Attention Conservation Notice: tl;dr. 9000 words trying to work my way through and in the process provide a reader’s guide to the techno-growth stagnation arguments of Robert Gordon, Tyler Cowen, and Brink Lindsey. The arguments are powerful. The authors are very serious economists. I wind up skeptical, and optimistic–partly because I am a techno-optimist by nature, partly because I am a politico-optimist and I think the literature confuses the past generation’s failures in distribution and demand-management due to political dysfunction with failures in accumulation and innovation, and partly because I have a different more micro-incremental conception of the process of economic growth than does Robert Gordon.
I. Once and Future Ages of Diminished Expectations
Back in 1990 Paul Krugman wrote a little book–a very nice little book–called The Age of Diminished Expectations. The central point was that the long era of more than a century during which Americans could expect 2%/year growth on average in their real per capita incomes and standards of living was over. This era stretched back to the immediate aftermath of the Civil War. This era saw each generation attain a level of material wealth and well-being twice that of its predecessors: 2%/year growth for 35 years is a doubling. And, Krugman wrote, the slow growth from 1973-1990–during which real GDP per worker had been a mere 1%/year–was a harbinger of a new, more pessimistic future: an age in which Americans’ formerly-great expectations of the future would have to be diminished.
Two years after the book was published, the American economy entered a phase in which growth in output per worker roared back to 2%/year and stayed that way for a decade and a half. I was only one of those who thought in the early 2000s that we might even shift to a phase in which our new normal was 2.5%-3.0%/year in average real GDP growth–and that would be 3.0%-4.%/year in total real GDP growth.
But now, after the financial crisis, in the Lesser Depression, after four years of substantially sub-par recovery, there is great talk that the great era of 2%/year growth in output per worker is over, that America’s future expectations are diminished, and that our long-run economic future is, well, not-so-great. Three relatively recent contributions stand out: At the level of the 30,000 foot bird’s-eye view, Brink Lindsey has written a paper on “Why Growth Is Getting Harder”. In the mid-range, Tyler Cowen has his little book on The Great Stagnation: How America Ate All the Low-Hanging Fruit of Modern History, Got Sick, and Will(Eventually) Feel Better. At the level of detail, Robert Gordon has his working paper on “Is U.S. Economic Growth Over? Faltering Innovation Confronts the Six Headwinds”.
How should we assess argument? Is this the normal outbreak of “secular stagnation talk” that always comes to the fore during and after every deep cyclical depression? Or are things, this time, really different? What is the argument that we are beginning an era of “great stagnation”, anyway?
II. Lindsey: How Much Harder Is Growth Getting?
Let us start with Brink Lindsey: Why Growth Is Getting Harder:
￼For over a century, the trend line for the long-term growth of the U.S. economy has held remarkably steady… growth in real gross domestic product (GDP) per capita… of approximately 2 percent [annually]…. Consider the four constituent elements… (1) growth in… hours worked per capita; (2) growth in labor quality… (3) growth in… physical capital invested per worker; and (4) growth in so-called total factor productivity…. Over the course of the 20th century, these various components fluctuated…. The fluctuations, however, tended to offset each other…. In the 21st century… all growth components have fallen off simultaneously…. It is difficult to resist the conclusion that the conditions for growth are less favorable than they used to be. In other words, growth is getting harder. Consequently, policies that are more friendly to long-term growth will be needed if more robust growth is to be revived…
Brink’s first point is that of our 2.0%/year of real GDP per capita growth over the past century and a third, 0.3%/year has been due to rising labor force participation. That 0.3%/year is gone, and is not coming back. That pushes us down to 1.7%/year of real GDP per capita growth–an increase in the doubling time from 35 years to 41 years.
Brink’s second point is that we have stopped increasing our rate of investment in education:
The big story concerning U.S. labor quality during the 20th century was a dramatic rise in the years spent in school by American workers…. According to Harvard economists Claudia Goldin and Lawrence Katz, the direct effect of increased educational attainment on labor quality accounted for about 15 percent of the total rise in real GDP per capita over the period 1915–2005…. This source of growth, however, has been waning…. Mean years of schooling completed by American workers rose from 9.01 in 1940 to 12.46 in 1980-for an average growth rate of 0.81 percent per year. Between 1980 and 2005, by contrast, mean years of schooling completed increased further to 13.54—for an average growth rate of only 0.33 percent per year…
That, too, was responsible for 0.3%/year of our 2.0%/year of real GDP per capita growth. But unless we make a renewed push to increase education as fast as we increased it over the twentieth century–and even then only if more education has the same payoff for those currently not getting a higher education as for those they are–that 0.3%/year factor is likely to drop to 0.1%/year. That pushes us down to 1.5%/year of real GDP per capita growth–an increase in the doubling time from 41 to 46 years.
Brink also sees a decline in the pace of savings and investment–a decline in the pace of capital accumulation. I am not sure that that is really there on the “investment” side. Moreover, the issue is complicated by the secular decline in the relative price of capital goods. So let me put this issue off to the side. Depending on what you think of the relationship between profits and the marginal product of capital, between 0.6%/year and 1.0%/year of our 2.0%/year GDP per capita growth was due to capital accumulation.
That leaves between 0.4%/year and 0.8%/year of growth to the fourth of the growth-promoting factors Brink Lindsey considers: innovation. He writes of how:
Both Tyler Cowen and Robert Gordon argue that the productivity slowdown… reflects… exhaustion of the… great technological breakthroughs of the late 19th and early 20th centuries… electricity, including electric lighting, motors, and appliances; the internal combustion engine, including automobiles, airplanes, supermarkets, and suburbs; “rearranging molecules,” including petrochemicals, plastics, and pharmaceuticals; and communications, including telephony, movies, radio, and television…. Was the IT revolution that has transformed our lives in so many ways really only good for a decade of strong productivity growth? Or is the current TFP slump merely a breathing spell in a long-term resurgence? Gordon, for his part, is pessimistic…
So let’s mark the conclusion of Brink Lindsey’s essay as a valid fear that our rate of real GDP per capita growth going forward is likely to be closer to 1.5%/year than to 2.0%/year, and turn to a more detailed examination of whether innovation is likely to amplify or offset that fall.
And so let us turn to the mid-range…
III. Cowen: A Great Stagnation?
Tyler Cowen–the world’s leading economics blogger–has been pushing this “great stagnation” meme hard. He is attracted to this idea–the best expression of it comes from a billionaire, Elon Musk Peter Theil:
I was promised a flying car, and all I have is the ability to send 140 character message via Twitter. Where is my flying car?
In Elon Musk’s case this seems to me to be somewhat deceptive: he doesn’t have a flying car, he’s soon to have a flying spaceship. If you’re a billionaire in this world today you can do remarkable things. And your personal flying spaceship was something regarded as unrealistic fantasy even back in the golden age of science fiction in the 1930s. Yet soon Musk can have one.
It is certainly true that the progress of technology has taken a very different direction than people back in the 1930s, 1940s, and 1950s thought–then it was very much: bigger, faster, louder. But it would have to be demonstrated to me that it has slowed down in any truly meaningful sense–even though Elon Musk and Robert Gordon and Tyler Cowen want to argue that it has.
Now for most of America’s population the “great stagnation” hypothesis has plausibility. If you are one of us–well, one of you–not in the top 5%, it looks true to reality: You look around and you see that you have lots more electronic toys that the people who were in your slot in the income distribution 30 years ago. But you also see that your house is further away from where you want to be. You see that education costs more. You see that health care costs much more. You see that commuting has become a bigger pain. You see that vacations in what used to be affordable places a generation ago are now out of reach. And so on. It is only for those of us up in the top 5% that material well-being is clearly much better today than it was for those in the same slots in the income distribution 30 years ago. But for those of us lucky enough to be in the top 5%, it is like: “Wow!” And for those of us in the top 1%, it is “WOW!!” And for those of us–well, you (a different you this time) in the top 0.1%, it is “WOW WOW WOW WOW WOW!!!!!”
Combine the fact that right now we are undergoing a large cyclical downturn with the fact that the past generation has seen an extraordinary upward leap in income inequality, and you can believe in the “great stagnation”. But if you look forward to a future in which the economy has recovered to normal levels of cyclical activity as of 2020 or so, and if you look at what is going on at the top of the income distribution as well as in the middle and at the bottom, I find it harder.
Thus, I have been uneasy about the argument of Tyler Cowen’s The Great Stagnation, and have resisted grappling with it thoroughly, in large part because it seemed to me that he was framing the issues in a way that I do not think is completely fair. For example, let me quote an introductory passage at length:
Tyler Cowen: The Great Stagnation: How America Ate All The Low-Hanging Fruit of Modern History, Got Sick, and Will (Eventually) Feel Better:
America is in disarray and our economy is failing us. We have been through the biggest financial crisis since the Great Depression, unemployment remains stubbornly high, and talk of a double-dip recession persists. Americans are not pulling the world economy out of its sluggish state–if anything, we are looking to Asia to drive a recovery. Our last three economic recoveries, beginning respectively in 2009, 2001, and 1991, have been “jobless” in nature. Commerce recovered far more quickly than did employment. Median wages have risen only slowly since the 1970s, and this multi-decade stagnation is not yet over. Typical individuals in earlier generations reaped much greater gains than ours, as their living standards doubled every few decades. We’ve even given back some of the growth we thought we had.
A lot of the prosperity of the “noughties” was built on debt, inflated home prices, and economic illusions. Currently, we are struggling to re-attain the economic output of 2008, and even before the financial crisis came along, there was no new net job creation in this last decade. Moreover, we face a long-run fiscal crisis, driven by the increasing cost of entitlements, our heavy reliance on debt, and our willingness to let matters slide rather than face up to paying the bills. The problems extend to American politics. The Democratic Party seeks to expand government spending even when the middle class feels squeezed, the public sector doesn’t always perform well, and we have no good plan for paying for forthcoming entitlement spending. To the extent Republicans have a platform, it consists of unrealistic claims about how tax cuts will raise revenue and stimulate economic growth. The Republicans, when they hold power, are often a bigger fiscal disaster than the Democrats…
The first sentence sets the stage. The second and third sentences seem to me to have nothing to do with any long run “great stagnation”–unless, that is, we allow (which it looks as though we may well) what was in its origins an (admittedly very large) cyclical downturn caused by demand-management failures to turn into a downward structural break. Sentences likewise deal with the cycle and not the trend, and point to problems of short run cyclical labor-market adjustment rather than to problems of long-run productivity growth. And sentence seven–about the stagnation of median wages since the 1970s–is once again not a sentence about trend productivity, but rather about maldistribution. So even as late as “the growth we thought we had…” in the passage I am quoting, Tyler has not said anything about any great stagnation that alters the trend growth rate of the American economy downward.
Starting with “a lot of the prosperity…” we have an argument that we were not as rich as we thought we were, an argument which indeed seems to me to be true, but that is not an argument that future growth will slow but rather that past growth was misperceived because of irrational exuberance. Only with “no new net job creation in the last decade…” is there even a datum that I see as about a “great stagnation”. But then Tyler branches off into failures of American governmance, failures that seem to me to be no greater than the failures of American governance have always been. Yes, failing governance. Yes, overestimated wealth at the end of the 1990s and in the mid-2000s. Yes, a somewhat more sclerotic-appearing labor market as far as cyclical adjustment is concerned. Yes, a successful class war waged by the 1% on the rest of us–or, rather, if my household income in 2014 is what I think it will be, on the rest of you. Yes, a huge demand shortfall-driven business-cycle downturn.
Where in all of this is the promised “great stagnation”? I do not see it.
But now I need to have a view about what the likely pace of economic growth and transformation here in the United States will be over the next two generations, and so I need to grapple with the argument Tyler presents. So here goes…
Tyler’s very best point, it seems to me, is that Simon Kuznets’s declaration that we had shifted from an era of Malthusian near-stagnation to one characterized by growth at 2%/year in living standards as far into the future as the eye could see was an extraordinarily broad historical, theoretical, and empirical generalization to make from an evidentiary base that was very narrow in both time and scope. As plausible ex ante as the belief that we had gotten onto an exponentially-explosive trajectory was that we were riding the upward curve of a logistic–a logistic made possible by the availability of readily-available fossil fuel resources, the development of the tools of modern science and engineering, the coming of precision machining, and the fortunate fact that once living standards get above twice bare subsistence and women learn to read we go through the demographic transition: nearly all women want to have one child, most want to have two, but only a minority want to have three or more. Thus since 1870 or 1800 or 1720 or 1492, depending on how exactly you characterize or count:
In a figurative sense, the American economy has enjoyed lots of low-hanging fruit since at least the seventeenth century, whether it be free land, lots of immigrant labor, or powerful new technologies. Yet during the last forty years, that low-hanging fruit started disappearing, and we started pretending it was still there. We have failed to recognize that we are at a technological plateau and the trees are more bare than we would like to think. That’s it. That is what has gone wrong. The old understanding was that the world broke through a barrier with the Industrial Revolution of the eighteenth century and that we can grow economically at high rates forever. The new model is that there are periodic technological plateaus, and right now we are sitting on top of one, waiting for the next major growth revolution….
And his second-best point is that there is some reason to think that our rates of economic growth over the past generation have been overestimated by our official statistics:
I have come to fear that the productivity statistics, and the national income statistics, are misleading us. It’s quite possible that actual productivity and actual GDP haven’t been going up as much as the published numbers make it seem…. A simple example: In 2005, finance accounted for 8 percent of U.S. GDP, and that figure had been rising throughout the 2000-2004 “productivity boom” period. I know what the numbers say, but what was the financial sector really producing during those years?
How much is health care really worth?… We go to the doctor because we hope it will make us healthier. [But] the doctor doesn’t face the same market test as the apple does. We know right away how good the apple tastes, and if it’s bad, we’ll stop buying that brand or stop buying from that store. On the other hand, very often we don’t know for a long time, if ever, what the doctor did for us. In other words, the market is testing whether or not the doctor can give us hope and the feeling of having been taken care of, not whether the doctor really makes us healthier.
Are children better educated than before? Educational expenditures are now about 6 percent of U.S. GDP. But is all that extra money invested in education giving us much of a return?… Let’s turn to the latest 2009 report from the National Assessment of Educational Progress, which is typically considered the definitive source of answers to these questions. On the first page of a fifty-six-page report, I find this sentence: “The average reading score for 17-year-olds was not significantly different from that in 1971.” On the same page, a little further below, I find: “The average mathematics score for 17-year-olds was not significantly different from that in 1973”…
Say that the expansion of finance from 4% to 8% of GDP has been a waste, that the 18% of GDP we spend on health–twice as much in real dollars as other North Atlantic countries with better health outcomes–is half wasted, and that our 6% of GDP spent on education doesn’t buy us anything more than 4% did 30 years ago. Add those up, and reach the conclusion that GDP per capita growth over the past 30 years has been overstated by 0.5%/year. That would mean that the 1.8%/year of measured growth in the economy’s potential to produce real GDP per capita has actually been 1.3%/year. Apply the 0.5%/year reduction in labor quality and participation growth we got from Lindsey, and get 0.8%/year–and note that we are now up to a doubling time of not 35 years for real GDP per capita but 87 years.
And his third very good point is that it looks like this diversion into unproductive activities has come about because an awful lot of our innovative and entrepreneurial activity has been tragically misdirected:
Contemporary innovation often takes the form of expanding positions of economic and political privilege, extracting resources from the government by lobbying, seeking the sometimes extreme protections of intellectual property laws, and producing goods that [have value because they are] exclusive or status-related… twenty-five seasons of new, fall season Gucci handbags. The dubious financial innovations connected to our recent financial crisis are another (perhaps less obvious) example…. The well-known rise in income inequality. Labor and capital are fairly plentiful in today’s global economy, and so their returns have been somewhat stagnant. Valuable new ideas have become quite scarce, and so the small number of people who hold the rights… whether it be the useful Facebook or… [to create] dubious… mortgage-backed securities–earned higher… returns…. If one sentence were to sum up the mechanism… it is this: Recent and current innovation is more geared to [redistributive] goods than to [useful] goods. That simple observation ties together… growing income inequality, stagnant median income, and… the financial crisis…
Take the conclusion that GDP per capita growth over the past 30 years has been overstated by 0.5%/year and add it to the 0.5%/year reduction in labor quality and participation growth we got from Lindsey. We are now down to a real GDP per capita growth rate of 1.0%/year–and note that we are now up to a doubling time of not 35 but rather 70 years. The only ray of hope is that we cannot do as badly over the next generation as we have in the past in allowing what William Baumol calls “destructive entrepreneurship” to divert our productive capacity into zero- and indeed negative-sum industries.
But assume that something has gone wrong with our political economy, and that we will in the next generation find new ways to squander our growth on unproductive industries. Is that as bad as things could get? Is there reason to think that the growth rate could be pushed down even further?
Tyler Cowen says: “yes”. And here Tyler Cowen seems to me to be on shakier ground. He begins talking about how the fruit we can see is not low-hanging:
Who today is the marginal student thrown into the college environment? It is someone who cannot write a clear English sentence, perhaps cannot read well, and cannot perform all the functions of basic arithmetic….. Educating many of these students is possible, it is desirable, and we should do more of it, but it is not like grabbing low-hanging fruit. It’s a long, tough slog…. One might argue that we have ongoing and future low-hanging fruit in the form of limiting job market discrimination…. [But] when it comes to boosting the rate of economic growth by discarding discrimination, many of the most important advances lie behind us…. We’re trying to eke out gains from marginal improvements…. Today… life in broad material terms isn’t so different from what it was in 1953… drive cars, use refrigerators… turn on the light switch…. You don’t have a jet pack. You won’t live forever or visit a Mars colony. Life is better and we have more stuff, but the pace of change has slowed…. It would make my life a lot better to have a teleportation machine. It makes my life only slightly better to have a larger refrigerator that makes ice in cubed or crushed form…. That’s the truth behind our crisis today–the low-hanging fruit has been mostly plucked…
And he starts following Robert Gordon by contrasting today with the wonderful possibilities opened up by the technologies of the First and Second Industrial Revolutions that began in 1740 and in 1880:
The period from 1880 to 1940 brought numerous major technological advances into our lives. The long list of new developments includes electricity, electric lights, powerful motors, automobiles, airplanes, household appliances, the telephone, indoor plumbing, pharmaceuticals, mass production, the typewriter, the tape recorder, the phonograph, and radio, to name just a few… agricultur[al machinery]… highly effective fertilizers. A lot of these gains resulted from playing out the idea of advanced machines combined with powerful fossil fuels, a mix… fundamentally… we have… exploited to a remarkable degree…
So what do we conclude when we focus more deeply, and look at the guts of changing economically-useful technologies?
IV. Gordon: One Big Wave?
Robert Gordon sees the economic history of the past quarter-millennium as dominated by three and a half industrial revolutions: the First Industrial Revolution from 1750-1830 of coal, steam, textiles, and railroads; the Second Industrial Revolution from 1870-1900 of electricity, internal combustion, utilities, communications and entertainment, and chemicals including petroleum; the spin-offs from the Second Industrial Revolution from 1900-1970 of mass production, air travel, road travel and automobiles, and skyscrapers; and the Third Industrial Revolution from 1960-present of computers, the web, and mobile telecommunications.
Gordon sees the Second Industrial Revolution–plus its spinoffs–as the big enchilada, the impetus behind a one-time unique once-and-for-all upward leap in material well-being that had no previous parallel and that cannot be repeated: we can urbanize and globalize only once; we can free women from their role as household beasts of burden that draw water, clean apparel, and sew cloth only once; we can stabilize our indoor climate at a comfortable temperature no matter where we live only once. That upward leap is, Gordon believes, now over. If he had to bet, he would bet that growth in real GDP per capita over the next generation or so in the United States would not be 2.0%/year, not be 1.5%/year, not be 1.0%/year, but only 0.5%/year–a doubling time of 140 years rather than the 35 years that we thought and think is our birthright.
What is the logic underpinning these pessimistic conclusions of Gordon’s?
His underlying framework is that:
The innovative process [is] a series of discrete inventions followed by incremental improvements which ultimately tap the full potential of the initial invention. For the first two industrial revolutions, the incremental follow-up process lasted at least 100 years…. Many of these processes could happen only once. Notable examples are speed of travel, temperature of interior space, and urbanization itself….
The audacious idea that economic growth was a one-time-only event has no better illustration than transport speed. Until 1830 the speed of passenger and freight traffic was limited by that of “the hoof and the sail” and increased steadily until the introduction of the Boeing 707 in 1958. Since then there has been no change…. This may seem obvious about horses, outhouses, speed, and temperature, but once you accept that, you’re drawn into the central theme of this article: economic growth may not be a continuous long-run process that lasts forever.
It is against this background that Gordon paints his picture of 1870-1970 as a one-time, unrepeatable transformation of North Atlantic human life from sub-Third World penury and drudgery to our current prosperity, where all of our difficulties are First-World problems:
A lot of progress had been made by 1870… water and steam… telegraph…. But most aspects of life in 1870… were dark, dangerous, and involved backbreaking work… no electricity… insides… dark… smoky… bedrooms… unheated… family members carried warm bricks with them to bed… lack of running water… water for laundry, cooking, and indoor chamber pots had to be hauled in by the housewife, and wastewater hauled out… housewi[ves] walk[ed] 148 miles per year while carrying 35 tons of water…. Within the cities, steam power was not practical, so everything was hauled by horses… 20 to 50 pounds of manure and a gallon of urine daily, applied without restraint to stables and streets… 5 and 10 tons per urban square mile [per day]…. Life expectancy was only 45 years in 1870…. In 1900, 13,000 people died in railroad deaths, about a quarter of them railroad employees…. The growth of productivity… slowed markedly after 1970…. It seems increasingly clear that the one-time-only benefits… had occurred and could not happen again. Diminishing returns set in…
The natural reply would be to say that our productive potential is still increasing–we are still inventing, innovating, and investing–but we are devoting our energy to different areas of life, and solving problems that have not been solved rather than making very marginal improvements in areas that have been solved. There is, after all, a limit to how much better a toilet can get. And how much is it worth spending to get from San Francisco to London in two hours rather than in eight?
Robert Gordon says “not so”. In his eyes, the Second Industrial Revolution was a blockbuster bomb, while the Third Industrial Revolution is merely a firecracker:
The third revolution (IR #3) is often associated with the invention of the web and Internet around 1995. But in fact electronic mainframe computers began to replace routine and repetitive clerical work as early as 1960…. Initially computers shared with the steam engine, the internal combustion engine, and the electric motor the many-faceted benefits of replacing human effort, making jobs easier, less boring, and less repetitive. It may seem surprising that so many of the computer’s laborsaving impacts occurred so long ago…. In… its early years IR #3 was a pipsqueak compared to IR #2…. The Internet, web, and e-commerce did, however, make a difference. Productivity growth began to recover in 1996 and by 1999 the arrival of the “new economy” was heralded…. ￼I was (Gordon, 2000) among the skeptics and doubted that the “new economy” would have an impact comparable to the inventions of IR #2. With 12 additional years of data, it appears that my initial skepticism was appropriate…. The era of computers replacing human labor was largely over…. Attention in the past decade has focused not on labor-saving innovation, but rather on a succession of entertainment and communication devices that do the same things as we could do before, but now in smaller and more convenient packages…
And so he dismisses the past, present, and future impact of our generation’s electronics and telecommunications revolutions as small beer:
A thought experiment helps to illustrate the fundamental importance of the inventions of IR #2 compared to the subset of IR #3 inventions that have occurred since 2002. You are required to make a choice between option A and option B. With option A you are allowed to keep 2002 electronic technology, including your Windows 98 laptop accessing Amazon, and you can keep running water and indoor toilets; but you can’t use anything invented since 2002. Option B is that you get everything invented in the past decade right up to Facebook, Twitter, and the iPad, but you have to give up running water and indoor toilets. You have to haul the water into your dwelling and carry out the waste. Even at 3am on a rainy night, your only toilet option is a wet and perhaps muddy walk to the outhouse. Which option do you choose? I have posed this imaginary choice to several audiences in speeches, and the usual reaction is a guffaw, a chuckle, because the preference for Option A is so obvious. The audience realizes that it has been trapped into recognition that just one of the many late 19th century inventions is more important than the portable electronic devices of the past decade on which they have become so dependent…
And says, moreover, that we won’t even be able to match in the next generation what computerization and communications globalization has done in the past one:
[To] start by assuming that future innovation propels growth in per-capita real GDP at the same rate as in the two decades before 2007, about 1.8 percent per year… strains credulity… makes the heroic assumption that another invention with the same productivity impact of the internet revolution is about to appear on the near-term horizon…
And so, taking Lindsey, Cowen, and Gordon together, we arrive at future real GDP per capita growth in America of 0.5%/year rather than 2.0%/year–and a doubling time of 140 years rather than 35.
And now let me make my quibbles. I have eight:
A great deal of the discussion about how there has been a “great stagnation” over the past generation that is likely to continue into the future misread distribution and demand-management failures as innovation and accumulation failures.
A corollary: for the near-term, worries about “secular stagnation” are overwhelmingly either wrong or worries about how to solve the technocratic problem of mobilizing societal risk-bearing capacity and properly distributing our productive capacity among sectors and our output to individuals.
Better and more comprehensive measures of human freedom–both positive and negative–paint a much more optimistic picture of America’s future then do Lindsey, Cowen, or Gordon.
Real GDP per worker is profoundly the wrong indicator to look at.
A corollary: the fact that marginal utility of wealth and income may decline sharply is not a statement about a slowing pace of economic growth but rather about declining marginal utility of wealth.
The market continues to provide the old incentives to figure out how to make the things we spend a lot of money on more cheaply and efficiently.
There is a valid “great stagnation” worry, but it is overwhelmingly one of institution design rather than of innovation exhaustion.
But the “great stagnation” literature as it is currently constituted seems to me at least to guide our attention in the wrong direction–and to quite possibly stampede us into making policy decisions we really would not want to make if we thought more deeply and calmly.
Let me now develop these at greater length:
(1) A great deal of the discussion about how there has been a “great stagnation” over the past generation that is likely to continue into the future misread distribution and demand-management failures as innovation and accumulation failures. We saw this in how Tyler Cowen frame the argument of his the great stagnation. Yes, we have done an awful job over the past generation in two respects: distributing the fruits of economic growth fairly; and focusing our productive potential on industries that add value rather than subtracted. But these did not just happen: we made them happen–we chose them through our politics. We may will have politics as lousy over the next generation as it has been over the past one as far as equitable growth is concerned. But if we do, the fault, dear Brutus, is not in the stars but in ourselves. And to talk about how this is our destiny–about how the fault is in the stars–is, I think, substantially misleading.
Consider Brink Lindsey’s reasons to expect slowing real GDP growth: reduced labor force growth, reduced ability and/or willingness to invest in human capital, reduced savings and investment, and reduced total factor productivity growth. To the extent that these have the potential to have a negative impact on growth in our standard of living, the first three of these are all under our control. The labor force increased because as part of the process of economic growth going to work seeing a good thing for people do with their time. That economic growth in the future will lead people to think that they should do something else with their time is not, properly assessed, a minus. We have chosen to make it more difficult for the young today and in the future to expand investment in their human capital. We have chosen to save and invest less, and to consume and waste more. We can and should unchoose.
The only substantially worry some factor, in that it is not something where the road to unchoose it is obvious, is the fourth: the potential for a slowdown in innovation and technological progress. And there seems to me to be a little reason to be pessimistic about the fourth.
(2) A corollary: for the near-term, worries about “secular stagnation” are overwhelmingly either wrong or worries about how to solve the technocratic problem of mobilizing societal risk-bearing capacity and properly distributing our productive capacity among sectors and our output to individuals. We may fail to accomplish these tasks. But they are our most urgent ones. We should not let the intellectual interest of wondering about the long run shape of technological progress do too much to distract us from the urgent current public-policy problems of restoring full employment, reversing the pointless and destructive rise in inequality of income and wealth, and putting our hypertrophied negative-sum sectors like healthcare administration and finance on a diet.
(3) Better and more comprehensive measures of human freedom–both positive and negative–paint a much more optimistic picture of America’s future then do Lindsey, Cowen, or Gordon. The point, after all, of economic growth is to advance our material well-being as a means and not as an end unto itself. We seek to move from the realm of necessity to the realm of freedom: where we do not what we must to hold body and soul together but what we choose–so that we can become who we want to be, or perhaps who we really are.From tthis perspective Abraham Lincoln, Martin Luther King, Jr., and Gloria Steinem are at least as important actors in the story of economic growth as are Andrew Carnegie, John D. Rockefeller, and Bill Gates. And a look at the occupation, education, and wealth distribution of America today leads immediately to the conclusion that there is a lot more low-hanging fruit from the effects of reductions in discrimination still to be picked. Those of us with First World problems have been largely freed by economic growth from a life of being exhausted by boring, backbreaking work. But unless you think we are already at the Gates of Utopia, advances in material well-being still have at least half the task left to do, for breaking the chains of necessity leaves us only at the border of the realm of freedom.
(4) Real GDP per worker is profoundly the wrong indicator to look at. Real GDP measures how much the things we make cost–what we could sell them for on the marketplace, if we could sell everything at its current market price. We calculate real GDP measures because back in the old days Colin Clark and Simon Kuznets were casting around for a rough-and-ready summary measure of economic activity that they could estimate fairly accurately at their desks from publicly-available information. They settled on real GDP as the best they could do. But it is not really what we want.
What we are really interested in is material well-being; what people want and need, and whether they actually obtain it. We are interested in material production as a springboard to human flourishing: How much of people’s time are they spending in drudgery trying and barely succeeding (or failing) to acquire enough food that they are not desperately hungry, enough shelter that they are not desperately wet (or parched), and enough clothing that they are not desperately cold? And is the surplus over bare biological material necessity of a quantity and type such that it enables them to choose and live the kinds of lives they want to live? These are the questions that a human satisfaction material well-being estimate is for. And real GDP just does not cut it.
Rather than asking what the things we produce would cost on the market, we should be asking how much is the surplus we obtain–how much more things are worth to us than they really cost is in terms of time, effort, toil, and drudgery. But GDP is the sum of the costs of things as they are or would be valued on the market.
You see the problem.
It is, in fact, in many cases surprisingly hard to see the problem. At the level of the individual, often the surplus received via some piece of property is what you can sell it for on the market–that is, what is really its cost. But although this is true at the level of the individual, it is not true at the level of the economy as a whole.
Back in the 18th century this issue went under the names of the “diamonds and water paradox”: water is really cheap, yet without people to work at getting us our water we all die of thirst; diamonds are expensive, get without people to work at getting us our diamonds we use some other kind of stone for display and status. And for them the issue was further muddled because often it turns out that for an individual you get the most surplus from a good not by consuming it yourself but they selling it at its market price, in which case cost is or is related to surplus. We today can at least think about these issues clearly: the problem is that the GDP statistics do not.
Moreover, a focus on real GDP implicitly makes an extraordinary bizarre and totally false value judgment. It assumes that we value an extra $1 of income to an idle heir whose family hasn’t done a lick of real work in three generations as much as an extra $1 of income to hard-working parents at the poverty line. And we do not believe that, we do not do that, we do not value things and people that way.
Now as long as increases in real GDP do not proceed alongside big changes in income inequality, you can use changes in real GDP as a proxy for changes in some broader–better–measure of human flourishing and well being. But you cannot then justify a policy that increases inequality on the grounds that it increases real GDP. And you cannot prioritize policies that increased real GDP by empowering the rich over policies that increased real GDP less by empowering the poor. And as long as there is a roughly constant relationship between the surplus received by consumers and users and the costs that are measured in GDP, you can use changes in GDP to gauge changes in surplus.
The first problem is that the past generation has seen extraordinary swings in relative income inequality. The second problem is them as our economy swings toward information goods of one sort or another the relationship between surplus as value and cost as value is sure to be disrupted. So any discussion of the changing pace of real GDP growth needs to do much more then drop asterisks: it needs to highlight when a shift in the real GDP growth rate is accompanied by an upward or downward deep in inequality that greatly affects how the amount produced is transformed into human material well-being; it needs to highlight the changing relationship between the surplus that enters into human material well-being and the cost that enters into the real GDP numbers as well. Reflect that the value-added in Henry Ford’s Ford Motor Company’s value chain’s contribution to real GDP was half of the ultimate value to users of the automobile. Reflect that the value added in Google’s value chain’s contribution to real GDP is at most one-tenth of the ultimate value to users of the internet. I, at least, see low-hanging fruit there.
(5) A corollary: the fact that marginal utility of wealth and income may decline sharply is not a statement about a slowing pace of economic growth but rather about declining marginal utility of wealth. Robert Gordon makes much of asking his audiences whether they would rather give up Google or give up their flush toilets, and concluding that the Third Industrial Revolution is very weak tea compared to the Second in its effect on material well-being. Maybe. Maybe not. But even if this thought experiment comes out the way Gordon thinks it does, it strikes me that it is worth the drawing a distinction between the utility and the pace of economic growth. Economic growth, it seems to me, is all about making what we used to have to do cheaper in time and resources and in inventing new useful things we can do. And as long as there is still scarcity–as long as things still have prices–economic growth is measured by the reductions in the real prices of the commodities we produce and use. We want something. It is scarce. We figure out a way of making it more cheaply, so we can deploy some of the resources we used to acquire it to doing something else. That is economic growth. The fact that we cannot now imagine living without some things that are now extraordinarily cheap to acquire–so cheap that we are absolutely satiated with them, for few people have a demand for additional bathrooms in their house except for the few years they have teenagers in residence–is a statement not about any potential slow down the pace of economic growth but rather about the shape of human psychology. Or so, at least, it seems to me.
(6) The market continues to provide the old incentives to figure out how to make the things we spend a lot of money on more cheaply and efficiently. When we measure the pace of real economic growth per capita at 2% per year, we assert (a) that this year we can make the valuable stuff we made last year with 2% fewer resources, and (b) that the gap between what we make costs and what its value is to us is has changed little. That is what 2%/year economic growth means. The market economy thus provides enormous incentives for economic growth: figure out how to make things people spend a lot of money on more cheaply, and become rich.
The things we used to buy become cheaper, and so we can buy more of them if we wish. Moreover, our material well-being is also boosted by two other things: First, it is still by the invention of new commodities and new types of commodities that we then discover we really want and need. Second, it is boosted by the availability of things we want in me that we find out we can get for free–or, more generally, by a divergence between cost and value. (And, conversely, negative externalities— things like congestion, declines and safety, and pollution— tend to subtract from material well-being even though they do not enter into real GDP.)
The point is this: things that make the production of things we buy more efficient – that make them cheaper – raise real GDP. And the larger is the expenditure share of a commodity in what we buy, the more does an improvement in the efficiency of its production boost real GDP per worker. Now in a market economy there is an enormous amount of money to be made by figuring out a cheaper way to make a commodity that is a highchair of expenditure, and then to skim off a portion of the difference between lower-cost old price. Hence inventive activity will naturally direct itself to those places where it would tend to have the most good for economic growth. And few people today are willing to pay much for an improvement in toilet technology, or spend a large share of their income on their toilets already.
Those things that are the most technologically impressive are not the same as the things that are most materially beneficial.
If you look not at heroic inventions but at continuous improvements, not at the automobile is a shoot of America but my willingness to eagerly pay an extra hundred dollars for an iPad that weighs half as much as last year’s a lot of Robert Gordon’s worries about the end of the heroic age of the Second Industrial Revolution become less salient.
As Tyler Cowen puts it in one of his optimistic moods:
In a typical day, I might write two tweets, read twenty blogs, track down a few movie reviews, browse on eBay, and watch Clarence White play guitar on YouTube… I am interested and amused the entire time. More and more, “production”—that word my fellow economists have been using for generations—has become interior to the human mind…. No single bit from the Web seems so weighty on its own, but the resulting blend is rich in joy, emotion, and suspense…. The new low-hanging fruit is in our minds and in our laptops…. Innovation hasn’t ceased, but it has taken new forms and it has come in areas we did not predict very well. Yet we made our old plans and maintained our old institutions on the understanding that the new innovation would be a lot like the old, except that it isn’t…
(7) There is a valid “great stagnation” worry, but it is overwhelmingly one of institution design rather than of innovation exhaustion.And here we reach what I regard as the big issue. In the future we are going to want to spend a greater share of our incomes and attention in areas where the market system works less well: information goods, public goods, increasing-returns goods, pensions, health care, education. The market works less well in these areas. But our alternative modes of collective organization, product take some bureaucracy, not exactly cover themselves with glory in these areas either. Thus I suspect that not innovation exhaustion but rather institution design will be our big problem in keeping the pace of true economic growth going into the long-run future.
(8) The focus on real GDP growth and its possible–or likely–slowing is a setup to panic us into making policy decisions we really do not want to make. The “great stagnation” literature as it is currently constituted seems to me at least to guide our attention in the wrong direction–and to quite possibly stampede us into making policy decisions we really would not want to make if we thought more deeply and calmly. The chain of logic is that measures to reduce inequality have a cost in terms of reducing the growth rate of the economy–that the bucket of redistribution is, in the terms of Arthur Okun’s Equality and Efficiency: The Big Tradeoff, a leaky bucket–and that when growth is slower we can no longer afford to engage in redistribution. This seems to me to be the wrong way to conceptualize it: the evidence that the bucket is leaky is weak–or, rather, there are many buckets, some very leaky, some not leaky at all, some anti-leaky–and in any event whether we should tradeoff potential growth for other objectives is not something the depends on how fast growth is. Policies that make sense if underlying GDP per worker growth is 3% probably still make sense if underlying GDP per worker growth is 1%. Policies that don’t make sense if underlying GDP per worker growth is 1% probably still don’t make sense if underlying GDP per worker growth is 3%.
But my aim here is simply to lay down a marker as far as point is concerned: to enjoin you not to get stampeded into going someplace you really do not want to go.
VI. The Bottom Line
Where does this leave us? You can pay your money and take your chance. Me? I am still at an expected rate of trend real GDP per capita growth of 1.7% per year–slower than the past because of slower labor force growth as a share of the population, but that is not something to sweat. And because I am an optimist about our future politics, take that number and add on to it bonuses for improving income inequality, for a widening gap between cost and value as information goods require additional salience, and for a politics that shifts resources out of negative-sum and into positive-sum sectors. Thus I see that 1.7% as carrying with it a 3% per year rate of growth in properly-measured North Atlantic human material well-being.
And on top of that, I share the techno-optimism of Paul Krugman:
The Case for Techno-optimism: The debate over secular stagnation has shown some signs of getting confused with the debate over technological stagnation; it’s important to understand that they are not the same thing. What Bob Gordon (pdf) is predicting is disappointment on the supply side; what Larry Summers and I have been suggesting is that we may face a persistent shortfall on the demand side. It’s true that Gordon’s world might suffer from low investment demand, since investment demand depends more on the rate of growth than it does on the level of real GDP. Still, they are distinct concepts.
But what do I think about Gordon’s notion that the good times of progress are behind us? One honest answer would be that I don’t know, and can’t even make a good guess.
What I can say, however, is that my gut feeling remains that while Gordon may be right about the next decade or two, he’s likely to be very wrong beyond that. Or maybe it’s a bit more than my gut. I know it doesn’t show in the productivity numbers yet, but anyone who tracks technology has a strong sense that something big has been happening the past few years, that seemingly intractable problems — like speech recognition, adequate translation, self-driving cars, etc. — are suddenly becoming tractable. Basically, smart machines are getting much better at interacting with the natural environment in all its complexity. And that suggests that Skynet will soon kill us all a real transformative leap is somewhere over the horizon, maybe not this decade, but this generation.
Still, what do I know? But Brynjolfsson and McAfee have a new book — not yet out, but I have a manuscript — making this point with many examples and a lot of analysis.
There remain big questions about how the benefits of this technological surge, if it’s coming, will be distributed. But I think this kind of thing has to be taken into account when we try to imagine the future; I’m a great Gordon admirer, but his techniques necessarily involve extrapolating from the past, and aren’t well suited to picking up what could be a major inflection point.