Investors Not Barking

Detective: “Is there any other point to which you would wish to draw my attention?”

Holmes: “To the curious incident of the dog in the night-time.”

Detective: “The dog did nothing in the night-time.”

Holmes: “That was the curious incident.”

We’ve seen several centuries of continuing economic growth enabled by improving tech (broadly conceived). Some of that tech can be seen as “automation” where machines displace humans on valued tasks.

The economy has consistently found new tasks for humans, to make up for displaced tasks. But while the rate of overall economic growth has be relatively steady, we have seen fluctuations in the degree of automation displacement in any given industry and region. This has often led to local anxiety about whether we are seeing the start of a big trend deviation – are machines about to suddenly take over most human jobs fast?

Of course so far such fears have not yet been realized. But around the year 2000, near the peak of the dotcom tech boom, we arguably did see substantial evidence of investors suspecting a big trend-deviating disruption. During a big burst of computer-assisted task displacement, the tech sector should soon see a big increase in revenue. So anticipating a substantial chance of such a burst justifies bigger stock values for related firms. And this graph of the sector breakdown of the S&P500 over the last few decades shows that investors then put their money where their mouths were regarding such a possible big burst:

S&P500breakdown

In the last few years, we’ve heard another burst of anxiety about an upcoming big burst of automation displacing humans on tasks. It is one of our anxieties du jour. But if you look at the right side of the graph above you’ll note that are not now seeing a boom in the relative value of tech sector stocks.

We see the same signal if we look at majors chosen by college graduates. A big burst of automation not only justifies bigger tech stock values, it also justifies more students majoring in tech. And during the dotcom boom we did see a big increase in students choosing to major in computer science. But we have not seen such an increase during the last decade.

So the actions of both stock investors and college students suggest that they do not believe we are at substantial risk of a big burst of automation soon. These dogs are not barking. Even if robots taking jobs is what lots of talking heads are talking about. Because talking heads aren’t putting their money, or their time, where their mouths are.

GD Star Rating
loading...
Tagged as: , ,
Trackback URL:
  • https://entirelyuseless.wordpress.com/ entirelyuseless

    It seems pretty clear at this point that even if a time will come when most jobs can be performed by computers, we are not remotely close to that time.

    • Noumenon72

      Gee, I don’t know why we bothered reading this article bringing together evidence for one side of this disputed question when we could have just asked entirelyuseless whether it was clear or not.

  • IMASBA

    I agree, Robin: it is indeed mostly just talking heads who go on about displacement these days. Scientists and engineers talk about automation for sure, but it’s the talking heads who then conclude that new jobs won’t be created. Maybe things will be different 100 years from now, but in the immediate future I doubt automation will have much of an impact on the unemployment rate.

  • Dave Lindbergh

    To me, that graph shows a pretty consistent trend toward the tech sector increasing – not in a bubble, but over the whole 35 years.

    But I wonder if you’re right to look for a tech boom in the first place.

    Why do the gains from automation need to go to those who sell the robots? Won’t those who buy the robots benefit as much or more (esp. if the supply of robots is competitive)?

    If so, the relative size of the different sectors doesn’t tell us much.

    • http://overcomingbias.com RobinHanson

      Even if only half of the gains go to those who sell automation hardware and software, a big burst of automation would make a big burst of revenue for those sellers.

      • Dave Lindbergh

        Agreed. But why expect “even half” of the gains to go there?

        The automation industry, being relatively unregulated compared to most other sectors, is highly competitive. So shouldn’t we expect small profit margins there? (This goes to stock values, not sector sales, of course.)

        And if automation is displacing labor because of improved technology (rather than larger sales of automation stuff), then with a constant level of sales, we might still see a big labor displacement.

        No?

      • http://overcomingbias.com RobinHanson

        No. These industries are competitive enough that what mainly matters is the elasticity of supply and demand.

    • arch1

      “To me, that graph shows a pretty consistent trend toward the tech sector increasing – not in a bubble, but over the whole 35 years.”

      Dave, are you talking about the tech sector’s share of the overall market cap (i.e. the thing graphed)? Even if you ignore the bubble bursting (and I don’t know why you would), the only growth periods worthy of the name (1991-2000, and 2002-2014) comprise but half of the 40 years graphed.

      • Dave Lindbergh

        The graph is a bit hard to read, but it appears that in the late 1970s the tech sector was consistently about 10% of the total market cap (if that’s what’s being plotted – it’s not entirely clear).

        By the 2010s, it was consistently about 20%.

        So that speaks to a long-term relative increase for the tech sector.

      • arch1

        Well I *do* agree that there was a long-term relative increase. (Also that the graph is more than a bit hard to read, for this purpose:-)

  • arch1

    I guess this makes me lean and ferret-like but I can’t tell from your article whether you’re a) suggesting that investors and college students will be just as mistaken this time around as they were in 2000, b) suggesting that this time they’ll be right (for some unstated reason), or c) agnostic on what the non-barking implies.

    • http://overcomingbias.com RobinHanson

      They weren’t necessarily very mistaken in 2000 – if they only estimated a substantial chance of a big automation boom soon. Their lower estimates now are info that you should also estimate low chances.

      • arch1

        It seems to me that the “chance of a big automation boom soon,” to the extent it can be meaningfully discussed, is much *higher* now than in 2000, since the enabling technologies (not the only factor, but an important and gating one) have grown exponentially. Which suggests that investors and college students almost certainly *were* substantially mistaken in 2000, presumably because they were driven more by irrational dot-com euphoria than by informed assessment (explicit or implicit) of the above chance.

        I agree that their lower estimates now should nudge mine down, but I feel that way more because their estimate has decreased as they have become less clueless and less euphoric, than because of any track record I should respect. That said it seems quite possible they’re *still* clueless for all practical purposes.

  • http://juridicalcoherence.blogspot.com/ Stephen Diamond

    Something like 40% of the world’s workers are unemployed. This is typically attributed to automation. Is there even an alternative theory? Essentially, the automation of agriculture has created more workers than industry can absorb. We have a growing international lumpen-proletariat.

    So it seems other-worldly eliding the question of whether automation has caused and will cause unemployment with the question of whether we’re on the cusp of replacing human labor completely.

    • IMASBA

      Where did you get the 40% figure? Most countries don’t even have a good idea of how big their informal work force is. There are definitely alternative theories on why some countries have high unemployment. Very common are recent population explosions, lack of education opprtunities and lack of a competitive economy that workers cannot overcome by moving because of borders, usually these factors go together.

      • sflicht
      • IMASBA

        That source uses “everyone over the age of 15” as a definition of “labor force”. You can’t compare that to regular unemployment figures.

      • sflicht

        It’s obviously not comparable to the standard “headline” unemployment rate. (Arguably LFP is a more meaningful number for understanding the economy!) But Stephen Diamond’s basic observation is correct, although like Robin I disagree with his claim that automation is a “standard” explanation for the decline in LFP.

      • IMASBA

        Stephen implied that 40% want a job but can’t get one. LFP ignores retirees, students, homemakers, trust fund babies, prisoners, etc… Those groups form the vast majority of the 40% and are influenced much more by demographics, culture and welfare systems than by automation.

      • sflicht

        I don’t think Stephen implied that, but I agree with you about what constitutes the 40%.

      • http://juridicalcoherence.blogspot.com/ Stephen Diamond

        There are definitely alternative theories on why some countries have high unemployment. Very common are recent population explosions, lack of education opprtunities and lack of a competitive economy that workers cannot overcome by moving because of borders, usually these factors go together.

        But why do we have all these problems throughout the world at the same time? Isn’t it driven by the advance of technology (automation)? Today’s “globalization” is driven by new (automated) technology. What you’re saying is essentially, I think, that we could imagine a society where automation doesn’t lead to high unemployment. But that is not to the point of whether we can count on “The economy … consistently [finding] new tasks for humans, to make up for displaced tasks,” as Robin cheerfully opines. “The economy” is the existing set of economic relations, not a theoretical reification.

      • IMASBA

        Corruption in one country is not the result of technological progress in another country. The same goes for having a culture that bars certain groups from education or employment or that demands a civil war every couple of years, or that encourages people to have 6 children.

        Even in a technologically stagnant world those factors would create a division between rich and poor countries.

      • http://juridicalcoherence.blogspot.com/ Stephen Diamond

        A division between rich and poor countries has long existed. But mass unemployment in the Third World is relatively new (decades), having to do with the transformation of agricultural societies under the impact of a global economy – and coping with the intransigent demands of international financial institutions.

      • IMASBA

        In those decades the populations of these countries have exploded. Sure, technological progress the problem of unemployment would be less, mostly because a sizeable portion of these people would have died of disease or hunger as infants. Such countries can solve their problems by using their low wage advantage to get technology and investment at a discount (and they don’t have to play nice with international rules, as China and India have demonstrated), but then of course there has to be peace and security, corruption needs to be manageable and fertility rates have to decline. The difference between Asia (sans the Middle East) plus Latin America on one side and Africa plus the Middle East on the other really shows the difference in outcomes between countries who try to get better results and those that don’t.

    • http://overcomingbias.com RobinHanson

      That is not remotely a “typical” attribution among economists. And unless the figure has been increasing a lot lately, it doesn’t speak to the chance of a big burst soon.

      • http://juridicalcoherence.blogspot.com/ Stephen Diamond

        That is not remotely a “typical” attribution among economists.

        Point taken.

  • Pingback: Under The Tree | EightAteEight

  • avid reader

    I know from your posts way back that you don’t care about spelling and style in general, it’s still du jour not de jure.

    • http://overcomingbias.com RobinHanson

      Fixed; thanks.

  • Bosda

    Nobody benefits from robots.
    Not even the people who own the robots.
    Large scale introduction of robotics could produce so much unemployment that the World Economy could tank.

  • RodneyFeeblenuts

    I just read that IBM’s Watson is being used to generate cross border tax planning strategies. Now that would be an impressive inroad into a complex, white-collar sector. To date, legal software has mostly eliminated the drudge tasks- sitting in a library, mass document review, marketing & invoicing. Strategy and planning is a core skill.

    Time to short the Big 4 accounting firms?

  • Mark Bahner

    Let’s look at something that’s virtually certain to happen…and in relatively short time. The first mass-produced cars and trucks will achieve Level 4 automation (completely autonomous driving) in less than 10 years. In 20 virtually all new cars and trucks will be fully autonomous. And in 30 years or less, virtually every vehicle on the road will be autonomous.

    That will likely obliterate brick-and-mortar stores of all types…Walmart, Target, Kroger, Publix, Home Depot, Lowes, Walgreens, etc. etc. Cashiers, salespeople, truck and taxi drivers, stocking people…close to 90% of all those jobs will likely disappear.

    So who benefits? Probably Amazon…but how many people will they even have with fulfillment centers completely automated? So if 90% of cashier, salesperson, taxi and truck drivers, and stocking person jobs are gone, that’s about 10 million jobs. What new jobs replace them?

    • consider

      I’d cut those time frames down a lot. You should eventually be able to fit an old car with automatic driving for a fraction of the cost of a new car.

      Wages in the translation industry have been pushed down in the past 5 years due to Google Translate and globalization, which attracts translators from poorer countries.

      Teaching at the high school and college level will be in less demand as students learn from top tier teachers through higher quality video.

      It isn’t that all jobs will be replaced or that new ones won’t emerge but labor will certainly look very different by 2025.

    • http://overcomingbias.com RobinHanson

      Care to offer concrete bets corresponding to your “likely” claim?

      • Mark Bahner

        “Care to offer concrete bets corresponding to your “likely” claim?”

        Sure.

        1) Contingent upon 80+% of the vehicle miles traveled in the U.S. in 30 years being by fully automated (“Level 5”)*** vehicles, I’ll bet up to $100 that 80+% of the brick-and-mortar Walmart and Target stores will have been shut down or repurposed to be warehouses (rather than retail establishments wherein people come to the store to buy things).

        But in 30 years, I’ll be over 85, so…

        2) Contingent upon 50+% of the vehicle miles traveled in the U.S. in 20 years being by Level 5 automation, I’ll bet up to $100 that 40+% of the Walmart and Target stores will have been shut down or repurposed.

        Per Walmart’s website, as of August 31, 2015, they had 5,249 stores in the U.S.

        Per Target’s website, they had 1793 stores at the end of 2013. So let’s say it’s at about 1810 stores today.
        P.S. ***My previous comment stated that Level 4 was fully autonomous. But it’s Level 5 that’s fully autonomous.

    • Lord

      No, not retail. Shopping offline is much easier than online for anything that requires immersive information and that won’t change this century. We haven’t begun to replace anything other than visual and textual information. Non shopping, that is, refilling of existing orders, online can be made to work but is hardly worth the effort.

      • Mark Bahner

        “Shopping offline is much easier than online for anything that requires immersive information and that won’t change this century.”
        What are some examples of things that require “immersive information,” and what percentage of store sales do you think require that “immersive information”?
        I’ll counter that the items most people would say most require a brick-and-mortar store…such as clothes and shoes…are actually much better handled by a fully autonomous vehicle delivery system, wherein customers receive several sizes to try at home, and return the unwanted items.
        Even with current technology, I’m pretty sure somebody on the order of Google/Amazon/Walmart/Target could set up a “virtual store” that could be navigated much like Google Earth Street View (though hopefully in a more intuitive manner!), and displayed on a big-screen LED TV. Technology like Oculus Rift would probably be even cooler for virtual shopping.

      • http://juridicalcoherence.blogspot.com/ Stephen Diamond

        It was easier to find interesting books at Borders Ann Arbor than today at Amazon. Disagree?

      • Mark Bahner

        “It was easier to find interesting books at Borders Ann Arbor than today at Amazon. Disagree?”
        Oh, I agree. But Ann Arbor is hardly what I’d call a “typical” city. I worked in downtown Detroit for a year (back when the first Back to the Future came out :-)). One weekend I went to Ann Arbor (I think partially to go to the UM library).
        I saw a city of 100,000+ people, pretty densely populated, almost entirely by well-educated people in their late teens and early twenties. From that very striking impression, I’d say if a brick-and-mortar bookstore can’t make it in Ann Arbor, it basically can’t make it anywhere in the world.
        And even then…I think if Amazon really tried to create a “bookstore experience” with a website, they could. Already, they recommend books based on whatever other books a person has looked at or bought. (I’m sure brick-and-mortar Borders could do that too, but do they?)
        P.S. About the only books I buy are far-from-mainstream stuff like Tyler Cowen’s The Great Stagnation…which I see is available at Amazon for $0.01 (plus $3.99 S+H). How can brick-and-mortar Borders compete with that? Especially when computerized book depositories and computer-driven vehicles will cut the S+H charges in half?

  • kurt9

    This analysis is spot-on and is a badly needed antidote to all of the silly posturing I see on facebook about how automation is about to eliminate all of the jobs. You would expect investors to pour their money into the companies responsible for the automation. Largely, they are not. Young people would also be more into pursuing careers in real engineering fields as well as computer science/IT. Again, this appears not to be the case.

    Either the argument that automation is about to eliminate most jobs is a load of unsubstantiated hype or that the vast majority of students and investors believe it to be a load of hype (or vaporware, as we say in the industry).

    It is issues such as this is why the idea of betting markets as a replacement for talking head pundits is such a brilliant idea. People who make predictions and pontificate on the basis of those predictions should be willing to bet their money on their predictions. People unwilling to do so make clear they have no personal conviction in the kind of rhetoric they spew forth, and can be safely ignored.

    • http://juridicalcoherence.blogspot.com/ Stephen Diamond

      This analysis is spot-on and is a badly needed antidote to all of the silly posturing I see on facebook about how automation is about to eliminate all of the jobs.

      I’m out of the Facebook loop. What image or program are automation doomsayers promoting? (It’s not, I wouldn’t think, exactly what gets called an sjw agenda.) What are they posturing as?

      [Who are the main parties to this tug of war?]

      • kurt9

        I get a lot of postings on my facebook page from those claiming to be “transhumanists” but who are really leftists advocating for some variant of the dole (universal basic income schemes). Most of these postings argue on the basis that automation and robotics will eliminate most jobs for humans.

        I really need to clean up my facebook page.

    • Mark Bahner

      “This analysis is spot-on and is a badly needed antidote to all of the silly posturing I see on facebook about how automation is about to eliminate all of the jobs.”

      Well, I don’t do silly posturing on Facebook (or anywhere else). But my blog does contain a prediction that approximately half of the top 15 jobs in 2012 will be gone by 2044, due to computer intelligence. In particular danger I think are tractor trailer truck drivers (>90% reduction), cashiers (~90% reduction), and material movers (e.g. loading dock workers…>85% reduction).

      “You would expect investors to pour their money into the companies responsible for the automation.”

      The company with the 2nd largest market capitalization in the world, Google, is gobbling up robotics companies.

      “Either the argument that automation is about to eliminate most jobs is a load of unsubstantiated hype or that the vast majority of students and investors believe it to be a load of hype (or vaporware, as we say in the industry).”
      I think the more likely is that automation is going to eliminate very many jobs, but 10-30 years into the future…and most students and investors don’t look 10-30 years into the future.