Why New York’s Apartments Sit Empty

In 2023, New York City's Housing and Vacancy Survey counted just 33,210 apartments available for rent. A year later, state registration records listed 47,720 vacant rent-stabilized apartments. The apparent contradiction raises an obvious question: why are so many apartments recorded as empty in a city facing an acute housing shortage? The answer is more complicated than the numbers suggest. The two figures measure different types of vacancy, and neither tells us how many apartments are being deliberately kept off the market. By 2025, statewide registered vacancies had risen to roughly 59,000, about 95% of them in New York City (NYC HPD, 2024; NYC RGB, 2025; NYC Comptroller, 2026).

There is a credible economic reason some units stay empty. A landlord facing an expensive renovation on an apartment with a low regulated rent may struggle to recover the upfront cost. Rising insurance, maintenance and financing expenses can make that calculation harder. But the scale matters. In 2023, the Comptroller counted 26,310 stabilized apartments vacant but unavailable for rent for all reasons and estimated that fewer than 2,000 low-rent units were held off the market because their owners could not make necessary repairs (NYC Comptroller, 2024).

The question is not whether rent stabilization has costs. It is whether those costs explain a meaningful share of empty apartments, and which policy would bring those homes back without giving up their affordability. The evidence points toward a problem concentrated in parts of the older, lower-rent housing stock, not a citywide explanation for every registered vacancy.

A City Short on Homes, with Thousands Listed Vacant

The 2023 Housing and Vacancy Survey found just 33,210 rental apartments available for rent citywide. That is a net rental vacancy rate of 1.41%, among the lowest on record. For rent-stabilized units, the rate was only 0.98%. Meanwhile, New York State Homes and Community Renewal registration files recorded 47,720 vacant stabilized apartments in NYC in April 2024, or 4.7% of that registered stock. The registrations rose further in 2025, according to the Comptroller’s statewide summary (NYC HPD, 2024; NYC RGB, 2025; NYC Comptroller, 2026).

These numbers appear contradictory only if they are treated as the same measure. The survey counts homes available for rent at the time of the survey; the registration data record whether a stabilized apartment was marked vacant on April 1. A unit could be between tenants, undergoing work, tied up in a dispute or unavailable for another reason. Some vacancies are persistent. Many are not. “Ghost apartment” is a compelling description, but it is not a category tracked by the state.

The Comptroller’s 2024 analysis illustrates the distinction. It found that stabilized apartments vacant but unavailable for rent fell from 42,860 in 2021 to 26,310 in 2023. The subset of affordable apartments unable to return because owners could not make repairs was estimated at fewer than 2,000. That narrower figure is not a full count of every economically constrained apartment, and it predates the 2025 increase in registrations. Still, it challenges the assumption that every registered vacancy is a lost rental opportunity (NYC Comptroller, 2024).

When the Renovation Math Stops Working

A vacant apartment presents a straightforward trade-off. Renovating it requires money up front, often before any rent comes in. Leaving it empty produces no rental income, but building taxes, insurance and maintenance do not disappear. The decision turns on the cost of the work, the rent legally collectible afterward, financing costs and the owner’s expectations. A rent ceiling can change that calculation, particularly when an old, low-rent apartment needs extensive repairs. It does not follow that every owner is better off keeping an apartment vacant.

The Rent Guidelines Board’s 2026 Income and Expense Study shows why the building matters. Using 2024 financial reports from nearly 17,800 buildings with stabilized apartments, it found average monthly collected rent of $1,681 per unit, total income of $1,890, operating expenses of $1,203 and net operating income (NOI) of $688. Total income includes revenue beyond apartment rent, which explains why NOI is not simply $1,681 minus $1,203. These are building averages, not the incremental cost or profitability of renovating a particular empty unit (NYC RGB, 2026a).

The average also masks enormous differences. NOI averaged $529 per unit in fully stabilized buildings and $386 in fully stabilized buildings built before 1974. About 9.2% of buildings in the study had negative NOI. Yet average NOI increased 6.2% across the entire study sample between 2023 and 2024. The lesson is not that all stabilized housing is financially distressed. It is that some buildings have much less flexibility to absorb a major repair bill than others (NYC RGB, 2026a).

That vulnerability is becoming more relevant as costs rise. The RGB projected a 4.1% increase in operating costs for 2026 to 2027, including a 16.9% increase in insurance costs; these are forecasts, not actual results. NYU’s Furman Center identified 455,979 stabilized apartments in older, predominantly stabilized legacy properties, where the median monthly rent was $1,344. Buildings that depend almost entirely on those rents have fewer ways to offset rising expenses than mixed-income properties with higher-rent units (NYC RGB, 2026c; NYU Furman Center, 2026).

Physical conditions follow a similar pattern. The Independent Budget Office found Emergency Repair Program charges in roughly 36% of pre-1974 stabilized buildings with median rents below $1,500, versus 11% of non-stabilized rental buildings citywide. But most stabilized buildings had similar or better conditions than the non-stabilized comparison group. The evidence shows where repair pressure is concentrated; it does not establish that rent stabilization caused the damage or the vacancies (NYC IBO, 2026).

How Rent Policy Changed the Calculation

New York’s rent stabilization system predates the 2019 Housing Stability and Tenant Protection Act (HSTPA). Before 2019, owners could obtain statutory vacancy rent increases and, under certain conditions, deregulate apartments once rents reached a high-rent threshold. HSTPA eliminated the vacancy bonus and high-rent vacancy deregulation. It also restricted rent increases tied to Individual Apartment Improvements (IAIs), limiting how much renovation spending could be reflected in a unit’s legal rent. The changes strengthened protections for future tenants while reducing one way owners had recovered turnover costs (NYC Tenant Protection, n.d.; NYS HCR, 2025).

Lawmakers partially revised those rules in October 2024. The general IAI spending cap increased from $15,000 to $30,000, while certain apartments with qualifying extended vacancies or long prior tenancies became eligible for a $50,000 second tier. In buildings with 35 or fewer apartments, $30,000 in fully eligible first-tier work translates to a legal rent increase of about $179 per month, using the 1/168 formula. The qualifying $50,000 second tier can yield about $347 per month under the 1/144 formula. Larger buildings use different rates. These are permitted rent adjustments, not an assurance that a renovation is affordable or that an owner will earn back the investment (NYS HCR, 2025).

The city then froze ordinary guideline increases at 0% for one- and two-year stabilized leases beginning between October 1, 2026 and September 30, 2027. Other lawful adjustments may still apply. For tenants, the freeze offers immediate relief. For landlords considering expensive work, it makes permitted improvement increases and other financing options especially consequential. Neither effect proves that the freeze will produce more vacancies; that outcome would require evidence about actual renovation and leasing decisions (NYC RGB, 2026b).

What the Vacancy Figures Cannot Tell Us

The headline number is not an estimate of warehousing. The 47,720 registered vacancies in 2024 are not equivalent to the 26,310 unavailable apartments counted in the 2023 survey. The samples, dates and definitions differ, so subtracting one from the other would be meaningless. The 2025 increase deserves scrutiny, but it cannot by itself identify whether repairs, normal turnover, legal disputes or deliberate withholding caused the change (NYC RGB, 2025; NYC Comptroller, 2024, 2026).

A better starting point is how long a home remains vacant. The Independent Budget Office found that most stabilized apartments reported vacant in 2021 were not still vacant in 2022, although about 13,400 were listed vacant in both years. Persistent vacancies are worth investigating, but two years of registration data still cannot show whether an owner faced an unworkable renovation budget, planned a sale or was dealing with another obstacle (NYC IBO, 2023).

An example makes the conflict tangible. In December 2025, City Journal visited two stabilized apartments on East 6th Street in Manhattan, where reported legal rents were below $900 despite much higher rents in nearby market-rate units. An apartment-industry representative said one studio would require at least $100,000 in renovations. That is a reported estimate from an owner-side source, not an independently audited bill, and four apartments visited by a journalist cannot establish a citywide pattern. A landlord-commissioned HR&A study also argues that the 2019 law worsened these incentives. The Comptroller’s 2024 analysis reached a different conclusion about scale, finding no evidence of a broad vacancy increase caused by HSTPA through 2023. These views can both contain some truth: the mechanism may exist for individual apartments without explaining most vacant units. The 2025 data call for a fresh test, not an automatic causal verdict (Lehodey, 2025; HR&A Advisors, 2024; NYC Comptroller, 2024).

What would settle the question is apartment-level evidence: consecutive HCR vacancy registrations matched to legal rent histories, inspection records, permits, documented repair bids, owner financing conditions and eventual occupancy. That would separate apartments temporarily empty from units truly unable to return under current economics. Until that evidence exists, a claim that tens of thousands of apartments can be recovered by changing rent rules is not a reliable forecast.

What Could Realistically Return by 2027?

Bringing a habitable apartment back into use is faster than constructing a new building, and each additional affordable home matters to the household that rents it. But scale is crucial. Even a successful renovation initiative would not necessarily move citywide rents: the effect depends on the number restored, their neighborhoods, their legal rents and what would have happened without intervention.

The strongest public benchmark remains narrow and historical. The Comptroller identified fewer than 2,000 low-rent, repair-constrained units in its 2023 analysis. That estimate should not be presented as the total stock available to restore in 2027. The city’s Unlocking Doors Pilot, meanwhile, offers up to $50,000 in qualifying repair support for up to 200 apartments. The 200 figure is program capacity, not a count of completed renovations or reoccupied homes. There is currently no verified citywide dataset that supports a precise December 2027 restoration total (NYC Comptroller, 2024; NYC HPD, n.d.).

Table 1. The evidence for a 2027 restoration outlook. These numbers measure different things and cannot be added together or treated as a single forecast. Sources: NYC Comptroller (2024); NYC HPD (n.d.).

Unlocking Doors is an instructive model because it ties funding to results. Eligible apartments must have been registered vacant in 2023 and 2024, meet rent criteria and pass inspection. A qualifying CityFHEPS voucher household must then be offered a two-year lease. This connects rehabilitation money with an actual path to occupancy, while retaining affordability requirements. It is a limited pilot, however, not evidence that 200 apartments have already returned to the market (NYC HPD, n.d.).

The most defensible next step is a targeted expansion with stricter measurement, rather than a blanket return to vacancy rent resets. HPD and HCR should prioritize apartments with verified long-term vacancy and repair needs, require itemized work and independent inspections, and condition subsidies or affordable loans on lawful rents and documented re-occupancy. For severely distressed buildings, preservation financing or supervised ownership transfers may be more useful than adjusting an individual apartment’s rent. Any reform should be judged against its cost to tenants as well as its effect on owners.

By the end of 2027, the relevant result would be the net number of existing apartments made habitable and occupied because of a program, excluding new construction and homes that would have returned anyway. The city should publish the number screened, approved, repaired and occupied, the average public cost per additional occupied unit and whether affordability was preserved. Until those results and an updated vacancy baseline are available, a precise citywide number would imply more certainty than the evidence warrants.

The Bottom Line

New York’s empty-apartment story is striking because it sits beside an acute housing shortage. The economics of renovation help explain why some low-rent units may remain unused. They do not explain every vacancy, and the data do not yet tell us how many units would return under a different rent regime. The most credible policy response is to identify the apartments where repair costs really are the barrier, finance work that would not otherwise happen, and count the homes that become occupied. The question is no longer simply how many apartments are empty. It is which ones can be brought back into use, at what cost, and without making affordable housing less affordable.