Charted: Public Cost per Converted Apartment — NYC ~$386k, Calgary ~$63k, DC’s $5k Framing

When a city pays to turn empty desks into apartments, the politically useful number is not the ribbon-cutting unit count. It is public dollars per delivered home — and, when the program buys affordability, public dollars per income-restricted unit. Those two denominators tell different stories, and cities that mix tax abatements with cash grants make the comparison even harder.

Three programs that publish enough detail to rough out a ladder are New York’s 467-m Affordable Housing from Commercial Conversions exemption, Washington, D.C.’s Housing in Downtown competitive tax abatement, and Calgary’s Downtown Office Conversion cash grant. On a per-delivered-unit basis, the New York City Comptroller’s July 2025 fiscal note implies roughly $386,000 of present-value tax expenditure for each apartment in a 14,500-unit Manhattan-south-of-59th sample. Calgary’s completed cohort pencils closer to USD $63,000 per home once a CAD $75-per-square-foot grant is applied to the office floor area removed. D.C. frames a $41 million program against about 8,400 projected units — roughly $4,900 apiece — which is a useful political metric and a poor present-value twin of New York’s 37-year score.

The dashboard above is built for that instrument problem. Switch among Cost ladder, Cost × scale, Conversion vs ground-up, NYC geo split, and Instrument schedule. Toggle the cost metric (delivered unit vs income-restricted unit), pathway (conversion vs ground-up benchmarks), and city focus. The point is not that every city should match New York’s price tag. It is that “we subsidized conversions” is not a single product.

What “public cost” means when the check never clears

Cash grants are easy to count: Calgary pays against office gross floor area removed, with a published residential rate of CAD $75 per square foot and a typical per-property cap near CAD $15 million. Tax abatements are harder. The public cost is forgone property tax over decades, discounted to today. New York’s Comptroller reports both a tax expenditure (full residential tax minus tax paid under 467-m) and an opportunity cost (tax that would have been collected in a carefully chosen counterfactual). Those two present values are $5.6 billion and $5.1 billion on the same Manhattan sample — close in total, different in geography.

D.C.’s Housing in Downtown awards are competitive annual abatement amounts inside fiscal-year caps ($2.5 million per year in FY24–FY26, rising to $41 million in FY28). The Mayor’s office describes the $41 million authority as helping deliver about 8,400 downtown units. That framing is disclosed; a full 20-year present-value score for each conditional award is not yet sitting in a single public table. Treat the ~$4,900 per projected unit figure as program marketing math, not a Comptroller-style PV.

The per-unit ladder, with instruments labeled

Read the table left to right before ranking winners. Calgary’s grant is mostly paid up front against square footage removed; New York’s dollars arrive as a multi-decade exemption schedule that can run 25–35 years with a step-down; D.C.’s number is an annual-authority narrative. A desk that ranks only the rightmost column without the instrument column will mis-brief a council.

New York: expensive per apartment, pricier still per restricted unit

467-m requires that 25% of apartments be permanently income-restricted (on average at about 80% of AMI) and rent-stabilized. In the Comptroller’s sample, that is about 3,600 restricted units inside 14,500 total. Divide the $5.1 billion opportunity cost by those restricted units and the public price is about $1.4 million each — $1.8 million in Lower Manhattan and $0.9 million in the rest of Manhattan south of 59th Street.

That is not an accounting error. In Lower Manhattan, several large conversions were already underwriting before 467-m existed and later received benefits retroactively. For those buildings, the Comptroller’s opportunity-cost counterfactual is a fully taxable market-rate residential building, not a continuing office. The program is then buying rent discounts (and neighborhood mix) more than “inducing” the conversion itself. At 25 Water Street, the note puts opportunity cost near $538 million present value against 330 income-restricted units — about $1.6 million per restricted apartment, or roughly $408,000 if spread across all 1,320 units.

Outside Lower Manhattan, the counterfactual leans toward continuing office taxation growing slowly. Opportunity cost falls below tax expenditure in that geography, which is the mirror image of downtown’s pattern. The dashboard’s NYC geo-split panel shows both bars and the per-IR-unit line together so the geography of the fiscal story is visible.

Calgary: pay per square foot removed, not per bedroom created

Calgary’s program is blunt by design. The city wants surplus downtown office square footage gone. Paying CAD $75 per square foot of office removed ties the public check to the vacancy problem rather than to unit mix. Eight completed incentivized projects have converted about 925,000 square feet into nearly 800 homes (plus hotel rooms in the wider set). That is roughly 1,150 square feet of former office per home — a useful yield for translating a square-footage grant into a per-unit cost.

At full take-up of the published rate, the completed residential cohort implies about CAD $69 million of grant capacity, or roughly USD $63,000 per home. The broader 21-project book (2.7 million square feet toward 2,667 homes) pencils near USD $55,000 per home on the same method. The city also cites roughly CAD $14 million of estimated tax uplift so far and a private-leverage story on the order of four private dollars per public dollar. Those secondary effects matter for a fiscal scorecard; they do not erase the sticker price on the grant line.

Because the incentive is not structured as a long abatement schedule, Calgary’s public cost is front-loaded and comparatively easy to audit. It is also less explicitly tied to deep affordability set-asides than New York’s statute. Comparing Calgary’s ~$63k to New York’s ~$386k without saying “grant versus multi-decade exemption” is how briefings go wrong.

Washington, D.C.: competitive abatements inside a rising cap

Housing in Downtown offers a 20-year real property tax abatement for change-of-use projects that deliver at least ten residential units and meet affordability floors (10% of units at 60% MFI or 18% at 80% MFI). Awards are competitive; DMPED sets the annual dollar amount from pro formas rather than from a rigid square-footage formula (the formula language was stripped in 2024 amendments).

The first conditional cohort is concrete even when award dollars are not fully public: 1625 Massachusetts Avenue NW (157 units, at least 15 affordable), 1825 & 1875 Connecticut Avenue NW (525 units, at least 69 affordable), and 615 H Street NW (72 units, at least 8 affordable) — 754 units and 92 affordable in total. The program’s political math remains the $41 million authority versus ~8,400 projected units. Until each reservation letter’s annual abatement is published and discounted, D.C. belongs on the ladder as a framing point, not as a Comptroller twin.

Conversion subsidies versus ground-up public money

Office conversions are often sold as cheaper housing policy because the structure already exists. Public cost is a different question from hard construction cost. Rough ground-up benchmarks still help bound the conversation:

• New York 4% LIHTC equity capacity is often modeled near ~$231,000 per affordable unit; 9% LIHTC nearer ~$195,000.
• Broader city-financed affordable production is frequently cited near ~$714,000 average public capital cost per unit in recent advocacy syntheses of HPD-era figures.
• Legacy 421-a / 485-x forgone revenue is enormous in annual tax-expenditure reports (about $2.0 billion in FY2025 across hundreds of thousands of units), but that stock figure is not the present value of a newly vested tower and should not be pasted beside 467-m’s pipeline PV without a long footnote.

On that messy canvas, New York’s ~$386k tax expenditure per converted apartment sits above typical LIHTC equity-per-unit figures and below deep city-capital averages — while the ~$1.4M per income-restricted conversion unit lands in a different league entirely, because the denominator shrank to the 25% set-aside. Calgary’s grant-per-home figure looks cheap beside both LIHTC and city capital, which is exactly what a vacancy-removal grant is designed to look like when land and shell costs are already sunk in the private basis.

Caveats that should travel with every per-unit tweet

Counterfactuals dominate the New York score. If Lower Manhattan conversions would have happened anyway, opportunity cost is mostly the price of rent restrictions. If midtown conversions needed the exemption to pencil, opportunity cost is closer to forgone office taxes. The Comptroller is explicit that both statements can be partly true.

Present value is not annual budget pain. A $386k PV per unit spread over 30-plus years is not a $386k hit to next year’s expense budget. Cash grants reverse that timing.

Unit mix skews denominators. Conversion pipelines are studio-heavy. Cost per bedroom can exceed cost per unit when studios dominate, which is why the Comptroller also publishes per-income-restricted-bedroom figures (about $1.7 million sample-wide).

Currency and FX. Calgary figures move with the CAD/USD rate; this post uses 0.73 as a working conversion, not a Treasury fix.

Program framing ≠ audited PV. D.C.’s $4.9k headline is useful for scale talk and dangerous as a cross-city rank.

None of those caveats erase the core pattern: cities are buying conversions with different instruments, different affordability hooks, and public price tags that span roughly two orders of magnitude per delivered home once you put New York’s present-value abatement beside Calgary’s cash grant and D.C.’s authority framing.