Four Forces Watch: Larry and the Robots

Posted on by

Mike Konczal writes,

There’s been a small, but influential, hysteria surrounding the idea is that a huge wave of automation, technology and skills have lead to a massive structural change in the economy since 2010.

He goes on to say that Larry Summers has demolished this notion, by pointing out that we have not seen rapid productivity growth over this period.

This looks suspiciously like a straw man to me. I am about as big a believer as there is in the significance of structural change, but I do not see a “huge wave of automation” that has taken place over the past five years.

My thoughts:

1. What I do see are the four forces: the New Commanding Heights (demand for physical goods tapering off while demand for health care and education rises); demographic dispersion–what Charles Murray calls Coming Apart; factor-price equalization (American workers confronting stiffer foreign competition); and Moore’s Law (improvements in computers and communication).

2. These four forces have been operating for decades, and there was nothing peculiar about the 2010-2014 period. The New Commanding Heights force has been operating for more than 50 years. The demographic dispersion force got started about 50 years ago, but for the first 25 years or so the impact was small. Instead, the most notable demographic change from 1965 to 1990 was the increase in female labor force participation. Factor-price equalization had to await liberalization of the economies of China and India, which did not really get started until the 1980s and even then took a while to have an impact. Moore’s Law was articulated in the 1960s, but as recently as the late 1980s Robert Solow’s quip that we see computers everywhere but in the productivity statistics seemed apt.

3. The four forces cause the economy to move in the direction depicted in Neal Stephenson’s The Diamond Age. In that novel, we see Thetes, who work very little and live inexpensively (think of a consumption basket dominated by big-screen TVs). And we see Vickies, who work creatively and hard in order to consume unnecessary luxury services (think of high-tuition education, high-end precautionary medical care, and exotic vacations).

4. Larry Summers looks at the last twenty years and sees a “secular stagnation” in aggregate demand, interrupted by the dotcom bubble and then the housing bubble. Instead, I look at the last 15 years and see The Diamond Age starting to become reality. The housing bubble gave Thetes the impression of being wealthier than they really were, and when it popped they had to adjust to reality. (Although one could argue that, illusions aside, they did not lose home equity, because they did not have any in the first place.) The process of adjusting to reality can take time–look at Greece.

5. Consider the statement, “If we had more aggregate demand, then more non-high-skilled people would have jobs and wages would be higher.”

I do not believe that statement, Instead, I believe that nothing short of direct intervention in labor markets (government make-work jobs and wage subsidies, or you could hope for an impact from changes in means-tested programs that reduce implicit marginal tax rates) will change macroeconomic outcomes. But as long as jobs and wages are lower than what Summers/Krugman/Sumner think they should be, there is no way to falsify the statement that “if we had more aggregate demand….” The alleged lack of jobs and wages can be viewed from their perspective as proof that there is insufficient aggregate demand. There is no measure of “sufficient aggregate demand” that exists independently from the desired result of a larger wage bill. Indeed, Sumner would say that the wage bill is a measure of aggregate demand that can be targeted by the Fed.

9 thoughts on “Four Forces Watch: Larry and the Robots”

  1. Are these no-growth, Baumol’s diseased sectors actually rising or is the tide going out revealing the rocks?

      • I’m working on it in my series (though, I won’t be getting into PhD levels of mathematics). One way to look at it is that rates of return on homes (defined as net operating surplus on rent) tracked along with real interest rates for decades, with a small premium, until 2007. By 2007, the implied return on homes was down to about 3%, 1% or so higher than long term tips, similar to real mortgage rates. But now, homes return about 5% while long term real bonds are very low. You can use that implied yield to estimate what the price should be. I hope to eventually do that. But, I think that the lack of home investment is partly what has driven bond rates down, so we can’t just use them as a benchmark, without some adjustment.

  3. This is the 21st century. Higher consumption levels in the US do not necessarily lead to higher demand for labor in the US. If consumption levels in the US rise, that is at least as likely to increase employment and wages in Mexico, India, Brazil, etc., as it is in the US.

  4. Since the ‘new commanding heights’ (eds & meds) are in non-tradeable services, how long can they avoid being affected by the same price-factor equalization processes (as the incomes of a large fraction of their customers converge with the developing world)? Is the disaster in law schools applications an early sign?

  5. There has been significant automation over the past five years, and there will be even more over the next five years. Much of the lawyer’s job has been automated. Professors don’t grade homework anymore (or at least I don’t)–it’s all graded by computers. Books are published directly by authors–intermediate services have been automated out (or replaced by the gig economy).

    In the near future truck drivers will lose their jobs. Many medical jobs will be automated. Cars will increasingly drive themselves (though automation in city traffic is probably still a few years off). Professors will be displaced by computers or on-line. 3D printing radically changes manufacturing and retailing, reducing the number of employees.

    I also disagree about meds & eds. Both of those industries will be partially automated, prices will collapse, and workforces will shrink. WRT eds, Bryan Caplan is probably the guy to listen to (though he does exaggerate). Much of these services will move off-shore, either through medical tourism, or through the internet.

    India is becoming the new powerhouse in medicine and medical research. What jobs remain will tend to move there.

    Productivity is a poor measure of this effect. Productivity is revenue/(employee-hours). If both the numerator and the denominator decline, then there is no change in productivity.

  6. There is a strengthening of the ongoing division between more high end creative jobs and more low end service jobs with fewer mid level jobs and a smaller middle class.

    Aggregate demand really depends on whose demand. More at the high end will just lead to increased asset prices and higher priced high end services. More at the low end will indeed lead to more jobs and services at the low end. Tax cuts that favor the wealthy won’t work because those are wrong sorts of investments to increase employment.

Comments are closed.