Productivity numbers send mixed—maybe wrong—signals

While the recent numbers from the Bureau of Labor Statistics (BLS) show an increase in the annual rate of productivity of 3.1% during the third quarter, some economics experts are questioning whether that number will have any long-term impact on manufacturing output. I would question whether that number is even realistic, given the sluggish economy. An article in the November 4 issue of the Wall Street Journal notes that this jump “may be a blip” given that the “broader trend remains consistent with a decade-long decline.”

Rethink Robotics
Image courtesy Rethink Robotics/flickr.

Aligning with other articles I’ve read over the past week, the WSJ says that “many economists expect the rate of output gains to slow again in the final three months of the year.” Comparing it with the number from a year ago, the WSJ article says that the third quarter of 2016 was nothing to write home about, in that it was “flat.” In fact, prior to the July-through-September period, “productivity had fallen for three straight quarters,” writes reporter Eric Morath, “the longest streak of negative readings since the 1970s.”

Why pay so much attention to productivity numbers? Because, writes Morath, “productivity gains have been sluggish during much of the expansion that began in mid-2009” and these gains “are a critical ingredient in determining the rate of growth for worker pay, consumer prices and the economy as a whole.”

But are the numbers correct? Some economists argue that productivity may be mismeasured and the statistics aren’t capturing the effects of innovations such as machine technology and robotics/automation that have become so commonplace on today’s production floor.

I hardly go into a plastics processing or mold manufacturing plant these days without seeing automation everywhere on the production floor. A report just released by the Robotic Industries Association shows that robot orders and shipments in North America set new records in the first nine months of 2016. A total of 23,985 robots valued at $1.3 billion were ordered from North American companies in the first nine months of this year, an increase of 7% in units and 3% in dollars over the same period in 2015, which also set a record.

Given the increased usage of robotics and other automation, how will that impact the productivity numbers? In July of 2015, Georg Graetz and Guy Michaels published a paper, “Estimating the Impact of Robots on Productivity and Employment,” in Robohub. The pair set out to discover the impact of robots on the average manufacturing worker by analyzing their effect in 14 industries across 17 developed countries from 1993 to 2007.

“We found that industrial robots increase labor productivity, total factor productivity and wages,” said Graetz and Michaels in their study. “While they don’t significantly change total hours worked, they may be a threat to low- and middle-skilled workers.”

To estimate the impact of robots, Graetz and Michaels took advantage of variations across industries and countries over time. “We checked the robustness of our results by including a large number of controls and by considering alternative ways of measuring the robot input,” they said. “A consistent picture emerges in which robots appear to raise productivity, without causing total hours to decline.”

They then asked themselves, “could it be that higher productivity causes a larger increase in robot use, rather than the other way around?” and developed a novel instrument variable strategy. “Our instrument for increased robot use is a measure of workers’ replaceability by robots, based on the tasks prevalent in industries before robots were widely employed,” they explained. “Specifically, we matched data on tasks performed by industrial robots today with data on similar tasks performed by U.S. workers in 1960, before robots were used. We then computed the fraction of each industry’s working hours in 1980 accounted for by occupations that subsequently became prone to replacement. Our industry-level ‘replaceability’ index strongly predicts increased robot use from 1993-2007.

“When we use our instrument to capture differences in the increased use of robots, we again find that robots increased productivity, and detect no significant effect on hours worked.” The pair concluded that in further results, “we find that industrial robots increased total factor productivity and wages. At the same time, we find no significant effect of robots on the labor share.

“Preliminary post-2007 data show the number of robots has continued to swell, and the set of tasks they can perform has expanded. Both of these trends indicate that robots will continue to play an important role in improving productivity.”

The conclusion that I reach from all of this is that when we see the productivity number increase at the same time that we see fewer workers in the manufacturing sector than we did two decades ago, we realize that higher productivity is being achieved both by fewer workers and more robots and automation. So Morath is probably right when he notes that some economists are failing to capture the effects of technology/automation in their measurements.

Paul Ashworth, an economist at Capital Economics, commented on Morath’s article: “The longer this slowdown goes on, the less confident we are of another technology-related acceleration. As it stands now, the slowdown in productivity growth, together with the aging of the population, suggests that the U.S. economy’s potential growth rate is well below 2%.”

It’s difficult to believe that a growth rate like that is a sign of a booming economy. Perhaps it behooves us not to get too excited over numbers.

So, here’s a question for you: Has your use of robots/automation increased over the past 10 years? If so, how has it affected your productivity: Has it increased significantly with fewer workers? Or has it not affected your productivity or number of employees? Share your thoughts with the PlasticsToday audience by posting your comments below.

Comments (0)

Please log in or to post comments.
  • Oldest First
  • Newest First
Loading Comments...