Rent Control

In a vote on Thursday 25th June, the Rent Guidelines Board (RGB)—six of nine members appointed by Mamdani—voted to freeze rents on about one million apartments, or about 27% of NYC housing stock.

This, as well known, was one of Zohran Mamdani’s key policies which helped elect him on the platform of cost of living. Under the previous term of Eric Adams, landlords received about 12% in cumulative rent increases as their net operating incomes rose by around 30%¹—with income growing much faster than the cost increases the rent hikes were meant to offset, undercutting the case that landlords need further increases just to stay afloat.

The single dissenting vote on the RGB came from an Adams appointed economist Arpit Gupta, who argued that a frozen nominal income against rising costs (insurance and fuel to name some) pushes exposed buildings towards disrepair or abandonment rather than heading towards an equilibrium profitable to tenants. His proposed alternative, which I think makes sense, is means-testing rent freezes.² This would allow the vulnerable low-income tenants stability whilst not sacrificing income from renters who are able to pay, lowering the toll on a building’s revenue. Another proposal he made was pairing upzoning with the distressed stabilised stock, allowing infill, additional floors, or rear additions to generate new market-rate units that cross-subsidise the stabilised ones, on a model already being persued. This helps solve the funding gap through supply rather than rent roll alone. The threshold for this imaged scenario of abandonment is yet of course unknown—Mamdani has pledged consecutive years of rent freezes—so it will be interesting to see at what point these supposed negative effects may take place, even possibly after a single year.

A study on the implementation of rent control in San Francisco from 2019 found that landlords newly covered by SF’s 1994 rent control expansion reduced their rental housing supply by 15%, driving a city-wide rent increase of 5.1%.³ Whilst controlled units became cheaper for tenants who already held them, who may very well be able to afford them in any case, the market as a whole tightened for everybody else. Compared to comparable stock between 1994 and 2016, covered properties were eight percentage points more likely to convert to condo or tenancy-in-common.

Rent control also often leads to a mismatch in tenants and the properties they rent—once a tenant has a below-market lease, obviously leaving it becomes very expensive, leading to them continually occupying it even if it doesn’t fit their lifestyle. For example in Cambridge MA, when rent control ended in 1995, freed-up units saw substantial value increases as they were renovated/reallocated to households who desired the size and location available (about $1.8 billion was directly contributed to Cambridge housing stock by 2004).⁴ Rent control could be far more successful if it fairly matched tenants and properties, instead of allowing such inefficient allocation. In NYC for instance, this could lead to empty-nesters in large stabilised apartment in good locations having every incentive so stay regardless of whether downsizing would suit them better, whilst young families queue for market remnants despite suitable stock existing. That is to say, for very low earners rent control can be very lucrative and in fact increase their housing supply, though the previous issues still stand.⁵ It is key to remember that ‘rent control’ is not a single policy but one where design and context both matter.

Vienna is probably the most interesting case of rent control, in that it isn’t only rent regulations helping the market. About 60% of Vienna’s population lives in municipal/subsidised housing—the city is a landlord in its own right, layering the private market compared to simply placing caps on the private market. Even in Vienna though similar distortions show up, with long-term tenants in inherited units paying a fraction of what newcomers have to pay; units are barely means-tested such that the well-off professionals can sit on cheap, well-located property indefinitely. Such is the extent of this that there exists off-contract “Ablöse” payments, sometimes in the range of tens of thousands of euros, that act as a shadow market for outsiders trying to get in to the system. Even this well-funded (Austria spends 0.25% of GDP on social housing)⁶, supply-backed version of the model produces insider/outsider splits, so it can be surmised that a freeze bolted on to much thinner public housing stock, as in NYC, would produce an even sharper split, blocking people from moving to the most productive city on Earth.

Landlord conduct is also worth talking about, since where vacancy decontrol exists, the financial logic pushes grossly towards getting troublesome or long-tenured tenants out, since a vacated unit resets to market rate. Academic work on "rational eviction" in San Francisco, and harassment-and-neglect studies out of Los Angeles, has tracked landlords using eviction, non-renewal, and deferred maintenance as substitutes for the rent increases they want. Due to the natural greed of the landlord, the wider the gap between frozen in-place rent and achievable market rent, the stronger the incentive to manufacture a vacancy by whatever means available.

Modelling commissioned for the Bay Area found that of every feature tested, vacancy decontrol and new-construction exemptions produced the largest effect on whether a rental project stayed financially viable.⁷ NYC's stabilised stock already has some of these features baked-in (pre-1974, six-plus-unit buildings), but a freeze applied uniformly across the entire stock, with no further income-targeting and no differentiation between mixed buildings and fully-stabilised distressed buildings, is quite ill-considered.

The basic redistributive case for some form of rent regulation is sound and not seriously disputed, protecting sitting tenants from extraction in a hyper-commodified market with high switching costs is a big welfare gain, and the NYC data on landlord income outpacing the increases nominally justified by costs is a very good reason to be sceptical of arguments from heavily biased ownership groups. But a blanket, non-means-tested freeze, repeated annually, without supply response, is the worst possible version of this policy. Extending SCRIE/DRIE-style means-tested freezing to all low-income stabilised tenants rather than freezing indiscriminately gets the people who actually need protection the stability they need without starving every building's revenue regardless of who's living in it. Learn Vienna’s lesson that regulation works when it's sitting on top of real supply growth (i.e. a large scale building project) and not substituting for it. Without caveats, a freeze buys current tenants a year or two of stability and probably costs the next decade of tenants a meaningfully worse market to enter. The bit I'm quite interested in is how many consecutive freeze years it takes before the supply response shows up in a market like NYC. Mamdani's multi-year pledge means New York will be an interesting data point in the future.