The Vacancy Economy

The 2023 New York City Housing and Vacancy Survey, conducted by the U.S. Census Bureau in partnership with the city's Department of Housing Preservation and Development, found 230,200 housing units across New York City that were vacant but not available for rent or sale. Of these, 58,810 were classified as held for seasonal, recreational, or occasional use — a category that captures pieds-à-terre, corporate apartments, and investment units maintained by owners who do not reside in them. This figure represents a decline from 102,900 such units in the 2021 survey, though the drop likely reflects methodological timing rather than a structural reversal: the 2021 survey captured pandemic-era displacement patterns, while the 2023 survey captured a market in which many owners had returned to intermittent use.

Manhattan's rental vacancy rate dropped from 10.01 percent in 2021 to 2.33 percent in 2023 — a compression that would suggest extreme scarcity of available housing. Yet the overall citywide vacancy rate of 1.41 percent coexists with tens of thousands of units that are neither occupied nor offered for rent. The 2023 survey counted only 33,210 units available for rent across the entire city of 2.357 million housing units. The distinction matters: a market in which available rental units represent 1.4 percent of total stock while over 230,000 units sit vacant and unavailable is not a market experiencing simple undersupply. It is a market in which a substantial fraction of the housing stock has been withdrawn from its residential function.

The number of units held for occasional use expanded from 9,282 in 1987 to nearly 75,000 by 2017 before spiking further during the pandemic period. This growth tracks the financialization of Manhattan real estate: as property values escalated, the economic logic of holding units as intermittently occupied stores of value strengthened relative to the logic of offering them for full-time rental. A unit that appreciates at eight to twelve percent annually while sitting empty generates a better risk-adjusted return for a wealthy owner than the same unit occupied by a tenant who imposes wear, management overhead, and legal complexity.

The stretch of West 57th Street between Sixth and Eighth Avenues in Manhattan — colloquially designated Billionaires' Row — contains the highest concentration of ultra-luxury residential towers in the Western Hemisphere. The corridor includes One57 (completed 2014), 432 Park Avenue (2015), 220 Central Park South (2019), 111 West 57th Street (Steinway Tower, 2021), and Central Park Tower (2021). These buildings were designed and marketed not primarily as residences but as stores of value with views. Their floor plans, pricing, and ownership structures reflect this function. Units at 220 Central Park South have traded at prices exceeding $100 million; Ken Griffin's 2019 purchase at that address set a U.S. residential record at approximately $238 million.

Occupancy data for these buildings is not publicly reported in a consolidated form, but investigative reporting and property records reveal a consistent pattern: a large proportion of units are owned by individuals and entities whose primary residences are elsewhere. Reports from The New York Times and other outlets have documented that roughly half of the units in several Billionaires' Row towers are either unoccupied at any given time or occupied only sporadically. Resale performance varies sharply by building — One57 has seen average resale discounts of approximately 25 percent from original purchase prices, while 220 Central Park South has recorded resale gains of 20 to 40 percent — but both patterns are consistent with asset-storage behavior. Owners at One57 who purchased early and hold at a loss are engaging in the same financial logic as owners at 220 CPS who purchased and hold at a gain: the unit is a position in a portfolio, not a home.

The property tax implications are substantial. New York City's assessment system generates tax revenue from these units based on assessed value, but the revenue per unit is substantially lower than the infrastructure cost imposed by the building's existence in a dense urban environment. The towers consume fire, police, water, sewer, and transit capacity that was built for a residential population. A 90-story tower with 50 percent occupancy imposes approximately the same infrastructure demand as a 90-story tower at full occupancy — the elevators, lobbies, mechanical systems, and street access exist regardless of whether the residents are present — while generating civic participation, local commerce, and neighborhood vitality from only a fraction of the potential population.

Several proposals for a pied-à-terre tax in New York City have been introduced in the state legislature since 2014. None has passed. The proposals have typically targeted an annual surcharge of 0.5 to 4 percent on the assessed value of non-primary-residence properties above a threshold (generally $5 million). The political economy of these proposals is instructive: they are blocked not by popular opposition — polling consistently shows majority support — but by the lobbying power of the real estate industry, which treats any tax on vacant luxury property as a precedent that could extend to other property classes.

The 2020 U.S. Census counted approximately 140 million housing units nationwide, of which 9.7 percent — roughly 14 million units — were vacant. By 2023, the Census Bureau's American Community Survey estimated approximately 14.9 million vacant units; by 2024, the figure had risen to approximately 15.1 million, representing 10.3 percent of total housing inventory. The Census Bureau's own analysis notes that seasonal housing accounts for a significant share of this vacancy — vacation homes, snowbird residences, and similar properties concentrated in Florida, Arizona, resort areas, and rural counties with low year-round populations.

But seasonal use does not account for the entirety of the vacancy figure, nor does it account for the vacancy patterns observed in high-demand urban markets where housing costs are highest. The Census classifies vacant units into categories: for rent, for sale, rented or sold but not yet occupied, for seasonal use, for migrant workers, and "other vacant." The "other vacant" category — units that are neither offered for rent or sale nor classified as seasonal — numbered approximately 7 million in 2020. These are the units most likely to represent asset storage: properties held by investors, estates in probate, units deliberately withheld from the market in anticipation of higher future returns, and properties whose owners have no economic incentive to rent them because the transaction costs and management burden exceed the marginal income relative to the appreciation gain from holding.

The geographic distribution of vacancy is the inverse of what a naive supply-and-demand model would predict. Markets with the highest housing costs — San Francisco, New York, Los Angeles, Boston, Seattle — have significant vacancy rates in luxury and above-median segments, while simultaneously registering rental vacancy rates at or near historic lows in segments affordable to median-income households. This bifurcation is the signature of a market in which housing serves two distinct functions for two distinct populations: shelter for wage earners and asset storage for capital holders. The two functions compete for the same physical stock, and the asset-storage function dominates in any price segment where capital returns exceed rental yields.

The mechanism by which vacant investment-held units inflate prices for occupied housing operates through the comparable-sales assessment system. Real estate appraisals and property tax assessments are based on recent transaction prices for comparable properties. When a unit at 220 Central Park South sells for $50 million, that transaction enters the comparable-sales database and raises the assessed value of surrounding properties — including buildings occupied by long-term residents. The assessment increase raises property taxes for owner-occupiers, raises the cost basis for developers evaluating new construction in the area, and raises the benchmark against which landlords set rents for occupied units nearby.

This transmission mechanism operates at every price level, not only the ultra-luxury tier. When an investor purchases a brownstone in Brooklyn for $3 million as a store of value and leaves it intermittently occupied, the transaction raises comparables for the block. When a foreign buyer acquires a Miami Beach condominium for $2 million as a hedge against political risk in their home country and visits three weeks per year, the transaction raises comparables for the building. The aggregate effect of thousands of such transactions across a metropolitan area is a systematic upward bias in assessed values that is disconnected from the income levels of the people who actually live in the market.

Academic research on the magnitude of this effect is limited by the difficulty of constructing a credible counterfactual — it is not possible to observe the same market with and without financialization-driven vacancy. However, studies of markets where vacancy taxes or foreign-buyer taxes have been implemented provide natural-experiment evidence. Vancouver's Empty Homes Tax, implemented in 2017 at an initial rate of 1 percent (subsequently raised to 3 percent of assessed value), resulted in approximately 5,355 fewer vacant units between 2016 and 2021 than would have been expected without the tax. The tax demonstrably reduced vacancy but did not significantly affect average rents — a finding consistent with the hypothesis that the primary price effect of vacancy operates through the comparable-sales assessment channel rather than through direct supply withdrawal.

France implemented a vacancy tax in 1999; the United Kingdom introduced council tax premiums on long-term empty homes in 2013, subsequently increased to 100 percent premiums (double council tax) in 2019 for homes empty more than two years. Spain's Catalonia region and the city of Jerusalem introduced similar measures in 2015. In each case, the policy reduced vacancy at the margin but did not fundamentally alter the asset-storage incentive structure, because the tax rates remained far below the annual appreciation rates that make vacancy economically rational for capital holders.

The Vacancy Premium is the aggregate price increment imposed on every housing market participant — buyers, renters, businesses, and public entities — by the conversion of a fraction of the housing stock from residential use to asset storage. It operates through three channels simultaneously. First, the comparable-sales channel: investment-motivated transactions at above-fundamentals prices raise assessed values market-wide, increasing property taxes and cost bases for all participants. Second, the supply-withdrawal channel: units held vacant or intermittently occupied reduce the effective available stock, tightening the market for the remaining participants who require housing as shelter. Third, the development-cost channel: inflated land values in markets with significant investment demand raise the minimum viable price point for new construction, ensuring that new supply enters the market at price levels accessible only to above-median buyers and renters.

The Premium is not directly quantifiable because the counterfactual market — one in which housing functions exclusively as shelter — does not exist in any major American city. But its magnitude can be estimated by comparing markets with different levels of financialization. The UN Office of the High Commissioner for Human Rights has documented the global pattern: cities that function as destinations for international capital flows — New York, London, Vancouver, Sydney, Hong Kong, Singapore — exhibit housing-cost-to-income ratios dramatically higher than cities of comparable economic output that do not serve this function. The gap between these two categories of cities provides an upper bound on the Premium.

The Vacancy Economy is self-reinforcing. As prices rise, the capital-gains incentive for holding vacant units strengthens, attracting additional investment capital to the market, which raises prices further. The cycle requires no coordination and no intent — it is the emergent behavior of rational actors responding to a market in which housing appreciation consistently outperforms alternative investment classes on a risk-adjusted, tax-advantaged basis. The individual investor who purchases a Manhattan condominium as a store of value is not conspiring to raise housing costs; they are making an economically rational decision within a system that has converted residential real estate into the most accessible and tax-favored asset class available to global capital. The Vacancy Premium is a structural feature of this system, not an aberration within it.