The Frozen Housing Market and What It Means for Self-Storage
Harvard's 2026 State of the Nation's Housing describes a housing market frozen at a three-decade low. Moving is the biggest single driver of storage demand. The report, translated into storage terms: 14 charts and the full PDF.
Stay in touch
Follow the business, or ask about investing.
Both paths go to a principal rather than a sales desk.
Interested investors are also added to the market letter. Choose Newsletter only if you want the letter without being contacted about opportunities.
4.1M
existing-home sales, a 30-yr low
11.2%
of households moved, a record low
$3,100
all-in monthly cost, median home, up from $1,700 in early 2020
65.2%
homeownership, down 2 yrs running
321K
2026 net immigration (projected)
Storage demand starts with a move. The classic rental triggers (moving, downsizing, divorce, death, renovation) all sit downstream of the housing market. The most common and most measurable one is moving.
So I read Harvard's annual State of the Nation's Housing as a demand report for storage. The 2026 edition came out in June. It describes a housing market that has frozen.
Here's what it says, with the data laid out so you can check my work.
What this is
Personal research, not investment advice. Verify the numbers against the primary sources linked at the bottom.
1. Housing turnover is frozen
Existing-home sales: about 4.1 million in 2025. Flat versus 2024. The report calls it a 30-year low, first hit in 2023 and stuck there since. As recently as 2021, 6.1 million homes sold.
A home that doesn't sell is a move that doesn't happen. A move that doesn't happen is a unit that doesn't get rented.
Chart
Existing-home sales are stuck at a three-decade low
Existing single-family + condo sales, millions of homes, annual
The only comparable stretch since 2000 was 2010–11, the foreclosure era. Sales have now sat near 4 million for three straight years, at two-thirds of the 2021 pace.
Source: NAR Existing Home Sales, annual, via FRED (EXHOSLUSM495S); 2025 per Harvard JCHS 2026. JCHS Figure 17 ("Home Sales Remain Near Three-Decade Lows") shows the same series from 1995.
UPDATE · JUL 2026
Since the report's 2025 annual figure: May 2026 existing-home sales printed 4.17 million SAAR, up 3.2% from April and about 3.5% from a year earlier. Off the bottom, but still a frozen market by any historical measure. (NAR via FRED, May 2026 data)
Mobility: 11.2% of households moved in 2024. Lowest in the 20-year history of the American Community Survey.
The decline is concentrated in young homeowners. 13.8% of under-45 owners moved in 2021. 9.8% in 2024. Those are the first-time and trade-up buyers who generate the most moves.
Chart
Young homeowners are moving far less
Share of homeowners under age 45 who moved in the prior year
Overall household mobility hit a record-low 11.2% in 2024.
Source: Harvard JCHS 2026, tabulations of US Census Bureau American Community Survey.
Household growth slowed for a third straight year: 2.0 million new households a year in 2020–21, 1.1 million in 2025.
The report's lead explanation is jobs. The economy added 116,000 jobs in 2025, the fewest in a non-recession year since 2002. Young adults without jobs don't form households. They double up.
Chart
Household growth has slowed for three straight years
Annual growth in US households, millions
Back inside the 2010s range after the pandemic surge. Fewer new households means fewer first-move-in events.
Source: Harvard JCHS 2026, Figure 11 (Census Housing Vacancy Surveys). 2010–19 values read from the published figure (±0.05M); 2020–21 shown as a 2019–21 average due to pandemic data-collection issues.
Chart
Job growth fell off a cliff in 2025
Annual change in US nonfarm employment, millions
116,000 jobs added in 2025, the fewest in a non-recession year since 2002. No job, no new household, no move.
Source: US BLS Current Employment Statistics via FRED; per Harvard JCHS 2026, Figure 12. 2015–23 values approximate (read from the published figure); 2024–25 as stated in the report.
It's not just jobs pushing household formation later. New CDC birth data shows early-20s birth rates fell below late-30s rates in mid-2025, for the first time on record. By Q1 2026: 52.4 births per 1,000 women aged 20–24, versus 54.8 for women 35–39.
A new baby is a classic storage trigger. Families starting later means that trigger fires later, into smaller homes, with more accumulated stuff.
Chart
Americans start families later
Births per 1,000 females by age group, quarterly
The lines crossed in mid-2025. For the first time, women in their late 30s are having children at a higher rate than women in their early 20s.
Source: CDC quarterly provisional birth data, via John Burns Research and Consulting (Eric Finnigan), data through Q1 2026, published July 2026. Intermediate values approximate, read from the published chart; Q1 2026 endpoints as labeled (52.4 / 54.8).
Immigration: net international migration fell from 2.7 million in 2024 to 1.3 million in 2025. The Census Bureau projects 321,000 in 2026, about a third of the 2001–19 average (~900K). That pulls a real slug of future household growth out of the system.
Chart
Immigration is collapsing as a demand driver
Net international migration to the US, millions of people
A projected 321,000 in 2026 would be roughly a third of the 2001–19 average.
Source: Harvard JCHS 2026, tabulations of US Census Bureau Population Estimates Program (2026 figure is the Census projection).
2. Affordability is doing the freezing
All-in monthly cost on the median-priced home: $3,100 in Q4 2025, versus $1,700 in early 2020 (JCHS: 3.5% down, 30-year fixed, plus mortgage insurance, property insurance, and property tax). Income needed to afford it: over $120,000, versus $68,700 in 2020.
Prices are up 54% since 2020. The median existing home runs nearly 5× median income. The 1990s norm was 3.2.
Chart
The all-in cost of owning the median home nearly doubled
Total monthly owner cost on the median-priced US home
Income needed to afford this payment: over $120,000, vs $68,700 in 2020.
Source: Harvard JCHS 2026 (ongoing homeowner costs). Assumes 3.5% down, 30-year fixed, plus 0.55% mortgage insurance, 0.35% property insurance, and 1.15% property tax.
The lock-in: rates ran above 6.5% in 2024 and 2025, but about half of outstanding mortgages still carry a rate below 4%. Moving means roughly doubling your rate. Most owners don't.
For storage that's a direct hit. Lock-in suppresses exactly the owner-to-owner moves that fill units.
Chart
The rate lock-in, in one line
Average 30-year fixed mortgage rate, annual
Rates jumped from ~3% in 2021 to ~6.5%+ in 2023–25. With about half of outstanding mortgages below 4%, moving means roughly doubling your rate. So owners don't.
Source: Freddie Mac Primary Mortgage Market Survey via FRED (MORTGAGE30US), annual averages; 2026 is the year-to-date average (weekly prints have ranged ~6.0–6.5%). Lock-in share from the National Mortgage Database via JCHS 2026.
UPDATE · JUL 2026
The 30-year fixed printed 6.49% the week of June 25 and has ranged roughly 6.0–6.5% all year. Lock-in fully intact. (Freddie Mac PMMS via FRED)
Price growth is cooling fast: 20.7% in 2022, 4.0% in 2025, 0.7% by February 2026. 41 of the 100 largest metros are posting outright declines. That's the market straining to clear at payments buyers can't make.
Chart
Home-price growth has gone from frenzy to flat
Annual home-price appreciation, US
Prices haven't fallen nationally, but 41 of the top 100 metros saw outright declines in early 2026.
Source: S&P CoreLogic Case-Shiller US National Home Price Index via JCHS 2026 (the index was renamed S&P Cotality Case-Shiller in 2025; JCHS Figure 16 uses the former name).
UPDATE · JUL 2026
The report's price data ran through February. April's print: +0.8% YoY nationally, and prices have now fallen in real terms for 11 straight months. The spread is wide: Chicago +6.5%, Seattle −2.3%. (S&P Cotality Case-Shiller, April 2026 data, published June 30)
Homeownership: 65.2%, down a second straight year. Under-35: 37.0%, from 39.0% in 2022. More would-be buyers renting longer. That reshuffles where storage demand comes from even as it shrinks the move-driven core.
Chart
Homeownership has drifted lower since the 2020 blip
US homeownership rate, annual
Down a second straight year to 65.2%. The under-35 rate fell from 39.0% (2022) to 37.0% (2025).
Source: US Census Bureau Housing Vacancy Survey via FRED (RHORUSQ156N); 2020 affected by pandemic data-collection issues. Under-35 figures from JCHS 2026.
3. Supply is pulling back
Single-family starts fell 7% in 2025. Total starts slipped 1%. Builders are pivoting to smaller homes, smaller lots, and build-to-rent, now 11% of single-family completions, nearly 3× the historical average. Multifamily is past its peak, with completions and units under construction falling.
Smaller homes mean less room for stuff. That part cuts storage's way.
Chart
Construction is pulling back, led by single-family
US housing starts, thousands of units (annual)
Single-family starts fell 7% in 2025; total starts slipped 1%.
Source: US Census Bureau, New Residential Construction via FRED (HOUST1F; multifamily = HOUST − HOUST1F).
UPDATE · JUL 2026
Since the report's cutoff: May 2026 starts fell 15.4% to 1.18 million SAAR, the lowest since May 2020. Single-family hit an eight-month low of 882,000. The pullback is accelerating, not stabilizing. (Census New Residential Construction, June 16, 2026)
Vacancy is rebounding too. Rental vacancy hit a 40-year low of 5.6% in Q2 2022; it's back to 7.3% as of Q1 2026, near the 1990s norm of 7.7%. For-sale vacancy went from a 70-year low of 0.7% in 2023 to about 1.1%.
More slack in housing means less urgency behind every move.
Chart
Rental vacancy has rebounded off its 2022 record low
US rental vacancy rate, annual (Q1 2026 latest)
From a 40-year low of 5.6% in Q2 2022 back toward the 1990s average of 7.7%. Housing is loosening from the supply side even as demand freezes.
Source: US Census Bureau Housing Vacancy Survey via FRED; 2022 low and Q1 2026 per Harvard JCHS 2026, Figure 3 (1990s average: 7.7%). Intermediate years approximate — spot-check against the report's figure data.
4. What it means for storage
The move-driven core of demand is contracting. Sales at a 30-year low. Mobility at a record low. Household formation back at its 2010s pace. Immigration falling away.
That matches what operators are reporting: occupancy off its highs, limited pricing power on new move-ins.
My own tracking says the same, with a twist. The Storbord Demand Index level is low, about 39 out of 100 as of early 2026. But momentum has firmed over the past year, from 50.2 in May 2025 to 58.9 in June 2026.
The momentum score weights the reasons people rent storage (per the SSA 2024 demand study) and scores each proxy's YoY change against a calibrated neutral. 50 means normal. The move to 59 comes almost entirely from three inputs.
Existing-home sales flipped sign: roughly −2% YoY a year ago, +3.2% now. That's the largest input in the score (30.5% weight) and nearly half the gain by itself. Business applications accelerated from +6% to +17% YoY, against a ~3% historical norm. Consumer spending went from neutral (+5.0%) to modestly supportive (+6.3%). The main drag: births, down 4% YoY.
If it seems odd that momentum is rising while every housing chart above points down: the index keys on rate of change, not level. The levels are frozen. The freshest monthly deltas are improving.
Demand is weak. It may have stopped getting weaker.
Chart
Storbord's read: demand momentum is firming off a low base
Storbord Demand Index — momentum sub-score (0–100), monthly
Shown on the full 0–100 scale so the move isn't overstated. This is a firming off a low base, not a boom. Level says 'frozen.' Momentum says 'maybe past the bottom.' Both can be true at once.
Source: Storbord Demand Index (composite of public demand indicators). Level sub-score ≈ 39 (Q1 2026); momentum shown here. No May '26 point: the index skips months without a new source print.
Chart
What's driving the move above neutral
Contribution to the June 2026 momentum score, points above or below a neutral 50
Contributions sum to ≈ +8.9 points over neutral. The comparatives a year ago: home sales about −2% YoY, business applications +6%, spending +5.0% (neutral). The sign flip in home sales is nearly half the move.
Source: Storbord Demand Index component data, June 2026 run. SSA 2024 reason weights × per-proxy YoY vs calibrated neutral. 8 of 13 demand reasons covered (no live proxy for remodeling, vacation-home, college, or climate-control demand); confidence: silver. The January 2026 business-applications print (+36% YoY) was flagged as base-effect noise and not used.
The offsets
Not all downside. Four things in this report support storage demand: smaller new homes (less room per household), young adults doubling up, build-to-rent growth (renters move ~22.6% a year vs ~5.1% for owners; JCHS 2026, ACS), and prolonged renting.
5. What I'm watching
Delayed moves don't disappear. They stack up. Lock-in owners, doubled-up young adults, renters waiting to buy: a growing backlog of households that would move if they could. When rates break the stalemate, that backlog releases faster than most models expect.
Three numbers
(1) The 30-year mortgage rate. The master switch on the lock-in. (2) Existing-home sales. A sustained move above 4.5 million signals a thaw. (3) Mobility and household formation in next year's data. Building backlog or releasing it.
Housing is the master variable for storage demand. Right now it says frozen. The job is to spot the thaw before it shows up in occupancy.
Read the full report
Harvard Joint Center for Housing Studies · June 2026 · free/public
View the report page (figure data in Excel & PowerPoint)Sources
Sources
- Harvard JCHS — The State of the Nation's Housing 2026 (full PDF)
- Harvard JCHS — report landing page (figure data, Excel & PPT)
- Harvard JCHS — Ten Takeaways from the 2026 report
- CDC quarterly provisional birth rates, via John Burns Research and Consulting (Eric Finnigan, LinkedIn, July 2026)
- FRED — 30-Year Fixed Mortgage Average (MORTGAGE30US)
- FRED — Homeownership Rate (RHORUSQ156N)
- FRED — Housing Starts: single-family (HOUST1F) & total (HOUST)
- FRED — Existing Home Sales (EXHOSLUSM495S)
- US Census Bureau — New Residential Construction, May 2026 (June 16 release)
- S&P Cotality Case-Shiller — April 2026 index announcement (June 30)
