So I’m flipping through my blogroll and I come across this post, which impresses me enough that I open it in a new window to blog about it.
The gist of the post is that Target now does, with one employee and a robot, a job that once required two employees (herding carts — look, you should really go read the original). Dr. Boudreaux pulls out the following nugget of wisdom, which is what I’d originally intended to blog about:
It’s especially important for those persons who support minimum-wage legislation to realize that employers can almost always, at the margin, substitute away from human labor and toward mechanized or electronic “labor” — that is, capital. Mythical indeed is the notion that employers must hire a given, or minimum, number of low-skilled workers. As the cost of hiring such workers rises, employers have greater incentives to substitute away from employing such workers.
In a mood for procrastination, I started skimming the comments, and holy shit man….
Sure beats the centipedes. Click the image link, not “centipedes”. Really.
Commenter “SheetWise” brings up the evolution of North American labour from a mostly agrarian to a mostly industrial work force, and from a mostly industrial to… what, the “knowledge economy”? With our public schools?
The historical movement of agrarian workers from 90%+ to < 5% is interesting, the assimilation of workers by concentration of production is interesting, and the continuing employment of technology is interesting.
Where I personally have a problem wrapping my mind around the progress I’ve seen in the last forty years is knowing if there’s an end-game that’s workable.
Today, existing computers are increasingly responsible for the development of newer and better computers — robots are increasingly responsible for building better and better robots. While these processes require human interaction — human labor is increasingly becoming dispensable.
There is a part of me that sees continuation of this process leading to the ultimate end — minimal labor and maximum leisure. What could be better? Or worse.
Edit: What to do with the people?
This is related but not identical to Too Many Malthusians. I expect that if you’d asked a farmhand in 1909 what he’d do if his job was replaced by a John Deere tractor attachment, he’d have had no idea — the option of, say, assembling V-8s for Chevrolet wouldn’t have occurred to him (not least because Chevrolet wasn’t around in 1909). Increased mechanization eliminated some jobs, but the lower cost of the goods it produced made other jobs viable.
Human labour isn’t becoming dispensable — it’s shifting. Fifty years ago, a semiconductor integrated circuit was an exciting laboratory toy. Forty years ago, it was a specialized component of a missile guidance system — which means it was being manufactured rather than assembled in an experimental process. Thirty years ago, it was a component of a general-purpose computer installation. Twenty years ago, we stopped talking about integrated circuits as such — we instead spoke of their function in any number of devices. ICs have gone from mysterious devices full of potential to mundane appliance components, and created a whole new class of jobs that didn’t exist in 1959.
“What to do with the people?” is, of course, a non-question: this isn’t a real-time strategy game where one’s computerized (see? ICs again!) minions wait patiently for the player to build a farm or a factory. It’s an admission of underdeveloped imagination and a fundamental lack of confidence in other people: “I can’t see, right now, how to solve the problem — therefore we need to think really hard about how to fix it.” But if we increase the sophistication of robot-manufactured goods, more and more sophisticated products will become mass-producible. (For example: Sous Vide machines, or iPhone/Android-class mobile devices.) If we increase the sophistication of the manufacturing robots themselves, we give people the opportunity to build stuff in (and sell stuff from) a small shop that would once have required a huge staff and an enormous facility. (Suppose, for example, that one of the drivetrain engineers on the late lamented Dodge Viper programme gets some CNC welding and pipe-bending rigs and starts churning out high-flow exhaust systems out of his garage comparable to the ones on the Viper GTS-R.)
Commenter johndewey puts it eloquently and concisely (emphasis added):
Machinery (technological progress) increases the real value of workers (increases their productivity), and employers can then hire more of them. That’s because there are always more value-producing opportunities available for workers as their productivity increases.
Minimum wage by itself does nothing to increase the real value – to increase the productivity – of workers. Some value-producing opportunities are not pursued because the price of low-skilled labor is artificially raised too high. That is, raised above the level at which the value-producing opportunity can produce value.
And commenter Economiser adds a heaping plate of perspective:
Imagine an average European in the year 1000 AD. That person working full-time has his “necessities” met, as far as he’s aware of them. A typical American in the year 2009 who wants that standard of living could work for probably 2 weeks out of the year and have the rest of his time devoted towards leisure. Of course, that person would be considered horribly poor by modern standards.
No one does this because innovation has made many items into “necessities” that were previously unavailable or unknown. And that trend will most likely continue. If we could flash forward 1000 years, we could probably buy a 2009 middle-class American lifestyle for virtually no labor, but we’d also be considered strikingly poor compared to society at the time.
The whole damn comment thread is worth reading. Seriously, go over there.
