Replacement hiring makes up most of the 1.7 million US trade openings expected each year to 2035

The report's two figures measure different things. The 1.7 million is a yearly flow of openings ^[1], while the 600,000 is the total of net new positions across the whole stretch to 2035 ^[2]. If all 600,000 arrived in a single year, they would still be only about 35% of that year's openings ^[11]. Spread across the period, they are a much thinner slice, and most of the hiring replaces people who leave.

Against an 18 million-strong trades workforce ^[4], 1.7 million openings a year is turnover of about 9.4% annually ^[12]. The 600,000 net additions grow the workforce by about 3.3% in total ^[13]. Data-center demand from the AI boom ^[18] sits inside the smaller of the two numbers. "You have to think about this like a bathtub with a faucet and a drain," Matt Sigelman, president of the Burning Glass Institute, told Fortune ^[5]. "We also have a very leaky pipeline," he said ^[6].

The drain shows in the age data. The report's age split ^[3] works out to a little over two workers aged 55 or older for each one under 25 ^[14]. Among tool and die makers, nearly half the workforce is over 55 ^[10]. Mike Rowe, host of Dirty Jobs, gave a similar ratio at a Ford Pro Accelerate panel: "For every five tradespeople who retire this year, two replace them," he said ^[19].

Higher costs and longer schedules for infrastructure projects are a plausible result of all this, but Fortune's account of the report does not include a wage forecast, a project budget or a delay estimate ^[1]. The closest thing to a price is one employer's training budget ^[7]. Technicians average $80,000 to $90,000, according to Fortune ^[8], so at the $85,000 midpoint General Motors' nearly $200 million equals about 2,350 technician salaries for a year ^[15].

The case for a long shortage weakens if young workers arrive faster. Fortune reported that Gen Z frustration with college costs is steering more young people toward trade work ^[17]. A rising under-25 share would add workers faster than the current age split implies. Better retention would also cut the 1.7 million itself, because a worker who stays is a seat that does not need refilling.

I think the shortage outlasts the data-center cycle, since most openings come from an age profile that moves slowly. The case against is that the report's 11% under-25 figure ^[3] is a snapshot taken just as younger workers began turning to the trades ^[17].