Economic Growth

Is the Great Stagnation Over?

Yes. Get ready for the Roaring Twenties.


In his The Great Stagnation, George Mason University economist Tyler Cowen asserted back in 2011 that the United States' economy had been essentially in a long productivity stall since about the 1970s. Yesterday, at an online panel session hosted by the American Enterprise Institute (AEI), the general consensus of the participants was that the Great Stagnation is coming to its end. The panelists were Tyler Cowen, MIT economist Catherine Tucker, University of Houston economist Dietrich Vollrath, and AEI economist Michael Strain. The session was moderated by AEI economics analyst James Pethokoukis.

Cowen argued 10 years ago that the previous set of general purpose technologies—machines and factories powered by fossil fuels and electricity—had run their courses, at least in the United States and other developed economies. When eventually nearly everyone had a car, electric appliances, and indoor plumbing, technological improvements were being made just at the margins. The result was a significant slowdown in the rates of productivity growth and incomes.

The online crew assembled by AEI expressed some optimism that a whole bunch of new technologies were on the verge of jumpstarting our sluggish economy. Strain declared himself very confident that the Great Stagnation is not permanent. He suggested that entrepreneurs were even now exploring how to adapt a whole suite of new technologies—batteries, vaccines, artificial intelligence, driverless trucks—to their best economic uses and whose benefits will become increasingly evident over the coming decade.

Tucker chimed in that it takes a while for entrepreneurs and innovators to figure out how to profitably apply ideas and new technologies. She noted that her fellow economists have been offering two excuses for why advances in digital technologies were not showing up in productivity figures. The first is that the enhancements were not being measured properly. The second is that the elaboration of general purpose technologies, e.g., steam and electricity in the 20th century, needs 20 to 30 years of experimentation before businesses can figure out how to really rev them up to boost productivity. She pointed out that people initially thought electricity was about the light bulb, but what really promoted economic growth was powering machinery in factories and electric appliances at home.

Cowen cautioned that many technological advances would doubtlessly improve human welfare but still might not show up in U.S. gross domestic product (GDP) and productivity statistics. For example, the new plug-and-play vaccine platforms may well result in highly effective vaccines for malaria and HIV, and that would be a huge boon for millions of people living in poor countries, but those benefits would be unlikely to show up U.S. GDP per capita statistics. He also pointed out that the recent significant advances in green energy production are occurring chiefly as a way to avoid the possible catastrophe of man-made climate change. Because climate change is a hidden counterfactual, replacing fossil fuels with solar, wind, and batteries would not necessarily lead people to feel as though their standard of living had risen.

Strain countered that the toll that infectious disease takes on the stunting of talents and skills in poor countries would be greatly ameliorated by rolling out cheap effective vaccines now made possible by messenger RNA technology. Over a longer time horizon, the U.S. and the rest of the world would significantly benefit from efflorescence of invention and entrepreneurship arising in regions whose development is held back by prevalent plagues.

Cowen noted that many of those poor disease-afflicted regions are not currently producing much in the way of new ideas or total factor productivity, so it might take 40 to 50 years for the beneficial effects of new vaccines to become evident. Until then, global economic growth could still just be the same old story of 2.2 percent annual increase.

Vollrath suggested that technological optimism based on technological advances was warranted, but noted that a lot of current innovations aim at replacing things, not adding a new product. Consequently, technological optimism should not necessarily be keyed off measured GDP and productivity.

Pethokoukis wondered if there might arise less tolerance for technological advances if people did not experience a tangible rise in living standards from them. Vollrath responded that during the current pandemic modern vaccines are incredibly tangible, as are digital technologies, like Zoom, that made work life possible for many Americans while maintaining social distancing. On the other hand, he wondered if people will accept that the productivity statistics aren't changing while nevertheless recognizing that, due to technological innovation, we are living fundamentally different lives. Strain suggested that the fact that we defeated the plague in just one year might induce in people more warmth toward abstract notions of economic and social progress through the process of creative destruction.

In response to the Pethokoukis question, Cowen mused that biomedical innovations might happen so fast that the average age of the U.S. population could rise faster than is currently projected. An older America might become a more status quo oriented society devolving into a tighter complacency and thus become less interested in technological innovation.

Strain speculated that technological progress might exacerbate economic inequality. As an example, he hypothesized that researchers will develop highly effective cancer cures that are also very expensive so that only the wealthiest could take advantage of them. On the other hand, Tucker countered that there is a body of research that finds many new digital technologies reduce inequality and cited research that found switching to digital medical records actually helped the less well off.

Pethokoukis observed that there has been a lot of media focus on how technologies supposedly increase social and economic inequality, enhance the surveillance state, and activate neoluddite concerns about unemployment arising from further automation. Tucker noted that prior to the pandemic she and her colleagues were concerned about a "techlash" accompanied by rising support for policies expressly designed to slow the pace of innovation, including bans, taxes, and stringent regulations on certain technologies.

Vollrath offered that such a techlash could stem from the fact that the marginal utility of many new innovations might not be as high as it was for our grandparents, who first obtained former luxuries like indoor plumbing, refrigeration, and personal automobiles. Getting an updated iPhone does not have the same weight as electrifying your house. The current changes are good, but comparatively not all that great when the earlier gains were so much more tangible.

Cowen, ever the contrarian, pointed out that newer technologies could also make things worse. He offered the example of how terrorists might use drones and crypto technologies to wreak havoc. He added that it's easy to imagine where innovation is bad.

Pethokoukis reverted to the question of whether we should worry about productivity growth. Vollrath responded that perhaps thinking about GDP and productivity is not a good way to measure growth. For example, inventing ways to use less stuff would help limit our impact on the natural environment, but it's harder to measure doing less harm and thus to characterize it as an improvement.

What about the impact of technological innovation on employment? Tucker noted that a lot of research on the effects of artificial intelligence (A.I.), specifically machine learning, finds that A.I. is a human-augmenting, rather than human-replacing, technology. Strain noted that automation has generally freed up people to do different and generally more productive work. Vollrath agreed that two centuries of automation have not reduced overall employment.

Cowen said that he didn't think that technological innovation was leading to job losses but worried instead that the internet was making people weirder in a QAnon sort of way. He observed that we have no good models of the psychological and governance impacts of pervasive digital technologies and that the old political equilibria we had is not coming back.

Looking to the future, Cowen suggested that instead of worrying so much about GDP per capita, let's talk about the capitas. He noted that many wealthy countries are on the verge of depopulating. Consequently, he argued that very important innovations would be those that lessen the burdens of rearing children. Vollrath observed that in general the more people the more ideas that are generated and that a shrinking population might slow down the pace of innovation.

Strain, however, expressed doubt about Cowen's suggestion that the burdens of rearing children is responsible for falling fertility rates in rich developed countries. He pointed out that many rich countries have deployed vast infrastructures to support child rearing. In fact, no pronatalist policies aimed at easing the burdens of parenting have yet yielded a sustained boost of a rich country's fertility rate above the population replacement threshold of 2.1 children per woman. Why? Because liberty and prosperity offer people many enticing life choices and pleasures other than parenthood.

Another Pethokoukis question to the panel was: If we look back in 10 years and find that the Great Stagnation has continued, why might that be the case? Would it mean new technologies are actually not all that good, or would it perhaps be the result of bad policies?

Cowen quipped that we had slunk back into our sloth. Strain said that his chief worry is that policies could be adopted that would prevent the emergence of new technologies while holding back new businesses and innovation. In addition, he suggested that an aging population could result in less demand that would spur businesses to do creative and innovative things. Vollrath expressed concern that the sort of cultural risk averseness embodied in the precautionary principle, rather than a failure of new technologies, could be responsible for a continuation the Great Stagnation. As an example, he cited the European Union's risk-averse and bureaucratically botched rollout of COVID-19 vaccines.

A viewer asked the panelists if new science was so complex that we need to expand the supply to scientists and innovators. Strain emphatically responded, yes, and lots more immigrants, too. He added we should hand out green cards to foreign university students who graduate with diplomas in STEM fields. When asked whether China or America would be more likely to lead the breakout from the Great Stagnation, Cowen basically summarized the consensus: I bet on America.