Many have noted that the slowdown in the growth of labor productivity is creating “missing GDP.” One explanation is that this value isn’t missing, but that we do not measure it correctly.
This time on Financial Sense, Chad Syverson, Professor of Economics at the University of Chicago Booth School of Business and a Research Associate at the National Bureau of Economic Research, explains why the mismeasurement hypothesis doesn’t hold water.
Trillion of Dollars 'Missing'
For about the last ten years, we’ve been experiencing a slowdown in the rate of growth of labor productivity, Syverson noted. Labor productivity is the output per worker, or better yet, the output per worker hour.
This is vitally important because labor productivity essentially sets the speed limit on economic growth in the long run, Syverson added. Over long periods, GDP per capita can’t go up any faster than labor productivity.
“If you have an extended period where labor productivity growth is slow, that means you’re going to have an extended period where economic growth is slow,” he said. “And we seem to be in that now.”
For context, Syverson noted that if we look at the last 70 years of productivity growth, we’re experiencing a decade-long slowdown again with rates slower than even before.
“We’ve seen 1.2 percent labor productivity growth per year on average over the last decade,” Syverson said. “Folks might think ‘it’s just a percent here and there,’ but you add these things up over several years, and it turns into real money very quickly.”
Had we not experienced the slowdown starting ten years ago, GDP would be about $3 trillion a year higher than it is, Syverson noted. In other words, there’s about $9,000 of income per person that’s missing because of this labor productivity growth slowdown.
“We’ve had a slowdown of about 1.5 percent per worker per year,” Syverson said. “Multiply that over ten years, and you get in the neighborhood 17 or 18 percent of GDP that’s ‘missing.’”
Scroll down to read more of his comments, or click to hear a preview of his interview below. Subscribers can access the full audio by clicking here or via podcast on their mobile device.
The Productivity Paradox and Mismeasurement Hypothesis
The issue is, we’ve seen massive technological progress in IT, which should lead to greater efficiencies and more output, but it hasn’t translated into higher rates of productivity growth. In order to resolve this paradox, one of the possible explanations put forward is that this perceived slowness is just an illusion caused by mismeasurement of our modern economy.
According to proponents of this argument, productivity growth hasn’t slowed down. Rather, our ability to measure economic growth isn’t adequate because of the nature of the new products and new technologies that have been developed.
“For example, if you think about Facebook, a lot of people spend a fair amount of time on Facebook and seem to enjoy it,” Syverson said. “Yet, they … don’t pay for it.”
Based on how we measure GDP, it’s the sum of all spending in the economy. So, if something is being used but not paid for, it’s not showing up in the GDP numbers, Syverson said. This makes it appear to some people that GDP hasn’t gone up much, but in reality, we’re able to enjoy great new products.
Mismeasurement Theory Doesn’t Stand Up
This attempt to explain the productivity paradox doesn’t work, according to Syverson. He points to four arguments he’s made in a recent paper to refute the idea.
First, if the mismeasurement hypothesis was correct, there should be a relationship between how much IT is being used and the observed slowdown.
To check this idea, Syverson collected data on historical labor productivity growth for 30 countries from the OECD and gathered several decades of productivity growth numbers for each.
He found that there is a slowdown in these countries, and in 29 of those 30 countries labor productivity growth slowed down at about the same time as it did in the US. However, he found that the size of the slowdown was not in any way related to the importance of IT products in that country’s economy.
“There’s no relationship between the size of the productivity slowdown … and the importance of IT in that country’s economy,” he said.
Economists had recognized the issue with IT products being utilized at effectively no cost. To find out if this would account for the paradox, these economists attempted to determine if certain “gatekeeper” goods, such as broadband access, would account for the missing value from near-free IT products.
The idea is to use some economic theory to infer the implied value customers are getting from all these online and digital products, Syverson said.
“When you look at these studies … you find most have valuations of $100 to $200 billion dollars total in the US,” Syverson said. “That’s nothing to sneeze at, but remember what we’re trying to explain here is in the neighborhood or $3 trillion of missing GDP. So you’re only explaining about one-twentieth of that.”
Is One Company Soaking Up All the Value?
Some have pointed out that it’s possible all the excess hidden valuation is being accumulated with one economic entity, Syverson said.
“Let’s suppose the mismeasurement hypothesis was right, and the IT-producing sectors of the economy have created this massive amount of output, $3 trillion, but we’re just not measuring it right,” Syverson said.
When we add up all these numbers, the entire information economy grew about a half a trillion dollars between when the productivity slowdown started and today, Syverson noted.
“The notion that we’re missing 85 percent of activity in the sector is a bit of a stretch,” he said. “It implies, by the way, productivity growth in that sector that is enormous and on the scale of nothing we have ever observed in any other sector before, and it just doesn’t seem to be plausible.”
The implied revenue growth would be six, seven or eight times what we see in the data, Syverson noted.
“It just implies an amount of missing growth that seems hard to believe,” he added.
If there is all this hidden value being created, Syverson noted, then we ought to see a lot of income for people who are getting paid to create these apps, to design and run these companies, but there aren’t many expenditures on the other side of the ledger to balance this out.
“If you take that idea and add it up to the nationwide level, we have two measures like that,” Syverson said. “We have gross domestic product—that’s just adding up all the expenditures people make—and on the other side of the ledger is gross domestic income—that’s adding up all the sources of income. Now, those two ought to be equal … [and if the mismeasurement hypothesis is correct] we ought to see gross domestic income be systematically higher than the gross domestic.”
If we look at the data, Syverson said, that has been the case on average since the productivity slowdown started, and is consistent with the mismeasurement hypothesis. However, the problem is, if we go back further in the data before the slowdown started, already income was higher than expenditures and had been consistently for seven years before the slowdown started, which isn’t what we would expect given the theory, Syverson said.
“Income has been higher than expenditures, but not because we’re paying people to make things that we’re selling for low prices or are giving away,” he said. “It’s because corporate profits are exceptionally high, which is not what you’d expect … if we’re in a world where we’re giving these things away at bizarrely low prices.”
The True Cause of the Slowdown
There are two explanations, according to Syverson, and they’re not mutually exclusive.
The first is that we had the first wave of productivity gains coming from implementing information and communication technologies, which accounts for the 1995 and 2004 speedup.
“The slowdown is just the natural consequence of the first wave of gains,” Syverson said.
This explanation closely agrees with Carlota Perez's work looking at the historical rhythm of technological revolutions, bubbles, and their associated slumps.
The second explanation only has some early evidence to support it, but the idea is that frontier companies haven’t experienced a slowdown in their productivity growth, and other average companies just aren’t implementing best technologies for some unknown reason, and that could also account for some of the productivity slowdowns, Syverson noted.
This is all extremely important, because if this productivity slowdown is sustained for another 25 years, incomes will drop by almost half.
That doesn’t mean there’s no room for optimism, however. There may yet be another wave of IT-driven productivity gains to come.
“We don’t see it in the data yet … but again, there’s hope that we know through historical experience that a second wave could come,” Syverson said. “It’d be great if things could turn around soon, and though there’s no guarantee of that, hopefully, it will.”