The Construction Problem Problem

I stepped foot on my first construction job in 2005 for a summer's worth of work. We were building an oil refinery and the job was massive, with thousands of labourers, tradespeople, heavy machine operators, and support staff. I was immediate taken aback by how much of the day was taken up by safety measures. Every morning, the entire site went out and did guided stretches. Then, a "tailgate talk" occurred with every crew going over the key work to be done for the day and the safety measures to be accounted for. Following that, every smaller team had to fill out safety cards with 5-10 (at least) hazards to be aware of throughout the upcoming tasks. The site speed limit was 25 km/h. Every time you got close enough to certain chemical storage tanks, you had to sign in at a kiosk, and sign back out again when you left. And this was all after a full week of safety training to kick it all off.

"This is too much", I said to myself. "We're wasting so much time filling out paperwork - I just want to get my work done".

Then a worker lost a hand while working on a suspended platform. Then a bunch of people went down with heat stroke during a particularly hot week. Then a guy got hit with a falling wrench and got knocked out. Then the safety guy told me a story about how a guy he worked with fell from a suspended walkway. His harness caught him, but he was in an area that was tough to access. The workers didn't have a plan in place and took too long to get him down. Once he stepped back onto the platform, he fainted - cardiac arrest.

Figure 1: The famous McKinsey graph (restyled by the Economist) that shows construction & downward productivity trajectory

I very quickly went from feeling like there were too many safety measures to feeling like there weren't enough. This was almost 20 years ago, and I still remember those events like they happened yesterday.

The venture community gets caught up with the well-covered 50-year productivity flatline of the construction industry. We see (and present) the same graph over and over again. The tech community has formed a consensus that the industry is stuck in the past and can’t seem to figure itself out. But maybe we’re looking at the wrong thing. 

Construction company mottos are actively telling us about their true values – not speeding up, but keeping workers safe:

  • "Nobody gets hurt"

  • "Everyone goes home safe"

  • "Zero incidents" 

The disconnect occurs because construction isn't dealing with the same set of concerns as other industries. Namely, it's been trying really, really hard not to kill and maim such a large percentage of its workers (that constitute 7% of the global working-age population, by the way). And this is why the comparing innovation in the construction space to something like the manufacturing space is frustrating. Manufacturing is famously known for automation, increasing worker productivity, and being relentless adopters of innovation, while construction is, buy and large, the opposite. When you take this picture into account, you instead see an industry that has actually made tremendous strides in the past 50 years, though not in a metric accounted for in productive output metrics:

Figure 2: US deaths per 100 workers, compiled from American Federation of Labor and Congress of Industrial Organizations data, found here


When you speed up an efficient process with minimal complexity, you get a large uptick in throughout, and only a marginal increase in adverse outcomes (defective products, for example). When you speed up an inefficient process with high complexity, the opposite happens - you get a marginal increase in throughput and a large uptick in adverse outcomes (death and injuries, for example). Such is the case of comparing manufacturing to construction. 

As a reminder, Occupational Health and Safety Act, which entitles workers to a safe workplace, was only passed in 1970. I don't think it's an accident that this dataset starts in 1970 - when legislation began holding bad actors to account – and I can’t imagine how bad the numbers were before.

So, when we're talking about innovating the construction space, let's not lose sight of a few things:

  1. Value added per hour worked doesn't tell the whole story of the massive improvement in this space over the last few decades. Investment and innovation have helped keep more workers safe - and that's a colossal achievement.

  2. This industry is still tremendously dangerous, arguably the most dangerous. When pitching new ideas for innovation, keep the human element in mind. Innovation cannot come at the cost of sending fewer people home safely at the end of the day.

  3. There is nothing that says construction cannot behave more like a manufacturing environment. Prefabrication is picking up steam in many areas (yes, even in large scale construction and industrial) and eliminates many of the factors that compromise workers out in the elements. It's hard to sprain your ankle in shin-deep mud when you're working on a concrete floor in a warm building.

Fascinatingly – and encouragingly – is that the manufacturing industry saw a material spike in productivity around 1980, just as safety measures pushed worker deaths down below 8/100k per year. Construction is nearing this mark right now. Could the consolidation in safety gains accrued over the last decade or so be setting the foundation for a step-function in productivity gains in the decade ahead? We think so.

So yes, the construction industry isn’t adopting new technology like crazy, but it is innovating at an impressive pace. When that energy is put to use in areas that align with its true priorities – chief among them, keeping workers safe and healthy – we see tremendous adoption:

The construction space is now ready for the step-function dollar and cents gains we’ve been waiting for (and given inflationary pressures the gains can’t come soon enough). But it will not come at the cost of the tremendous worker safety gains and risk reduction that the industry has achieved over the past 50 years.

Scott Kaplanis