Editor’s note, September 21, 2026: David Morales won the Democratic primary on September 9, beating Mayor Brett Smiley by just over 1,700 votes, 53 to 46 percent, with 23 percent of registered voters participating. In a city where 55 percent of voters are registered Democrats, he will likely be the next mayor, implementing his vision of rent control after he takes office on January 4, 2027. The author was interviewed on this topic for the Wall Street Journal’s coverage of the race; it did not seem to sway the electorate.
In April 2026, the Providence City Council passed a rent stabilization ordinance by a vote of 9 to 6. Mayor Brett Smiley vetoed it three days later. The override fell short in May. That would normally be the end of the story, except that David Morales, the state representative who championed the ordinance louder than anyone, is now running for mayor on a promise to bring it back and sign it. Polls have found 72 to 74 percent of residents support capping rent increases.
This essay asks the question supporters and opponents mostly talk past: what do people actually do once the cap is in place? Economists call the answers the margins of adjustment.
Everything below is built on the real terms of the 2026 ordinance and actual market facts: 59 percent of Providence households rent, rents are up about 40 percent since 2020 against 15 percent income growth, and a two-bedroom now runs $2,200 to $2,600 depending on whose survey you read. The people are invented, but their situations are drawn from the actual structure of the city’s housing stock: old triple-deckers, small landlords, not nearly enough new construction.
Most rent control arguments are about rent control in the abstract. The Providence ordinance is a specific machine with specific parts, and the parts matter. Here is what the version passed in April does. Terms marked with an asterisk come from the introduced text and Council summaries; the final amended ordinance contains further machinery (a fair-return standard, rehabilitation provisions) that press coverage did not detail and this essay does not model.
Nobody knows precisely, including the city. Testifying in February, Providence’s CFO estimated that fewer than 44 percent of apartments would be rent-controlled, and that was before subtracting the new-construction exemption, which the city cannot compute because it only started tracking new construction three years ago. Here is the arithmetic from public data.
Two consequences follow from the denominator. First, the ordinance is narrower than both its champions and its critics talk as if it were: something like a third to a half of Providence’s rental units would face the cap, concentrated in non-owner-occupied multifamily buildings owned by landlords like Gil, and in larger buildings owned by LLCs. Second, the city is legislating blind, and so is this essay: Providence has no rental registry, no tracking of renewal increases, and three years of construction data. Every coverage number you see in this debate, including the one above, is estimated.
Start with the effect everyone can see, because it is real and it is the reason the policy polls at 72 percent.
Under the cap, Marisol’s rent can rise at most 4 percent a year no matter what Gil would otherwise charge. When his renewal increases would have run above 4 percent, a wedge opens between what she pays and what she would have paid. That wedge is the point of the policy: a transfer from Gil to Marisol, delivered with none of the overhead of a housing voucher. The wedge is smaller than the gap between her rent and a newcomer’s asking rent, because sitting tenants get a discount everywhere, cap or no cap.
Drag the sliders and watch the wedge. If the cap shrinks rental supply, the asking rents newcomers face drift higher; flip on the supply response to see it.
Two things are simultaneously true. First: Marisol is better off. At the recent pace of Providence rent increases the cap saves her twenty-odd thousand dollars over a decade, counting only what the cap causes, and she is exactly who the ordinance’s sponsors have in mind. Council President Rachel Miller’s line, that a rent increase is “an eviction notice by another name,” describes a real thing that happens to real families. The best research agrees on this point: in San Francisco, covered tenants were about 20 percent more likely to stay at their address.
Second: Gil is not a corporation. He’s a man holding an asset whose income is capped while his costs (insurance, up 30 percent in three years in coastal New England; taxes; a roof) are not. The transfer doesn’t come from “landlords” as a class of abstraction. It comes from Gil’s pension, one month at a time. What Gil does about that is Chapter 3.
The wedge also creates a new problem. That gap between $1,450 and the market price is now attached to a physical apartment. Whenever it empties, someone gets a prize worth thousands of dollars a year. Who?
Prices do two jobs. They transfer money, which everyone notices, and they allocate scarce things, which almost nobody does. Cap the price below the market rate and demand exceeds supply. The apartment still gets allocated. Just not by price.
Under market pricing, the unit goes to whoever values it most, and the people who value it most are the ones with the fewest alternatives: Tenaya splits the rent with a roommate, Marcus offers prepaid months because his record fails every screen in town. Under the cap, the price is fixed, so the competition moves entirely to paperwork: credit history, income multiple, references. The landlord, who bears all the downside of a bad tenancy at a rent he can never reset (remember: 4 percent at turnover, once), picks the safest file every time, and the safest file is never a thin one.
A ceiling changes the currency of competition from money, which newcomers can offer, to accumulated paperwork, which they cannot.
Meanwhile a subtler misallocation compounds. Marisol’s neighbors, the Hendersons, raised three kids in their capped three-bedroom. The kids are gone. In a market, the Hendersons would downsize and free the 3BR for a family; under the cap, leaving means surrendering a below-market deal they can never get back, so they stay. Glaeser and Luttmer estimated that about a fifth of New York’s controlled apartments are occupied by the “wrong” household in exactly this sense. The cap pays people to sit still, on housing they’ve outgrown and on jobs they’d otherwise leave: Marisol herself will turn down a better-paying job in Worcester in 2029, because the raise is smaller than the deal she’d forfeit by moving.
Nobody repeals the housing supply by decree. It leaks away, one legal, individually reasonable decision at a time, along margins most people never think to watch. Below is one block in Elmwood with 10 lots. Play the decade under each scenario, or drag the year slider. Click any building for its story.
1. The maintenance margin. Gil can’t raise the rent, but nobody can make him renew the kitchen. A unit is a bundle of rent and upkeep; when the law fixes one, the other adjusts. Massachusetts rent-controlled units were measurably older, shabbier, and less renovated than their neighbors (Sims 2007).
2. The exit margin: condo conversion. A triple-decker is worth more as three condos sold to owner-occupants than as a capped income stream. In San Francisco, buildings subject to rent control saw a 15 percent decline in rental supply, much of it through exactly this door. Gil sells; three renter households’ units become two owner households’ homes; the block gets richer and the rental stock gets smaller.
3. The exemption margin: owner move-in. Eddie owns the two-family at number 20 and lives in Cranston. The day the cap takes effect, the cheapest renovation available to him is a mattress: move into unit 1, and the building is owner-occupied, 1–4 units, exempt. One rental unit exits the market so that the other can leave the cap.
4. The platform margin. Karen inherits her mother’s triple-decker at number 22. She can become a regulated landlord with a Rent Board file, or she can furnish two units and list them nightly to tourists and traveling nurses, entirely outside the ordinance.
5. The never-built margin. Dev and Priya had an accepted offer on the gut-job three-family at number 24: renovate, rent two units, live in neither. Under the cap their spreadsheet dies (capped income, uncapped renovation costs, no rent reset, ever), so they walk. The house sells to a flipper who makes condos. Three rental units that would have existed, never do. You will not see this on any eviction docket or Rent Board complaint.
6. The generosity margin. Walt, at number 18, has charged his longtime tenants $300 under market for years, because they’re good tenants and he’s a decent man. The 180-day lookback means his kindness is about to become a legal ceiling, permanently. Every landlord in Providence who reads the ordinance learns the same lesson at the same moment: never price below the maximum again, and get your increase in before the base-rent date. Cities that pass rent control reliably see a burst of pre-effective-date increases.
7. The unbundling margin. The 12-unit building on the corner, owned by a Boston LLC, stays rental. But parking is now $150 a month, laundry is app-metered, there’s a $95 “amenity and administration fee,” and the heat included in the lease becomes tenant-paid utilities at renewal. Rent is capped; the price of renting is not.
8. The screening margin. Same building: with a queue at every vacancy and no way to price risk, management tightens to 700+ credit, 3.5× income, and documented rental history. Chapter 2, institutionalized.
9. The spillover margin. Rosa and Manny live in their three-family at number 16, so they’re exempt. They didn’t ask for pricing power, but as covered units leak out of the market, the demand lands on whatever remains. Exempt-sector rents rise faster than they would have. The cap doesn’t lower the market price; by shrinking supply it raises it, about 5 percent in the San Francisco study, which is how tenants as a whole can end up paying more in a city that capped the rent.
10. The stayers. And in Marisol’s building at number 26, the cap does exactly what it promised: she’s still there in 2036, kids through school, savings account no longer empty. Any account of the block keeps her on the ledger too. The question the block poses is how many Tenayas, standing outside a smaller housing stock, paid for Marisol’s decade.
The new-construction exemption (10 years, or 20 for projects built with prevailing wages and apprentices) is supposed to hold developers harmless. It only half works: buildings live for 60 years, and spreadsheets can see past 20.
Notice three things. First, the asymmetry: at the ordinance’s actual terms the cap shaves half a point to a point off the return. That sounds survivable, and for a project that penciled at 9 percent, it is. But development is decided at the margin. The marginal project is the one that penciled at 8.2 against a 7.5 hurdle, and for that one, half a point is the whole ballgame. It dies in a Tuesday investment-committee meeting in Boston, and its 48 units are never missed because they never exist; Cranston and Pawtucket, same workers, same rents, no cap, get the building instead. Second, the labor-linked exemption is a real option with a real price: at a typical prevailing-wage premium, the extra ten years of exemption don’t pay for the extra construction cost in this model, so the 20-year track functions more as a labor standard than as developer relief. Third, the model is one developer’s spreadsheet, disclosed so you can disagree with it slider by slider.
You don’t have to take the model’s word for it. St. Paul ran the experiment.
Rent control is among the most-studied policies in urban economics, and the findings are unusually consistent: real protection for incumbents, paid for in supply, quality, and misallocation.
Note what the literature does not say. It does not say rents fall for everyone (they rise for newcomers). It does not say landlords are impoverished as a class (they adapt, exit, or convert). And it does not say incumbent tenants aren’t helped (they demonstrably are). The economist’s objection is that the help is financed by an invisible tax on the city’s future tenants. In San Francisco the accounting was almost exactly symmetric: $2.9 billion in gains to covered tenants between 1995 and 2012, and $2.9 billion in costs to the renters who came after them, paid through higher market rents on a smaller stock.
Price controls don’t create or destroy much value on day one; they reassign it, and the reassignment drifts over time. Green is gains and red is losses, both relative to the no-rent-control status quo. Hover any cell.
Anyone can tell a story; the discipline is attaching probabilities you’d be embarrassed to miss. Each outcome below gets two probabilities: one if Providence enacts the ordinance (say, in 2027 under a Morales administration), one under the status quo. The difference between the bars is the policy’s fingerprint.
The costs above shrink if: the cap rarely binds (market rent growth stays under 4 percent, which is exactly what happened in post-2023 St. Paul, where soft demand, not the ordinance, did the “stabilizing”); the Rent Board runs its pass-throughs generously; or enforcement is light. And the strongest steelman for the ordinance is distributional and temporal: the gains land now, on identifiable struggling households, while the costs land later, diffusely, on people who mostly don’t live here yet. A city is allowed to make that trade with open eyes.
If the goal is more people affordably housed, experience points to a boring answer: legalize more housing (Providence’s stock barely grows), subsidize the poorest directly, and let prices keep telling builders where homes are worth building. The 4 percent cap does something different: it makes the existing shortage more comfortable for the people already housed, while slowly shrinking the stock of homes available to everyone who comes after.