Rent control produces one benefit that is difficult to dispute.
If you already live in a protected apartment, it can keep your rent well below what the open market would charge. That’s valuable. For an older tenant on a fixed income, it can be the difference between staying in a neighborhood and leaving it.
Now ask what happens next.
The apartment that no longer fits
Imagine a couple who rented a three-bedroom apartment when their children lived at home. Twenty years later the children are gone. The couple would happily move into something smaller.
Except the three-bedroom unit is rent controlled, and the one-bedroom apartment available today costs more.
So the economically rational decision is to remain in housing they no longer need. In California that choice is reinforced by law: under the 1995 Costa-Hawkins Act, most controlled rents reset to market when a tenant moves out. Leaving means giving up the discount for good.
Housing hasn’t disappeared. It has become mismatched. Run your own numbers in the lock-in calculator.
What the best study found
The most cited modern evidence comes from San Francisco. In 1994, voters extended rent control to small apartment buildings (two to four units) built before 1980. Economists Rebecca Diamond, Tim McQuade and Franklin Qian compared tenants and landlords in those newly covered buildings with nearly identical buildings built a few years later, which stayed uncontrolled. Their results were published in the American Economic Review in 2019.
| Finding | Result |
|---|---|
| Tenants staying at their address | Roughly 10–20% more likely to stay, with the largest effects for older, long-time tenants |
| Rental supply in covered buildings | Down about 15% relative to comparable buildings |
| How landlords reduced supply | Condo and tenancy-in-common conversions, sales to owner-occupants, redevelopment |
| Benefit to covered tenants | About $2,300–$6,600 per person per year (1995–2012) |
| Citywide rents (model estimate) | About 5.1% higher because of the lost supply |
The 15% is a building-level effect for the covered buildings, not a 15% drop in San Francisco’s entire rental stock. The 5.1% citywide figure comes from the authors’ model in the 2018 working-paper version.
The important point is not that rent control provides no benefit. The researchers found substantial benefits for incumbent tenants. The question is who bears the offsetting cost. In the authors’ model, 42% of the welfare cost of higher citywide rents fell on people who did not yet live in San Francisco.
Follow the chain
If existing tenants stay longer, fewer units turn over.
If landlords remove properties from the rental inventory, supply falls.
If constructing additional housing is already slow and expensive, the shortage compounds.
San Francisco is not the only data point. When Cambridge, Massachusetts, ended rent control in 1995, property values rose substantially, and more than half of the roughly $1.8 billion gain came from spillovers onto neighboring properties that had never been controlled. After St. Paul, Minnesota, passed one of the country’s strictest rent-control measures in 2021, property values fell about 6–7%, and the tenants who gained most tended to have higher incomes.
What about neglected buildings?
Walk through some of the most valuable neighborhoods in Los Angeles and Beverly Hills and you will occasionally see rental buildings that look surprisingly tired for their zip code. That’s an observation, not proof of cause. Plenty of things determine how a building is maintained.
The evidence here is genuinely mixed. Researchers studying Cambridge concluded that rent control muted landlords’ incentive to invest. But in San Francisco, the dominant landlord response wasn’t neglect. It was the opposite: renovating, converting and redeveloping to get buildings out of the controlled stock.
So the useful question to test is this:
What happens when the market value of improving an apartment and the regulated return from renting that apartment move further and further apart?
Sometimes the answer is deferred maintenance. Sometimes it is a condo conversion. Either way, the rental supply the policy was meant to protect gets smaller.
Two groups, two answers
Rent control is easiest to understand if we separate two groups: the tenant who already has the apartment, and the person trying to find one.
A policy can help the first while making life harder for the second. Both facts can be true. A 2024 review of the empirical literature reached essentially that conclusion: rent controls are quite effective at slowing rent growth for covered units, and they come with a wide range of adverse effects elsewhere.
That doesn’t settle the policy question. Stability has real value, and some cities may decide it is worth the cost. But the honest version of the debate starts by naming who pays.
The Zesht box
Follow the incentive
- What behavior does this reward?
- Staying in the same unit for a long time, even after it stops fitting the household. For landlords: getting tenants to leave, or exiting the rental business through conversion or redevelopment.
- What does it discourage?
- Moving for a job, a better-sized home or a better location. Keeping older buildings in the controlled rental stock.
- What becomes more expensive?
- Uncontrolled and newly vacant rentals. In the San Francisco authors' model, citywide rents rose about 5.1%. Moving becomes costly for anyone holding a controlled lease.
- What becomes scarcer?
- Vacant, rentable units in older buildings. In San Francisco, the number of renters in controlled units in covered buildings fell 25–30% relative to 1994.
- Who receives the benefit?
- Sitting tenants, especially older and long-tenured households.
- Who ultimately pays?
- Owners of covered buildings, through lower returns and property values, and future renters, who face a smaller, pricier pool of available apartments.
- What happens when people have another option?
- Landlords with an exit (condo conversion, owner move-in, redevelopment, buyouts) use it. Tenants with a controlled unit rarely give it up. Searchers turn to the uncontrolled market.
Goal → mechanism → result
The scorecard
- 1The goal
- Protect renters from sharp rent increases and displacement, and give households stability in neighborhoods where they have built their lives.
- 2The mechanism
- A cap on annual rent increases for covered units. In California, rents on most controlled units reset to market when a tenant moves out (vacancy decontrol under the 1995 Costa-Hawkins Act).
- 3The incentive
- Tenants are rewarded for staying put. Landlords are rewarded for turnover, conversion to ownership housing or redevelopment into exempt buildings.
- 4The observed result
- In San Francisco's 1994 expansion, covered tenants were 10–20% more likely to stay at their address, and landlords of covered buildings cut rental supply by about 15%.
- 5The unintended consequences
- Less rental supply, higher market rents in the authors' model, apartments mismatched to the households living in them, and buyouts or owner move-ins aimed at newer tenants in fast-rising areas.
- 6What supporters say
- Renters cannot otherwise insure against rent spikes. Stability has real value, especially for older and long-time tenants, and the measured benefits to covered tenants are large.
- 7What critics say
- Costs are shifted to future renters and to owners, supply falls, and benefits are poorly targeted; in St. Paul, Minnesota, the tenants who gained most tended to have higher incomes.
- 8What the evidence actually shows
- Well-identified studies in San Francisco, Cambridge (Massachusetts) and St. Paul consistently find real benefits for incumbents alongside reduced supply or lower property values. The size of the effects depends heavily on how the law is designed.
- 9What remains uncertain
- How well these results transfer to looser statewide caps that exempt newer buildings and allow larger increases, the long-run effect on construction, and the size of citywide rent effects outside model estimates.
Questions people ask
Does rent control reduce housing supply?
In San Francisco's 1994 expansion, landlords of newly covered buildings reduced rental housing supply by about 15% relative to similar uncovered buildings, mainly by converting units to condos or tenancies-in-common and redeveloping (Diamond, McQuade and Qian, American Economic Review, 2019). The size of the effect elsewhere depends on how a law treats new construction, vacancies and conversions.
Does rent control help tenants?
Yes, it helps tenants who already live in covered units. In San Francisco the benefit averaged roughly $2,300 to $6,600 per covered person per year between 1995 and 2012, and covered tenants were 10–20% more likely to stay at their address.
Does rent control raise rents for everyone else?
The San Francisco researchers' model estimates that the lost rental supply raised citywide rents by about 5.1%. That figure is a model estimate rather than a direct measurement; the published paper says the lost supply likely drove up market rents in the long run.
Why don't people leave rent-controlled apartments?
In California, vacancy decontrol under the 1995 Costa-Hawkins Act lets landlords reset most controlled rents to market when a tenant moves out. A tenant who leaves gives up the discount permanently, which creates a strong incentive to stay even when the apartment no longer fits.
Sources
- The Effects of Rent Control Expansion on Tenants, Landlords, and Inequality: Evidence from San Francisco
- The Effects of Rent Control Expansion on Tenants, Landlords, and Inequality (Working Paper 24181)
- What does economic evidence tell us about the effects of rent control?
- Housing Market Spillovers: Evidence from the End of Rent Control in Cambridge, Massachusetts
- Robbing Peter to Pay Paul? The Redistribution of Wealth Caused by Rent Control
- Rent control effects through the lens of empirical research: An almost complete review of the literature
- Analysis of initiative to repeal the Costa-Hawkins Rental Housing Act
Cite or link to this article
See something wrong? Zesht changes the theory when the evidence changes. Send a correction · Our standards