In "Is U.S. Economic Growth Over?," a 2012 working paper for the National Bureau of Economic Research, the Northwestern University economist Robert Gordon argued that the country was in for 25 to 40 years of very slow growth. In particular, Americans in the bottom 99 percent of the U.S. income distribution could expect only 0.2 percent annual increases in their real per capita disposable incomes. This is dramatically lower than the 2 percent annual increase in incomes that occurred in the century before 2007. Gordon attributed this fall-off in growth to six "headwinds" and the slowing pace of technological innovation.
Gordon extends his analysis in a new study, "The Demise of U.S. Economic Growth: Restatment, Rebuttal, and Reflections." In this paper, also published by the National Bureau of Economic Research, Gordon revisits four of his six growth-slowing headwinds: demography, education, inequality, and government debt. (The other two are globalization and energy and environmental concerns.) He also attempts to bolster his argument that technological progress is stagnating.
With regard to demography, Gordon points out that the percentage of the population working has been falling, as have the number of hours per employee. A lower percentage of people engaged in productive work, he argues, cuts projected future growth rate by 0.3 percent, which he duly subtracts from the previous norm of 2 percent.
The second headwind is education. Economic growth over the past century, he writes, was boosted by 0.35 percent annually as the percentage of Americans graduating from high school and college rose. That increase in educational attainment has now stalled. As a consequence, Gordon calculates that future economic growth will be reduced by another 0.2 percent annually.
Gordon's third headwind is rising inequality. Citing data from the Berkeley economist Emmanuel Saez, Gordon notes that from 1993 and 2013 the household incomes of the bottom 99 percent grew at a rate of 0.34 per year while the average rate was 0.87 percent. Over that time, the household incomes of the top 1 percent grew at 3.3 percent annually. Since he is focusing on the economic prospects of the bottom 99 percent, Gordon deducts the 0.5 percent difference in income increases.
His fourth headwind is mounting federal, state, and local government debt. Paying back this money will require either substantial tax increases or steep cutbacks in transfer payments—or both. Gordon calculates that this debt overhang will reduce future growth by 0.2 percent annually. All four headwinds together, he concludes, will lower U.S. economic growth for the bottom 99 percent of households to 0.8 percent annually.
To illustrate the dire effects of this slowdown, Gordon calculates that per capita real income of $49,387 in 2007 would rise at a 2 percent rate to $200,273 by 2077. At 0.8 percent annually, average income would rise to only $86,460 by 2077. Gordon then points out that since 2007 that average per capita incomes at a 2 percent growth rate would now be $56,243, and at 0.8 percent at $51,800. Instead, third quarter 2013 per capita income was only $50,022, just $635 dollars higher than it was in 2007, implying an annual growth rate of 0.18 per year since 2007.
As if those figures were not discouraging enough, Gordon then considers the effects of technological innovation on economic growth. Using productivity data from 1891 to 2012, Gordon calculates that the increase in output per hour averaged 2.36 percent until 1972, then dropped to 1.38 percent from then to 1996. The creation of the Internet and the spread of information technologies boosted the annual productivity increase to 2.54 percent between 1996 and 2004, but the figure has since dropped to just 1.33 percent.
Since Gordon expects the pace of innovation to continue to lag, he subtracts -0.6 percent from future growth to arrive at a slack annual growth rate of just 0.2 percent for the disposable household incomes of the bottom 99 percent. If he's right, that would mean that real per capita incomes of most Americans would average only $56,800 in 2077 in constant dollars, a mere $7,413 more than in 2007.
Must Americans reconcile themselves to such drastically reduced economic prospects? Maybe not. Let's take a look at some data that challenge Gordon's gloomy forecast.
Is demography economic destiny? Not necessarily. In a 2012 Journal of Population Economics study, Harvard demographer Klaus Prettner looked at the effects of an aging population on economic growth. He found that the positive impact of longevity increases on per capita output more than balance out the negative influence of lower fertility rates. "Our main conclusion is that currently ongoing demographic changes do not necessarily hamper technological progress and therefore economic prosperity," Prettner writes. On the other hand, it must be acknowledged that the annual growth rates in developed countries such as Japan, Italy, and Germany with aging populations have hovered around 1 percent for the past couple of decades.
What about education? Two trends may get the U.S. educational system out of its stall: competition and technology. There is evidence that charter schools and other forms of school choice are improving educational results and even future incomes. And while it isn't clear yet what effect new information technologies will have, educators are exploring how they can be deployed to enhance learning both in schools and in non-traditional settings.
Does inequality slow growth? A lot of economists believe that it does, but there is not complete agreement on this issue. It's notable that Japan, Germany, and Italy all have greater income equality than the United States, yet their average economic growth rates since 1990 have been significantly lower.
Will the government debt overhang slow growth? Yes, and likely by more than Gordon estimates.
Finally, is America stuck in a permanent technology slump? Gordon has been arguing this case for more than 10 years. His central claim is that the sustained increases in productivity prior to 1972 were fueled by the advent and elaboration of three general purpose technologies—electricity, internal combustion engines, and wireless communication. We have now topped out since industry and households have been thoroughly electrified and there are now 800 automobiles for every 1,000 Americans and 2.5 televisions per household. Gordon argues that there is no similar suite of general purpose technologies in the offing to fuel future economic growth.
Gordon does acknowledge that the development of the Internet and pervasive digitalization boosted productivity, but he counters that "we have already experienced the digital revolution for the past 40 years during which the most fruitful applications of electronics have already occurred." He dismisses "foreseeable" developments in genomic medicine, robotics, artificial intelligence, 3D printing, Big Data, and driverless automobiles as slight improvements on old technologies.
This is doubtful. Consider Gordon's assertion that "future advances in medicine related to the genome have already proved to be disappointing." It is true that many researchers had hoped that decoding the genetic makeup of human beings would have generated new cures faster than it has, but medical progress is likely to speed up soon. For example, using results garnered from cheap whole genome sequencing, physicians will be able to employ medical search engines enabled by Big Data and artificial intelligence to precisely diagnose and design very specific treatments for individual patients. Physicians will likely to be able to repair broken genes, seed failing organs with refreshed stem cells, or even use 3D printing to create replacement organs. Surgical robots will guide doctors in installing the newly printed livers and kidneys. Driverless vehicles might even deliver the replacement organs.
Gordon somehow misses the fact that computation is the most general-purpose technology ever invented. Ever more augmented intelligence will enhance and add value to nearly any imaginable product or service. But while the future of technological innovation probably isn't as bleak as Gordon suggests, he has given us fair warning of the doleful direction in which the economic headwinds are blowing.