News & insights

Most of Europe's housing was built in a single generation, during the thirty-year boom that followed the Second World War. Two-thirds of what we live in today comes from that period, and it is now turning eighty - the age at which keeping a building going starts to get hard.

That forces a decision, at continental scale, on every owner of a postwar block: refurbish it, or knock it down and rebuild. It is a decision made against a housing crisis that already has governments across Europe promising ever more ambitious building programmes.

The instinct is to refurbish. Europe can't afford to lose housing stock, and keeping a structure sounds cheaper and greener than replacing it. But both halves of that instinct are now questionable.

The crossover point

Take cost first. ARGE, the German building research institute, recently compared replacing an ageing multifamily building against deep modernisation. Replacement came out at roughly €2,290 to €2,420 per square metre of gross floor area including VAT, against €2,490 to €2,630 for modernisation. In most cases, knocking down and rebuilding already works out cheaper.

That crossover is a benchmark, not a rule, and the exact point shifts building to building. But it frames a decision facing an enormous pool of housing. Germany has around 17.75 million dwellings built between 1950 and 1979. Spread over a thirty-year cycle, roughly 600,000 a year reach the point where someone has to decide what happens next.

Very little of that stock is actually being replaced, and the refurbishment most of it receives is thinner than the word suggests. In Germany, 94% of energy-focused modernisation - insulation, windows, heating, roof and facade - is partial. Only 6% is comprehensive. This is piecemeal patching, not a rethink to deliver homes that work for twenty-first-century life.

There is a deeper problem the cost comparison excludes. A building from 1955 was designed for the life, and the weather, of 1955: low ceilings, small rooms, and no allowance for the heat Europe now experiences. A new building can be designed from the ground up for how people live and for the climate as it actually is - orientation, shading and ventilation chosen for today, broadband and sensors wired cleanly into open walls during the first fix. Refurbishment works around what is already there. Replacement starts from the right question.

The timber answer to the carbon objection

The strongest case for refurbishing is carbon - the embodied energy locked into materials before a building is even heated. Keeping a structure will always cost less carbon than building a new one, so on materials alone, refurbishment wins.

But materials are only half the story, and timber narrows even that half sharply. Agora Industrie puts median embodied carbon for a multifamily building at 106 kg CO₂e per square metre in engineered mass timber against 290 in reinforced concrete - around 65% lower. Once you count emissions across the whole life of the building rather than just its construction, the gap closes almost entirely: a German lifecycle study puts a timber rebuild practically level with refurbishment. The difference is that the rebuild holds that standard for eighty years, while the refurbishment faces its next intervention within thirty.

So why is timber not already the default? Because it costs more than concrete on a like-for-like build, and the carbon argument is aimed at the regulator, not the developer. Even where it wins with the regulator, the developer runs the numbers and reverses the decision on cost. Price is the lever that has to move. Bring a timber structure below concrete rather than merely close to it, and the environmental case and the economic case finally point the same way.

Move the crossover

This is where cost stops being a constraint and becomes the whole game. Imagine construction costs fell by 20% - not as a promise, but as a scenario worth taking seriously. Buildings that were borderline, or that failed the replacement test outright, start passing it.

ARGE's own transformation scenario already sees replacements roughly doubling, from around 114,000 a year to 224,000 by 2045, on current cost structures. Our own modelling suggests a reduction of that scale could tip a far larger share of the refurbishment pool - as much as half - into rebuild, on the order of 300,000 additional homes a year. Treat that as a working scenario rather than a forecast. The point holds regardless of the precise figure: a cost reduction of this size moves buildings across a threshold, rather than just shaving margins.

And the market number is the least of it. It means more people housed to a twenty-first-century standard rather than a patched 1955 one. It means more varied, more liveable cities, instead of postwar layouts frozen in place for another generation. And it means money spent once on something that lasts eighty years, rather than again and again on patches designed to last thirty.

Integration, not invention

How do you take 20% out of the cost of building? Not by inventing anything. Every piece already exists: robotic fabrication, design automation, on-site robotics and structural timber that stores carbon. The task is integration - combining proven components into one process, the way the smartphone combined a camera, a phone and a computer that had each existed for years.

That is what All3 is built to do. Our design software generates buildings that are compliant and, just as importantly, compatible with what our factory and on-site robotics are designed to produce and assemble. A decision made at the design stage stays buildable all the way to site, rather than being reworked when someone discovers it can't be made efficiently. That integration, more than any single machine, is where the saving comes from.

Europe will spend the coming decades working through the largest housing renewal wave in its history. The open question is not whether it rebuilds, but what standard it rebuilds to.