The Office Market Isn't Dying. It's Dividing — and the Divide Decides Whose Dividend Survives

The obituary was written years ago, and it is still how most people talk about offices: remote work killed the building, vacancy hovers around 19%, the whole asset class is a value trap that income investors should avoid. The data tells a more interesting story. The U.S. office market is no longer one market. It is splitting into two — and for anyone holding or sizing an office REIT, which half you own now decides whether the dividend is durable or a trap.

Two markets hiding behind one vacancy number

The national headline vacancy of roughly 19% is real, but it is not evenly spread. Prime office space — the newest, best-located, most amenitized buildings — sits at about 12.7% vacant, and that number is falling at 80 basis points a quarter, the steepest pace since 2018. The gap between prime and everything else has widened to 640 basis points, the largest on record. CBRE's framing is blunt about where the problem actually lives: the mismatch is not a permanent collapse in demand, but an oversupply of dated, functionally obsolete buildings. One estimate puts 90% of all U.S. office vacancy inside the bottom 30% of the building stock.

The stock is shrinking fast. Since 2024, four out of five tracked markets have seen less office supply, as landlords demolish or convert tired Class B and C buildings into residential and other uses. For the first time since CBRE began tracking in 1988, removals outpaced new completions in 2025. That is a meaningful turn: the glut that crushed office real estate is now being worked off from the bottom, while almost nothing new is being built at the top.

Why the top can raise rents and the bottom cannot

This is where pricing power reasserts itself. Hybrid work did not end demand for offices; it compressed demand into fewer, better square feet. Companies that used to spread across floors now want one excellent, amenitized, transit-connected building that makes the commute worth it — and they will pay for it. Trophy and Class A buildings command an average rent premium of roughly 50% over Class B in major markets; in Houston, quoted rents on trophy space rose 5.2% year over year against 2.6% for the market overall. The very best new buildings are reportedly leasing at a near-60% premium versus a roughly 40% historical norm.

The old buildings have no such lever. A 40-year-old tower competing with a flood of vacancies cannot raise rents without losing the tenant, which is why the pricing-power filter I look for in every income business applies here with unusual force. The buildings built before the 1990s have absorbed roughly 236 million square feet of negative demand since the pandemic, while assets built since 2015 have added about 127 million. The moat at the top of the market is being rebuilt; the base is still being hollowed out.

The recovery is real, but it cascades from the top, slowly

Office fundamentals are a leading-indicator business, and prime space is the leading indicator everyone should watch. In Manhattan, prime vacancy peaked at the end of 2022, a full year before overall vacancy did; nationally, prime turned late in 2024 before the broader market improved. CBRE expects prime vacancy to return to pre-pandemic levels by the end of 2027 and argues the tightening will "cascade" into the next tier of buildings as the best space runs out — historically, demand spills over once prime vacancy falls to roughly the 10–12% zone.

Keep the caveats honest, because they decide the timing. This is a thesis, not a fact. The recovery is uneven: leasing is still about 21% below pre-COVID norms, major laggards like Chicago and Washington, D.C. trail by 35–40%, and San Francisco is an exception with prime vacancy still near 14.4% even as AI firms lease aggressively. The path works only if prime vacancy keeps falling. That is the single live variable, and it is worth checking on the way in rather than assuming.

Which half you own decides the dividend

For an income-focused investor, the split changes how office REIT yields should be read. In this market the highest headline yields tend to sit on the wrong side of the divide, precisely because investors are pricing in impairment. City Office, a landlord with heavy exposure to secondary commodity space, trades around $7. Alexandria Real Estate, a life-science-office owner, cut its dividend 45% in December 2025 to conserve roughly $410 million as lab demand faded. Canada's largest office REIT, Allied Properties, slashed its payout 60%.

Even the quality leader is not immune. Boston Properties, the largest publicly traded owner of premier Class A offices in the country, had extended its dividend every year for a quarter century before cutting it 29% in September 2025, to $0.70 a quarter, to free up roughly $50 million per quarter. The cut came as a funding reset, not as a collapse — its Q2 2026 occupancy rose 100 basis points on strong gateway-market leasing — but it is the clearest evidence that portfolio quality alone does not immunize a payout. Coverage and balance sheet matter, and Boston Properties carries net debt of about $15.6 billion, debt-to-equity near two to one on book equity of roughly $8 billion, and free cash flow after capital spending of only a few hundred million dollars.

That is the practical rule the two-tiered market hands you. Screen for prime, recently built, gateway assets with real rent momentum, but then check the coverage ratios and leverage just as hard. A high yield on a commodity landlord is a warning sign, not a bargain; a moderate yield on a prime owner with a covered payout and manageable debt is the better risk/reward, accepting a slower, cyclical recovery in exchange for a dividend more likely to survive it.

The "death of office" was always too neat. The market is dividing, supply is finally shrinking, and pricing power is coming back to the buildings that deserve it. The opportunity is real but early, uneven, and unforgiving of leverage — so own the prime half, check the balance sheet, and size for the volatility the rest of the way down the cascade.