As I read this passage in Jeffrey Sachs’s Boston Globe op-ed on income inequality:
Consider technology. Throughout modern history, ingenious machines have been invented to replace heavy physical labor. This has been hugely beneficial: Most (though not all) American workers have been lucky to escape the hard toil, drudgery, dangers, and diseases of heavy farm work, mining, and heavy industry. Farm jobs have been lost, but with some exceptions, their backbreaking drudgery has been transformed into office jobs. Farm workers and miners combined now account for less than 1 percent of the labor force.
In the 1900 federal census 90% of men listed their occupations as “farmer” and in 1935 there were 8 million farms in the United States. Technology has changed that. Although U. S. agricultural production is considerably greater than it was a century ago, the total number of farms and farmers has dwindled. Only about 1% of American workers are farmers.
We do not have 89% unemployment among men. It isn’t 49% or even 29%. It’s closer to 5%. How can that be?
What has happened is that other jobs, including jobs that couldn’t even be imagined in 1900, have taken the place of all of those jobs as farmers. Although fewer workers are involved in primary production enormously more workers are involved in secondary production. Jobs like information technology, healthcare, education, and UPS delivery.
Is it possible that today is different? That today’s jobs that become obsolete through technology won’t be replaced by anything? Why, yes it is. But the evidence for that is scanty and there are much more compelling explanations for what’s happening including the rise in occupational licensing requirements and other barriers to entry, deadweight loss, and immigration.
However unlikely, how you would go about mitigating the risk of widespread, massive technological unemployment is a reasonable question. The strategy I’d propose is to deal with the factors mentioned above which would reasonably be expected to produce the results we’re seeing before taking any sort of strenuous measures to deal with widespread, massive technological unemployment that may never occur.
I wouldn’t really expect it to happen suddenly and with no change in direction. I think changes occur now much faster than in the past. People certainly change jobs now much more frequently than in the past. Given the rate of change, I think it exceedingly likely that at some point in the cycle we will be destroying jobs faster than we are creating them, and that this could go on for a while. So even if the paradigm has not shifted over a longer period of time, meaning you are correct, for shorter periods, which could last for months-years, we could see large numbers out of work because of changes in technology. Other than build a better safety net and better access to the kinds of education needed to have the skills and knowledge needed when the new jobs do come along, not sure we can do much.
Steve
Technological advances allow the population to increase. Contrary to Malthus and any modern Malthusian predictions, the number of people that can be sustained through future technological advances will continue to increase. This increase has direct and secondary requirements.
At one time, Boeing was testing the use of Virtual Reality eyewear for their workers to eliminate the need to lug around paper manuals (or something similar). They would have been able to overlay a heads up display onto the part of the airplane they were repairing or assembling, but I am not sure whatever happened with it.
The reason jobs require a college degree is because a high school diploma is worthless and because there are too many applications submitted for one job. Increasing requirements is a method of culling the pool. Decreasing requirements is a method of enlarging the pool.
Technology advances allow low skilled labor to do the jobs that once required highly skilled labor. Advances that allow better control for more precision can result in fewer rejects, and/or it can open up an entire new product capabilities.
Computing has become easier not harder. How many people know how to write code in Assembler/COBOL/FORTRAN, to use a punch card machine, to wait for the printout from their code run, or to read the output (green and white paper)?
If it seems easy today, try optimizing your Autoexec.bat and config.sys files. To the majority of today’s technologically skilled workers, that could have been written in ancient Greek. Technology for geeks is different than technology for the public. For geeks, Win7 is still using a lot of old systems, but for the public, they have nice shiney non-geeky interfaces. (I have not gotten into Win10, yet.)
The people who need to be worried are the skilled people. They are too expensive. The problem with unskilled labor is not expense. It is reliability and production quality. Robots can replace them and fix both of those problems, and for skilled labor, unskilled workers can be used with technology supplying the skills they do not possess.
A 3D printer allows somebody with no skills to become a sculptor for a few hundred dollars, and at some point, only the experts will know the difference. The person creating the 3D template will only need to know enough to work the application, and the rest will require creative skills.
Other than truly creative types, everybody can be replaced or augmented through technology. Technology “levels the playing field” for the less capable.
steve:
That could well be right. Basically, I think that planning for the end of work is premature.