The coverage dilemma
Every rent control policy has to answer a fundamental question: what does it cover? The answer determines which of two bad outcomes you get.
Universal coverage
Controls apply to all or nearly all private rented properties. Every landlord faces constrained returns. The financially marginal ones sell up. The rental market shrinks. Fewer homes, same demand — and the market tightens further with each passing year as the pool of available homes contracts.
Partial coverage
Controls apply to some properties but not others. The rent-controlled homes stay controlled — a fixed pool of lucky tenants paying below-market rents. But all the unmet demand that cannot access controlled homes concentrates into the uncontrolled segment. Rents spike there. The city divides into big winners and big losers, and the gap only grows over time as the two markets diverge further.
This is not a theoretical concern. It is what has happened every time rent controls have been tried in major cities, and the pattern repeats regardless of the design of the policy.
The zero-sum reality
The clearest illustration of the dilemma comes from San Francisco. When strict rent control was extended to small multi-family buildings in 1994, existing tenants in those buildings saved $2.9 billion in rent over the following sixteen years. But renters who moved to San Francisco after controls were introduced paid an extra $2.9 billion over the same period — exactly what existing tenants saved.
Rent control did not reduce the total cost of renting in San Francisco. It redistributed it: from future renters to current ones, from those who came later to those who happened to be in place when controls were introduced. This shows how rent control does not help all renters — only a few. Anyone who needs to move loses out.
And that is the hidden trap. Under rent control, anyone who ever needs to move — for a new job, for more space, to leave a relationship — faces the full force of the uncontrolled market. Their controlled rent is not portable. Once they leave, it is gone. The result is that tenants become trapped: staying in homes that no longer suit them, because the cost of leaving is prohibitive. Rent control does not give renters security. It gives them a cage with a low rent.
Rent control changes the system for rationing scarce homes from price to luck — specifically, the luck of who happened to be renting when controls were introduced. It does not make housing more available. It makes access to good housing a matter of tenure luck rather than need, freezing in place whoever was there first and penalising everyone who comes after.
The knock-on effects nobody talks about
When people cannot move without losing their home, the consequences extend well beyond the housing market. Workers cannot relocate for better jobs, or to the places where their skills are most needed — slowing economic mobility and productivity across the city. Families living in homes that no longer fit them — a couple who cannot afford to upsize when they have children, an older person rattling around in a large flat they would willingly exchange for something smaller — cannot adapt their housing to their lives.
In more serious cases, the inability to leave a rent-controlled home can make it harder for people to leave dangerous situations. Someone in an abusive relationship who knows that leaving means re-entering an unaffordable open market faces a higher barrier to getting out. This is not a hypothetical. It is a direct consequence of a system that makes the cost of moving prohibitively high.
These are the costs that never appear in the studies measuring whether rents fell in controlled units. They are real, and they fall on the people least able to absorb them.
What happened in the cities that tried it
New York prioritised rent control over housebuilding for decades. The result: rent-controlled luxury apartments in elite Manhattan neighbourhoods have rented for decades at under £100 a month, while average one-bedroom apartments without rent control in working-class parts of Brooklyn cost over £1,300 a month. The policy created enormous benefits for long-term incumbent tenants — and severe shortages for everyone trying to enter the market. New York City has lost 152,000 units of rent-stabilised housing stock since 1993 as landlords converted to condominiums.
Stockholm has had near-universal rent control since the 1940s. The result — with a near-zero vacancy rate — is a queue, not a market: residents wait decades — sometimes their entire working lives — for a rent-controlled flat. The black market for tenancy transfers is large and well-documented. New arrivals to the city have no realistic prospect of accessing the controlled sector and must rely on a small, expensive uncontrolled segment.
Berlin introduced a rent brake in 2015 limiting increases to 10% above local averages, then a full rent freeze in 2020. The freeze was struck down by Germany's constitutional court in 2021 after 18 months. During that period, rents in the controlled sector fell modestly — but rents in the uncontrolled sector, where new builds and some exempt properties sat, spiked significantly. The policy did not achieve its stated goal of stopping gentrification. When the freeze was lifted, rents in the previously controlled sector rose sharply.
The most rigorous modern study of rent control, using a natural experiment from San Francisco's 1994 reform. Rent control reduced rental housing supply by 15% as landlords converted properties to condominiums or sought planning permission to redevelop sites. The policy protected sitting tenants but transferred wealth away from future renters, who faced a smaller and more expensive market. Long-run rents rose citywide. Read the paper →
The most comprehensive cross-national study of rent regulation, covering over a century of evidence across 16 countries. Stricter rent regulation consistently reduces rental supply in every context studied. Harder caps produce larger and faster landlord exits. The effect is present across wartime and peacetime, booms and downturns, different welfare state models and planning systems. Read the paper →
When Massachusetts lifted rent controls in 1995, rental housing supply increased significantly — confirming that controls had been suppressing it. This natural experiment runs the evidence in reverse: removing controls produced the increase in supply that economic theory predicts controls would reduce. Read the paper →
Where we stand
Rent controls are bad policy. The cities that have made housing affordable — Vienna, Helsinki, Auckland, Minneapolis — did so by building significantly more homes. That is what Glasgow needs — not a system that entrenches the luck of who happened to be renting when controls were introduced, or lengthens the waiting list for everyone who comes after.
Want to model how many rental homes would leave Glasgow's market under different rent control scenarios? Adjust the cap level, years in force, and coverage — with the evidence behind every number.
Open the calculator →Wealthier renters looking to buy
As landlords sell up, more properties enter the sales market. The people best placed to take advantage are those with savings or mortgage access. Rent controls quietly transfer housing wealth upwards.
Sitting tenants
Renters already in a controlled property — incumbent tenants — pay below the market rate in the short run. But the protection comes with a hidden cost: moving means losing the controlled rent, so tenants become trapped in homes that may no longer suit them. And if the landlord eventually sells, they re-enter a market that has shrunk around them.
Renters in the cheapest homes
The properties most likely to leave the market are those at the bottom: older stock with the thinnest yields. When they go, the cheapest options disappear first — the exact opposite of the intended effect.
New renters entering the market
Young people, those moving for work, and anyone leaving temporary accommodation face a smaller market with less choice and higher asking rents on uncontrolled properties. These are the people rent control was supposed to help most.