Going off the deep end

The arguments supporting Luddism, the belief that technology is eradicating jobs and creating unemployment, are getting subtler — but they’re still wrong.

Imagine you and your family are visiting a swimming pool, and the life-guard comes to have a talk with you. Your kids, he tells you, are at one end of the pool, and they are splashing all the water towards the other end.

You apologize for their disturbing the other bathers and tell him you’ll speak to them.

“That isn’t the problem,” the lifeguard continues. “The real problem is, they will pile all the water up at the other end of the pool.”

This is a rather confusing complaint. “What?”

“They are all at one end, and they’re splashing toward the other end. If they keep doing that, all the water will be piled up at that far end, and where they are standing will be dry.”

You check to see if the man is serious, but he appears to be. “But it’s water,” you point out. “It flows, it moves.”

“Of course, when someone splashes it. But there’s no one at the other end splashing back, so we are very concerned that they’ll empty their end.”

Now you begin to question the lifeguard’s sanity. “Have you ever seen that happen in the past?”

“We’ve had some concerns. A few months ago, some professional swimmers were practicing their kicks very vigorously.”

You try unsuccessfully to keep the sarcasm out of your voice. “Did the water all ‘pile up’ then?”

“No,” he admits reluctantly. “But that doesn’t mean it can’t happen.” Now it’s his turn to sound sarcastic. “Why? Do you believe there is some magic, invisible force keeping the surface of the pool flat and level?”

Well, yes, you do. Even if you don’t have a subtle understanding of gravity and fluid dynamics, it’s certainly your experience that water does “seek its own level”, that there is an invisible force keeping the surface flat and level.

Economics can be seen the same way. If “water”, the supply of labor (or anything else) piles up in one place, while there’s a “dry area”, demand freed up by technologically driven savings, in another, the two automatically find each other and cancel each other out. Well, not “automatically” — the buyers and sellers spend a great deal of effort finding each other, but they will do so and they don’t need help from the professional hand-wringers.

I’m going to push this swimming-pool analogy a little bit in order to rebut the second generation of Luddite arguments.

A friend of mine indignantly pointed out that Blockbuster, the video-store chain, had laid off thousands of people when it went under, but Netflix, the technological innovator that destroyed Blockbuster, only hired a few hundred. Blogger Martin Ford makes a more sophisticated version of the same argument by pointing out that wages for jobs running machinery are dropping.

Both these people are misunderstanding the situation.

To go back to pool analogy, it’s a fact of fluid dynamics that if you splash water, some quantity of the splash comes right back at you (try it). But that isn’t why a pool of water will stay level. Even if you invent some smooth, highly efficient technique of splashing water, you still cannot pile up water at one end of the pool.

Similarly, automation typically isn’t free: it needs people to design the machines, build them, install them, and operate them. That fact (or usually-true fact) is irrelevant to the always-true fact that technology doesn’t cause unemployment.

Even if some new labor-saving technology is utterly free and requires no expertise or labor to use, even if it creates no demand for labor in the particular industries it’s used in, even if it utterly eradicates any need for labor in its area of application, it still will not cause unemployment. If a technology (or anything else) frees up demand (by saving money, the point of the technology) and it frees up supply (by saving labor, the complaint of the Luddites), the supply and the demand will still inevitably find each other and cancel each other out.

Several of my commenters put their collective finger on a more plausible, if lesser, possible defense of Luddism. Imagine the splashing children in the analogy were replaced with a powerful outboard motor. If you watch an outboard motor run you can see that while it is actually running, it does create a “valley” in the water right behind the prop and a corresponding “hill” a little further away.

My commenters, of course, did not torture the analogy the way I have, to make their point. They just said, in effect, what if this time, it’s different. What if the economic dislocation that is caused by technological improvement becomes so powerful, so thorough, so fast, that it overwhelms the countervailing economic forces that up until now kept the country at near-full employment — just as the powerful thrust of the outboard’s prop pushed water up faster than gravity could pull it down.

This phenomenon isn’t (unlike most of the other propositions of the Luddites) utterly fictitious, but you can see how weak it is. Even in the modern world, technological change occurs over years and decades; finding new jobs can occur in days or weeks.

And unlike the outboard motor example, technological doesn’t only eliminate jobs. It also improves the efficiency with which new jobs are created and once created, found.

A Midlands textile worker in 1830 might very well curse the mechanized loom that put him out of work, but he should also have blessed the newspaper that could tell him about new jobs and should have blessed the new steam ships and steam locomotives that would carry him to new factories and carry the products of those factories to new markets. A travel agent today might be upset that online ticketers destroyed his vocation, but he should remember that online educators allow him to train for a new career, and online social networks and publishers allow him to find a job in that new career.

This balance, between the disruption caused by technological change and the improvements in calming the disruption, isn’t automatic the way the supply and demand are automatically in balance, but the experience so far is that improvement side is actually winning: it’s getting easier to find new jobs faster than old jobs are being chewed up. It might change in the future, but it might not.

There’s a third, much more interesting argument, about what might happen as technology gets better and better, but this column has already run on too long. I’ll just say, the third argument is right.

Advertisements

7 thoughts on “Going off the deep end

  1. “The arguments supporting Luddism, the belief that technology is eradicating jobs and creating unemployment, are getting subtler — but they’re still wrong.”

    On the face of it, that statement is actually factually correct.

    See, automation does ‘destroy’ jobs. But that is not the issue that freaks everyone out, really. It is how many NEW jobs will be created to replace the old ones and in what timeframe and with what effort required.

    “it’s getting easier to find new jobs faster than old jobs are being chewed up.”

    Not for many people in the span of time this is happening (which is getting shorter and shorter). Within the next 10-15 years, truck drivers, airline pilots, warehouse workers, shipping crews at the ports, much retail staff, etc. will be replaced by machines or enough automation will be in place where fewer workers are needed and it will happen all at once and unbelievably fast.

    To the average joe’s sense of perception, there will be tens of millions of people thrown out of work ‘immediately’ but the economy will not be able to create replacement jobs for them fast enough. Nor will many ‘new’ jobs that can compete against machines be low-skilled, if at all. Low-skilled workers can not suddenly become brain surgeons or software engineers, if if aptitude was not a required factor in all that.

    In short, it could very well become nasty before Utopia arrives — if it arrives at all.

    • It is how many NEW jobs will be created to replace the old ones

      How much water will slosh back from the high side of the pool to fill up the hole on this side?

      and in what timeframe and with what effort required.

      As I say, currently the creation of jobs is a much faster process than the destruction of them.

      Within the next 10-15 years, truck drivers, airline pilots, warehouse workers, shipping crews at the ports, much retail staff, etc. will be replaced by machines or enough automation will be in place where fewer workers are needed and it will happen all at once and unbelievably fast.

      I for one do not regard “10-15 years” as “all at once” or “unbelievably fast”.

      The time between the point when the incumbents can no longer deny the situation and the point when they actually lose their jobs may be short, but that just depends on their own ability to deny the obvious.

      Attention all truck drivers, airline pilots, warehouse workers, etc.: you are currently screwed. Find a new line of work. You have 10 years.

      Or you can wait. When all the airliners are automated and their pilots are laid off, the demand for hotel clerks and chambermaids and so forth will increase (more people will travel). So Mr. Pilot, what’s it going to be then, eh? Spend the next 10 years learning a different, high-paying skill (auto-pilot design and maintenance comes to mind) — or fluff pillows and change bedsheets for the people who can afford to fly to Waikiki, now they aren’t paying your obsolete ass $200 an hour. Up to you.

      To the average joe’s sense of perception, there will be tens of millions of people thrown out of work

      And of course tens of millions of previously unemployed will be put back to work.

      And Average Joe might very well riot, seeing as he no longer has a job but Average Bob does. Let’s not encourage him by pretending he has a legitimate grievance.

    • I don’t see truck drivers being replaced easily or soon. Too many variables in delivery systems, too many small businesses that are regional/local. In fact, trucking is one of the areas that has had unexpected growth over the past decade.

      • You are assuming that a machine isn’t flexible enough to deal with variations in delivery systems and so forth, and that assumption is currently correct. Once a computer is sophisticated enough to reason about these problems…

  2. I for one do not regard “10-15 years” as “all at once” or “unbelievably fast”.

    Sorry, my bad. What I meant was that no later than 10-15 years from now, we will have reached a point commercially where the widespread adoption of automation technology will enable the wholesale layoffs of millions of people in a relatively short period of time from that point forward. Think of it as a state of ‘critical mass’, if you will.

    As I say, currently the creation of jobs is a much faster process than the destruction of them.

    Uh…really? Because all the job creation stats I’ve seen show that new jobs are not being created fast enough nor in enough quantity relative to the destruction of old jobs. Also, most forms of ‘destruction’ is taking place in simply the phasing out or not hiring of existing job types as automation enabled companies to simply expand production w/o hiring as much as they needed to in the past.

    “The time between the point when the incumbents can no longer deny the situation and the point when they actually lose their jobs may be short, but that just depends on their own ability to deny the obvious.”

    Doesn’t matter. They will outvote you and me 30:1 at least. So Neoludditism will become a staple force in politics just as food stamps and Obama phones are today.

    Also, none of this is ‘obvious’ to most people. Most automation advances and deployments is rather subtle…until it is no longer. Then it is just simply too late to react according any rational planning basis.
    Most individuals follow any of this, remember. Even if they did, most don’t even know enough about basic economics nor current automation technology trends to follow any analysis and debates like you and I are engaging in right now. So I would have to say that what you consider to be obvious is not so to all those people in the career sectors I mentioned.

    So Mr. Pilot, what’s it going to be then, eh? Spend the next 10 years learning a different, high-paying skill (auto-pilot design and maintenance comes to mind) — or fluff pillows and change bedsheets for the people who can afford to fly to Waikiki, now they aren’t paying your obsolete ass $200 an hour. Up to you.

    Why should he? Why would ANYONE rationally put in the investment in time, money and opportunity costs that such a 10 year investment in retraining would require when what they are retraining for will be automated even before that 10 year period is over. Even if it is not, there will be enough people competing for that high-skill job to make it pay a lot less than you think it would. Note: By then, it will be obvious that any job will be at risk of technological obsolescence to most people.

    It would be like buying a house with 10 year mortgage (never mind a 15 or 30 year one) in an environment where there is a strong chance your home would be legally seized by another. People will adapt accordingly in their choices.

    Let’s not encourage him by pretending he has a legitimate grievance.

    Who’s pretending? Acknowledging the facts of the situation vs. holding on to shiboleths that fit one’s own personal worldview instead is even less rational than Joe ANYBODY rioting over this, one can forcefully argue. I personally agree that Joe Sixpack technically as a matter of fact does not have a legitimate grievance. Same for banks who loaned out mortgages that they shouldn’t have and the people who signed up to pay those mortgages. Yet, no matter who was technically right both groups got bailed out and are still getting bailed out one way or another, correct?

    There’s what you and I may think should go down. And then there is how things truly do go down. 🙂

    And of course tens of millions of previously unemployed will be put back to work.

    No, they won’t. They will become a drain, not a productive resource. Sort of like all that surplus labor the Mexicans shipped north of the border because their economy had no work for them, albiet caused by completely different reasons. Point is, there can be such a thing as surplus people who do not produce but are rather a train on productive assets regardless of what individuals chose to do or not do in their careers.

    So, there will only be enough jobs total for people to fluff pillows at the hotels and even those will replaced by robots. Ergo, there will not be enough jobs. And yes, I realize that ‘new’ kinds of human-done jobs we can’t even conceive of will also be created as well. I just maintain that it won’t happen fast enough nor in enough quantity to make the difference you seem to have faith (grounded in very solid historical data I acknowledge) that it will.

    We are entering Ancient Rome Land…specifically the point where slaves from the conquered territories flooded Roman labor markets and destroyed the Roman middle class in the process. Only this time, the slaves will be robots and they will only get better and cheaper over time vs any possible human competitors.

    In a world where human labor is only worth anything of value for jobs what machines can not do — yet in that same world the machines will be replacing those jobs at an astounding rate — only ownership of capital (owning the robots, so to speak) will make a difference for anyone.

    So instead of even trying to save the jobs we should instead find ways to democratize capital ownership in a gradual and non-confiscatory way.

    • Ergo, there will not be enough jobs.

      Dennis, I’m not sure you’ve understood the first thing — the only thing — I’ve said. Technology cannot destroy jobs, in the net. It’s arithmetically impossible. Whenever a job-category is destroyed or reduced, the demand is freed up to create more jobs. It was true in Ancient Rome, it was true in Georgian Britain, and it’s true now.

      People might believe the Luddite myth, and that belief might make them dangerous, but it remains a myth.

  3. Technology renders upper-body strength highly over-rated meaning more jobs for women!! Woot woot! Equalization of the sexes baby yeah

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s