American Labor’s Real Problem: It Isn’t Productive Enough | Hacker News Reader

American Labor’s Real Problem: It Isn’t Productive Enough

wsj.com

18 pointsceohockey6040 comments

Boeing moved production from Seattle, where they had people with experience building airplanes, to North Charleston, South Carolina, and cut pay. Of course they now have people who don't know what they are doing.

There's a more general problem. Who goes to college to learn how to set up a factory? Where do you go to learn that? That's a difficult skill, and it helps to have done it a few times as a junior before you're in charge. What happens if you don't know how is that you have to let the people who sell you the machinery set up your factory. This works for simple stuff, like a bakery, which you can buy turnkey. More complex stuff, not so much. Worse, much of the knowledge doesn't transfer. If you know how to design a production line for metal furniture, that may not translate well to batteries. One of the things that auto manufacturers have painfully discovered is that a battery plant is partly a chemical plant, which is not something automakers know much about.

Machine design is quite hard, not that many people are good at it, and it pays less than coding webcrap. The US used to have people who were good at this, and most of them were laid off in the 1980s.

I've seen people who had never built a factory try to build one, and it did not go well.

> Who goes to college to learn how to set up a factory?

There's a field called "industrial engineering":

https://en.wikipedia.org/wiki/Industrial_engineering

"Industrial engineering is an engineering profession that is concerned with the optimization of complex processes, systems, or organizations by developing, improving and implementing integrated systems of people, money, knowledge, information and equipment. Industrial engineering is central to manufacturing operations."

In the university I attended, the engineering school had a degree program for it.

Also, I'd presume that depending on what the factory produces, you'd have electrical, chemical, mechanical, aerospace, etc. engineers involved in designing and building the factory.

School still isn't the same work. There's a lot more intricacies in the real world than you can ever be taught in school.

Of course. As in any field of work, new graduates get entry level jobs and learn by working alongside more experienced practitioners. (E.g., someone who just graduated with a CS degree isn't likely to know how to ship robust, scalable and maintainable software - that's something you learn from experience.)

Yes. Building something complicated requires formal education plus years of experience attempting to build the complicated thing.

This degree is quite popular in Israel. Because it's relatively easy and probably interesting. You learn many subjects(people, money, knowledge, information and equipment) but not in that much depth.

graduates of that degree work in working factories, manage people and optimize processes. I don't think they go into deep technical subjects.

That's very different from what Animats described.

It's related. If you have a working factory, data collection and optimization can make it better. If you have a vacant lot and money, that's a different problem.

I'm reading Ashlee Vance’s “When the Heavens Went on Sale”. This describes most of the smaller space launch startups, which he visited at length, including the launch sites in Outer Nowhere. Each of them had to start manufacturing rockets from nothing. How they approached this, who succeeded and who failed, and why, is worth a read.

> Who goes to college to learn how to set up a factory? Where do you go to learn that?

Industrial and/or manufacturing engineering.

Some endeavors require historical knowledge, aircraft engineering being one of them. If you break the chain, you lose it; policies that were developed over decades are discarded. "Why do we need to do it this way, this other way is faster and cheaper!" No one is around to explain why any more, so the lesson is learned the hard way, again.

The real culprit is false accounting. If it's awkward to put on a spreadsheet, it doesn't get counted as being valuable.

There's an episode of This American Life about IIRC a sausage factory that gets moved, and the sausages for some reason do not taste the same. The tacit knowledge embedded in the old factory is an example of this effect. You think you know how something works, so you make decisions based on that, and it turns out your spreadsheet was wrong.

The same thing is the case with your example. It's easy to count the salaries and the cost of the new site, along with the expenses for the machinery.

Outsourcing is another one of these things, and problems are compounded when it's not just counting the costs/benefits correctly, but doing do at an appropriate time. You send your multi-year coding project abroad, spreadsheet costs are lower, manager who did it gets promoted, project falls apart.

> For example, American manufacturers use far fewer robots than their competitors, in particular in Taiwan, South Korea and China, according to the Information Technology and Innovation Foundation, a Washington think tank.

Just handwaving this aside is ridiculous. Productivity is a function of capital investment in technology. If all the free capital is going into share buybacks, $5 billion in the last year, no wonder productivity is flat.

https://www.levernews.com/automakers-hand-billions-to-shareh...

Whenever you encounter an executive who complains they can't get qualified people, ask how many people they have in training classes right now. Large companies used to have large training programs, to go along with a stable workforce, and a good pension plan. A few still do. Yes, after they finish training you have to pay them more.

Agree. Everyone wants the fullstack dev, no one wants to train the junior engineer.

Nowadays, it seems that companies don't train or pay for training.

Our macro policy pumps assets and dumps exports. Our manufacturing workers must be more productive in the colloquial sense to be as productive in the economic sense. The headline weaponizes this nuance to read like "our workers are lazy." This is intentional, and it is ugly.

Why does this argument never apply to executive compensation?

https://www.businessinsider.com/uaw-strike-automaker-ceo-pay...

Executives don't need to be productive, they just need to promulgate the bizarre idea that people higher up in the hierarchy inherently deserve more pay for some reason.

Executive compensation is a drop in the bucket compared to the total salaries for all the rank-and-file employees.

According to your article the average hourly labor cost at GM is $60. Assuming 2,000 hours per year, that is $120,000 per employee. For 92,000 U.S. employees [0], the total is about $11,000,000,000.

Another way to look at it: If you distributed the CEO's compensation ($30,000,000) among GM's U.S. employees (92,000 [0]), you'd give each of them about $326.

[0] https://www.gm.com/company/usa-operations

This is so often completely ignored and IMO it’s the real reason why executive pay is so high. Companies are willing to bid up executive pay but not worker pay because worker pay comes with a huge multiplier and executive pay doesn’t.

An executive can also completely tank your business in a way that a line worker cannot.

So it's still a bit of a paradox why you would pay anyone so much over the odds. How confident can you really be that this one guy you found is going to increase profits by 10%? And if you do find this guy, why isn't there another guy with similar experience offering the same tradeoff?

That's exactly why they are paid so much. It's more important that they don't fuck things up than "demonstrate productivity".

That's what I'm saying doesn't make any sense. How do you know they won't? Plenty of people have messed up a business, who's to say your guy with pedigree won't?

Why isn't their compensation two directional then? Why not give them both options and have them sell puts that company can exercise. So if company loses 10% of valuation they will pay company some money...

Executive compensation is a function of profit/capital allocation for shareholders away from labor.

This doesn’t show the full picture. It’s not about redistributing their salaries. Have a look at the excess returns procured BY labor and where that went.

You know why.

A Murdoch rag has an anti-worker pro-ownership class article. News at 11.

If automotive CEOs don’t want their assembly line employees to demand 40% raises, maybe they shouldn’t take 40% raises themselves. Since they have more than enough money to live on in multiple homes and boats they should have no problem making themselves an example of austerity for their employees.

Maybe if companies put time and effort into training and retaining employees they wouldn't be fighting brain and productivity drain.

I've seen this happen enough in software companies to know it's already a problem. They want 10x engineers but aren't willing to pay even 2x rates and sure as fuck aren't willing to give them good raises to keep them around.

Trying to browbeat labor into being more productive isn't going to work. Unless they really want the return of how unions used to behave a century ago.

The cult of American ignorance doesn’t help as well as the idea of manufacturing as “real” work.

Americans sympathize with people who complain that a Mexican took their job rather than telling them that if someone with 4 years of education can do what they do with 12 for 1/8th the pay, they are utterly pathetic.

Non-paywall: https://archive.ph/sGdQ4

Thanks for this, but infinite captcha for me.

Typical WSJ hit piece. When capitalist lack critical thinking skills and cry that the rate of their wealth isn’t increasing fast enough, one can be sure WSJ writers will provide soothing words. Even if they are false.

>Pay is ultimately tied to productivity

This must be why CEO pay, relative to worker pay, has increased so much over the last few decades!

Well, CEOs are increasing their performance at all important job metrics, at a much faster pace than any unimportant "worker" has done, in all of history.

Just look at stock price, dividends and/or executive compensation. They are obviously much more important than the people who actually make the product, provide the service or interact with the customer. For any actually important metric, CEOs add more value to the people who determine what value the CEO adds more than any other part of the business.

All hail our true lords, NASDAQ and DJIA.

Reply on news.ycombinator.com