When robots and automation have taken over not only the agriculture and manufacturing jobs but even the high-level service jobs, who will drive consumption? Will the economy stagnate? These are the questions posed by Martin Ford in his challenging, important and well-researched book, Rise of the Robots: Technology and the Threat of a Jobless Future.
At one point in his book Ford cited an often-told story, “about Henry Ford II and Walter Reuther, the legendary head of the United Auto Workers Union, jointly touring a recently automated car manufacturing plant. The Ford Motor Company CEO taunts Reuther by asking, ‘Walter, how are you going to get these robots to pay union dues?’ Reuther comes right back at Ford, asking, ‘Henry, how are you going to get them to buy your cars?’”
The assumption of economists has always been that creative destruction in the economy will result in overall economic growth and job growth. When some jobs are destroyed, other, better jobs, and more of them, will take their place – created by the very disruptive technologies that destroyed the old ones. When agriculture was mechanized, the manufacture of machines provided enough jobs to employ the displaced workers. When horses and wagons were crowded out by automobiles, horse-and-wagon makers went to work on Henry Ford’s assembly lines.
But what author Martin Ford wonders is whether this pattern – dependable throughout industrial history– is about to end. Robotization not only of manufacturing but of service professions, even financial advice and investment management, may not create enough new jobs to compensate for the ones that were destroyed – and there is no prospect that it will.
Robots are taking over faster than you think
Much of Ford’s book is taken up with accounts of how service professions are being automated by artificially-intelligent machines and robots, which can do the jobs that service workers do, not only just as well, but often better and more reliably. For example before pattern-recognizing machines, very well-paid radiologists read cancer scans for signs of disease; but now machines are brought in as “second opinion” devices, often even first opinion, and their conclusions are just as good.
Ford related one anecdote, apparently true, about two people who saw doctors in entirely different locations, a man in Marburg, Germany, and a woman in Denver, for very similar, severe symptoms. The woman in Denver wound up being given a heart transplant. But in Germany they solved the riddle, because “one of the doctors had seen a February 2011 episode of the television show House. In the episode, the show’s protagonist, Dr. Gregory House, was faced with the same problem and made an ingenious diagnosis: cobalt poisoning resulting from a metal prosthetic hip replacement.”
The point of this anecdote is that while doctors had to rely on serendipity to come up with the right diagnosis, a computer – or the cloud of servers – that stored all the case histories ever recorded and could riffle through them at lightning speed would surely have come up with the answer in an instant. When that happens, and it is by no means in the distant future, what will come of the medical profession? It may be reduced to technicians tending machines that make the diagnoses and prescribe the proper treatment, delivery of which in most cases will itself be automated, be it a pharmaceutical treatment – or even a surgical operation. Will the technicians themselves even be needed?
As to food service, the last refuge of job-seekers with university degrees who can find no other, that industry is fast being automated too. Ford cites the example of a 262-restaurant chain in Japan called Kura in which “robots help make the sushi while conveyor belts replace waiters.” Customers order using touch screens, and after they eat the system clears and cleans the plates and tabulates the bill. The restaurants are remotely monitored from centralized facilities.
Sectors that you wouldn’t believe could be automated
The success of IBM’s Watson computer in beating the champion at the television game of Jeopardy! in 2011 ushered in the realization that computers could do things we thought only humans could do, such as understand – and use – “natural language.” IBM has made Watson’s capability available for many other applications too. The exploration of what is possible may have only begun.
For example, according to Ford, Forbes magazine uses technology from a company called Narrative Science to produce automated articles on sports, business and politics. The automated system actually writes the story, by taking data that has been compiled and transforming it into interestingly readable text, indistinguishable from copy written by a human journalist.
Furthermore, robots can learn; from that learning they can even design and create new robots to do related jobs. If robots and humans work in tandem, with the human using the robot to do the more routine work, the robot can absorb data about what the human is doing and gradually routinize that, too, teaching itself to replicate the human’s job – or designing a new robot to do it. There seems virtually no end to the ways in which robots imbued with artificial intelligence and machine-learning algorithms can replace humans.
The new high-tech corporation uses fewer workers
Ford provides the obvious examples of major companies that need only a fraction of the number of employees formerly required to generate a billion dollars in profits. Google, while employing only 38,000 people in 2012, had a profit of nearly $14 billion; General Motors’ highest employment was 840,000 in 1979 but its profit was only $11 billion (adjusted for inflation), 20% less than Google’s with more than 20 times the workforce. WhatsApp had a workforce of 55 when it was purchased by Facebook in 2014 for $19 billion – an incredible $345 million per employee. Similarly, YouTube, with 65 employees, was purchased by Google in 2006 for more than $1.5 billion, and Facebook acquired Instagram for $1 billion in April 2012 when it employed only 13 people. Value, it seems, is not generated by numbers of employees; very few are needed to create large amounts of capital if the company corners a market.
Ford highlights – as did Thomas Friedman in his 2007 book The World Is Flat – the winner-take-all nature of Internet competition. When one company emerges the winner in a competition it tends to dominate almost completely. This is because people use it because they know other people are using it – this is known to economists as the “network effect.” Once again, this creates massive value without regard to or need for a large number of employees. The relative losers in the competition may still creep along, but their rewards will be only a pittance. The result is a very small number of very big winners, and a very large number of losers and small winners.
Impacts on the economy
We have begun to realize what impacts these developments have on the economy only fairly recently. The impacts have been accelerating, but some of them have been masked by other phenomena. For example, one might expect that if jobs are being destroyed, then consumer spending would decrease. But in the 1990s and until the late-2000s although income inequality was increasing, consumer spending also increased. This increase, however, was powered mostly by increased consumer debt, especially housing debt. After the deep recession of 2008-2009, the subsequent recovery in consumer spending was comprised largely of spending at the top of the income distribution.
We have not only been seeing the effect on the U.S. of offshoring of the workload to lower-income countries like China, Mexico, Bangladesh, Indonesia and many other developing-world nations, but of the increased automation and robotization of production and even of services. In the future, the robotization will become the larger factor. Even in China, where the cost of labor is low compared to the U.S. but increasing, new factories are being designed to make extensive use of robots.
The result is increasing corporate profits, increasing returns to capital – e.g. the soaring stock market – and decreasing returns to human labor. The prediction of Thomas Piketty, author of the surprise best-seller Capital in the Twenty-First Century, that the return to capital will increasingly outstrip the return to labor is proving true.
What does this mean for the future?
The resulting rise in corporate profits from deep cuts in workforce cost is undoubtedly what is powering the stock market. But as a recent much-cited article by Nicholas Eberstadt in Commentary magazine points out, things are much worse in the U.S. than most people in America’s “elite” classes – those who are comfortably employed, or otherwise financially comfortable – are aware. As Eberstadt says, “the Great American Escalator, which had lifted successive generations of Americans to ever higher standards of living and social well-being, broke down” around the year 2000, “and broke down very badly.”
The reason is a combination of the replacement of American labor by automation as well as by cheaper offshore labor. But in the future it will be primarily the replacement of labor by robots – not only American labor, but all labor.
If that is the case, then the Great American Escalator that has historically lifted up the working class may not be the only thing broken. If the laboring class that makes up the vast majority of the population cannot earn, then it cannot spend. Eventually this has to negatively affect corporate profits.
Can the top one percent – those who will continue to earn vast amounts, because either they were lucky winners in the winner-take-all competition, or they were those who provided the capital, to which returns are large – alone power the economy? Can they spend enough to keep the economy moving at a brisk pace?
This seems unlikely. As Ford says, “A single very wealthy person may buy a very nice car, or perhaps even a dozen such cars. But he or she is not going to buy thousands of automobiles.”
Wealthy people spend a much higher percentage of their incomes on investments, as opposed to consumption, than do middle-class people. But if those investments are in ventures that ultimately will not be able to earn a high return because the economy lacks consumers, then those investments will perform poorly. The whole economy could slow down dramatically. It may yield to the “secular stagnation” that has been talked about and forecasted a lot recently.
What is the solution?
If not in exchange for their labor, spending power will have to be placed in the hands of the vast majority of consumers by some other means. Ford suggests that a basic income guarantee (BIG) for all citizens may be necessary. Ford proposes as justification for that basic guarantee that taxpayer money has funded much of the basic research that allowed the development of information technology. Therefore, they might all, in a way, be regarded as shareholders in the benefits that have accrued and therefore entitled to a distribution.
It may be difficult to imagine how such a BIG could be accomplished politically, but Ford explains that the idea was advocated by libertarian-darling and Margaret Thatcher’s intellectual guru, philosopher-economist Friedrich Hayek. Ford quotes Hayek as saying, “it is unfortunate that the endeavor to secure a uniform minimum for all who cannot provide for themselves has become connected with the wholly different aims of securing a ‘just’ distribution of incomes.”
The design of such a BIG is problematic – how should it be distributed and paid for, should there be a means test and how should it be phased in as earned incomes rise, and so on. Some possible details are interestingly discussed by the economist-interviewer Russ Roberts with his guest Michael Munger, in Roberts’s Econtalk interview.
What could be the impact on financial advisors?
Financial advisors may think that giving financial advice is the ultimate service requiring the personal touch, but they have recently seen a proliferation of robo-advisors spring up to challenge that assumption. If automation, artificial intelligence and natural-language speaking robots develop as Ford predicts they will – and are already developing – the personal touch may not be as unrelinquishable as many advisors think. If so, the result will be a rapid falling-off of demand for advisors, and a winner-take-all tendency for the best – or most widely accepted – robo-advisor to get all the clients.
Advisors need to think about this and about how they will respond to this challenge – if indeed they even can.
Michael Edesess is a mathematician and economist, a senior research fellow with the Centre for Systems Informatics Engineering at City University of Hong Kong, chief investment strategist of Compendium Finance and a research associate at EDHEC-Risk Institute. In 2007, he authored a book about the investment services industry titled The Big Investment Lie, published by Berrett-Koehler. His new book, The Three Simple Rules of Investing, co-authored with Kwok L. Tsui, Carol Fabbri and George Peacock, was published by Berrett-Koehler in June 2014.
Read more articles by Michael Edesess