Office Submarkets: Urban vs Suburban, Gateway vs Sun Belt

Introduction

The single most useful fact about US office right now is that the consensus of three years ago has inverted. Through 2018-2022, the received wisdom held that gateway markets were the structural losers, hollowed out by remote work and tech-sector retrenchment, while the Sun Belt was the structural winner riding corporate relocations and inbound migration. By Q1 2026, the opposite is closer to true. Midtown Manhattan prime trophy space ran at 2.9% vacancy, tighter than at any point since before the pandemic, while Austin, the poster child of the relocation boom, sat above 22% and led the nation in year-over-year vacancy increases. The axis that ended up mattering most was not gateway versus Sun Belt at all. It was building quality.

That does not make geography irrelevant. Office still divides along two distinct axes, urban versus suburban and gateway versus Sun Belt, and both carry real information. But the divergence within each axis now dwarfs the divergence between the axes, which is why a national vacancy figure (roughly 17.6% as of spring 2026 on one widely cited tracker, and a record 21% on Moody's stricter measure) tells an analyst covering the major office REITs or an office lender almost nothing actionable. The work happens at the metro and submarket level, often down to the individual building tier.

The deepest demand shock landed on dense urban cores, and the reason is structural rather than cyclical. A central business district served the longest-commute workforce, exactly the population most willing to trade office days for hybrid flexibility once employers allowed it. These same districts carried large bases of older Class B and C stock, the lower rungs of the property quality ladder, that the flight to quality stranded, lacked the residential mix that newer live-work-play districts use to stay full, and ran higher operating costs that compressed landlord margins as rents softened. Suburban office escaped most of this: shorter commutes for hybrid workers, proximity to where people actually live, lower running costs, and inventory that skewed less toward the obsolete commodity tier.

CBD (Central Business District)

The dense urban core of a metropolitan area, built around high-density office concentration, mass-transit access, and historical primacy as the home of major employer headquarters. US examples include Midtown Manhattan, San Francisco's Financial District and SOMA, Washington DC's downtown, Chicago's Loop, and Boston's downtown.

The gap shows up cleanly in the recovery data. CBRE's tracking has US downtown office vacancy up 6.2 percentage points since Q1 2020 against only 3.6 points for suburban office, and through 2025-2026 it is suburban markets that have been tightening faster, with stronger rent growth and quicker vacancy reduction. An April 2025 federal policy shift reinforced the trend: an executive order revoked the longstanding preference for siting federal offices in CBDs in favor of cost-effective locations, which pushed CBD development sharply lower and tilted new construction toward suburban and mixed-urban sites.

Suburban office is not a uniform safe haven, though. The same flight to quality runs through it: suburban office in declining metros, the older rings around Detroit or Cleveland, faces the same structural erosion as the CBDs in those cities, and commodity suburban stock everywhere drifts down even as trophy suburban product in desirable residential submarkets stays full. The bifurcation mirrors the CBD pattern; it simply runs at a milder amplitude.

The gateway recovery is led almost entirely by the top of the quality stack, and Manhattan is the clearest case. Trophy and prime Class A space captured roughly 55% of all 2025 leasing in Midtown despite being a small share of total supply, prime availability fell to 2.9% in Q1 2026, and Class A asking rents pushed about 6% above their pre-pandemic level (Grand Central direct Class A averaging near $93/SF, with the most visible trophy floors fetching $265 to $320-plus). Manhattan's 2025 leasing total of 39.8 million square feet was the highest since 2019, and a wave of AI-firm expansion has been a meaningful driver of the surge. San Francisco, the deepest-distress gateway through the downturn, has begun turning as well, with overall vacancy down to roughly 28% in Q1 2026 from its 33.8% peak in Q3 2024 and leasing up around 43% year over year (about 3.4 to 3.5 million square feet, its highest first quarter since 2000). This is the dynamic the class A flight to quality plays out in concentrated form: trophy assets in gateway markets are arguably the single strongest office category in the country, while commodity assets a few blocks away are among the worst-positioned.

The Sun Belt's stall is a supply story more than a demand story. Developers delivered a heavy pipeline into the hot 2020-2022 window, underwriting it against double-digit office-employment growth, and that inventory is now competing for tenants just as the relocation wave has crested.

Other Sun Belt markets are not moving in lockstep, which is the deeper point about the category. Atlanta's metro vacancy actually improved year over year to 26.5% in Q1 2026, with even its CBD ticking down, while Miami and Tampa held up far better than Austin or Dallas thanks to lighter pipelines and steadier demand. Dallas, Phoenix, Charlotte, and San Antonio, the markets that built the most aggressively, are the ones still seeing vacancy climb. The Sun Belt label is a reasonable first-pass sort, but any analyst underwriting a specific asset has to drop to the metro and submarket beneath it.

The practical consequence runs through everything an office REIT or office lender does. A national vacancy headline, whether 17.6% or 21%, is fine for narrating a trend and useless for valuing a building. The dispersion is large enough that two assets in the same MSA, even at the same building class, can support valuations 30 to 50% apart purely on submarket and micro-location. Credible office work captures that variance explicitly rather than applying a metro average and hoping the asset behaves like the mean, a discipline the US office market in structural transition framework treats as the baseline expectation for the sector.

So the question to carry into any office analysis is not "how is office doing" but a stack of three: which metro, which submarket within it, and which quality tier within that submarket. The answer to the first barely narrows the range; the answer to the third often decides whether the asset is a 3% trophy or a 30% write-down candidate. That ordering, from least to most explanatory, is itself the most useful thing to understand about how office submarkets behave today.