Scott Alexander dismantles the comforting narrative that San Francisco's sudden visual calm is the result of a benevolent policy victory. Instead, he presents a stark, data-driven reality: the city didn't solve homelessness; it simply made it harder to see. For busy observers tracking the shift in urban safety, this piece offers a crucial correction to the headlines, arguing that the decline in visible suffering is largely an artifact of legal shifts and enforcement tactics rather than a resolution of the crisis.
The Illusion of Resolution
The core of Alexander's argument rests on a timeline that refuses to align with political credit. While Mayor Daniel Lurie and the Supreme Court's City of Grants Pass v. Johnson decision are often cited as the turning points, Alexander notes that the decline in tent encampments began months before these events took hold. He writes, "This timeline doesn't match the two factors most people credit with the decline - the Grant's Pass v. Johnson case... and Daniel Lurie taking over as mayor." Instead, he points to a specific federal court ruling in September 2023 that restored the city's ability to clear camps by exploiting a loophole: keeping a token number of shelter beds reserved just long enough to technically offer them, even if most individuals would decline.
This legal maneuvering allowed the administration to bypass the requirement of having ample shelter available before enforcement. Alexander observes that the subsequent Supreme Court decision merely cemented what local judges had already permitted. The result was a rapid removal of tents, but not necessarily a reduction in the population of unhoused people. As he puts it, "Most of the former campers are still homeless. They just don't have tents."
The aesthetic and safety improvements are real. But I think it speaks against the argument, common during the height of the crisis, that there was no tradeoff and actually enforcement was the truly compassionate option.
This distinction is vital. The author argues that the enforcement strategies employed—confiscating tents and possessions—did not convince people to seek help but rather incentivized them to hide. The tradeoff is explicit: cities gained visual order at the cost of making the lives of the unhoused more precarious. Critics might argue that removing blight is a necessary first step to restoring civic function, but Alexander suggests that without a parallel increase in housing supply, this approach merely displaces the problem into the shadows.
The Mechanics of Disappearance
How does a city appear to solve a crisis without solving it? Alexander details the mechanics of this disappearance. He explains that the post-Grants Pass environment gave police new levers of control. If a person has a tent, it can be seized. If they have other belongings, they can be jailed for a day or two, by which time their possessions are often stolen. This creates a powerful incentive for the unhoused to "lie low and avoid trafficked areas."
The author suggests that the perceived drop in homelessness is partly a measurement artifact. When people are forced to hide, they become harder to count. Alexander writes, "Hidden homeless people are harder to count than homeless people living in conspicuous tents. Therefore, the count is lower." This challenges the narrative of a policy success. The decline in the "point in time" counts may reflect a change in behavior driven by fear of enforcement, not a change in housing status.
Furthermore, Alexander identifies a secondary, economic factor: falling rents driven by decreased demand and population outflows in California. He notes, "The primary driver behind the falling rental prices in California is not an increase in housing supply, but rather a decrease in demand." This economic shift likely allowed some individuals to find housing with friends or family, but the scale of this effect is modest compared to the visibility shift caused by enforcement.
The Limits of Local Policy
The commentary also serves as a sharp critique of the political narrative surrounding Mayor Lurie's tenure. Alexander finds little evidence that Lurie's specific policies drove the changes. The mayor's plan to create 1,500 shelter beds was abandoned after only 100 to 200 were built. "He admitted this was too hard and gave up," Alexander writes, characterizing the administration's broader strategy as a rehash of previous mayoral plans.
Similarly, the crackdown on open-air drug markets and fentanyl has yielded mixed, perhaps negative, results. Alexander points out that preliminary data suggests overdose deaths may have increased, potentially because the disruption of established dealer-addict relationships led to more toxic, one-off transactions. "Some experts argue that the clearing of open-air markets shifts the dealer-addict relationship from an iterated game to a one-shot," he explains, leading to exploitative strategies by dealers.
I don't think we learned the sort of thing we hoped we might learn, the lever we could push to solve everything with no downsides.
The author also debunks the popular theory that cities are simply "shipping" their homeless populations to neighbors. While San Francisco has programs to bus individuals to other locations, the numbers are too small to account for the city-wide decline. Alexander notes that if this were the primary driver, neighboring counties like Alameda would show a spike in numbers, which the data does not currently support.
Bottom Line
Scott Alexander's analysis succeeds by stripping away the political spin to reveal a grim, mechanical reality: the visibility of homelessness is highly elastic and can be suppressed through enforcement, but the underlying crisis remains untouched. The strongest part of this argument is the rigorous timeline analysis that decouples the decline in tents from political milestones. Its biggest vulnerability is the reliance on preliminary data regarding overdose rates and the difficulty of proving where the "hidden" homeless population has gone. Readers should watch for the next annual count to see if the numbers rebound once the initial shock of enforcement wears off, or if the decline holds as a new, more hidden normal.