A Bit More on Technology, Jobs, and Wages

February 17th, 2014 at 12:21 pm

I couldn’t jam everything I’ve been thinking about these questions about technology, work and wages into a single piece, so just a few (tantalizing) bullet points on other dimensions of what I think is a fascinating line of inquiry.

–A lot of how people think about these issues is impressionistic, based on amazing advances in consumer electronics.  The absolutely remarkable things that this small, thin smartphone in my pocket (the quite awesome Samsung Galaxy 3, fyi) can do must be economically transformative, right?  That’s certainly the Tom Friedman view of the world, but it’s largely assertion that doesn’t always match up with the facts.

–That is, the argument that technology is driving wage or employment trends is a lot more elusive than you’d think.  As Larry Mishel points out here, the wage trends don’t consistently move the way the tech story predicts.  Research associated with economist David Autor initially made a compelling case that technology was displacing workers in the middle of the occupation scale.  But while that theory roughly fit the data in the 1990s (see first two figures here), it doesn’t work in the 1980s or, more importantly given the subsequent dispersion of technology, the 2000s (see Autor’s own Figure 1 here).

–There’s newer research suggesting that the demand for skilled workers has actually decelerated in recent years.  Beaudry, Green, and Sand present an exhaustive and rigorous statistical analysis of skill demands over the last three decades.  They look at tasks, jobs, and earnings, and find that the demand for skilled workers “underwent a reversal” around 2000.  The growth in the share of high-skill, high “cognition” (using Autor’s tasks framework), and high-wage occupations stagnated in the 2000s, where the share of college grads kept growing.  “That means,” as Green told me, “that the probability that a newly graduating BA will get one of these jobs is declining sharply.”  And as more highly skilled workers are displaced, they moved down the job scale, hurting the job prospects of less-skilled workers who now have to compete with those “…displaced from cognitive occupations.”

All of which is highly inconsistent with the idea that underlies all of this tech stuff: we’re surrounded by amazing devices with massive informational and communications capacity.  Surely, their ascendency has changed skill demands, wage trends, and reduced the employability of many.

In fact, there’s of course some truth to that, but it’s highly dynamic—like the wage trends, there’s evidence for the tech story in some periods but not in others—and it’s but one factor in play.  After all, we achieved full employment in the latter 1990s, including rising wages for the low- and middle-wage workers at a time when tech was booming (in fact, bubbling) and the skills of the workforce were little different than they are today.

Two more points.

First, while much of the above argues for assigning a lesser role to tech in driving economic developments, there’s also an important way in which economists understate the impact of tech on our lives: consumer surplus.  I don’t like to depend on anecdotes, but they abound in this space and I can’t imagine how I could be wrong about this.  EG, I just met this couple with a newly adopted (adorable) 10-year old girl from China.  Of course, she speaks Mandarin, but they’ve been successfully using Google translator, which is, of course, essentially free.  (Yes, there’s bundling and ads and internet costs, but I’m counting them in my assumptions re consumer surplus.)  I guarantee you they’d pay a ton for this translator, but they don’t have to.

To be clear, I’m not saying, as some do, that this somehow replaces wage and income stagnation.  It does not.  But the proliferation of consumer technology has made us better off in ways we don’t always count.

Second, why am I going on about all this—what are its policy implications?

Well, you know my methods, Watson.  It’s precisely because of the policy implications.  Simply put, the tech diagnosis leads almost exclusively to the education prescription.  Now, the problem with that is not more education, which is of course essential, especially for those locked out by inequality/opportunity barriers.  It’s the “exclusively” part.  The tech diagnosis shifts the emphasis from macro to micro, from institutional power imbalances to individual shortcomings, from demand-side to supply-side, from financial market oversight to “the problems with our schools,” even from stimulus to deficits.

Again, that doesn’t mean the tech story is wholly wrong by any means, though its advocates tend not to deal with challenges like those I’ve ticked through above.  But it is clearly incomplete and that has serious, negative policy implications.


Print Friendly

27 comments in reply to "A Bit More on Technology, Jobs, and Wages"

  1. Robert Buttons says:

    I’m not an economist, I have a background in science. But I find it extremely depressing that trained economists even are having the “does technology kill jobs debate?”. I would have thought the Luddites, like the creationists, would have been purged long ago from the realm of scholarly discussion.

    To put it simply, technology frees up workers to do other things.

    The basic goal of modern economics seems to be: “We need to figure out how to make sure all are citizens can slave away on an assembly line”. (A job is an end)

    It should be:

    “We need to figure out how to use technology, so we can work as little as possible” (and a job is only a means)

    • Fred Donaldson says:

      “To put it simply, technology frees up workers to do other things.”

      That includes being unemployed.

      • Robert Buttons says:

        Monsieur Bastiat has this to say:

        We are suffering from the ruinous competition of a rival who apparently works under conditions so far superior to our own for the production of light that he is flooding the domestic market with it at an incredibly low price; for the moment he appears, our sales cease, all the consumers turn to him, and a branch of French industry whose ramifications are innumerable is all at once reduced to complete stagnation. This rival, which is none other than the sun, is waging war on us so mercilessly we suspect he is being stirred up against us by perfidious Albion (excellent diplomacy nowadays!), particularly because he has for that haughty island a respect that he does not show for us .

        We ask you to be so good as to pass a law requiring the closing of all windows, dormers, skylights, inside and outside shutters, curtains, casements, bull’s-eyes, deadlights, and blinds — in short, all openings, holes, chinks, and fissures through which the light of the sun is wont to enter houses, to the detriment of the fair industries with which, we are proud to say, we have endowed the country, a country that cannot, without betraying ingratitude, abandon us today to so unequal a combat.

        If France consumes more tallow, there will have to be more cattle and sheep, and, consequently, we shall see an increase in cleared fields, meat, wool, leather, and especially manure, the basis of all agricultural wealth.

  2. Ruthmarie says:

    This is something that I have been screaming about for years to anyone who might actually listen. Everyone is wringing their hands about a “skills mismatch” and no one is looking at the obvious. Salaries in the over-touted “STEM” fields have been outsourced, insourced, and simply eliminated. In most STEM fields there is a very grim GLUT of scientists and engineers that is crushing opportunities and salaries.

    I have a doctorate in immunology and microbiology from a prestigious institution. I saw the writing on the wall in 2005 and QUIT. I now sell real estate and am an architectural photographer. The money is being sucked out of these fields as well. But I was not willing to use my hard-earned 7-year doctorate toiling 70 hours/week for under $40k a year in terminal post-doctum until I was old and gray. The opportunities have been sucked out thanks to industry creating a race to the bottom and austerity programs that are gutting the NIH and NSF.

    It takes the commitment of a society to create the atmosphere in which these fields (and the benefits society can reap from discovery) to flourish. Its a 3-legged stool – a commitment by academia, industry and government is required or it all falls apart. Industry is just chasing $$$, academics are squeezing the life out of their grad students and post-docs because THEY are squeezed and government is doing nothing to help and in fact is making things worse through sequester and other cuts.

    The scary thing is that all this wonderful technology we are seeing is about to dry up. The well runs dry when we don’t plant the seed corn. We stopped planting corn about 15 years ago….So the other shoe is going to fall on this scenario if we don’t do something.

  3. Benedict@Large says:

    Good piece over at Economix. I can recall back in the early 70s when I got into data processing that everyone (yours personally excluded) was a-buzz with how many jobs we’d be eliminating. Indeed, each new project was heralded by a report that (among other things) featured just this fact. I thought this silly even then for exactly the reasoning as Blinder uses: We’ve always had this productivity/machine growth, and it’s always eliminated jobs, except that we never seem to end up with any fewer jobs. In fact, from my experience back then (in the 70s and so), whatever we eliminated was far as jobs was always replaced by new jobs that paid more and were generally more satisfying (less tedious) for the people who did them.

    So for those who suggest a new machine age, I’d say the burden is on them. Our history over the many millennia quite firmly says otherwise.

    [On a personal level, I’m much more hostile than the above would indicate. I’ve absolutely buried myself in macroeconomics since the 2008 crash, and to my now educated view, the entire focus of macro (since then) is to make up excuses why they all missed the boat, and why their “stuff” isn’t really wrong after all. I’ve been around a lot of professions and academia in my lifetime, and I’ve never seen such a sorry bunch of excuse-makers as I’ve found in macroeconomics. Not all, of course. Just most of them.]

  4. readerOfTeaLeaves says:

    Great point that blaming ‘tech’ leads to limiting policy discussions to ‘problems with our schools’, rather than confronting the deeper, underlying policy problems. And IMVHO, these problems include our continued inability to address the Shadow Banking sector, as well as the fact that ‘debt’ has become a very profitable consumer product the past 30 years.

    I’ve never seen anyone make a timeline linking the changes in inequality and US economic performance with the emergence post-1980 of private equity and leveraged buyouts.

    A good timeline would probably show up some correlations: many of the ‘lost jobs’ in the 80s and 90s were lost to ‘leveraged buyouts’, although automation got the blame. And those leveraged buyouts occurred with (higher risk) junk bonds. That speculative economic activity: (1) increased the power of financiers/rentiers, while also (2) prompting businesses to adopt new accounting measures in order to try make themselves less appealing to hostile takeovers. I suspect this had led to some nefarious shifts in corporate governance.

    IOW: a lot of companies were purchased ‘with other people’s money’ and then laden with debt. It’s no surprise that layoffs would have followed as companies struggled under debt loads that their business models could not sustain.

    PE lies in the realm of Shadow Banking, and as PE, tax havens, off-books and shell corporations have mushroomed, the economy has suffocated under what might be called Enron Economics. (FWIW: Some of the money sloshing around PE originates in municipal and state pension funds — it would be simpler and more straight-forward to simply let the pension funds purchase and manage rentals within their jurisdictions, rather than giving a cut to PE and the banks.)

    ZIRP appears to feed leveraged buyouts, by providing more ‘liquidity’ to the finance sector than is probably wise or prudent in the larger scheme of things. It also avoids dealing with the problem of the rapidly morphing Shadow Banking System.

    I’ve also never seen a timeline that overlaid changes in US usury laws onto longer term economic data. But as Sen Eliz Warren documented brilliantly in ‘The Two Income Trap’, after a SCOTUS decision in 1978 allowed an interpretation of the law to mean that banks in one state could ‘export’ interest rates to another state, suddenly a whole lot of bank credit products were located in South Dakota — where rates were capped at 24%. (p 128)

    IOW, a bank in State A had a usury cap of 12% until 1978. After that, they could claim a base in South Dakota and charge as much as 24% interest. And they did. Small wonder that writing in 2003, Eliz Warren could state that ‘bank profits had more than tripled’ since 1970.

    In addition to digitization since the 1970s, we’ve had: (1) usury, which enlarged the REIT sector, and (2) the rise of Shadow Banking (PE, leveraged buyouts, junk bonds, tax havens). Because usury, PE, tax havens, and shell corporations are entirely created by a system of laws, those problems are amenable to policy solutions. Or rather, they would be if the US had a few politicians besides Liz Warren, Sherrod Brown, and Maria Cantwell with enough brains and backbone to devise economic guidelines that would enable the economy to be more productive.

    • smith says:

      You left out Senator Bernie Sanders, who actually introduced a law in 2009 to cap interest rates on consumer credit at 15% (from the current 30%). One should point out that the 30% rate removes moral hazard from the banks, they can offer risky credit because they count on making up for it with the 30% rates. It also guarantees that consumers will get into serious debt or there wouldn’t need to be 30% rates. Once you have a 30% rate, because only because your credit isn’t good to begin with or you have financial difficulties, how can you recover paying 30% in interest. You can’t start over with bankruptcy due to the 2005 bankruptcy law, which again removes moral hazard of banks making risking loans, and guarantees consumers will get into serious trouble. The banks increase lending beyond prudence, and none of this passes without Democrat support.

      • smith says:

        The Sander’s legislation was defeated garnering only 20 something votes, also one should mentions Dick Durbin and his quote about bankers “And they frankly own the place” (the Senate)

  5. smith says:

    I would like to offer a conflicting opinion
    “the tech story is wholly wrong”
    Here is why. In our economy, there is potential demand, especially for useless stuff we don’t need. Is anyone seriously suggesting everyone has everything they want? Of course not. In fact, looking at the level of poverty, and the distribution of income where the greatest number of people make the lowest amount*, it’s laughable to consider there might be a lack of potential demand. Thus the current problem of demand and any foreseeable future problem of demand is one of distribution (inequality). Proof of potential demand? Debt levels, as saving levels again hit record lows.

    Suppose people finally decided they didn’t need all the latest junk and potential demand didn’t keep pace with technology enabled capacity? Productivity increases should decline to zero as lack of demand dampens investment. But, if capital is willing to share the productivity gains with labor, then productivity can continue to increase. Rational management replaces labor only if technology yields a net gain. If the gain is shared, both capital and labor can win. Again the problem is one of distribution, but more specifically from the gains provided by technology, which can be monetary, and/or less time spent working.

    The benign technology scenario does say output is a useless measure. Just because large screen T.V. and computer prices drop, does well being go down? If households can get by with one wage earner instead of two, and many prefer that, is that a bad thing? If we over spend on health care with worse outcomes, do we really care if GDP is larger by 5%? (not a hypothetical, the actual case right now). Spending on environmental cleanup vs. preventing industry damage? Or defense spending matching the entire world’s (4% of GDP)?

    Finally, I have real problems with the graph in the Times, and comparison of time periods. The employment trend from 1959 is 2007 is pretty straight, use a straightedge on your screen to split the curve. Or use a straightedge from 1959 to 1991, it actually extends under the 2007 level. I would question comparing the 90’s which had exceptional employment growth spurring output, with 2/3 of the 2000’s which began with a recession. I would further add that the jobless recovery of the 2000s was due to stimulus that benefited mostly the rich (tax cuts), rising inequality, again broken record like, seems a problem of distribution.

    Of special note is what happened to tech jobs 2000 to 2005 following the tech crash.

    *the shape of income distribution is a child’s slide, instead of the expected bell curve

  6. Nick Batzdorf says:

    Smith says:

    “But, if capital is willing to share the productivity gains with labor, then productivity can continue to increase.”

    Exactly. And the next question is what it’ll take to convince capital to do that – which comes down to who owns the robots. We need policies that encourage shared capital ownership.

    But there’s no denying that the robots are coming, and this is an interesting study of how it’ll happen:


    • Robert Buttons says:

      “Policies” (presumedly govt mandates) that “encourage” (presumedly require) shared capital ownership will ensure the robots (and any other new tech) WON’T be coming.

      • Nick Batzdorf says:

        Right, Socialism bad. Got it.

        Look how successful the FHA has been for promoting home ownership. What if we had something like that to promote employee stock ownership?

        The next stage in the digital revolution could end badly or end well. Read that link I posted – 45% of current US employment is likely to be automated in the next couple of decades! There’s not going to be any demand for what the robots produce if only their owners reap the productivity gains. On the other hand, we could all live easier lives if the robots work for us.

        This revolution has both similarities and differences to the industrial revolution, in my opinion. It’ll certainly create some new avocations, but just shouting “free market!” won’t replace the ones that disappear.

        • Robert Buttons says:

          We have seen mega, massive, technological advancement in the last 200 yrs. But we have not seen mega, massive, unemployment.

          The orthodox economists’ monomaniacal focus on jobs is puzzling. We should be focusing on increasing our standard of living. When machines make the necessities of life so cheap that we live better doing menial labor then professional avocations of past eras, I say the economy is working. Further it is trivial if the CEO is earning ever increasing multiples of assembly line work, worker standard of living is much more important.

  7. purple says:

    I agree, the automobile is going to make a lot of Teamsters unemployed… and the spear is going to wipe out axe makers. Isn’t this the story of human beings ? Nothing new here.

    The internal combustion engine is still going strong and was a lot more transformative than a cell phone. And so was the washing machine.

    All that’s going on now is cruddy public policy and an indifferent, smug, elite.

  8. urban legend says:

    We needed a tech bubble in the 90s to achieve full employment, and a bigger and more widely distributed real estate bubble in the 2000s merely to get a reasonable but less-than-satisfactory facsimile of full employment before the Great Recession hit. We presumably know that the huge loss of manufacturing jobs, the worst in the world compared to other industrial countries even as their manufacturing employment more slowly eroded, was largely the result of a high dollar policy started in the late 1990s. But isn’t there a common denominator creating an employment gap in the generation-long slowdown in infrastructure spending, a slowdown aided and abetted by Reagan’s proclamation that “government is the problem,” along with the tax protest of the wealthy?

    It’s all well and good that the President has called for more infrastructure spending to create jobs. However, he doesn’t make much of an explanation that would make people feel it is essential, and so it just sits there. If the case could be made that our failure to keep up with the rest of the world in maintaining and modernizing infrastructure has cost us millions of high-paying, permanent and non-outsourceable jobs, year after year after year, is reducing our competitiveness as a country, and has been a primary reason why employment levels have been so bad for so long, then political pressure to push it through could be magnified.

  9. Kevin Rica says:

    How many janitors, busboys, roofers, or landscapers have lost their jobs to automation and how many have lost their jobs to illegal immigrants?

    Technological change is not a policy and is not readily fixed by policy.

    Self-destructive immigration policies based on Lysenko economics are fixable. But it takes intestinal fortitude.

  10. Mary L Robinson says:

    I find this focus on hardware – devices and robots – very disappointing. I guess software is invisible and hard to detect by the uninitiated. My career was spent automating tasks performed by middle-class workers. They have been eliminated in droves by software alone. How many people work now as airline reservation agents, travel agents, bank tellers, accounting clerks, classified ad takers, etc. etc. compared to 1990. These jobs were not eliminated until the web took hold in the 90s. During that time period, my company went from 1 admin person per 8 technical staff to 1 per 30. How many typists are employed now? How many in 1980? This has nothing to do with cell phones or robots.

    And yes, these were largely boring jobs and people can be redeployed elsewhere. But the pace of job destruction is far outpacing job creation. It will take some time to catch up, and in the meantime, people are hurting. Handwaving is not helpful.

  11. Dave says:

    I’ve been pushing an idea for a long time that the real problem in our economy is that we have a huge shortage of talented CEOs, financiers and opinion column writers. They don’t seem to be doing a very good job, and I think we need new legislation to bring in some guest workers to fill the skills gap, as well as beginning a shift to outsourcing these jobs to the much larger pool of talent in Asia.

    I don’t know what I get no traction with my idea.

    • Dave says:

      By the way, back in 2009 or so I pitched this idea to an old friend who was the president of a software offshoring company (in Russia). I was trying to point out the asymmetry of the power. His response, “you son of a …”

      Capitalism and greed are blinding and character destroying.

Leave a Reply

Your email address will not be published.