The 100-Year Building Problem: Why We Stopped Building Things That Last

A Dome With No Steel In It, Still Standing

In 128 CE, Roman engineers finished pouring the dome of the Pantheon. It remains, to this day, the largest unreinforced concrete dome on Earth — no rebar, no steel mesh, nothing but concrete holding concrete up. Nearly two thousand years of earthquakes, weather, and Roman traffic later, it's still there, still in use, still structurally sound enough to walk under without a hard hat.

Meanwhile, under the current European design code — the standard much of the modern world builds to — an ordinary building only has to be engineered to last 50 years. Not because that's the best we can do. Because that's the number the industry decided was good enough.

This isn't a story about ancient civilizations being smarter than us. In almost every technical sense, we understand concrete, loads, and materials science far better than the Romans did. This is a story about what happens when durability stops being the goal — and something else quietly takes its place.

First, the Myth That Needs Correcting

You've probably heard the claim that houses in Japan only last 30 years. It's one of the most repeated statistics in construction writing, and it's also, in the way it's usually presented, misleading — which makes it a good place to start, because the correction reveals something important.

The distinction matters because it points to the real driver: in Japan, homes are demolished young not because they physically fail, but because of economic depreciation. Research on Japanese housing transactions found that the structure's value — separate from the land it sits on — can lose as much as 50% of its worth within the first ten years, continuing to depreciate until around year thirty. A study of Japanese office building owners found that 74% preferred to rebuild within 30–50 years, even for reinforced-concrete structures explicitly designed for a 100-year service life.

So the "buildings don't last" story is really two separate stories tangled together: one about what materials can physically survive, and one about what a market decides is worth keeping. Let's take them one at a time — starting with the concrete itself.

Why Roman Concrete Outlasts Modern Concrete

For decades, the durability of Roman concrete was credited entirely to volcanic ash — pozzolana — shipped in from the Bay of Naples. That's part of the story, but in 2023 a team at MIT and Harvard found something that had been overlooked, sitting inside the concrete in plain sight the whole time: small white mineral deposits called lime clasts, long dismissed by modern standards as evidence of sloppy mixing.

They weren't sloppy. They were the point. Researchers cracked samples of concrete made with the ancient "hot mixing" technique — using highly reactive quicklime instead of the gentler slaked lime modern methods favor — and ran water through the cracks. Within two weeks, the cracks in the ancient-formula concrete had sealed themselves. A calcium-rich reaction inside the lime clasts had recrystallized as calcium carbonate, gluing the fracture shut before it could spread. An identical sample made without the lime clasts never healed at all.

Modern reinforced concrete does not have this property, and worse, it has a built-in weakness the Romans' unreinforced concrete never faced at all: steel.

This isn't a defect or a design mistake — it's a known, modeled, engineered-around process. Modern codes manage it deliberately, mainly through concrete cover (the depth of concrete between the surface and the rebar) and mix design. It's also exactly why the Romans' steel-free approach, while unsuitable for the tension-heavy structures modern buildings need steel for, will physically outlast almost anything we build with rebar today, corrosion-free by design rather than by maintenance.

The Number Nobody Tells You: How Long Your Building Is Actually Designed to Last

Every building built to a modern structural code has an explicit, numeric design life baked into its engineering — most people who live or work in these buildings have simply never been told what it is.

Fifty years is not a physical ceiling — plenty of 50-year-design buildings, well maintained, will comfortably outlive their paperwork, the same way a car engineered for 150,000 miles doesn't explode at 150,001. But it is a revealing number, because it tells you what the standard-setters considered an economically reasonable target to design and insure against, not the maximum a building could achieve if durability were the priority.

Why We Keep Choosing the Shorter Number Anyway

If concrete science and design codes can both support buildings lasting well over a century, the obvious question is why so much of what gets built doesn't get anywhere close. The honest answer is almost never structural. It's economic.

• Land, not structure, holds the value. In many markets — Japan's being the most studied example — a building depreciates like a car while the land beneath it appreciates like an asset. Once the structure's accounting value approaches zero, there's little financial incentive to preserve it, and every incentive to demolish and rebuild something that matches current tastes, codes, or density allowances.
• Post-war materials optimized for speed, not lifespan. Much of the housing stock built in rapid post-war booms — across Japan, Europe, and elsewhere — used newly available synthetic materials (vinyl siding, chemical adhesives) that were cheaper and faster to install but degrade over 20–30 years, compared to the natural materials and craft methods they replaced.
• Tax and depreciation schedules reward rebuilding. Depreciation-based tax incentives in several countries make replacing an aging structure more financially attractive than maintaining or renovating it — even when the structure remains physically sound.
• Fast-tracked urban growth prioritizes throughput over permanence. When a country needs millions of housing units quickly, the calculus shifts from "how long will this last" to "how many can we build this year" — a trade-off every rapidly urbanizing economy has made at some point, including, historically, Japan, and today, much of South and Southeast Asia.

None of these forces are about what's technically possible. They're about what gets rewarded. A structure engineered for a 100-year life that gets torn down at 30 isn't a failure of engineering — it's a decision, made somewhere upstream of the site, that thirty years of value was all anyone needed to capture.

The Bill Nobody Puts on the Invoice

Every demolished building takes its embodied carbon with it — the emissions that went into producing its cement, steel, and materials in the first place, released for nothing once the structure is rubble. At a global scale, this adds up to a genuinely large environmental line item that almost never appears on any individual project's budget.

Researchers modelling this trade-off have made the underlying math explicit: extending average building lifespans by even 10–20% would meaningfully cut the sector's environmental burden, precisely because so much of a building's total lifetime emissions are locked in during construction, not accumulated slowly during use. A building that stands for 100 years amortizes that upfront carbon cost over twice as long as one demolished at 50 — the same embodied emissions, doing twice the useful work.

What Engineers Are Actually Doing About It

None of this means the industry is standing still. Several genuinely promising threads are converging on the same goal — buildings that last longer without costing dramatically more:

Self-healing concrete research, directly inspired by the Roman lime-clast discovery. Engineers are now experimenting with modern concrete formulations that reintroduce reactive lime compounds, aiming to give contemporary concrete some of the same crack-sealing behavior — potentially extending service life and reducing the maintenance cycle that currently drives early demolition.

Corrosion-resistant reinforcement in critical, hard-to-maintain locations. Stainless steel rebar costs 6–10 times more than ordinary steel, which makes it impractical everywhere — but engineers increasingly concentrate it specifically at the joints and exposure zones most prone to corrosion, extending a structure's weakest points without paying that premium across the entire building.

Design-for-disassembly. Rather than assuming every building will eventually be demolished and landfilled, some architects now design structural connections (bolted rather than welded, modular rather than monolithic) specifically so that a building's components can be unbolted, salvaged, and reused when its current use ends — turning demolition waste into a supply chain instead of landfill.

Performance-based durability codes, not just prescriptive ones. The next generation of concrete codes is moving toward directly specifying and verifying actual service-life performance — how long a specific mix, cover depth, and exposure condition will really last — rather than relying only on generic tables that don't reflect a specific project's real environment.

What This Means If You're Building Something Right Now

The Pantheon wasn't an accident, and neither is a building that's torn down at year thirty. Both are the predictable result of what their builders optimized for. If you're planning a project today, the honest question worth asking isn't just "what will this cost me to build" — it's "what am I actually optimizing for, and does my material and quality choice match that answer?" A slightly higher-spec concrete mix, better cover depth, or corrosion protection at the right joints can be the difference between a building that needs major structural repair at year 35 and one still standing, structurally unremarkable, in year 90.

That decision starts at the estimating stage — long before the first foundation is poured.