Go to any booming urban center these days, whether in Boston or Washington or Atlanta, and one thing is certain: Rental costs are frighteningly high.

In San Francisco, perhaps the poster city for housing sticker shock, the average one-bedroom rents for about $3,300 a month. That’s an increase of about 50% since 2011.

Can rent control be a solution?

Rent control has a bad reputation among many economists, who often argue that it makes things worse. If landlords can’t charge market-level rents, goes the theory, they will gradually take their properties out of the rental market by converting them to condos or Airbnb rentals. That can create acute shortages in rental housing, which can drive prices higher than they would have been otherwise.

But a new study of San Francisco, by Stanford Graduate School of Business professors Rebecca Diamond and Tim McQuade, and Stanford economics PhD student Franklin Qian, finds that rent control creates both winners and losers — even among renters themselves.

When San Francisco expanded rent control in 1994, it provided huge savings to many tenants — especially older ones — who got in at the start and stayed put. But it created almost equally big losses in the form of higher rents for tenants who came later and may have accelerated gentrification.

To Diamond, the study validates much of the skepticism about rent control. But she argues that the surprisingly big benefits suggest that cities should look for alternative forms of protecting renters — such as tax credits or subsidies that offset soaring rents.

“Right now, the government already provides lots of different insurance — unemployment insurance, disability insurance, and insurance to help securitize mortgages,” Diamond notes. “But we don’t provide any formal insurance against rent increases.”

Tracking Tenants

Diamond says there is surprisingly little precise data about exactly what happens after a city imposes or expands rent control. Who are the winners and losers? How do landlords respond? What happens to the buildings covered by rent control? How do the benefits stack up against the costs?

Those are urgent issues for many local governments. Seven cities in the Bay Area have some form of rent control; five additional Bay Area cities placed rent control measures on the 2016 ballot, with two passing. State lawmakers in Illinois, Oregon, and California are considering laws that would give cities more freedom to impose rent controls.

To get at some answers, the researchers tracked exactly what happened to tenants, buildings, and prices after San Francisco expanded rent control in 1994.

The analysis draws on new data from a consumer identity management firm called Infutor, which provides rich historical property and household data throughout the United States. These detailed data provided the researchers a unique lens to measure how many San Francisco residents lived in rent-controlled units, how long they stayed in them, and how much money they saved in below-market rents.

Using municipal records linked to the Infutor data, the researchers were also able to analyze what happened to the supply of rental housing. They could tell how many landlords did indeed convert their properties into owner-occupied condos or other forms of housing that didn’t fall under rent control. Many landlords paid their rent-protected tenants to leave, and the average buyout price in San Francisco was more than $30,000. The researchers also took note of new construction and the total supply of rental units.

The big winners were people who were lucky enough to be under rent control back in 1994. Those renters saved between $2,300 and $6,600 a year — a total of $2.9 billion in benefits from 1994 through 2010. Many of those beneficiaries were older people who had deep roots in their neighborhoods.

By contrast, people who came later faced higher upfront rents and a growing scarcity of rental housing. The number of rent-controlled units declined by 25% between 1994 and 2010, and the total stock of rental housing (some of which isn’t under rent control) declined about 5%.

Overall, the researchers estimated, the renters who came later paid about 5% more than they would have in the absence of rent control. Coincidentally, the total of those additional costs through 2010 was $2.9 billion — essentially the same as the savings reaped by other tenants.

One side effect of all this, Diamond says, is that San Francisco became even more skewed toward upscale housing. Since new construction is exempt from rent control, some landlords chose to demolish their old rent-controlled housing and build new high-end apartments in its place. Likewise, converting rental units to owner-occupied condos inherently shifted the mix of residents toward people with more money.

Diamond argues there may well be a better way. A city could protect renters by offering subsidies or tax credits to offset at least some of their rent increases. If landlords don’t have to absorb the entire cost of rent control, she suggests, they won’t be under as much pressure to pull out of the rental market.

“If there’s a way to give tenants some protection against rent increases, without having to make the landlords pay for it, we might be able to have the benefits of rent control without creating the distortions that make things worse for others.”