Arnold Kling  

Notes from What I've Been Reading

PRINT
Nyman's Non Sequitur on Health... The Extra Burden of Childless ...

As you know, it's Inventing Ourselves Out of Jobs? America's Debate over Technological Unemployment, 1929-1981., by Amy Sue Bix. It came out in 2000. I ordered it after seeing it referred to in A Great Leap Forward, the book I just reviewed.

One of my take-aways is that in many industries the productivity improvements were of an order of magnitude. Some excerpts are below the fold.

After 1940, the issue of technological unemployment was largely dropped. The second world war put a premium on productivity. The advent of the suburban middle class in the 1950s took the edge off of the labor vs. capital adversarial relationship that had permeated popular consciousness in the decades before the war.

Among economists, the Keynesian theory of aggregate demand crowded out interest in technological unemployment. I think it merits another look.

p. 12:

Between 1895 and 1915...adoption of increasingly fast stone-planing machines, along with pneumatic tools and other new devices, had displaced about 50 percent of the nation's stonecutters...one operator using the Owens bottle machine could reportedly equal the production of eighteen traditional glass-blowers.

The classical economists debated technological unemployment. p. 27:

French economist Jean-Baptiste Say's 1803 Treatise on Political Economy declared that "industrious human agents" must "needs be thrown out of employ" whenever new devices appeared...Say quickly added that economic laws could "wonderfully reduce the mischiefs" of displacement...as output rose, prices of good would drop, stimulating an increase in consumption which could ultimately provide jobs for mroe peole than ever...

but the 1819 book New Principles of Political Economy challenged the belief that mechanization brought rising consumption and hence stimulated employment. Swiss economist Jean Charles Leonard de Sismondi [cautioned] that markets could easily become saturated with goods.

...[p. 28] A country might as well have only one inhabitant, its ruler, who "by constantly turning a crank, might produce, through automata, all the output of England."

also on p. 28,

David Ricardo...eventually reversed his initial thinking...His 1817 version of The Principles of Political Economy endorsed an optimistic faith...Like Say, Ricardo felt that over the long run, growing demand should compensate for any "inconvenience" of temporary job loss due to installation of new technologies. In a revised edition five years later, Ricardo declared such assumptions "erroneous." He had become convinced that mechanization indeed proved "often very injurious" to workers

p. 35:

The Bureau of Labor Statistics reported, for example, that ditch-digging machines with one operator and assistant could perform as much work as forty-four hand laborers.

p. 54:

A singel wheat farmer using the latest tractors could, in three hours, complete work that would have taken fifty-seven hours a century before.

p. 66:

One cotton planter who had purchased new tractors and cultivators subsequently eliminated 130 out of 160 sharecropping families...Between 1930 and 1935, the farm population of one Texas panhandle county had fallen 24 percent.

p. 67:

farm employment had dropped from 12.2 million to 10.6 million between 1909 and 1939, as the number of tractors grew from 10,000 to 1.3 million...By the most conservative estimates, each tractor eliminated 150 man-hours of labor a year.

p. 81-82:

Large cigar companies adopted rolling machinery...The resulting industry shake-up, combined with declining cigar consumption and the economic downturn, caused at least fifty-six thousand cigar-making jobs to vanish between 1921 and 1935.

p. 115:

locals expressed outrage at a public works contractor who brought in $75,000 worth of heavy machinery and finished a road job using just eight men, while a comparable site nearby employed forty men using shovels.

p. 120-121:

A less hysterical look at the evidence suggested that mechanization might account for 10 to 15 percent of total joblessness at most, [Meredith] Givens announced, and the remaining cases resulted from the problem of "idle machines rather than the busy machines."

p. 229:

United Auto Workers president R.J. Thomas...[said that] Technical innovations in the industry had increased productivity by 134 percent [between 1923 and 1939] but surging consumer demand had stimulated a 212 percent leap in production which sent employment up 30 percent...even during the Depression years of 1929 to 1938, technological change in the industry resulted in a 17 percent gain in per capita productivity.

p/ 232:

By 1940, rising concern about the international situation had started to affect the way Americans regarded mechanization...In the Depression, labor-saving machines seemed to present a threat; in a time of global crisis, they came to represent a virtue.



COMMENTS (8 to date)
mike shupp writes:

Keynesian thinking may have won over economists after 1940, but I don't recall that it much affected non-economists then, or even perhaps now. My recollection is that as late as 1970, business writers and pop sociologists and labor economists were still batting the idea of computer-related unemployment around, with the general feeling being that some huge crisis was just over the horizon. I remember horror tales about American corporations becoming as overstaffed with middle class managers as Tokyo department stores with greeting ladies.

What changed things amongs the hoi polloi? To some extent, increased respect for Keynesian economics, due to the success of the Kennedy tax cuts. For another, the slowdown in productivity in the 1970's, connected with distinctly non-technological energy shocks and inflation. And for another, rising employment among "pink collar" women workers which covered up a fall of employment among blue collar males -- be sure and read some Marvin Harris.

Kenneth A. Regas writes:

Dr. Kling, perhaps you could clarify your point. I should think that you, of all economists, would say that technical advances can temporarily cause unemployment, but in the long run displaced workers can always find new ways to produce value - forming new patterns of specialization and trade.

We (you?) can say that macroeconomics is powerless to address the long-term level of unemployment, because such unemployment results mostly from social norms and government practices that are not the subject of traditional macroeconomic analysis. These include unemployment insurance, welfare, family support of the unemployed, stigma associated with taking "low" work, and so on. Any society that has a high chronic unemployment rate is a society that, as a matter of expressed preference, would rather support indolence than demand that the indolent find something constructive to do. Desperately poor countries don't have unemployment. Even begging is a kind of employment.

Today we read about the beached white male, emblematic of technical progress undermining people who imagined themselves immune from such things. These are not unemployable people. They are people who strongly resist the truth of their recently reduced value in the labor market. Only those who cannot earn the minimum wage can be said to be "technologically unemployed." Every one else is "voluntarily unemployed," or perhaps "unemployed by unwillingness to accept reduced employment prospects." Yes?

Larry writes:

I'm beginning to think we're on the verge of a once-in-history change. The robot/AI flavor of technology is eating jobs that are far removed from the jobs that tractors and other "conventional" labor-saving devices consumed. The trend is clear. Unless we crash into some fundamental limit, this century is likely to see the end of conventional work. Robots/computers will be able to perform all tasks better and less expensively than humans. The wages that a rational employer would pay will be insufficient to support a person/family. Then what?

Simplistically, either we'll have to create a synthetic culture where we pay people to do things that automatons could do better/cheaper, or we'll shift to a leisure culture. How do we allocate goods and services in such a world? Can the human psyche adjust without destroying itself? A third alternative is that we'll hybridize with technology and continue as before...

Adam writes:

Bix's book sounds like a great history and resource. It seems to fit the thesis that the great depression was prolonged by the dislocation of unskilled agricultural labor.

WWII seems to have retrained the US workforce in the skills needed for the new industrial age--cooperative work in teams and the knowledge needed for the petroelectric age (electricians, mechanics, machine operators, electronic specialists, industrial logistics, aerospace, Hewlett and Packard in Silicon Valley, etc).

jb writes:

People are only capable of doing jobs that require a basket of skills and capabilities( IQ, charm, etc) that they can meet or exceed. For example, an autistic could never be a good fashion designer. And a low-iq person will never be a good software developer.

That's a given, yes? We can all agree that a sufficiently low-IQ person can never provide complex analysis services - they are essentially good for little more than manual labor with a modicum of judgement.

And it is exactly those jobs that are being automated away these days. Cashiers, bank tellers, repetitive manufacturing... These jobs are going away, and there are no good jobs for these low-IQ people to take.

That's a given too, yes? That there is a threshold below which people cannot provide useful labor.

My assertion, then, is that this IQ threshold (I'm using IQ as a shorthand here, read "IQ, charm, analysis, creativity basket" is rising over time as we get smarter and more sophisticated about our automation technology.

And if I'm correct and that threshold is rising, at some point that threshold of uselessness will exceed the capabilities of the average person. Not next week, but I suspect that it may occur in our lifetimes.

Evan writes:
And it is exactly those jobs that are being automated away these days. Cashiers, bank tellers, repetitive manufacturing... These jobs are going away, and there are no good jobs for these low-IQ people to take.

That's a given too, yes? That there is a threshold below which people cannot provide useful labor.


So far the action of Say's Law has prevented this from happening. Otherwise we'd all have been unemployed by mechanized farming. When the supply of something goes up (labor in this case), it becomes cheaper. This motivates entrepeneurs to find new uses for it, increasing demand.

It might be possible that low-IQ people will end up like horses, where their labor is less valuable than the goods and services required to feed and house them. But I'm not sure that's happened yet, if there weren't any jobs for low IQ people we wouldn't see so many of them risking the wrath of the law to move to the 1st world.

The other reason I'm not sure that's happened is because humans, unlike horses, have the ability to drastically alter their skillset through learning. It's sort of like if horses had had shapeshifting powers. I think the market for shapeshifting horses might have survived the automobile.

medwards writes:

So, what did these displaced workers end up doing? Did they find new jobs, or did they remain unemployed and eventually on the dole? If it's the first, then there's nothing to do but wait. If it's the second, then we better increase the dole.

roystgnr writes:
It might be possible that low-IQ people will end up like horses, where their labor is less valuable than the goods and services required to feed and house them.

This could happen to high-IQ (and high-creativity) people, too. The first copy of anything intellectual may be expensive, but the marginal cost of future copies is trivial. The corn farmer is going to have to harvest just as much corn next year, but once the software developer writes this year's program or the engineer designs this project's product or the producer finishes this movie, it's time to find something different to work on... and that "something different" had better either be so dissimilar that it doesn't compete with the previous work or so much better that it wins the competition handily.

So far that's been pretty easy (new intellectual inventions have a way of enabling even-more-and-better inventions), but it seems optimistic to expect that that will always be the case.

Comments for this entry have been closed
Return to top