The latest quarterly data on apartment absorption came in surprisingly strong, and it’s time to take a fresh look at the trajectory of demand. When the first quarter readings came in soft, we started to once again hear the narrative about structural weaknesses. We were reminded that the job market is wobbling. Recent college graduates cannot find work. A record share of young adults has moved back in with mom and dad. Consumer confidence is in the basement. And on top of all that, the industry buried itself in the largest wave of new apartment supply since the 1970s. By any reasonable reading of the setup, renter demand should have buckled, right?
It didn’t. And the stubbornness of that demand is, to me, the most important and most underappreciated story in rental housing right now. Granted, a lot of the absorption has been at the expense of free rent up front, and rents are still soft in most markets, but it is a relief to see that the supply is getting mopped up. Concession burnoff will come next.
The second quarter numbers are better than most expected
Let’s start with the headline figure. Both RealPage Market Analytics and CoStar reported net absorption topping 250,000 apartment units in the first half of 2026 — a number RealPage economists confirmed in the firm’s Mid-Year Multifamily Update as "more than 250,000 additionally occupied units than we had going into this year." That is down from the frenzy of last year, but it still ranks among the strongest first halves in the history of either dataset — and beats every single pre-pandemic year on record. The second quarter did the heavy lifting: in its 2nd Quarter 2026 Data Update, RealPage reported that "the nation absorbed more than 187,000 units in 2nd quarter, at a pace that was notably above average for this high-performance time of year," pushing occupancy to 95.5%.
We are making headway against the flow of new unit completions. Net absorption outpaced new supply by roughly 100,000 units so far this year, as calculated by economist Jay Parsons. That gap is what finally let occupancy claw back a little of the ground it lost during the 2023–2025 supply deluge. CoStar reported that stabilized occupancy improved 20 basis points in the second quarter — the best quarter-over-quarter gain since 2021. Vacancy rates have been declining for four straight months, the first such streak since 2021.
CoStar shared the time series data graphed below for this article, and it makes it clear that demand is hanging tough, and the trend currently is in an encouraging (upward) direction.
Put plainly: we added more than a quarter-million net new renting households in six months, despite low consumer confidence. Sunbelt markets are showing more resilience in aggregate demand than other parts of the country, so your mileage may vary.
For those who aren’t “seeing it”
Some operators are pushing back on the narrative of improvement if they are not seeing the improvement in their own projects. Bear in mind that much of this demand is flowing into recently built communities still grinding through prolonged lease-up, concentrated in the high-supply Sun Belt and Mountain West metros. If you own stabilized product in a submarket getting hit with new deliveries next door, the national absorption number will feel like it belongs to someone else. It largely does — for now. We are still absorbing the aftershocks of the biggest completions wave in half a century, and until that inventory stabilizes, the demand gets spread thin and it stays a renter’s market. Also, the sheer number of projects in lease-up right now means less demand per project, because the large demand “pie” is being chopped into a lot of small slices. But that means something important – once the excess inventory has been absorbed, there will still be strong demand, and new projects leasing in 2028 will absorb rapidly.
But reduced move-outs to buy — while they do lift retention — cannot manufacture 250,000 net new renter households. Most would-be buyers who stay put were already renting. The additions are real.
I have noticed recently that some analysts have come out with downbeat-sounding reports, but those came out before the strong 2Q data were available. The uptick in April-June was noteworthy.
The quality of the demand is key
For those who build and finance rental product, the composition of this demand matters as much as the headline count, and it is quietly encouraging.
The renters signing these leases are not overextended. The bulk of them are spending somewhere in the neighborhood of 21% to 22% of income on rent. Rent-to-income for market-rate apartment renters is back to pre-COVID levels, and it got there the healthy way: wage growth has outrun rent growth for more than 40 consecutive months. That is the opposite of a demand pool stretched to its breaking point. It is a renter base with room to absorb rent increases when pricing power eventually returns.
There is also a structural story underneath the cyclical one. Renters now account for roughly 80% of all U.S. household formation, according to Arbor. And the "kids living at home" statistic that gets cited as a headwind is, in my view, deferred demand rather than lost demand. The National Multifamily Housing Council's Sharon Wilson Géno pointed out at the Harvard Joint Center for Housing Studies discussion of the State of the Nation's Housing 2026 report that roughly 2.5 million people aged 25 to 35 — about a third of that cohort — are living with their parents. That is not demand that evaporated. It is demand sitting in a coiled spring.
The supply cliff is what makes this a setup, not just a data point
I spend most of my time on the supply side of the ledger, and this is where the demand story turns actionable.
The wave that flooded the market is receding fast. Quarterly completions in the first quarter of 2026 came in 53% below the Q3 2024 peak. Starts have fallen even harder. CoStar pegged first-quarter starts as low as 55,000 units — the lowest quarterly level since 2011. The under-construction pipeline has been cut roughly in half from its early-2023 peak. When you deliver only about 150,000 units in a half-year — close to pre-COVID norms — against 250,000-plus units of absorption, vacancy has nowhere to go but down.
That is the equation operators have been waiting on. The goal for now isn’t for absorption to stay at peak levels; it's simply that absorption not fall faster than supply. As long as supply keeps dropping faster than demand, occupancy keeps healing. Strong first-half absorption paired with a collapsing delivery schedule is the classic precondition for pricing power to return in the back half of the year — modestly and unevenly, market by market, but return nonetheless. We model it as a “concession burnoff” forecast in our site-specific studies.
Let’s look at some of the markets around the country, scaling the data so that the largest ones don’t swamp the rankings. Looking at 2Q absorption as a percentage of the inventory in the market, Raleigh, Charlotte, Austin, Denver, Nashville, and Phoenix top the list.
I would temper the enthusiasm in one direction only. This is a recovery, not a boom. There is still a lot of lost ground to make up on both occupancy and rent, and Yardi Matrix and others rightly caution that plenty of markets — Tampa and Austin among them — are still mid-wave with substantial inventory to lease before they turn. Demand could still soften if the labor market deteriorates further. It is fair to question whether this pace of absorption will persist, but one can’t dispute the strong absorption we have recently recorded.
Determinant of Rental Demand: Renter Household Formations
When trying to forecast future demand for apartments, it is useful to take a look at the number of new renter households being formed, each of which represents demand for an additional rental unit. Renter household formation ran unusually hot in 2024. According to Harvard’s Joint Center for Housing Studies (JCHS), the renter population jumped by 848,000 in 2024 and continued growing into early 2025 — a pace well above the 524,000 annual average that renter households averaged from 2000 to 2025, and one that pushed the share of households that rent to nearly 35%, approaching the peak levels of the 2010s.
Looking ahead, the Center’s projections show renter growth cooling relative to 2024, but still remaining strong: under the base scenario, renter households are projected to add just 299,000 per year from 2025 to 2035. The low-homeownership scenario, which assumes persistent affordability barriers to buying (highly likely, in this author’s view), keeps renter growth near its historical norm at 523,000 per year.
The way I see it, things will probably come out somewhere in between the base scenario and the low-homeownership scenario, so likely in the range of 450,000 per year for the next five years, and 300,000 per year for the following five.
What it means for those who are building rentals
The signal for developers, capital partners, and build-to-rent operators is consistent across every dataset we all trust. The through-line is the same one I keep returning to in our own work: demand for rental housing in this country has proven far more resilient than the pessimists modeled, and the supply that scared everyone is getting mopped up on schedule. That combination doesn't guarantee a landlord's market tomorrow. But it is exactly the foundation you'd want under a multi-year recovery. The smart move right now is not to explain away the strength — it's to underwrite the setup it creates.