Is Technological Improvement a Market Failure?

In Reply to a Mises Myth Regarding Free Markets

“Technological improvements in the farm sector particularly caused much of the material progress in the 20th century, because it enabled the agricultural sector to maintain productivity with fewer workers…Yet, Stiglitz argues that technological improvements are a market failure…

Does he really? It seems to me that the market failure here is not the emergence of new technology per se but the time horizon over which new technologies are deployed. History shows us very clearly that markets fail to correct for technology-driven labor dislocations quickly enough to deter the kind of demographic backlash that ultimately cause them to recede into Recession or worse.

What free market proselytes of the Mises camp actually mean when they claim that the future need not be managed is that “what happens in-between doesn’t matter so much as what happens in the end.” This is not only intellectually sloppy, but ethically dubious. The extent to which this kind of reasoning attempts to avoid all possibility that market intervention might sometimes be a good thing reveals what they truly believe. They believe that humanity should merely accept sociopolitical instability as an inevitable byproduct of modern living. At least that’s less bad than industrial policy (or Communism), right?

What happens in-between doesn’t matter so much as what happens in the end.

Alleviating people of the burden and misery of rote work is surely a good thing. Ignoring the real social costs of change, however, on mere faith that markets tend to equilibrate over time is foolish if on the path to equilibrium this implies civil instability or economic recession.

Take, for instance, the rapid rise of Uber in 2014 and the taxi revolts which spread like wildfire over Europe that summer. Though labor markets do ultimately adjust to changes in patterns of economic specialization and trade, I would be careful to invoke Ned’s pejorative here. Do Luddites really despise technology, or are they fearful that they and their families are likely to be excluded from the economic transition they sense is fast approaching?

Insomuch as the long game of technological improvement looks hopeful, we really should not discount the possibility that pitchforks might be just around the corner.

The quotes continues:

“Yet, Stiglitz argues that technological improvements are a market failure…even though most of the transition from agriculture to manufacturing occurred 20–30 years prior to 1929.”

There is a very simple explanation for this. The rate of change of structural unemployment is a function of technological change with a time lag. In fact, it was in precisely this period 1889-1914 which saw the Second Industrial Revolution come to heel. Major technological advances during this period included the telephone, light bulb, phonograph and the internal combustion engine. It was a period of growth for pre-existing industries and the expansion of new ones, such as steel, oil and electricity. Electric power was utilized to amplify mass production to a scale simply not imaginable to those who had lived during the First Industrial Revolution, those of whom who were more at home with the Spinning Jenny than AC power. This, as Christoper Westley of the Mises Institute correctly points out, all occurred 20–30 years prior to 1929.

Critical new technologies seem to take between 20–30 years to affect sustainable patterns of specialization and trade, which then fundamentally alter aggregate demand and as a consequence the structural demography of the population over which the new technology is distributed. Mass unemployment seems to be the predominant medium-run result. So why the time lag?

A simple yet compelling explanation comes from Peter Turchin, whose work specializes in cultural evolution and cliodynamics — a field of mathematical modeling and statistical analysis that explores the dynamics of historical societies. The time lag between the discovery of new technologies (a good proxy for which is rising income inequality) and a trend reversal in socioeconomic well-being occurs because despite the market leverage that comes with adopting new technologies, existing employers “may be constrained by prevailing social norms of fairness (Turchin, 2016),” which is simply to say that it is more difficult to suppress wages or eliminate old jobs than it is to form a stranglehold on technology that nobody else yet possesses (that is, to prevent more productive and hence higher-paying jobs from being duplicated across the economy and amongst competitors).

Figure 1.0 —Dynamics of relative wages (wages relative to GDP per capita). Calculations by the author based on wages (production workers in manufacturing) and GDP data from MeasuringWorth (Johnston and Williamson 2013, Office and Williamson 2013). The units of the y-axis are workers’ annual compensation as percent of GDP per capita. — Ages of Discord (Turchin, 2016).

The Third Industrial Revolution occurred between 1978–2008. It focused on advancements such as the personal computer, the internet, and information and communications technology (ICT) in general. If we assume that the brunt of the ITC Revolution ended approximately 15–20 years ago, c. 1997–2002, then the size and power of today’s newest tech firms, such as Facebook, Google and Amazon is surely to be expected. Given that they grew out of garages, with the full force of the latest technology at their disposal, they had no jobs to cull and few social norms to evade.

Figure 2.0 — Dynamical relationship between immigration and economic inequality. — Ages of Discord (Turchin, 2016).

The last time that relative wages (wages as a proportion of GDP) declined was in the early 1900s, just prior to World War I and again from the 1820s until the outbreak of the US Civil War. Do relative wage measures predict major military conflicts? Well, technology-biased inequality certainly drives wage suppression from a position of leverage, and this is often accompanied by lobbying efforts to ramp up immigration policies that further favor corporate greed (Figure 2.0).

Over time, the domestic wage immiseration caused by these conditions leads to periods of sociopolitical instability and sometimes war, in which labor tends to be destroyed. When people are destroyed, wages (in relation to GDP) tend to increase because there are fewer people per unit-capital (note also that since wealth is Pareto-distributed and death in war time is relatively stochastic that average wages will rise faster than expected if the process were non-random (Taleb, 2016), i.e., the elimination of one wealthy estate releases substantially more wealth than the average economic casualty).

This isn’t exactly a rosy picture. If you look again to Figure 1.0 above, you’ll notice that most major US military conflicts occur just after a free-fall in relative wage growth (or after periods of wage stagnation), and coincide with periods in which relative wages were rising (The Barbary Wars, War of 1812, 1800–1820; US Civil War, 1861–1865; World Wars I and II, Korean War, Vietnam War,1920–1960). Basically, inequality froths and bubbles to a critical point beyond which conflict is inevitable. Only after the carnage and bloodshed does everything return to normal.

We are now at the precipice of the Fourth Industrial Revolution, which will see the emergence of a nexus between informatics and physical stuff, i.e. robotics, artificial intelligence, nanotechnology, quantum computing, biotechnology, The Internet of Things, 3D printing and autonomous vehicles. However, before novel forms of productive employment spring forth, we will likely experience the negative repercussions of the industrial reorganization which took place at the tail end of our own contemporary Third Industrial period. It is now nearly 2018, a full decade after the end of the last great structural transition. What will the become of the next 20 years?

The Battle for Paris, circa 2027? https://autodo.info/pages/w/world-war-3-concept-art/

References

Nicholas Taleb, Nassim. (2016). Stochastic Tail Exponent For Asymmetric Power Laws. SSRN Electronic Journal. 10.2139/ssrn.2834165.

Turchin, Peter. (2016). Ages of Discord: A Structural-Demographic Analysis of American History. https://www.amazon.com/Ages-Discord-Peter-Turchin/dp/0996139540