Domo arigato, Mr. Roboto. Mata au hi made. Domo arigato, Mr. Roboto, Himitsu wo shiri tai. You’re wondering who I am (secret secret I’ve got a secret). Machine or mannequin (secret secret I’ve got a secret). With parts made in Japan (secret secret I’ve got a secret). I am the modern man. Thank you very much Mr. Roboto. So beings the 1983 song, written by Dennis DeYoung, and recorded by the band Styx.
Robotics are one of the technologies that businesses use to increase productivity. Not unlike the spinning looms and tractors of the past, robotics can replace human labor by automating repetitive tasks. If used correctly, robots can be a way to replace labor with capital, something economists call capital deepening. Capital deepening is a traditional way that firms increase their productivity; and increasing productivity is extremely important in driving the economy forward. True economic growth comes about when firms do more with less stuff, or when they move stuff from a low value activity to a higher one. Technology, like robotics, can do this by replacing people with machines, freeing those people to do higher value things.
Henry Ford once said Improved productivity means less human sweat, not more. Ford was no economist, but he knew productivity when he saw it. Productivity is a measurement of the total output in an economy divided by the labor hours (or sweat) used to produce it. It is a measure of how much stuff the average person makes in an hour. The number of labor hours worked in the US economy has increased by 2.0 percent so far this year, while output increased by 5.0 percent. This means that productivity grew by 2.9 percent in the first half of 2018, the largest quarterly gain since the first quarter of 2015.
Even with the good start to 2018, over the past few decades, productivity in America has been falling, and falling quite dramatically. As one might expect, surveys of economists are all over the place when it comes to explaining this productivity gap. Many economists have discussed how much of the investment being made today is financial rather than mechanical; others have suggested that the economy has shifted from a productive manufacturing base to a less productive service economy; while still others have claimed that a mature economy with mature industries just does not have any room to increase productivity further. It is likely that all of these ideas do factor into the decline in productivity, but none really address the basic mathematics of the productivity equation.
The idea that investment is going into financial vehicles (buying back stock) rather than building machines, or performing R&D does not mean that investment is not occurring. Overall the nation’s capital stock (the country’s inventory of buildings, machines, fork-lifts, patents) has not been falling. The growth rate has waned, from about 47 percent during the 1950s to 27 percent in the 2000s, but this is a false comparison. When the growth rate in the labor force is accounted for, the trend goes away, and growth in capital per worker becomes more volatile decade over decade, or year over year. For example, the country experienced virtually no capital deepening during the 1970s. In sum, capital stock continues to rise year after year.
A shift to services does tend to initially impact output per worker, but not to an extent that it would stop productivity growth. A restaurant can replace workers with machines much like a factory can. Service businesses can easily produce more with less workers by adding capital (say computers over index cards) or by modifying processes.
Finally, the idea that a mature economy does not have room for productivity growth is simply nonsensical. The claim assumes that an economy is stagnant. This would mean that both investment and labor cannot move from one industry to another. But economies grow by moving resources from less productive enterprises to more productive ones – from horses to cars, and from cars to airplanes, and airplanes to transporters.
If the standard reasons don’t really apply, why might productivity growth be slowing? The chart presents an index of both employment (a proxy for hours worked) and real output. As it shows, output falls rapidly during recessions, but then picks up as the economy grows. Employment, on the other hand, is steady with much smaller swings over the economic cycle. But output growth has fallen below employment over the last two business cycles. Simply put overall sales have not grown all that rapidly. And I think I know the reason.
Since the 2001 recession, there has been a huge growth in digital communications, and in services that – well that simply lose money. Amazon started moving into products other than books in 2001. Social media came on-line in the early 2000s with MySpace being founded in 2003 and Facebook in 2004. Google set up shop in Menlo Park in 1998, and Napster, the precursor to Pandora and iTunes was first released in 1999. What all of these companies have in common is that they provide services not for money, but for information, and as younger generations enter the marketplace they consume more and more of these information-paid services. These payments do not show up in the output figures and steady employment and falling output can look like falling productivity even though the actual amount of stuff produced per worker is rising.
Additionally, there are many firms, like Uber and Amazon, that lose money on every sale. If a taxi ride costs $10 and Uber sells one for $5, measured output falls, even though the stuff produced (the ride) remains the same. If digitization and technology are in fact decreasing output, maybe more is really being done with less stuff, and the economy is more productive. Maybe this is Mr. Roboto’s secret.