For Useful Computers, You Have To Fire People

We have innovation on many different levels, from materials to hardware organization to software to marketing to organization design to overall regulation. Many seem to think the lowest hardware levels are key, and other levels mostly go along for the ride. But not only is software as important as hardware, neither help much without good firm practices:

US productivity growth accelerated after 1995 (unlike Europe’s), particularly in sectors that intensively use information technologies (IT). Using two new micro panel datasets we show that US multinationals operating in Europe also experienced a “productivity miracle.” US multinationals obtained higher productivity from IT than non-US multinationals, particularly in the same sectors responsible for the US productivity acceleration. Furthermore, establishments taken over by US multinationals (but not by non-US multinationals) increased the productivity of their IT. Combining pan-European firm-level IT data with our management practices survey, we find that the US IT related productivity advantage is primarily due to its tougher “people management” practices. … American firms have higher scores on “people management” practices defined in terms of promotions, rewards, hiring, and firing. (more)

GD Star Rating
Trackback URL:
  • DW

    And fire the ones that can’t use computers first.

  • Poelmo

    American multinationals don’t believe in labor rights, neither do European multinationals but the former can’t be touched by European governments, while the latter can.

  • Poelmo

    The worst thing is it’s management that gets a bonus when productivity goes up, while it’s the programmers who started working harder and got their pay cut (at least their hourly rates), just read the stories from the creators of LA Noire, to get a sense of how bad it gets. But hey, that’s what we want to teach our kids, isn’t it? Don’t try hard in high school because you’re better off going to business school instead of studying computer science anyway since firing people/pushing people to work harder for less (cost cutting decisions a monkey could come up with) pays better than actually creating something useful.

    • Miley Cyrax

      Your sour grapes are showing. The ROI for any business outside of the M-7 or perhaps even Harvard/Stanford/Wharton is pretty questionable. And if you could had gotten into those b-schools, you were likely quite brilliant and hard-working in the first place.

      • Miley Cyrax

        And by “you” I mean “you” in the general sense, not you, personally Poelmo. I’ve also found the relatively low compensation for computer programmers to be pretty bizarre, maybe even undesirable. I suppose it’s just a hint that verbal and social skills matter more than we may personally prefer.

      • Poelmo

        I’m pretty sure every computer sciences graduate has the intellectual capacity to study business at an M-7 university, and you can buy your way into those schools.

      • Poelmo

        It also goes deeper than just sour grapes. If economists, can ask workers to work harder for less to make the economy more productive, even though none of the gains would flow back to the workers, then why can’t I ask employers to reward their employees more fairly, so people won’t get disillusioned, useful work is encouraged and the economy would also get better, but this time there will be gains for the workers?

      • Sandra Pierce

        Miley, what is the “M-7”?

  • So, does this imply that computers are no longer complements to human labor, but replacements for human labor? In other words, that the Luddite thesis is correct?

    • Carl Shulman

      Unless the complementary effect is in the exact same jobs, you need to fire people from the obsolete job, with free labor being reallocated to the jobs benefiting from complementary effects.

      • Poelmo

        Maybe that’s what happens in an economist’s fantasy world, but in the real world there is absolutely no guarantee new jobs will be created to replace the lost jobs. Do mind, I’m not saying we should halt technological progress to preserve jobs (I’m not even sure we could do that and survive, since we need new technology to solve the world’s major problems), I’m just saying we should acknowledge that jobs aren’t unlimited: we either need to shorten the workweek to make sure everyone can get a job, or we need to accept growing unemployment as a fact of life and reform social security accordingly.

      • xxd

        For some reason I can’t reply to poelmo so I’ll reply to this post instead. While technically true, the economist 101 model of the results of creative destruction don’t take into account lag.

        There is often lag between the new industry destroying jobs and the old industry workers gaining the skills to enter the new industry. Also the new industry may be small and destroy jobs rapidly but it take some time for new jobs to be generated to fill the gap.

      • Konkvistador

        Imagine for a second we cap put a human brain in a box. Now imagine we can hook up that box to a good robot body that can.

        Now imagine we can make and maintain that box for a fraction it costs to employ a flesh and blood human.

        Which jobs do humans do in this world?

        And do humans loose jobs *only* in this extreme example? No. We shall gradually see the least competent humans made more and more unemployable and uneconomical. This will have social consequences.

      • If you define “competent” to mean things that cannot be done by a machine, that is one thing.

        But if you define “competent” to be things that cannot be done by a machine and that I am willing to hire a human to do, it is quite another thing.

        There are many things that cannot be done by machines at present. Teaching children, taking care of infants, caring for elderly demented patients.

        Unfortunately, children, infants and the demented are unable to appreciate that they need teachers and caretakers and don’t have the wealth to hire them if they did, and people who do have wealth don’t appreciate that people need to be hired to do such things.

        There is a good paper that discusses where much of the “value” comes from in hostile corporate takeovers. It is “Breach of Trust in Hostile Takeovers” by Andrei Shleifer and Lawrence H. Summers. It comes from breaching implicit contracts.

        Of course, one person’s implicit contract is another persons entitlement. It is really “increased efficiency” to fire high health care cost workers and replace them with healthier ones? It would seem that some of the increased efficiency observed with “tougher” people management is the willingness to breach implicit contracts. That isn’t a generation of wealth, it is simply a transfer.

    • Poelmo

      They’re both, just like they always have been and always have been. Sure the percentages may shift but ultimately it’s simply counterproductive to expect all computer related labor to ever fit into one stereotypical box.

  • Poelmo

    Above comment is directed @gwern.

  • Poelmo


    Can’t directly reply to you either (there’s probably a reply limit of 1 to discourage lengthy discussions).

    I know about lag but the economist 101 model is still BS: basically they’re assuming that if some new manufacturing technology made 9 out of 10 jobs obsolete then the 1 in 10 of us that still have a job would all have 9 personal butlers. This is absurd for two reasons: most people have no desire for infinite opulence and even if they did chances are they wouldn’t have the means for it since employers will most likely not pay the 1 in 10 ten times as much, instead they’ll put 9 wages in their own pockets and they probably won’t use that to hire 90 personal butlers (it’s well established that rich people spend a smaller percentage of their income ).

    • Scary Robot

      What you describe as “absurd” has essentially happened in the (now) developed economies of the world over the last century — real GDP per capita has increased 8-10x. This has failed to result in mass unemployment. It’s true that most people have little need for nine butlers, but rich civilizations invent new things for people to want — indoor plumbing, electricity, radios, telephones, cars, air travel and air freight, personal computers, smartphones, advanced medicine, etc. and jobs are created to provide these new things. There’s no reason to expect this to stop, at least not until we have cheap human-level or better AI and robotics.

      • Poelmo

        You fail to mention that developed economies between 1900 and today have:

        1) pulled children out of the workforce (not just 14 year olds who got to high school, but also 20 year olds who go to college)

        2) experienced aging of their populations, so the able bodied adults are now a smaller fraction of the population than they used to be

        3) shortened the workweek and added vacations

        4) exported to other markets (though that’s being reversed now, and resulting in unemployment)

        5) experienced two periods of 4-5 year wartime economy as well as occasional high tax regimes (but those are things of the past) and instated social security to force redistribution of wealth so workers could hire more services if they wanted to in the first place, since, as I noted, even if people want to hire butlers they can’t hire them if a tiny elite highjacks the profits of increased productivity (and less people will be counted in the unemployment figures).

        The only trend going against this is the inclusion of women into the workforce (though it’s a myth that almost no women worked in 1900 and even today many women work part-time or not at all). I don’t deny new jobs are created by new industries but I am saying this is not 100% over long periods of time, if only because natural resources are finite, and we should accept that and adjust the system accordingly (shorter workweek and/or more social security).

        I regret that I don’t have the numbers but I want to bet you that for a developed country, say Germany, the total number of hours worked by the population, divided by the population, has decreased between 1900 and today.

  • Shouldn’t the expected result in econ 101 of people getting fired not that they stay unemployed, but that their next job is at a lower wage? Total efficiency for society goes up, but I don’t think any standard theory guarantees that the losers from technology/trade competition have to come out ahead overall.

    My question for Robin: do you think we have an inefficienctly low rate of fires even in America, for some of the reasons Caplan specifies here? And if so, are there any steps we should take to correct that?

  • Pingback: deep change in tech and economy | Pearltrees()