Rent Control and What the Economists Actually Found in San Francisco

In November 1994, San Francisco voters approved a ballot measure called Proposition I. It was not glamorous. It closed a loophole in the city's 1979 rent ordinance, which had exempted small buildings of two to four units where the owner lived on site. Under Prop I, tenants in those older small buildings, built before 1980, would now have their rent increases capped like everyone else's. Thousands of renters, many in the kind of Victorian flats that fill postcards of the city, gained protection overnight. Two decades later, three economists at Stanford realized that this small, specific vote had created something rare in social science: a natural experiment. Buildings on one side of a date line were covered; nearly identical buildings on the other side were not. What they found has been quoted by both sides of the rent control debate ever since, usually only half of it.

In short: A Stanford study of San Francisco's 1994 expansion found that rent control did what tenants wanted: covered renters were much more likely to stay in their homes and neighborhoods. It also found that landlords responded by converting or redeveloping covered buildings, cutting the rental supply and pushing up rents for everyone else. Here is the argument: rent stabilization is a real protection worth having, and it has to be paired with building far more homes.

A ballot measure that became an experiment

To understand why economists were so excited, consider how hard it usually is to study rent control. Cities that adopt it tend to be cities with hot housing markets, so comparing them with cities that do not is comparing apples to pears. Even within a city, rent-controlled buildings are often older and different from newer ones.

Prop I created a cleaner comparison. It applied to small multifamily buildings built before 1980. Small buildings built in 1980 or after were not covered. A two-unit building finished in 1979 and one finished in 1980 are about as similar as buildings get, and only one got rent control. By tracking tenants and properties on both sides of that line over many years, researchers could estimate what rent control itself did.

The economists were Rebecca Diamond, Tim McQuade and Franklin Qian. They used a data set that followed individuals' addresses over time, along with records on the buildings themselves. Their paper was published in the American Economic Review in 2019, one of the most prestigious journals in the field, and it remains one of the most cited studies of rent control.

What the study found for tenants

Start with the finding that rent control's supporters cite. Tenants in buildings that gained rent control were substantially more likely to remain at their address years later than similar tenants in uncovered buildings. They were also more likely to stay in San Francisco at all. The effects were largest for older tenants and for people who had lived in their homes longer.

That matters more than it might sound. Staying put means a child does not change schools mid-year. It means a grandmother keeps her doctor, her church, her bus route and the neighbors who check on her. It means a worker keeps a short commute. Displacement has costs that do not show up in rent data: lost social ties, disrupted schooling, longer commutes and the stress of moving. The study found that rent control prevented a lot of that, and the authors treated it as a real benefit worth measuring.

The researchers also noted that this protection functioned as a kind of insurance. Rent control does not lower rents so much as it limits how fast they rise for a tenant who stays. In a city where market rents can jump sharply, that stability is valuable, especially to people who cannot easily absorb a sudden increase.

What the study found for landlords and the city

Now the finding that rent control's critics cite. Landlords of newly covered buildings responded by reducing the amount of rental housing they offered. Some converted units to condominiums or to a form of shared ownership called tenancies in common. Some redeveloped or remodeled buildings in ways that took them out of rent control. Some moved in themselves or sold to owner-occupants. By the authors' estimate, the rental housing supply in covered buildings fell by about 15 percent compared with uncovered ones.

Because those buildings left the rental market, the remaining rentals across San Francisco became scarcer, and the authors estimated that citywide rents rose by about 5 percent as a result. In other words, the policy helped the tenants who had it partly at the expense of future renters who did not. And the condos and redeveloped units tended to house higher-income residents, which the authors argued pushed the city further toward gentrification, the very trend rent control was meant to slow.

Put the two findings together, and you have a policy that protected incumbents and made the market tighter for newcomers. That is a genuine trade-off, and pretending one half of it does not exist does not make it go away.

How each side reads the same paper

Opponents of rent control read the study as confirmation of what economists have long warned: price controls reduce supply. They point out that surveys of economists have consistently found broad skepticism of rent control, and they argue that the right answer to high rents is to build more housing, full stop.

Many tenant advocates read it differently. They point out that the landlord responses the study measured, condo conversions, owner move-ins, redevelopment, are things a city can regulate. San Francisco and other cities have limits on condo conversions and evictions, and the effects might have been smaller under stronger rules. They also stress that the authors themselves did not conclude that rent control should be abolished. The paper argues that the insurance it gives tenants is valuable and suggests that a better design might pay for that insurance through public funds, such as subsidies or tax credits, rather than placing the cost on individual landlords, which gives them a reason to exit the rental business.

Both sides are reading the study accurately. They are weighting the findings according to what they care about. The useful move is to stop arguing about whether rent control has costs and benefits, since it plainly has both, and to start designing policy that keeps the benefits while shrinking the costs.

The state laws that shape the fight

San Francisco does not write its rent rules alone. Two California state laws set the boundaries.

The Ellis Act, passed in 1985, lets landlords leave the rental business entirely by withdrawing all units in a building, a process that has been used to clear tenants from rent-controlled buildings before they are converted or sold. Tenant groups have long sought to restrict it; landlord groups defend it as a basic right not to be forced to remain a landlord.

The Costa-Hawkins Rental Housing Act, passed in 1995, blocks California cities from applying rent control to housing built after 1995, or after an earlier local cutoff, and to most single-family homes and condos. It also bans vacancy control, so when a tenant leaves, the landlord can reset the rent to market. California voters have been asked three times to repeal or roll back Costa-Hawkins, in 2018, 2020 and 2024, and have said no each time.

In 2019 the state took a different approach with the Tenant Protection Act, known as AB 1482, which capped yearly rent increases on many older buildings statewide at 5 percent plus inflation, up to a maximum of 10 percent, and required just cause for many evictions. The same year Oregon passed the first statewide rent stabilization law in the country, with a cap tied to inflation and new buildings exempt for their first fifteen years. Both laws reflect a design lesson that economists and advocates increasingly share: exempt new construction so the cap does not discourage building, and set the limit high enough to rule out gouging without freezing rents.

Where progressives have made this harder

Here is the uncomfortable part for the left. San Francisco's housing crisis was not caused by rent control. It was caused mostly by the city and region adding far more jobs than homes for decades, under some of the most restrictive zoning and approval processes in the country. Many of the same activists who fought for tenant protections also fought market-rate housing projects in their neighborhoods, sometimes successfully. Every blocked building added to the scarcity that made rents spike and made rent control both more necessary and more costly.

That is not an argument against tenant protections. It is an argument that tenant protections and housing production are partners, not rivals. A city that limits rent increases but forbids new homes will have a small group of protected tenants and a growing crowd of people locked out. A city that builds aggressively and protects existing tenants from sudden shocks can give stability to the people who live there now without making the city closed to the people who want to come.

A better deal for tenants

What would a policy look like that took the Stanford findings seriously from a progressive perspective? Here is one version of the case:

1. Stabilize, do not freeze. Caps tied to inflation, with a ceiling, protect tenants from gouging without making landlords want to exit.
2. Exempt new buildings for a period of years. That keeps the incentive to build, and the buildings eventually join the protected stock.
3. Close the exits. Strong limits on condo conversions and owner move-in evictions, and fair relocation payments when tenants must leave, reduce the supply loss the study found.
4. Spread the cost. Public money for rental assistance and social housing shares the burden more broadly than asking individual small landlords to carry it.
5. Build, build, build. None of this works without many more homes of every kind.

There are good-faith people across the spectrum who would disagree with parts of that list. The point is to argue about design, not slogans.

The tenants behind the coefficients

Economics papers speak in percentages, but the people in this study were real. Many of the tenants protected by the 1994 vote were working-class and middle-class San Franciscans: bus drivers, nurses, artists, restaurant workers, retirees who had lived in the same flat in the Mission or the Richmond for decades. When the city's tech booms drove market rents sharply higher, rent control was often the only reason they could stay.

The study found that the protection mattered most for older tenants and long-term residents. For a retiree on a fixed income, the difference between a capped increase and a market increase can be the difference between staying in a neighborhood she has known for forty years and leaving the city entirely. Economists measure that loss as a change of address. The people living it experience the loss of a church, a doctor, a corner store and the friends who notice when they are sick.

That is why tenant advocates bristle when rent control is dismissed as simple economic illiteracy. The benefit is real and concentrated in people with little power in the market. The question is not whether it is worth protecting them. It is how to protect them without making the city harder for the next nurse who needs an apartment.

Who owns the small buildings

One detail in the San Francisco story is easy to miss. The buildings Prop I covered were small, two to four units, and many were owned by individuals rather than large companies. Some owners lived in one unit and rented out the others. For these small landlords, a cap on rent increases could squeeze a thin margin, especially when property taxes, insurance and repair costs climbed.

That helps explain the supply response the study found. A small owner facing years of capped rents had a strong reason to sell to someone who would live in the building, convert units to ownership or redevelop. Large corporate landlords with diversified holdings might absorb the same rules more easily.

This is one reason some housing economists who are skeptical of traditional rent control favor public alternatives, such as direct rental assistance or tax credits that compensate owners for keeping rents stable. Progressive advocates often answer that public funding has never been large enough, and that tenants cannot wait for Congress. Both points are fair. A serious agenda can pursue tenant protections now while fighting for the public money that would make them less costly to the small owners who carry much of the burden today.

What you can do

• Learn your local rules. Rent stabilization, just-cause eviction and condo conversion rules vary by state and city. Your city's rent board or housing department can tell you what applies to you. This is general information, not legal advice; a tenant union or legal aid office can help with a specific situation, and the organizations directory lists housing and legal aid groups in every state.
• Follow the money in your budget. Housing is the biggest expense for most renters. The inequality page puts real wages and the minimum wage next to the wealth shares of the top and bottom, which helps explain why rents rise faster than paychecks.
• Track the wider cost-of-living fight. Rent control, zoning and wages are all part of the same debate about whether ordinary people can afford to live where they work. Follow it on the affordability topic page.
• Call your state legislators. Many of the most important rules, like Costa-Hawkins in California, are set in the state capitol rather than city hall. Find your state legislators by address and tell them you want tenant protections and more homes together.