Pennsylvania's Warehouses Went Up Before WWII Ended. Nevada's Before the iPhone.

Stand in the middle of Pennsylvania's median small-bay industrial building and you're standing in 1940. The walls are load-bearing brick, thick enough to still be holding yesterday's temperature. Tall steel-sash windows in grids of small panes throw daylight across the floor, because when this shop was built, daylight was the lighting plan. The clear height is modest — a forklift mast has room, a modern racking system doesn't. Out back there's the faint scar of an old rail siding. The city filled in around the block decades ago; there's nowhere left to expand and there never will be. A machine shop still runs inside it.

Now fly to Nevada and stand in its median building. It went up in 1996, and everything about it says so. The walls are tilt-up concrete panels — poured flat on the slab and craned upright in an afternoon. Clear height jumps ten feet. A wall of reflective storefront glass fronts a small office; steel roll-up doors line the dock behind it. Outside is surface parking, an arterial road, a graded pad, and the same tilt-up building repeated in every direction out to the edge of the desert. It was built on spec, for a tenant it hadn't met yet.

Fifty-six years separate those two rooms. Same asset class — industrial, under 200,000 square feet — doing the same job for the same kind of tenant, and you'd never guess they were cousins. That 56-year spread, from Pennsylvania to Nevada, is the widest in the 22 states we measured, and it's the clearest evidence I've found that there isn't one American industrial market. There are two.

The Sun Belt boom is the story the sector has told for a decade — Texas, Florida, Arizona, the cranes, the rent growth. It's a real story, and it's mostly a rent story. Run the vintage instead — the year each building actually went up, across roughly 600,000 dated small-bay records in 22 states — and a different map surfaces underneath. It doesn't split into hot markets and cold ones. It splits into old and new. And both halves, read closely, point at the same word: scarcity. This is the third pass over the same universe — how old the stock is, then who owns it, and now where the old and the new actually sit — plus the twist that the difference between them may matter less than the map makes it look.

What counts as "small-bay" here

One definition before the numbers do any work. Small-bay in this study means industrial property under 200,000 square feet — the whole family that trades under a dozen names: small bay, shallow bay, light industrial, warehouse, flex office/warehouse, micro-bay. Both buildings above qualify. So does most of what your town's tradespeople rent.

That's who occupies it — not one or two tenant industries but hundreds of them: machine shops and welders, caterers and gyms, auto shops, and the e-commerce seller shipping out of a single unit. When a state's median building predates the interstate highway system, that's the roof a large slice of the working economy still operates under today.

Five numbers that hold the whole thing

56-year spread in median vintage — Pennsylvania (1940) to Nevada (1996). Same asset class, two different centuries.
Pennsylvania's stock is overwhelmingly pre-1970 — 76.9% of it went up before 1970, and only 2.3% since 2015.
California — the largest coastal market — has the tightest new supply of all 22 states: median vintage 1973, and just 2.1% built since 2015.
Texas is the youngest big market and still barely replenishing — median 1991, only 14.5% built since 2015. The best-building large state in the country is under 15%.
The 1982 national median from study #1 lands almost exactly on Oklahoma (median 1982) — the hinge between the old half of the map and the new one.

Which states hold the oldest small-bay stock?

The legacy industrial belt of the Northeast and Great Lakes; the newest sits in the Sun Belt and Mountain West. Pennsylvania's median dates to 1940 and New York's to 1953 — both older than the interstate highway system — and the gradient runs cleanly down from there to Texas (1991) and Nevada (1996). All 22 states, oldest to youngest:

| State | Records with year built | Median year built | Built since 2015 | Built before 1970 | |---|---|---|---|---| | Pennsylvania | 11,182 | 1940 | 2.3% | 76.9% | | New York | 24,543 | 1953 | 3.6% | 69.8% | | Wisconsin | 8,301 | 1961 | 4.6% | 59.8% | | New Jersey | 6,748 | 1965 | 2.8% | 57.4% | | Connecticut | 16,651 | 1966 | 3.3% | 54.4% | | Ohio | 14,441 | 1969 | 4.7% | 51.0% | | Massachusetts | 44,467 | 1970 | 4.8% | 47.9% | | Illinois | 20,067 | 1972 | 2.8% | 45.4% | | California | 52,816 | 1973 | 2.1% | 45.0% | | Michigan | 16,052 | 1975 | 1.6% | 40.8% | | Minnesota | 5,525 | 1976 | 5.1% | 41.1% | | Indiana | 16,778 | 1977 | 4.6% | 38.2% | | Maryland | 10,087 | 1979 | 4.5% | 36.0% | | Washington | 6,665 | 1980 | 5.8% | 31.9% | | Oklahoma | 9,594 | 1982 | 8.0% | 20.8% | | Tennessee | 5,324 | 1983 | 6.4% | 25.0% | | North Carolina | 21,894 | 1985 | 7.6% | 27.9% | | Arizona | 18,177 | 1986 | 8.6% | 18.3% | | Georgia | 10,394 | 1987 | 7.1% | 20.9% | | Florida | 158,094 | 1987 | 8.1% | 17.1% | | Texas | 115,185 | 1991 | 14.5% | 13.7% | | Nevada | 5,906 | 1996 | 7.2% | 12.3% |

I'd hold any single state number loosely — a statewide median is exactly the kind of average I distrust, and the real story lives at the metro and parcel level. But the gradient is impossible to miss: where land built out longest ago and industrial zoning is most contested, the stock is oldest, and the least new supply is coming. It squares with what the brokers see nationally — only about 5% of shallow-bay inventory has gone up since 2010, and more than 80% predates 2000 (CBRE, March 2026). Our state map is that fact, disaggregated.

Which states are actually building anything?

The Sun Belt and the Mountain West — and even they are barely keeping up. Rank the 22 by share built since 2015 and the leaders are Arizona (8.6%), Florida (8.1%), Oklahoma (8.0%), North Carolina (7.6%), Nevada (7.2%), and Georgia (7.1%). Texas runs away with it at 14.5% — roughly double the next state. Then look at what "runs away with it" means: the single best-building large industrial market in America has replaced fewer than one building in six since 2015. Everywhere else, it's closer to one in twelve.

The national pipeline says the same thing. There's about 23 million SF of sub-100,000-SF space under construction nationwide — under 0.3% of total industrial stock (NAIOP, Spring 2026). New small-bay isn't being withheld out of spite; it just doesn't pencil. Small buildings mean more demising walls, more doors, more tenant improvements per square foot, and they want expensive infill land — so developers build big boxes, or houses, instead. Even the young markets are adding supply at a trickle.

Old is not the same as broken.

Vintage measures a birthday, not a condition. A well-kept 1965 masonry building three miles from downtown beats a shiny 2024 box forty miles out — for the welder who has to drive to it every morning, and for the owner who could never rebuild it at today's land and labor. Age tells you the building can't be reproduced. It doesn't tell you the building has stopped working. Most of it hasn't.

Then why is the oldest stock the fullest?

Because working buildings get occupied regardless of their birth year. Sub-100,000-SF industrial is about 41% of all U.S. industrial inventory — the single largest size tranche in the country — and it runs the lowest vacancy of any size band, around 4.4% (Newmark, 4Q24). So the oldest, least-replaced product in the sector is also the most fully leased. Read that twice. Pennsylvania's 84-year-old median stock isn't a museum; it's fully occupied infrastructure the modern economy still runs on — and can't afford to rebuild.

The two industrial Americas

One America — the Northeast and Great Lakes — runs on small-bay buildings that are, functionally, irreplaceable infrastructure from another century. Pennsylvania (median 1940), New York (1953), Wisconsin (1961), New Jersey (1965), Connecticut (1966), Ohio (1969): every one carries a median older than the moon landing, and in each, half or more of the stock predates 1970. You can't buy this at replacement cost, because there's nothing to replace it with.

The other America — the Sun Belt and Mountain West — is younger, and at a glance it looks like the opposite market. It isn't. Nevada is practically still under warranty next to Pennsylvania — a 1996 median — and it has added all of 7.2% of its stock since 2015. Nearly 86% of Texas's dated small-bay was up before 2015. These aren't markets drowning themselves in new product. They started their construction cycle later, hit the same wall everyone else hit, and stopped.

So the Sun Belt boom the sector underwrote as a rent story is, at the small end, a vintage story with the clock set forward a few decades. Same shortage. Later start date.

So what does the vintage map actually say?

I'll keep this a read, not a recommendation. The intuitive take is that old markets are tired and new markets are the opportunity. I think the data points the other way. The oldest markets carry the deepest replacement-cost moats — you physically cannot rebuild Pennsylvania's or California's infill small-bay at today's land and construction costs, which is why 45 to 77% of their stock predates 1970 and still fills up. That scarcity is already in the land price.

The young markets? They're about one decade of underbuilding away from the same place. Texas, the best builder of the group, is already turning over under 15% of its stock per era, and the forces that froze small-bay everywhere else — land cost, disappearing industrial zoning, financing that prefers one big box to forty small suites — are tightening in Phoenix and Dallas too. It's why capital keeps drifting down-market: all-industrial cap rates sat in the mid-5% range in late 2025, and more than 70% of 2025 industrial sales volume was in deals under $100 million (per NAIOP and BKM). The historical discount small-bay trades at versus institutional big-box — my read, a higher cap rate, maybe 50 to 100 basis points wider — has been narrowing as that money hunts smaller assets.

Whether a market reads 1940 or 1996, the direction of travel is identical: older, tighter, harder to replace. What the map can't tell you is which individual building is functional and which is genuinely finished. That's the entire discipline of this sector — a parcel question, not a state-average one.

Frequently asked

Which U.S. state has the oldest small-bay industrial stock? Pennsylvania, in the states we studied — its median small-bay building was built in 1940, and 76.9% of the stock predates 1970. New York is next-oldest, with a median of 1953, per SpanVor's records as of July 2026.

Which state has the newest small-bay industrial? Nevada, with a median vintage of 1996, followed by Texas at 1991. But "newest" is relative: even Nevada has built only 7.2% of its stock since 2015, and nearly 86% of Texas's stock predates 2015.

Is the Sun Belt building enough new small-bay? No. Texas leads all 22 states at 14.5% of stock built since 2015 — under one building in six — and most Sun Belt states sit near 7–9%. Nationally, only about 23 million SF of sub-100,000-SF space is under construction, under 0.3% of stock (NAIOP, Spring 2026).

Does an older building mean a worse one? No — vintage measures when a building went up, not its condition. Sub-100,000-SF industrial runs the lowest vacancy of any size band (~4.4%, per Newmark), so the oldest, least-replaced product is also the most fully occupied.

How old is U.S. small-bay industrial overall? The national median is 1982, with just 7.3% of the stock built since 2015 — see our first study. This state map disaggregates that national figure into its two halves.

Methodology

Figures were computed across SpanVor's national industrial property universe — built from primary public records (county assessor rolls, permits, and land records) and classified industrial-only. This study covers the 22 states with at least 5,000 small-bay records carrying a recorded year built — nearly 600,000 dated buildings in all. "Small-bay" is the full under-200,000-SF industrial family — small bay, shallow bay, light industrial, warehouse, flex office/warehouse, and micro-bay (see the definition above). Median vintage and the "since 2015" and "before 1970" shares reflect the records that carry a recorded year built. Figures are current as of July 11, 2026 (Central Time), and describe the SpanVor platform's coverage.

Key takeaways

The median small-bay building spans 56 years by state — Pennsylvania (1940) to Nevada (1996). There isn't one industrial America; there are two.
The Northeast and Great Lakes run on irreplaceable stock — Pennsylvania (76.9%), New York (69.8%), Wisconsin (59.8%), and others have most of their small-bay built before 1970, and it still fills up.
Even the "new" markets barely replenish — Texas leads at just 14.5% built since 2015; California, the biggest coastal market, sits at 2.1%.
Old is not obsolete — sub-100,000-SF industrial runs the lowest vacancy of any size band (~4.4%, per Newmark); the least-replaced product is the most occupied.
The read: the oldest markets hold the deepest replacement-cost moats, and the newer ones are one decade of underbuilding from the same scarcity. The direction of travel is identical.

See it in your state. This is the national view; the useful version is your corridor, resolved to metros and parcels — which buildings are old, which are functional, and who owns them. Start at spanvor.com/state, then drill into your metro — every market on this map is mapped to the parcel.

One last thing, since you read this far: the code SpanvorBlog takes 25% off a SpanVor Pro subscription — where the parcel-level data behind posts like this one actually lives.

Written by Jason Probert, Founder of SpanVor — Industrial Property Intelligence.