perverse flip side to rising rents: they may also be exactly what is needed to get some stalled housing projects back into construction.

Apartment development is ultimately a financial equation. Developers and lenders look at the income a completed project is expected to generate and compare it with the cost of acquiring land, building the project, financing construction, operating the property, and insuring it. Over the past several years, many of those costs have moved in the wrong direction. Construction costs remain high, borrowing costs have increased, and insurance has become more expensive. The Business Times specifically identifies interest rates as another challenge for projects trying to make the numbers work.

For many stalled projects, the problem is therefore not that they cannot be approved. It is that expected revenues have not been sufficient to justify the cost of actually building them.

That means higher rents, undesirable as they are from an affordability standpoint, can improve development feasibility. If developers can reasonably underwrite higher future rents, the projected value of a completed apartment building increases. That can narrow the gap between development costs and expected revenues, make lenders and equity investors more comfortable committing capital, and potentially allow projects that have been sitting on the sidelines to move forward.

In other words, today’s rent increases could eventually produce a supply response.

There is an important caveat. A jump in asking rents does not automatically make every project feasible. Developers and lenders are generally underwriting rents several years into the future, when a project will actually be completed, rather than simply assuming that today’s unusually rapid increases will continue indefinitely. Construction, financing, and insurance costs may also continue to rise. But sustained rent growth can materially improve the economics of projects that were previously just below the threshold of feasibility.

That creates one of the stranger dynamics of the housing market: the worsening affordability problem can eventually help create the financial conditions necessary to build more housing.

San Francisco May Be Entering the Next Phase of Its Housing Cycle

Housing markets tend to lag economic changes, and development responds particularly slowly.

The first phase of San Francisco’s recent cycle was the pandemic decline, when demand weakened and rents fell. The second was the development slowdown, as weaker rents, rising construction costs, and higher interest rates made new projects increasingly difficult to finance. We may now be entering a third phase, in which strong employment growth and sharply rising rents begin to restore the economics of apartment construction.

The timing will not be immediate. Even if a project becomes financially viable today, it may take years to secure financing, complete construction documents, obtain permits, and build the project. But if rents remain elevated, some developments that were previously shelved could begin moving again.

That is why the current surge in rents could eventually prove to be more than a symptom of the housing shortage. It could also become part of the mechanism that brings additional supply to market.

The policy question for San Francisco is whether the city will allow that market signal to translate into actual housing production. If improving rents finally make a project financially feasible, the city should be looking for ways to reduce the remaining barriers between an entitlement and a groundbreaking. Faster permitting, predictable fees, sufficient density, reduced carrying costs, and clear development standards become especially important at the moment when projects are close to penciling.

There are signs the city has begun to move on the piece it controls most directly: process. A 2025 report from the Board of Supervisors’ Budget and Legislative Analyst found that San Francisco still took a median of 280 days and three rounds of review to issue a housing building permit, roughly three times as long as Austin and more than twice as long as Seattle or Washington, D.C. But the same report found that projects submitted after the start of 2024, under Mayor Daniel Lurie’s and former Mayor London Breed’s permitting reforms, reached permit issuance in a median of just 114 days. That is a meaningful improvement, and it matters most precisely now, when a faster, more predictable path to groundbreaking can determine whether a marginal project moves before financing conditions shift again.

San Francisco cannot control interest rates, construction prices, or insurance premiums. But it can influence how much additional cost, delay, and uncertainty the development process adds. The more those locally controlled costs can be reduced, the less rent a future project needs to generate before it becomes financially feasible.

The Housing Shortage Did Not Disappear. Demand Did.

San Francisco’s pandemic-era decline created the impression, for a time, that the city’s longstanding housing shortage had somehow been solved. It had not. Demand had temporarily collapsed.

Now demand is returning, and the structural shortage is becoming visible again. The depth of that structural gap is easy to understate: by one estimate cited by Axios, San Francisco remains on the order of 140,000 homes short after decades of underbuilding, a deficit that would be difficult to close even if population growth stayed flat. The current AI cycle did not create that hole, it is simply the latest force exposing it.

The Business Times describes the immediate catalyst as aggressive AI hiring combined with scarce inventory. But the larger lesson goes beyond the current technology cycle. San Francisco has one of the strongest urban economies in the country, and periods of economic expansion will repeatedly create housing demand. A city cannot control when the next industry grows, when workers decide to return, or when capital markets improve. What it can influence is whether its housing supply is capable of responding when those things happen.

A median two-bedroom rent above $6,000 is therefore more than another milestone in San Francisco’s famously expensive housing market. It is evidence of what happens when a city experiences an economic rebound without a corresponding housing rebound.

Paradoxically, though, those record rents may also help set the stage for that housing rebound. If higher future revenues are finally sufficient to offset elevated construction, financing, and insurance costs, projects that have been approved but dormant may begin to pencil again. The question then becomes whether San Francisco can convert improving development economics into actual construction quickly enough to make a difference.

The AI boom may be providing the spark. The shortage of housing is supplying the fuel. But rising rents may finally provide enough economic incentive to start building again.