Questioning Construction Productivity

What’s the productivity benchmark for a guitar-shaped building? (Miami Herald).

This week’s post was inspired by a LinkedIn post from Eric Koehler. Hello Eric.

Eric is a sharp, thoughtful guy and someone I respect. He was also a guest on the Concrete Logic Podcast way back in the early days. Episode #032. Audio only, the way podcasts should be. That is a discussion for another post.

Eric shared an article by Brian Potter titled “Stagnant Construction Productivity Is a Worldwide Problem,” which looks at how construction productivity has changed across countries using KLEMS data.

It is a thoughtful piece and, to Brian’s credit, he is careful to point out how messy and assumption heavy productivity data really is.

Reading it made me want to take some time to question what we are really comparing when we say manufacturing productivity has improved by X percent over the last 40 years while construction productivity has flatlined or declined by X percent.

That framing is common. It is also where things start to get shaky.

The comparison is not that manufacturing and construction use the exact same productivity metric. The comparison is that manufacturing shows steady improvement over time, while construction does not. The natural follow up question is why.

That is a fair question. But it assumes the underlying work is evolving in comparable ways.

Manufacturing productivity improves through repetition, standardization, controlled environments, and design for production. The output may change slowly, but the process is relentlessly optimized.

Construction productivity is measured against a moving target.

Even within concrete construction, the work today is not the work of 40 years ago. A concrete framed residential tower with 60 or 80 repetitive floors is fundamentally different from a university arts building with irregular geometry, mixed structural systems, transfer levels, architectural concrete, and constantly changing details. Both involve concrete. Both involve labor. But the efficient labor profile for each is completely different.

That is why even something that sounds more grounded, like cubic yards of concrete placed per labor hour, starts to fall apart when you look closer. Yards per hour might look great on a repetitive high rise with flying forms and stable crews. The same metric will look terrible on a complex, one off building where concrete is only part of the structural system and every level is different.

Same industry. Same material. Very different work.

Over time, construction has taken on more complexity, not less. Concrete today carries higher loads, tighter tolerances, more embeds, more penetrations, more coordination with other trades, and higher consequences for getting it wrong. Mix designs are more sophisticated. Reinforcement is more congested. Post tensioning is common. Performance expectations are higher.

Codes reflect that shift. Concrete construction today operates under far more detailed and demanding standards than it did decades ago. ACI 318 was updated again this year, and like every previous cycle, the newest version increases design checks, detailing requirements, and documentation expectations. Those requirements translate directly into labor hours that did not exist in earlier eras.

Another thing that almost never shows up in productivity charts is how much quality requirements have expanded, especially on certain types of projects. I learned more about NQA-1 requirements this week for nuclear related work, and it is a good example of how far things have moved.

Under NQA-1, the work is not just about placing concrete. It is about documentation, traceability, material controls, inspection hold points, personnel qualifications, and verification steps that did not exist on most projects decades ago. Every one of those steps takes time. None of them place a single extra cubic yard.

The same is true for U.S. Army Corps of Engineers projects. Corps work comes with layers of quality requirements that fundamentally change how production happens. More inspections. More testing. More submittals. More approvals. More pauses in the work. That is not inefficiency. That is the job.

When productivity metrics treat all labor hours the same, they ignore whether those hours were spent placing concrete or satisfying mandatory quality controls. A project executed under NQA-1 or Corps requirements will almost always look less productive on paper, even though it is delivering a much higher level of reliability, durability, and risk reduction.

That is a trade most owners are intentionally making. But the productivity charts never acknowledge it.

So when we say construction productivity has not improved at the same rate as manufacturing, we should be honest about what that statement is really saying. It is not just that construction failed to optimize. It is that construction absorbed decades of added complexity, regulation, risk, and performance expectations, and did so while largely maintaining output.

There are also improvements that do not show up cleanly in productivity trendlines. Concrete construction today is safer. Failures are less tolerated. Rework is more expensive. Schedules are tighter. Quality expectations are higher. Structures are expected to last longer and perform better.

If we are delivering more complex buildings, under stricter rules, with better safety outcomes and longer service lives, calling that flat productivity feels incomplete at best.

I do not think the right takeaway is that construction productivity is broken while manufacturing figured it out. I think the takeaway is that productivity improvement in construction is far more context dependent, project dependent, and complexity dependent than the charts suggest.

If high productivity is truly the goal, we should probably start by designing for it, measuring it honestly, and acknowledging how much the work itself has changed.

Otherwise, we are just comparing trendlines and wondering why two very different worlds do not move the same way.