New Technologies, Automation and the Future


In recent postings, I have joined many others in speculating on what effects the information revolution will have. Already, its impact has been tremendous. Its effects include:

  • The elimination of a huge number of jobs lost in both service and production industries;
  • Social media has completely transformed the manner in which we interact/communicate;
  • Online shopping is growing rapidly, and
  • Digital communication has resulted in an acceleration of technological discoveries.

So what is coming? Since 2002, the McKinsey Global Institute (MGI) has been tracking changes resulting from the information revolution. This is needed because many politicians have little idea of what is happening. For example, Trump says he will bring back jobs by taxing imports. This will hardly solve the problem. A recent MGI study concluded:

“Trade and outsourcing explain only about 20 percent of the 5.8 million manufacturing jobs lost during the 2000-10 period; more than two-thirds of job losses can be attributed to continued productivity growth, which has been outpacing demand growth for the past decade.”

In its most recent study, MGI attempted to determine the labor force implications of technological developments. In what follows, I summarize their methodology and conclusions along with my own commentary.


MGI used the state of technology in respect to 18 performance capabilities to estimate the technical automation potential of more than 2,000 work activities from more than 800 occupations across the US economy, and then broadened our analysis across the global economy.


MGI started by developing seven categories of “automatable” human activities. Table 1 provides its estimates of automatable shares for each category. While 81% of “predictable physical” can be automated by adapting currently available technologies, MGI estimates that only 9% of “managing” is automatable. The third column is the category’s share of the US automatable total. MGI has concluded that the categories most susceptible for automation are “predictable physical” and the “collecting and processing of data.” Together, these three activities constitute 51% of total employment and $2.7 trillion in wages.

Table 1. – US: Time Spent on Automatable Activities

Source: McKinsey Global Institute, “A Future That Works: Automation, Employment, and Productivity”

MGI then looked at the automation potential for more than 2,000 work activities. They reduced this down to 19 industrial sectors for the US economy. They then broadened the analysis to the global economy. The results from this exercise are summarized in Table 2.

Table 2. – Automation Potential for 19 Sectors

Source: McKinsey Global Institute, “A Future That Works: Automation, Employment, and Productivity”

Of course, it is easy to second-guess MGI’s estimates. For example, I believe a far greater share of educational activities than 27% can be automated. Why? Because a good deal of education occurs via lectures, and definitely there are better and worse lecturers. In light of this, why not get the best lectures on all topics and provide them digitally to classrooms? Classroom teachers would then only have to discuss and answer questions on the lectures. Of course, such an action would be vigorously resisted by the teachers’ unions.

And to that point, MGI argues there are five key factors that will influence the pace and extent of automation:

  • Technical feasibility, since the technology has to be invented, integrated and adapted;
  • The cost of developing and deploying solutions, which affects the business case for adoption;
  • Labor market dynamics, including the supply, demand, and costs of human labor as an alternative to automation (see education discussion above);
  • The economic benefits, which could include higher throughput and increased quality, as well as labor cost savings.
  • Regulatory and social acceptance will affect the rate of adoption even when deployment makes business sense.

MGI believes that when all of these factors into account, it will take decades for automation’s effect on current work activities to play out fully.

Demographics, Productivity Growth and Labor Force Adjustments

Over the last four decades, productivity gains have been labor saving. In fact, these gains have wiped out a huge number of middle class jobs. However, if we look out over a longer time frame, the picture is somewhat different. Economic growth is constrained by the labor supply and its productivity. We know that for most developed countries, populations and as a consequence labor supplies are falling. That means that economic growth depends exclusively on productivity growth. For most developing countries, economic growth will benefit from both rising populations and growing productivity. MGI points out:

“…peak employment will occur in most countries within 50 years. The expected decline in the share of the working-age population will open an economic growth gap: roughly half of the sources of economic growth from the past half century (employment growth) will evaporate as populations age. Even at historical rates of productivity growth, economic growth could be nearly halved. Automation could compensate for at least some of these demographic trends. We estimate the productivity injection it could give to the global economy as being between 0.8 and 1.4 percent of global GDP annually, assuming that human labor replaced by automation would rejoin the workforce and be as productive as it was in 2014. Considering the labor substitution effect alone, we calculate that, by 2065, automation could potentially add productivity growth in the largest economies in the world (G19 plus Nigeria) that is the equivalent of an additional 1.1 billion to 2.3 billion full-time workers.”

Table 3 provides MGI’s estimates of the full time equivalent (FTE) GDP gap in 2065 of assuming today’s productivity and current population projections and today’s GDP/FTE. So, for example, the US GDP/FTE would be 4% less than it is now because of the labor force decline. MGI concludes for developed nations:

“Automation can provide the productivity boost required to meet economic growth projections that they otherwise would struggle to attain without other significant productivity growth accelerators. These economies thus have a major interest in pursuing rapid automation adoption.”

Table 3. – FTE Gap Between FTE Projections and Number of FTE’s Needed to Maintain GDP per Capita as Percent FTE’s in 2014

Source: McKinsey Global Institute, “A Future That Works: Automation, Employment, and Productivity”

While this is an interesting exercise, I remain far more concerned about labor supply surpluses resulting accelerating labor-saving automation than slower economic growth resulting from labor supply shortages.

Investing – Who Produces Automation Equipment?

Automation is clearly accelerating so it is reasonable to ask who the producers are. A list is provided in Table 4. Many of the companies have been around for a long time and getting into automation is an evolution from their main focus. Examples include Amazon, Boeing, Honda, Hyundai, Kawasaki, Lockheed, Samsung and Toyota. “Purer” automation plays include ABB Robotics, AeroVironment, Fanuc, iRobot, Intuitive Surgica, Swisslog and Titan. A relatively large number of these firms are Japanese.

Table 4. – Leading Automation “Producers”

Source: Robotics Business Review

I answer this with a quote from Peter Lynch. Peter Lynch’s track record was truly remarkable. As I have noted elsewhere, he beat the S&P 500 (^GSPC) in 12 of the 14 years he ran Fidelity’s Magellan Fund (FMAGX). And on average, if the S&P gained 5%, Magellan gained 22.5%! Even in the two years the S&P outperformed Lynch, the spread was very small. Lynch argues you should:

“Use your specialized knowledge to home in on stocks you can analyze, study them and then decide if they’re worth owning. The best way to invest is to look at companies competing in the field where you work.”

With rapid technological change, the automation business is inherently risky. And I don’t work in the industry.


It is hard to keep in mind just how rapidly technologies are changing. And that, coupled with how long it takes people to adapt, makes the future very uncertain. Most people are quite slow in adapting to the new technologies – just look at how long it has taken buying on line to become significant. There are reasons to be tremendously optimistic in the fields of health and energy. But dealing with jobs lost to automation remains a real problem.

The content above was saved on the old Morss Global Finance website, just in case anyone was looking for it (with the help of
Scroll to Top