Manhattan leased just under 23 million square feet of office space in the first half of 2026. That is the strongest first-half leasing performance since 2002, a twenty-four year benchmark, and the borough’s availability rate finished the second quarter at 13 percent.
For a market that spent four years being written off, that is a real reversal. The question worth asking from Dallas is what produced it, because most of it does not transfer to North Texas, and the part that does is not the part the headline points at.
What actually drove the number
Law firms and AI companies did most of the work, and neither represents a broad return to the office.
Law firms and professional services took 30 percent of second-quarter activity, with law firms accounting for five of the ten largest leases signed in the quarter. Simpson Thacher & Bartlett took 916,000 square feet at 570 Fifth Avenue. Cleary Gottlieb took 475,000 at One Liberty Plaza.
The second group is newer. AI companies leased 800,000 square feet in Manhattan in the second quarter alone, more than the sector took in all of 2025. That is an industry expanding from a small base with capital that does not behave like normal tenant capital.
The finding here is concentration, not weakness everywhere else. Second-quarter leasing ran 11.02 million square feet, so what sits outside those two groups is substantial on its own. What is unusual is how much of a record half came from two tenant types competing for a narrow band of buildings.
The supply side is the part people miss
Availability at 13 percent did not happen because demand rose enough to fill the market. A significant part of it happened because space left the market permanently.
New York removes roughly 2 million square feet of its weakest office stock every year through conversion and demolition, and the pace has accelerated sharply. Office-to-residential conversions ran 1.6 million square feet in 2023, 3.3 million in 2024, and 5 million in 2025, the highest total in two decades. More than 16,000 residential units are planned from converted office buildings for 2026 and beyond.
That is the mechanism. When a market takes its worst buildings out of the denominator, the vacancy rate improves even if leasing stays flat. New York has policy built specifically to encourage it, in the form of the City of Yes zoning changes and the state’s 467-m tax incentive for commercial conversions.
Where DFW actually stands
North Texas is running a different experiment, and the numbers show it.
DFW marketwide office vacancy sat at 25 percent in the second quarter, down from a peak of 26.9 percent at the end of 2024. Net absorption was positive at 939,000 square feet, a real improvement after negative 1.2 million square feet in the first quarter. Leasing volume was 1.2 million square feet, down about a third quarter over quarter, with more than 92 percent of it in Class A space.
Average asking rents reached $34.79 per square foot marketwide, with Class A at $39.46. In Fort Worth specifically, the first quarter brought 109,000 square feet of positive absorption, the strongest start since 2018 and the third consecutive positive quarter, with asking rents up 3 percent year over year to $29.47.
And about 3 million square feet is under construction.
That last figure is the sharpest difference between the two markets. Manhattan is subtracting inventory while sitting at 13 percent availability. DFW is adding inventory while sitting at 25 percent vacancy.
The Class B number is the one worth studying
The DFW Class B figures do not match the national story at all.
Class B space absorbed 626,000 square feet in the second quarter, against negative 359,000 square feet in the same quarter a year earlier. Class B vacancy fell 220 basis points year over year. Every market narrative for the past three years has been flight to quality, and in Fort Worth that still holds, with trophy and Class A accounting for nearly 85 percent of positive absorption.
But the Dallas Class B number says something else is happening underneath. Tenants who were priced out of new construction, or who never needed it, are transacting again. For an owner holding a 1980s building in a decent location, that is a more relevant signal than anything happening on Park Avenue.
It is also fragile. One quarter is one quarter, and the leasing volume that produced it was down sharply from Q1.
Price is still the argument
The reason any of this matters to a DFW owner is that the two markets compete for the same relocating tenants, and the gap between them is the pitch.
Manhattan marketwide numbers are dominated by trophy towers, and that makes them a poor comparison for the tenants DFW actually recruits. The more useful figure is what a small or midsize company is quoted. Tandem Space, which tracks small-tenant activity in New York, puts the median asking rate for startup-scale Manhattan suites at $64 per square foot a year, with neighborhood rates running from $35 in the Garment District to $93 in Tribeca. Anyone building a relocation case can pull current Manhattan office asking rates by neighborhood and compare them against a DFW Class A quote of $39.46.
That is a real gap and it is not going to close. Texas has been winning that argument for a decade and the Manhattan recovery does not change the arithmetic.
What it does change is the quality of what a tenant gives up. Two years ago there was slack at the top of the Manhattan market, with the City Comptroller putting Manhattan CBD availability at 18 percent in mid-2024, on a narrower geography than the borough-wide figure above. On either measure the direction since has been the same. With law firms taking whole towers, the discount a relocating tenant could once get on good Manhattan space is gone. That strengthens the DFW case.
What actually transfers
Three things transfer.
The supply lesson is the big one. Manhattan’s availability improved substantially because the market retired its weakest buildings, and it had policy support to make that happen. DFW has 25 percent vacancy and a meaningful quantity of functionally obsolete space, and no comparable mechanism for removing it. Conversion is harder here, since suburban floor plates were never designed for residential and the incentive picture is nothing like New York’s, but the owners running the numbers now are ahead of the ones waiting for the market to absorb the space on its own.
The second is that recovery arrives one sector at a time. Manhattan’s rebound is a law firm and AI story. DFW’s is corporate relocation, financial services, and now, tentatively, Class B users. Owners who know which specific tenant type their building serves will read their own market better than owners tracking the metro average.
The third is about attendance. Manhattan’s demand rests partly on employers requiring more days in the office, with the most recent Partnership for New York City survey showing 57 percent of Manhattan office workers at their desks on an average weekday and a quarter of employers planning to require more. That pressure operates in Texas too, and space demand follows attendance policy rather than leading it.
The honest comparison
Manhattan at 13 percent availability and DFW at 25 percent vacancy are two different markets, with different supply pipelines and different tenant bases, and North Texas should not expect the same recovery to arrive on the same terms.
What the Manhattan numbers do show is that its improvement came from removing inventory as much as from adding tenants. That is the lever North Texas has barely touched, and it is the one worth arguing about locally.
