Practical analysis for investment professionals
16 June 2014

The Robot Revolution: Innovation Begets Innovation

Posted In: Economics, Future States

AGCO Corporation (AGCO), a farm machinery manufacturer in Hesston, Kansas, recently installed its 28th robot. Ford Motor Company (F) just installed 600 robots in just one facility, its Flat Rock Assembly Plant plant in Michigan. A Philips (PHG) plant in the Netherlands “employs” 128 robots to assemble electric shavers. Even Panera Bread (PNRA) announced that it will begin replacing its cashiers with automated kiosks by 2016. In the United States alone, companies bought about 22,500 robots in 2013. Worldwide, that figure is estimated at 162,000, and the installed base of robots is estimated at some 10 times these amounts. In many cases, robots perform the job better, faster, and cheaper than their human counterparts — and can work 24 hours a day, 365 days a year.

Clearly, technology is permeating our lives. But will it put us all out of work?

Recently, Blackrock CEO Larry Fink went on record suggesting that the trend in technology will lead to global unrest as the developing world must now grapple with jobs being lost to technology. Likewise, MIT professor Erik Brynjolfsson has suggested that the sluggish employment growth of the past 10–15 years has been a direct result of technology. Even the media is on board. The Financial Times recently highlighted how robots are escaping the factory and will change the world. Are they right?

To understand these issues, we must first go back in time to an unlikely source: the French economist Frédéric Bastiat. In 1849, Bastiat wrote a paper titled “That Which Is Seen, and That Which Is Not Seen,” in which he illustrated that an economic choice creates not only one effect (which is readily seen) but also a series of effects (which are unseen).

When a business manager invests in capital equipment that can do the job of five workers, the employees who had been doing that work are displaced. But we must also consider the unseen effects and weigh them side by side with the seen effects. Maybe the business managers then decide that their new capital investments provide their firm a durable advantage over competitors. Consequently, they keep their prices the same and enjoy greater profits as costs per unit decline. In such a scenario, the business owners, managers, and remaining employees achieve more profit and/or greater pay.

This incremental savings, whether it accrues to the business owners or employees, must go somewhere. Either these people save it or spend it. If they save it in a bank, the bank then lends it out to be invested in other productive projects. If they invest it in the stock or bond market, they must buy stocks or bonds from somebody who in turn is faced with the same choice — do I save it or spend it?

Alternatively, if the business is competitive, the auto manufacturer will lower prices to better compete in the marketplace. Who wins? The customer. Not only does the customer pay less for the product, but also that same customer can deploy that incremental savings into other goods and services.

Likewise, the introduction of robotics at the automotive plants today leads to immediate savings in the cost of production per vehicle. It’s not clear where the savings will be allocated today, but it must go somewhere (banks, purchases, etc.). There is, of course, no limit to the creative ways the incremental savings from productivity improvement might be spent. This explains Marc Andreeson’s recent Twitter rant on why robots won’t lead to mass unemployment.

In a business setting, labor competes with capital. The higher the cost of labor, the more attractive are capital investments in productivity (i.e., robots). Labor has little in the way of upfront costs but relatively large ongoing, hourly costs. In other words, when GM hires an assembly-line worker, there is typically no large upfront cost to “acquire” an employee, as they only need to pay him or her on an hourly basis as that employee performs work (excluding any union contract issues, of course). In contrast, technology has high upfront acquisition costs (e.g., the new robots installed by GM cost about $800,000 each), but relatively low ongoing, variable costs (energy, maintenance, and depreciation). So, in 2012 when Ford Motor Company installed 700 new robots at a Ford Motor plant in Louisville, Kentucky, it was plain to see that human assembly-line workers would no longer be needed to apply paint and sealer to vehicles. But this is far from the end of the story.

Obviously, we’d like to know how many employees might be let go, but what really matters is what happens to the overall economy. If Ford Motors sheds 100 jobs as a result of installing robots, but the robot manufacturers add a combined 150 jobs, overall employment grows. The most important question is: on balance, are jobs created or lost?

But we don’t need to rely on economic theory. Over the stretch of history, there are numerous examples of large-scale improvements in productivity reducing the need for labor. Perhaps no other innovation was as massive and transformative as the introduction of assembly-line manufacturing in the period from 1910 to 1930. The advances in assembly-line manufacturing dramatically reduced costs on a wide range of goods, especially cars, changing virtually every facet of modern life, from work to transportation to living arrangements (suburbs) to leisure.

To help gauge the impact of assembly-line manufacturing, we can look at the automobile industry during the period in which the assembly line was introduced. Registered cars owned in the United States increased from a few hundred thousand in 1910 to 8.1 million in 1920 to 23 million by 1930. This corresponds to roughly 1.8% of households to 34% of households to 77% of households, respectively.

Nevertheless, unemployment was largely unaffected. Between 1910 and 1920, as assembly-line manufacturing was first ramping up, unemployment averaged about 5.1%. Between 1920 and 1930, as assembly-line manufacturing exploded across the US economy, unemployment averaged about 5.5%. So, despite a world war, a major recession, and a dramatic wave of labor-reducing technology, unemployment decade by decade was essentially the same.

Today, the automobile industry that is at risk of losing jobs to robots is the very same industry that was at risk of losing jobs to assembly line-manufacturing in yesteryear. Moreover, automobile units grew much faster in the early 1900s than robots today. Between 1910 and 1930, auto sales increased at a rate of 24% per year compounded (with the 1920–30 period compounding at 11% per year).

In contrast, the robotics field has been growing much more slowly. Between 1993 and 2013, robot unit sales globally have increased by 5.9% per year. Over the 2003–13 time frame, robot sales increased only marginally faster, at 7.3% per year.

Maybe robots are much more productive than the assembly-line concept was? Nope. The assembly line reduced labor requirements by 80%, whereas today’s robots have a similar productivity improvement (though in all fairness, it is hard to measure precisely, as robots are made for many different applications).

Yet, in the 1910s and 1920s, despite the fact that automobiles and assembly line manufacturing had indisputably much greater growth and similar productivity gains, unemployment remained steady. Why? What could possibly explain this phenomenon? In short, the productivity improvement freed capital to pursue other endeavors. This is the history of markets. Innovation begets innovation. In fact, the price of Ford’s standard four-seat vehicle fell from $850 in 1909 to $260 by 1920. So, clearly the cost savings were passed on to the consumer as automobile buyers in the 1920s could now spend the $620 in savings on other needs. Is it a mere coincidence that the Hollywood entertainment industry and the hotel/motel industry exploded in popularity in the 1920s? Was it because nobody had created these things before? Or was it because these things became more affordable to large numbers of people as the middle class exploded in size? Productivity improvements meant that people had both more time and more money to enjoy leisure activities. These effects are unseen but no less powerful.

As we stand and look to our future today, labor-saving productivity in the form of robots will increasingly shape our world. The question is not simply how many jobs will be lost, the question is how creative we can be in deploying our incremental savings — which can create new jobs, new industries, and new opportunities.

History has demonstrated time and again that productive innovation leads to an expansion of overall economic activity. Moreover, despite the hundreds of thousands of robots in use today, I have yet to see even one robot start its own company.


Please note that the content of this site should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute.

Photo credit: ©iStockphoto.com/RicAguiar

About the Author(s)
Ron Rimkus, CFA

Ron Rimkus, CFA, was Director of Economics & Alternative Assets at CFA Institute, where he wrote about economics, monetary policy, currencies, global macro, behavioral finance, fixed income and alternative investments, such as gold and bitcoin (among other things). Previously, he served as SVP and Director of Large-cap Equity Products for BB&T Asset Management, where he led a team of research analysts, 300 regional portfolio managers, client service specialists, and marketing staff. He also served as a Senior Vice President and Lead Portfolio Manager of large-cap equity products at Mesirow Financial. Rimkus earned a BA degree in economics from Brown University and his MBA from the Anderson School of Management at UCLA. Topical Expertise: Alternative Investments · Economics

8 thoughts on “The Robot Revolution: Innovation Begets Innovation”

  1. David Botbol says:

    Yesteryear’s productivity gains indeed were passed on to the middle class in the up until the 70s. There remains to be seen that the latest batch of robot-induced incremental increase in profitability will also filter thru down the middle class. I say : the jury is still out …

    1. Hi David! Thanks for your comments. It’s never immediately clear where the benefits of productivity will accrue (owners, employees, customers, etc.). Who benefits from first order effects depends upon whether or not the business is competitive. However, the second and third order effects send that capital out into the economy and it goes to work somewhere. As long as the money goes into savings or spending (not mattresses), the benefits will accrue to the economy at large, helping all classes (middle class included). The point I am making is that any adversity in the labor market is very likely due to other issues, not robots/automation. Thanks again!

  2. LAURENCE BRODY says:

    while the price of autos went down in the 1910’s and 20’s there was a recorded DEPRESSION in 1921-23

    “To be sure, the 1920–1921 depression was painful. The unemployment rate peaked at 11.7 percent in 1921. But it had dropped to 6.7 percent by the following year, and was down to 2.4 percent by 1923. After the depression the United States proceeded to enjoy the “Roaring Twenties,” arguably the most prosperous decade in the country’s history. Some of this prosperity was illusory—itself the result of subsequent Fed inflation—but nonetheless the 1920–1921 depression “purged the rottenness out of the system” and provided a solid framework for sustainable growth.
    As we know, things turned out decidedly differently in the 1930s. Despite the easy fiscal and monetary policies of the Hoover administration and the Federal Reserve—which today’s experts say are necessary to avoid the “mistakes of the Great Depression”—the unemployment rate kept going higher and higher, averaging an astounding 25 percent in 1933. And of course, after the “great contraction” the U.S. proceeded to stagnate in the Great Depression of the 1930s, which was easily the least prosperous decade in the country’s history.
    The conclusion seems obvious to anyone whose mind is not firmly locked into the Keynesian or monetarist framework: The free market works. Even in the face of massive shocks requiring large structural adjustments, the best thing the government can do is cut its own budget and return more resources to the private sector. For its part, the Federal Reserve doesn’t help matters by flooding the shell-shocked credit markets with green pieces of paper. Prices can adjust to clear labor and other markets soon enough, in light of the new fundamentals, if only the politicians and central bankers would get out of the way.”
    http://www.fee.org/the_freeman/detail/the-depression-youve-never-heard-of-1920-1921/

    1. Hi Laurence,

      Too few people make a clear and cogent defense of free markets when there is indeed one to be made. As you illustrate in your response, policy proscriptions all too often create unemployment, not markets. And certainly not innovation. Even the most selfish and stingiest among us helps grow the economy simply by placing his or her money in the bank (or buying stocks or bonds) as that capital is lent out to productive investment. Only policy can slow down the natural flow of money. Thank you for your comments!

  3. Gianluca Tabbita says:

    Once strong AI comes of age I don’t see any opportunities left for humans in the job market. It doesn’t matter what industries or jobs we can create, we’ll be able to automate them from the very beginning. Maybe some jobs will be spared, those at the upper level of the pyramid pool, the very best of our AI experts, robotics engineers, and so on, but these jobs are in their nature few and sparse, the largest slice of of the jobs pool is the easiest to automate, i.e. drivers, clerks, teachers, and so on. At the end I don’t see any other chance for the economic cycle to survive but to implement an universal basic income to get the demand going.

  4. Ashok says:

    Someone gotta press the button.. Still someone’s gonna need to make robots.

  5. Gianluca,

    (sorry if this is a repeat, my last reply appears to be lost in cyberspace). Great comments. Thank you. Of course, it is possible you are right. It is possible that this time really is different. It is possible. But not likely. Perhaps the sci-fi movies are right that AI/robots will do just about everything for us. But isn’t it human needs that are getting met? When the paradigm shifts to robots selling to robots, maybe humans become an afterthought. But until then, bots can produce more and more of our work and dramatically reduce costs on a wide range of goods and services. And, as has happened so many times in the past, new “needs” are created and met. Have you considered a world where almost everyone is an entrepreneur because there is so much opportunity? Have you considered a world where someone makes a nice living being a maven whose thoughts and opinions are highly valuable to other humans for one reason or another? Have you considered a world where leisure time and human interaction dominate our lives more than work? What might happen if we all only needed to work one day per week? Maybe everyone will have a personal bot that acts as an agent on his or her behalf. The future is yet unwritten. But one thing we know with certainty, as productivity increases, it frees up dollars to do other things. What things? That is the question.

    1. Neil Salmond says:

      The ultimate scarcity is pleasant proximity to other people, as realised by good walkable urbanism. (The other pleasant environment is full nature, which is just a question of not messing with it.)

      Surplus capital from the robot revolution therefore flows to that scarcity (via banks, who have a license to print money by lending for mortgages) which is why we see land among walkable urbanism being bid up to the sky.

      The most egregious policy barrier is parking minima, with Euclidean single-use zoning and AASHTO anti-human public space design rules a very close second and third.

      The surplus capital from the robot utopia should see humans as greco-roman philosophers, dining, wining and dancing in beautiful towns of villas. Instead it’ll probably see us gorging on chips lost in an oculus rift world and robotaxi-ing among a sprawling wasteland of cheap tilt-ups.

      Demand zoning reform now!

Leave a Reply

Your email address will not be published. Required fields are marked *



By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.

Close