Skip to content

Home Publications

Part of a Collection:

Data Brief: Legacy 90%+ Rent-Stabilized Properties

Photo of Buildings and streets near Midtown Manhattan, New York City.

This data brief is the first in a two-part series on segments of the rent-stabilized housing stock that provide deep affordability but are experiencing increasing financial and physical pressures. It follows our Overview Brief that explores how rent stabilized homes fit within New York City’s housing landscape. Here, we focus on legacy 90%+ rent-stabilized properties.

What are Legacy 90%+ Rent-Stabilized Properties?

We use the term “legacy properties” to refer to pre-1974 multifamily buildings with six or more units that contain at least one rent-stabilized apartment. Within this group, “legacy 90%+ properties” are those in which at least 90 percent of units are rent stabilized. 1  These buildings make up a substantial share of the city’s lower-cost private rental housing and play an important role in maintaining affordability.

Legacy 90%+ properties were not developed with public subsidies and are not governed by income-restriction requirements. 2 This distinguishes them from newer rent-stabilized developments that received public financing or tax incentives in exchange for providing income-restricted units. It also distinguishes them from other legacy properties that contain a meaningful share of deregulated units and therefore generate some revenue from market-rate apartments. In contrast, legacy 90%+ properties rely almost entirely on rent-stabilized units for operating income.

Legacy 90%+ properties also provide some of the city’s most affordable rents. The median collected rent in this segment is estimated at $1,395 per month in 2025, roughly the level affordable to a household earning around 40 percent of Area Median Income (AMI) in 2025. 3  While tenant income data for this specific segment are not publicly available, many households living in these buildings are likely to have incomes below the $60,000 median income for all rent-stabilized renter households reported in the Housing and Vacancy Survey (HVS). 4  This occurs against the backdrop described in the overview brief, where renter incomes have grown more slowly than rents and rent burdens remain widespread, especially among lower-income households.

We estimate that legacy 90%+ properties contain about 456,000 rent-stabilized units across 16,600 buildings, representing 47 percent of all stabilized units citywide. These buildings make up the largest portion of the broader legacy segment, which includes roughly 617,000 stabilized units in total. In other words, the 90%+ subsegment accounts for nearly three-quarters of rent-stabilized units in legacy properties.

Throughout this data brief, we compare legacy 90%+ properties with two other legacy subsegments: buildings where 36 to 89 percent of units are stabilized (“36–89%”) and buildings where more than 0 but up to 35 percent of units are stabilized (“>0–35%”). 5  These comparisons illustrate how differences in the building-level composition of rent-stabilized units shape rent levels, revenues, operating conditions, and location patterns across these subsegments. Together, these distinctions offer policymakers insight into where financial pressures and preservation needs are most acute.

Where are Legacy 90%+ Properties Located?

Legacy 90%+ properties are distributed across the city but are more heavily concentrated in Northern Manhattan and the Bronx, including Washington Heights/Inwood, Kingsbridge Heights/Bedford Park, Fordham/University Heights, and Highbridge/Concourse. Brooklyn clusters are also prominent, particularly in Flatbush and Midwood (Figure 1). 

This spatial pattern resembles the broader rent-stabilized stock, but with important differences. Legacy 90%+ properties are more common in outer-borough neighborhoods and far less prevalent in Manhattan below 96th Street or in central parts of Brooklyn and Queens. Relative to 90%+ buildings, >0–35% properties are heavily concentrated on the Upper West Side and Upper East Side, and 36–89% properties are especially clustered in Washington Heights/Inwood.

Figure 1:

The Typical Unit Count of Legacy 90%+ Properties 

Legacy 90%+ properties tend to be small to mid-sized multifamily buildings. The median building in this subsegment contains 16 units, only slightly larger than buildings in the other legacy subsegments. More than 80 percent of legacy 90%+ buildings contain fewer than 50 units (Figure 2). 

This size distribution affects how these data should be interpreted, for two reasons. First, the definition of legacy 90%+ properties, requiring that at least 90 percent of units be rent stabilized, excludes six-to-nine-unit buildings that may have only one market-rate unit. Although these buildings fall below the threshold for this analysis, they may operate similarly to those included in the segment, which is important for policymakers considering how to set program eligibility criteria.

Second, the Rent Guidelines Board (RGB) relies on income and expense information from the Department of Finance’s (DOF) Real Property Income and Expense (RPIE) filings to inform its annual adjustments to rent-stabilized rents. Buildings with ten or fewer residential units are not required to submit RPIE data, unless they contain a commercial space. 6 As a result, more than 40 percent of legacy properties (10% of units in the segment) are not included in the underlying financial data. These smaller buildings may face different operating conditions and cost structures, so the figures that follow should be interpreted as reflecting only the portion of the segment captured in RPIE filings.

Figure 2:

Rent Levels in Legacy 90%+ Properties

Legacy 90%+ properties contain some of the lowest rents within the legacy stock. The median imputed rent in this subsegment is $1,344 per month, compared with $1,455 across all legacy properties (8% lower). 7 Put differently, half of the roughly 456,000 stabilized units in legacy 90%+ buildings have average rents below $1,344 per month, a level generally affordable to a household earning around 40 percent of AMI in 2025. 8

The rent levels described in this section are derived from DOF’s Notice of Property Value (NOPV) data, which incorporate income and expense information that owners report through RPIE filings and that the DOF adjusts as part of its annual assessment process. These adjusted figures serve as the foundation of the city’s multifamily property tax system, and are also the data that the RGB relies on in its annual deliberations. 

Differences in rent levels across the legacy stock reflect how market conditions interacted with rent-regulation laws prior to the 2019 Housing Stability and Tenant Protection Act (HSTPA). Between 1993 and 2019, state law permitted permanent deregulation of units once legal rents exceeded designated thresholds (Deregulation Rent Threshold or DRT). 9 Because the trigger was a unit’s legal rent, deregulation occurred more often in neighborhoods where market rents rose above the threshold, while units in lower-rent areas generally remained stabilized. As an example, the monthly DRT in 2019 was $2,775.

To illustrate how these dynamics shaped the legacy stock’s current rent distribution, we examine the three subsegments described above. As Figure 3 shows, buildings with fewer stabilized units (>0 to 35%) have a median monthly rental income, per unit, that is 78 percent higher than those in legacy 90%+ properties ($2,386 versus $1,344), while those in the middle segment (36–89%) are about 26 percent higher ($1,698).

Figure 3:

So far, we have used the median rent to describe the “typical” building in each subsegment. But relying on the median alone can obscure the variation within legacy 90%+ properties, from buildings with low average rents to others with much higher average rents. 10  

Average monthly imputed rents in legacy 90%+ properties are roughly normally distributed, with a modest right tail of buildings where the average per-unit rents approach $2,400, about 13.5 percent lower than the deregulation threshold in effect when HSTPA passed in 2019 (Figure 4). 11 At the same time, a substantial share of buildings have much lower rents. The most affordable portion of the legacy 90%+ stock, the bottom 5 percent of buildings, has average rents of about $1,000 per month, a level considered affordable to households earning roughly 30 percent of AMI.

Figure 4: 

Legacy 90%+ properties contain apartments with rent levels comparable to publicly subsidized affordable housing, making them a distinct and important component of New York City’s affordable rental stock. Yet the same low rents that define this segment also constrain the revenue available for day-to-day operations and capital reinvestment. These rent levels create a narrow revenue base within which owners must manage rising operating costs. The next section examines how these constraints appear in recent trends in operating costs and net operating income (NOI).

Because the trends in this section occurred during a period of elevated inflation, this brief reports changes primarily in inflation-adjusted (real) terms unless otherwise noted. Nominal values reflect the dollars reported at the time, while real values adjust for changes in consumer prices and indicate purchasing power. When nominal income grows more slowly than inflation, the real resources available to operate buildings may decline even if nominal NOI increases. Therefore, real trends provide a clearer picture of financial pressure.

Gross Income, Operating Expenditures, and Net Operating Income

Real revenue trends provide the clearest starting point for understanding the financial pressures faced by legacy 90%+ properties. Between 2019 and 2025, median gross income per unit in this segment declined by about 9 percent, after adjusting for inflation (Figure 5).

We next describe the costs to operate buildings. In this brief, “operating expenditures” or “expenditures” refer to the amounts owners report spending to operate buildings, as captured in the Department of Finance’s assessment data. These figures approximate actual spending, but can differ from owners’ internal expense records because DOF applies standardized adjustments as part of the tax assessment process. These figures also do not include property taxes.

Over this period, operating expenditures excluding property taxes remained roughly stable or fell modestly in real terms. Even with relatively stable expenditures, however, declining gross income caused real NOI to fall, and the decline was sharper in legacy 90%+ properties than in other legacy subsegments. Figure 5 shows that legacy 90%+ buildings experienced the steepest erosion in NOI. 

These patterns indicate tightening operating margins that may limit the ability of owners to invest in building upkeep. Because NOI represents the resources available after operating expenses, declines in real NOI reduce the cushion available to plan for repairs, build reserves, or manage unexpected increases in costs. These pressures are further compounded by the fact that legacy 90%+ buildings have no market-rate units to offset slower revenue growth. As a result, rising costs cannot be cross-subsidized with higher-rent apartments.

Figure 5:

Inflation and Operating Cost Pressures

While the data just discussed indicates that owners kept what they were spending on the building’s operations and maintenance steady in real terms in recent years, that likely obscures the pressures owners face from the costs of operating their buildings. Data from the Rent Guidelines Board show that the Price Index of Operating Costs (PIOC) increased by about 33 percent between 2019 and 2025, compared with a 24 percent increase in the Consumer Price Index (Figure 6). 12 Several major cost categories increased faster than inflation: utilities rose by 31 percent, maintenance by 39 percent, and insurance by 150 percent, while property taxes, labor, fuel, and administration increased close to or slightly below the overall rate of inflation. 

As defined by the RGB, the PIOC “measures changes in the cost of purchasing a specified set of goods and services (market basket) paid by owners in the operation and maintenance of buildings that contain rent stabilized units in New York City.” 13 It reflects the broader cost environment facing these properties, though it does not capture what owners actually spent. The PIOC nonetheless provides essential context for understanding the pressures legacy 90%+ properties face as real revenues decline. 14

Figure 6:

Expenditure Patterns in Legacy 90%+ Properties

Despite rising underlying costs, Department of Finance data show that owners of legacy 90%+ properties reported operating expenditures that declined slightly in real terms. Operating expenditures excluding property taxes increased by nearly 20 percent in nominal terms over this period (based on data collected in 2017 and 2023), but declined by 3.3 percent after adjusting for inflation (Figure 5). 15 Compared to other legacy subsegments, the decline was greatest in the 90%+ segment.

Maintaining flat or declining real expenditures during a period when several major operating inputs rose faster than inflation suggests that owners may have reduced or shifted spending within building operations. These patterns are more consistent with deferred or reallocated maintenance than with reduced cost pressures. Owners may delay maintenance, repairs, or replacement cycles to absorb less flexible costs such as insurance or utilities. Some adjustments may not immediately affect building quality, but declining real expenditures during a period of rising costs warrants attention because it may indicate emerging deferred maintenance. 16

Property Taxes

Property taxes are not included in the RPIE-based expenditure figures discussed above, but they represent a substantial additional cost for stabilized properties and therefore merit separate consideration. In legacy 90%+ buildings, real per-unit tax obligations rose slightly faster than inflation between 2019 and 2025, increasing from $2,843 to $3,082 (Figure 7). 

Because operating expenditures excluding taxes have remained roughly flat while gross income has declined, rising tax obligations further narrow the margin between income and total operating costs. This tightening margin implies that the resources available for maintenance, repairs, and day-to-day operations may be increasingly limited.

Figure 7: 

Across the legacy 90%+ segment, the financial data show a consistent pattern: real revenues have declined, and operating expenditures excluding property taxes have remained flat even as expenses in several major cost categories increased more rapidly than inflation. As a result, NOI has fallen, and also remains lower than in other pre-1974 buildings given relatively lower rental income. While these trends do not indicate widespread distress, they point to a tightening financial environment that may constrain the ability of some owners to sustain current levels of upkeep or reinvestment. 

Additional Indicators of Emerging Financial and Physical Pressures

Other indicators point to financial pressure in some legacy 90%+ properties. Legacy 90%+ buildings were more likely to appear on New York City’s 2025 tax lien sale eligibility list than other legacy properties, which identifies buildings with significant arrears on property taxes or related charges such as water and sewer fees. 17 Slightly more than 4 percent of legacy 90%+ properties met the criteria (about 1 in 25 buildings, the highest share among the segments examined). A similar share of properties in the 36–89% segment was also eligible, suggesting that cash-flow constraints may be more common in buildings with higher shares of stabilized units.

Table 1: 

Physical Indicators of Distress

Changes in housing-code violations also offer insight into day-to-day building conditions. Violation rates increased across all property types in recent years, reflecting both the New York City Department of Housing Preservation and Development’s (HPD) expanded enforcement activity in 2023 and 2024 and higher levels of tenant complaints. 18 Legacy 90%+ properties have historically exhibited higher rates of code violations than other subsegments of all legacy properties, and the gap appears to have widened in recent years. Between early 2021 and 2025, the violation rate for legacy 90%+ properties increased by 47 percent, compared with 22 percent for buildings with smaller shares of stabilized units (Figure 8). 

The difference is also visible when examining specific categories of violations. Class C (immediately hazardous) violations, as well as leak and mold infractions, rose more sharply in legacy 90%+ properties. 

Figure 8:

Figure 9:

Figure 10:

Market Activity as Indicators of Financial Pressure

Market activity provides another perspective on how buyers and sellers value the income potential of legacy properties. For legacy 90%+ buildings, median sale prices per square foot have returned to roughly their 2011 inflation-adjusted levels after declining following the 2019 regulatory changes and through the pandemic (Figure 11).

Figure 11: 

Annual sales volumes have remained below pre-pandemic levels, extending a downward trend that began in the mid-2010s. Transactions were volatile during the pandemic period, declining in 2020 and then rebounding in 2021, but overall activity has been subdued since the late 2010s (Figure 12). 

Figure 12: 

Taken together, these data suggest several dynamics.

First, the sales price trends noted above do not indicate a market dominated by distressed selling. Instead, many owners appear to be holding properties and reassessing values in light of slower revenue growth, rising operating costs, and the post-2019 regulatory environment, which together require a recalibration of long-term expectations for stabilized assets.

Second, the muted pace of transactions likely reflects a broader environment of uncertainty. Prospective buyers have become more cautious about underwriting rent-stabilized assets, and several major lenders and investors have scaled back their activity in this segment. Signature Bank’s 2023 failure placed more than 2,200 rent-stabilized or rent-controlled multifamily loans into FDIC receivership, removing a major lender from this segment. 19  Owners seeking to sell or reinvest may also face difficulty securing financing or recapitalizing at levels that make transactions feasible, particularly given that recent market reports point to markedly curtailed lending for rent-stabilized buildings. 20

Third, fewer transactions limit comparables and slow price discovery, creating uncertainty about current asset values, future revenue prospects, and the financing terms that are likely to be available.

This picture contrasts with what we observe in the other legacy subsegments. Properties with fewer stabilized units (>0–35%) have experienced sustained real price growth since 2011, while buildings in the 36–89% segment have seen more moderate gains. Legacy 90%+ properties, where nearly all units are stabilized, have recorded the weakest price performance of the three groups.

Taken together, these trends suggest that investors are valuing legacy 90%+ properties at a discount relative to other legacy buildings, reflecting expectations of more limited rent growth and rising operating costs. While lower values may ease speculative pressure, they can also make it more difficult to refinance debt or plan for major repairs. In that sense, the market signals observed here are consistent with the financial constraints documented in earlier sections.

Implications and Conclusion

Legacy 90%+ properties occupy a distinctive place in New York City’s rental market. They provide some of the city’s most affordable rents to generally lower income households. The analysis above shows that the quality and stability of this segment may be at risk because financial margins have narrowed. Revenues have not kept pace with inflation, and several major operating costs have risen faster than prices. Yet owners’ reported operating expenditures have remained flat in real terms. As a result, net operating income has declined, reducing owners’ capacity to absorb rising costs or invest in long-term maintenance.

The roughly 456,000 stabilized homes in legacy 90%+ buildings represent a critical share of the city’s affordable housing. Replacing these units with newly subsidized housing or with income support for tenants would require substantial public resources. For that reason, it is important that the City and State address the challenges posed by rental revenues that lag inflation and operating costs that are rising in real terms across the broader multifamily market, while keeping affordability for current tenants at the center of any policy response. As policymakers, practitioners, and researchers consider potential interventions, the specific pressures facing the legacy 90%+ segment should remain a key focus.
 

Next in Series: Data Brief: Government-Subsidized, Income-Restricted Rent-Stabilized Properties

  1. In most of these buildings, 100 percent of the rented units are rent stabilized. We have used a threshold of 90 percent, however, to avoid missing buildings in which there are units for the superintendent or other staff or management that are not stabilized. (Super or other non-rental units should still be registered as “temporarily exempt,” it is unclear if these units are included in DOF tax records.) The 90 percent threshold means that buildings with six to nine units may be excluded from the legacy 90%+ subsegment even if just one apartment is not stabilized. Those buildings may face similar conditions to legacy 90%+ buildings. But as we note later in the text, data the Rent Guidelines Board relies upon does not report the income or expenses in purely residential buildings of 10 units or fewer units, so we do not know for sure whether those small buildings are most like our legacy 90%+ category, or more like other subsegments of the legacy stock that have some market rate income. Policymakers should consider whether to extend any policy applied to stabilize the legacy 90%+ buildings to small buildings with just one non-rent stabilized unit given that uncertainty. ↩
  2. We include properties that currently receive, or previously received, J-51 tax benefits to support renovation or conversion to multifamily use. ↩
  3. New York City Department of Housing Preservation and Development. (n.d.). Area median income. https://www.nyc.gov/site/hpd/services-and-information/area-median-income.page ↩
  4. In a recent focus group, we heard from a local non-profit organization that the average income of people who call a hotline in search of legal or financial assistance is less than $30,000 a year. ↩
  5. One reason why we use the 35 percent rent-stabilized cut-off is that it aligns with the eligibility threshold for Major Capital Improvement (MCI) increases; buildings with fewer than 35 percent rent-stabilized units may not apply for an MCI. ↩
  6. For more information on RPIE statements, please see the technical appendix in the Overview Brief in this series. ↩
  7. Median values are calculated from building-level average (imputed) rents. That is, we take the median of per-building average rents, which summarizes the typical rent level across properties rather than individual units. Average rents are estimated by dividing total collected residential income by the number of residential units among properties with no commercial square footage. Because imputed rents reflect collected residential income, building-wide vacancies, nonpayment, or other sources of lost income can lower these estimates. As a result, imputed rents should be interpreted as a lower-bound estimate of average unit-level rents, and actual legal or preferential rents may be modestly higher in some buildings. ↩
  8. NYU Furman Center Senior Policy Fellow Mark Willis testified before the Rent Guidelines Board in Spring 2025 that roughly 200,000 units in legacy properties faced greater financial pressure, based on an estimated 400,000 units in the Legacy 90%+ segment at that time. We have since updated the estimate of this segment to 456,000 units, which implies that approximately 228,000 units fall below the median imputed rent. Readers familiar with earlier estimates will see that the underlying conclusion remains the same. Buildings with lower rents in this segment face greater financial pressure. See Mark Willis, “Testimony of NYU Furman Center Senior Fellow Mark Willis Before the Rent Guidelines Board,” The Stoop, NYU Furman Center, https://www.furmancenter.org/thestoop/entry/testimony-of-nyu-furman-center-senior-fellow-mark-willis-before-the-rent-guidelines-board. ↩
  9. New York State Division of Housing and Community Renewal. (2018, September). Deregulation rent and income thresholds. https://hcr.ny.gov/system/files/documents/2018/10/deregulationrentincomethreshold.pdf ↩
  10. Moreover, two buildings with the same average rent can have very different distributions of actual rents, ranging from cases where all units rent for the same amount to cases where the rent distribution is highly skewed, with some units with very low rents and others with much higher rents. ↩
  11. Some buildings may have units with legal rents above the former deregulation threshold because certain regulatory conditions, such as the presence of an active J-51 tax benefit, prohibited deregulation even when legal rents otherwise exceeded the threshold. ↩
  12. The Rent Guidelines Board’s Price Index of Operating Costs (PIOC) is a measure constructed to track year-to-year changes in the typical costs of operating and maintaining rent-stabilized buildings in New York City. It is based on a market basket of goods and services that includes labor, maintenance supplies and contracts, fuel, utilities, insurance, taxes, and administrative expenses. The index estimates how these inputs change in price over time and is used by the RGB to inform deliberations about annual rent adjustments. The PIOC is not an audited accounting of what buildings actually spend, nor is it a measure of total operating expenditures or property-specific cost changes. It does not capture capital improvements, financing costs, debt service, or variations in cost structures across different segments of the rent-stabilized stock. Because the PIOC reflects cost inflation for a standardized market basket rather than observed expenditures, it provides guidance on broad cost trends but does not indicate whether owners’ actual spending kept pace with inflation or whether any observed financial distress is caused by these cost changes. ↩
  13. New York City Rent Guidelines Board. (2023, April 20). 2023 Price Index of Operating Costs. https://rentguidelinesboard.cityofnewyork.us/wp-content/uploads/2023/04/2023-PIOC.pdf ↩
  14. When allowable rent increases are set below the rate of inflation, the purchasing power of rental income declines in real terms even if nominal rents rise. For a discussion of how recent Rent Guidelines Board adjustments compare with inflation and the implications for real revenue, see Mark Willis, “Testimony Before the Rent Guidelines Board,” NYU Furman Center (2025), available at: https://www.furmancenter.org/thestoop/entry/testimony-of-nyu-furman-center-senior-fellow-mark-willis-before-the-rent-guidelines-board ↩
  15. As described in our Overview Brief, we rely primarily on Notice of Property Value (NOPV) data from the Department of Finance (DOF) for information on property-level income, expenditures, and net operating income. Because the NOPV DOF data does not allow us to isolate residential income from commercial income, we restrict our sample to properties without commercial square footage. NOPV expenditures do not include property taxes. DOF releases NOPV data two years after collection, and the figures have been adjusted and trended forward by DOF. For this reason, the NOPV data is not an identical reflection of the financial status of a property as originally reported. Our analysis relies on NOPV data based on RPIE filings covering the years 2017 to 2023 and released between 2019 and 2025. When adjusting for inflation, we use the Consumer Price Index for the year of release. DOF does not rely solely on owner-reported values when preparing assessment roll data. The agency evaluates reported expenses against benchmark expense ratios that vary by building characteristics and substitutes standardized assumptions when reported values fall outside expected ranges. As a result, some expense figures in the public data may differ from the costs owners actually incur. ↩
  16. The PIOC measures the cost of maintenance inputs and shows these rising faster than inflation between 2019 and 2025. DOF data reflect reported spending and show flat or declining real expenditures. This divergence suggests that owners may have reduced or deferred maintenance rather than experiencing lower cost pressures. ↩
  17. Lien sale eligibility is not a definitive indicator of financial distress. Properties can appear on the eligibility list for reasons that range from temporary cash-flow disruptions to administrative delays, and some owners resolve arrears before a sale occurs. We rely on the 10-day lien sale list because it identifies properties that remained eligible immediately before the scheduled sale and therefore reflects those at highest risk of formal enforcement action. ↩
  18. Spauster, P. (2024, October 16). Housing violations in NYC jumped 24% this year. We mapped them by neighborhood. City Limits. https://citylimits.org/housing-violations-in-nyc-jumped-24-this-year-we-mapped-them-by-neighborhood/ ↩
  19. Federal Deposit Insurance Corporation. (2023, July 13). Multifamily loan portfolio frequently asked questions. https://www.fdic.gov/bank-failures/multifamily-loan-portfolio-frequently-asked-questions ↩
  20. Aviram, D., & Kaminsky, J. (2025, July). The Mamdani rent freeze and its impact on New York City’s multifamily market. Columbia Business School, Paul Milstein Center for Real Estate. https://business.columbia.edu/milstein-center-research-lab/mamdani-rent-freeze-new-york-city. See chart titled: Rent-Stabilized Loan Originations in New York City.  ↩

Related Posts