It takes many years to become a skilled welder, and nuclear is the hardest class of welding to get certified for.

A few months at the local community college can get you to the point where you can do the most basic welding jobs, but these low skill jobs are the ones that either left the US or got automated, and the few such jobs that remain are mostly done by mediocre welders who should have retired years ago. If you want to do the high skilled welding that pays well, you need to essentially support yourself for years as you develop those skills on your own - and most people who have the means to do so have better opportunities in other fields.

Realistically, employers need to eat the cost of training new people if they want the pipeline to open back up again, but it's more economical to poach a good welder after some other sucker puts their resources into them, so you get a classic tragedy of the commons.

I think there's an underestimated resistance to changing the labor rate. If a company quotes a typical job at 80 man-hours with $15/hr welders, and a year later you want to hire welders and can only find them at $20, they're more likely to complain there's a welder shortage and decline jobs because they don't have enough welders to complete them than increase their labor rate. Plus, if you increase the rate for a new guy, and your $12/hr hire from a decade ago hears the new hire is making $20, he's going to be pissed, so you'd have to increase costs across the board or suppress discussion of salaries.

I see this with those profiles of factory owners complaining about Those Dang Kids not wanting to work, while omitting the fact that they offer too little relative to other options. A lot of people running places like this don't want to accept they're also in competition with retail now.

People look at the small difference between pay at the factory or mine (dangerous) and pay at Walmart (not quite as dangerous + employee discounts) and make the rational choice.

This explanation doesn't really hold water. Some company is going to eventually win the contract by bidding with $20/hr welders right? It's not like the client is just going to give up on their project.

> This explanation doesn't really hold water.

During the 2010-2019 US recovery wage stickiness was noticed, even with everyone complaining about not being able to find qualified workers. Companies were very reticent about increasing wages, often preferring to go with signing bonuses.

It is suggested that this reticence/stickiness was because after 2008-2009 GFC many places were stuck with high wages (from the previous boom), and getting people to cut their salaries was almost impossible.

* https://www.investopedia.com/terms/s/sticky-wage-theory.asp

* https://en.wikipedia.org/wiki/Nominal_rigidity

No, I think you do see clients give up. There are certainly more infrastructure maintenance projects in the US than governments are willing/able to pay for. I’m not sure the market can bear much higher costs for construction.

There was a rather famous story about Aceribo being abandoned due to lack of upkeep and high cost of repairs.

It takes literal years of daily practice for a welder to get good enough to do the high quality specialty materials pipe fitting welds that would make up so much of nuclear plant construction. I am not exaggerating. A good welder/fitter team can work pretty much wherever they want, they are in incredibly high demand. There aren't more of them because not that many people have the baseline talent to be able to achieve the ability, regardless of how much practice they put in.

Source: used to weld at the NASSCO shipyard in San Diego.

Could you expand on the "baseline talent"? This isn't the sort of hand-eye coordination that would be required to hit a major-league fastball, is it? (Somehow those employers are able to fulfill their skilled-labor needs.) How would an individual welder determine whether she has the talent that would make those years of practice worthwhile? If that is not possible, why don't interested organizations set up some testing programs?

Hand eye coordination isn't really the issue.

When welding, you put an incredible amount of heat into the material. This causes it to move in real time. A welder needs to predict how it will move ant actively account for this so that the final piece winds up in the right condition. The tolerances depend on the application - for furniture who cares if it's out a bit, for a coolant line to a reactor you need it to be damn near perfect.

Then there's heat penetration - adding heat to the material changes its material properties. Go too slow and steel becomes excessively weak, go too fast and you don't form a strong bond. Again, big differences between "probably won't break if you sit on it" and "this will kill a dozen people if it fails"

Finally, there's just plain old margin for error. If you're welding together consumer products, it doesn't matter if you scrap one every now and then. However if you're trying to join together two multi-million dollar custom fabrications, you need someone who's going to walk up to this and get it on the first try every time.

I'd compare it less to being a major league baseball player and more to being a professional musician - you can't determine that someone could never do it, but some people are certainly born with a knack for it and you can't simply churn out more on a whim.

Thanks for the detailed and insightful answer! It seems that welding is what takes over when engineering can't go any further.