Discussion: (3 comments)
Comments are closed.
A public policy blog from AEI
Not only is America experiencing an historically weak recovery, but there are deeper signs of trouble. A big one: Productivity growth — a key indicator of technological innovation — has been basically flat since the Great Recession, according to official statistics.
Are things ever going to get better? Or as singer-songwriter Merle Haggard once put it,“Are the good times really over for good?”
I discuss the long-term prospects for the US economy and American workers in this week’s podcast with economist Robert Gordon of Northwestern University. His research asking whether U.S. economic growth is “almost over” has been widely cited, and he was named by Bloomberg as one of the nation’s ten most influential thinkers. He is also author of the new book “The Rise and Fall of American Growth: The US Standard of Living since the Civil War.”
Here are excerpts of our talk, which you can listen to in full over at Ricochet.
Pethokoukis: We are going to talk about the book, “The Rise and Fall of American Growth,” which you say is built around two big ideas. The first big idea is that the period from 1870 to 1970 was a special century when it comes to innovation, productivity and economic growth, a period – a special century both unique in human history and unrepeatable because the achievements could only happen once. So that’s the big idea number one.
Big idea number two: some inventions and innovations are more important than others, and that special century after the Civil War was made possible by a unique clustering in the late 19th century of what you call “the great inventions.”
If I can use the old Merle Haggard song, “are the good times really over for good?” I think you might say that maybe they are – that we’re not going to have that special century again.
GORDON: Well, the idea that it was a special century is pretty obvious if you go back before 1870 because we had virtually no economic growth from the Roman Empire through about 1700. The best research suggests that in England, where they have the best statistics, from 1300 to 1700, growth was only at a rate of 0.2% a year, enough that would take four centuries for the standard of living to double.
Starting around 1870, we had sufficient growth at roughly 2% a year; that is, not 0.2 but 2.0, and that’s enough to allow the standard of living to double every 30 years or so. So since the Civil War, it’s become commonplace to expect that each generation will be twice as well off as the generation that came before.
The reason this happened was that the world was ripe with opportunity for people to invent new things in the decade starting in the 1870s. The two really big inventions, the greatest of all were electricity — which of course made so many things possible, from the elevator to home appliances to power machine tools on the ground and power machine tools that people held in their hands — and the internal combustion engine, which made possible not only motor transport – cars, buses and trucks – but also allowed man to finally achieve his longtime dream of flight.
That special century went far beyond just a couple of inventions. We had the spread throughout urban America of an old idea called running water, so we had running water that came to the house and waste pipes that took it away – something that did not exist in a world of 1870, dominated by outhouses and poor housewives having to carry water into their homes in pails.
And we had perhaps the most valuable – if we try to quantify it, the most valuable of all improvements, which could only happen once, was the conquest of infant mortality. In 1890, the percentage of newborn babies who died in the first year was 22%. By 1950, only 60 years later, that percentage was down below 1%. And think of all the lives that were saved and all the value that was created. Some economists have calculated that the ending of infant mortality alone was worth more than all the other new consumption goods invented during that period.
So that’s the first big idea that the revolutionary century after 1870 was special and had never happened before.
And now, the second big idea is that it didn’t happen again. We’ve had plenty of inventions since 1970 but it’s been focused on the narrow sphere of entertainment, information, and communications technology. That means everything associated with the television, including time shifting through VCRs and DVRs; computing, going from the mainframe through mini-computers and personal computers through to the laptop and the smartphone; and the mobilization of communication, moving from the landline phone to the dumb mobile and now the smart mobile phone.
Those innovations are everything that we talk about today, but in perspective they’re just a small slice of what human beings care about. If we looked at food, clothing, shelter, transportation, entertainment, motion pictures, we have achieved relatively slow progress since 1970. So that’s the second idea – that the progress we’ve achieved has been more narrowly focused in a smaller part of the economy.
So those are the two big ideas and they lead to a forecast for the future, which is that despite all of the technological hoopla that we hear about, we’re not going to be having a return to the special century.
You begin by talking about what life was like in 1870. And to most people, it would seem to be unimaginable to have to live your life as they did in 1870. And then by 1940, life really began to look a lot like it does now. So life by 1940 was vastly different than it was in 1870.
Well, I would extend it up from 1940 to 1970 because the real modernization of the house, the modern kitchen with cabinets and electric appliances built in, was something that was really made possible by the consumer prosperity after World War II. The refrigerator and the washing machine had been invented in the 1920s. About half of urban families had them by 1940. But the real accomplishment of rolling out the modern lifestyle took a few decades into the postwar era.
So what changed that we had century after century of virtually no economic growth, no economic progress and then we did? It looks like thousands of years of the flat line and then a line that goes up exponentially. So what is your theory why we suddenly had that takeoff?
I think it’s because we had a free society with an extremely inexpensive patent system. Anybody could go and get a patent. That created the incentive to invent things without fear that your ideas would be stolen. We know, for instance, that Alexander Graham Bell achieved his invention of the telephone by arriving at the patent office only three hours earlier than his competitive rival, Elisha Gray. So people were actively inventing, competing with each other, and many smaller inventions were taking place at the same time.
Alexander Graham Bell’s telephone in 1876 predated the electric light bulb of Edison by only three years. And the first workable internal combustion engine of Karl Benz, the German, came only 10 weeks after the electric light bulb. So this was a very exciting time. And to the extent that these inventions happened in America, I credit the patent system as one of the main causes.
And, of course, people had wanted to communicate by telegraph before. Once the telegraph was invented, naturally the mind went to the next step of how to get the voice to go over the wire instead of the dot and dash of the Morse code. So there were lots of opportunities there. Man had wanted to fly for a long time, but the internal combustion engine was light enough so that it made possible putting an engine on top of a pair of wings. And so it went.
So we’re talking about about slower GDP growth. It’s not that the economy is not going to grow. It will just grow slowly and [because of] that slow growth combined with income inequality, it will certainly feel like stagnation. Is that right?
Since 1970, we’ve had much slower growth in productivity; that is, the growth in how much each worker hour produces. And that was interrupted by a relatively rapid decade, between 1995 and 2005, what we call the dot-com decade, the decade in which the Internet was introduced along with search engines and e-commerce. And we observed productivity temporarily pick up during that period.
But my forecast for the future is that it’s going to look very much like the last 40 years. It’s not a slowdown in growth or productivity. The problem is that output per hour is just one of the essential underpinnings of the process by which economic growth makes it down to the ordinary family.
And to give you a couple of numbers, I’m predicting that productivity growth will be 1.2% over the next 25 years – not that dissimilar to what it has been over the last 45 years, with the exception of that one single dot-com decade when we did better.
But the average American is not going to be able to enjoy that 1.2% growth for two reasons. The first reason is that the population is aging and many people are moving from active work into retirement, the so-called Baby Boom retirement phenomenon. This means that the number of hours of work per member of the population is going down. And that means that the output each hour is producing has to be spread across more people, many of them retired, and so that the 1.2% that I started with for output per hour gets reduced down to 0.8 for output per person. That’s because the hours per person are declining.
And then, as you mentioned, inequality prevents the average American from gaining the full fruits of our economic progress: as much as half of the growth in income has been going into the pockets of the top 1% of the income distribution. And so if you look at the median income, the income of the American who’s in the middle at the 50th percentile, that income has been growing about 0.4% slower than for the average as a whole, which includes the top 1 %. So you take off that 0.4 for Baby Boom retirement and another 0.4 for inequality and you’re down from 1.2 to a much, much slower growth of 0.4. And that’s where things start looking very different from the past.
You’re not forecasting necessarily that productivity growth is going to collapse. It’s more-of-the-same, and that combined with demographic factors ends up working out to slower GDP growth than what we’ve had.
And it’s important – it’s important to emphasize that productivity growth is not going to be zero. A 1.2% productivity growth is plenty to encompass the possibility of many of the inventions that people are talking about: the gradual arrival of robots, the filtering of artificial intelligence into more and more of the economy, the driverless car and truck and bus that are on the horizon. I’m not saying that technological change is coming to a halt.
We, after all, during these slow growth periods since 1975 had enormous numbers of individual inventions. The big mobile phone transitioned into the small, miniature mobile phone, transitioned into the iPhone with all of the applications that come with it, combining a GPS unit, a camera and a communication device and a web search engine, all in the same little compartment.
All of that invention that we have enjoyed took place in an environment since 1970 when productivity growth was not all that fast. And much of what I’m forecasting for the next 25 years is a continued evolution rather than revolution.
You are often called the techno-pessimist — even if I’m sure you don’t think that describes your views — while other people will be called the techno-optimists, the difference being that they think that these inventions will have a much greater impact on our measurable living standards than what you do.
That’s partly because I’m looking over the entire economy and I’m seeing sector after sector where we’re not finding much influence of technology at all. Take the grocery store. We have checking out with bar code scanning, very similar to what we’ve had for the last 20 or 30 years with most payment taking place with credit cards instead of cash or check. We made enormous progress in the checking out process at a supermarket, but that all happened 20 and 30 years ago.
I play a game called “Find the robot” where I’d look everywhere in my daily life, not just in the supermarket but throughout retail, checking in for a doctor’s appointment, throughout the educational industry, looking for examples of modern intelligence, artificial intelligence or robots taking over human jobs.
And I find that the techno-optimists are just greatly exaggerating the span – the sphere of life that will be impacted by these innovations that are indeed gradually happening.
But wouldn’t they argue, though, that’s their exact point? That you have these huge swaths of the economy which so far seem impervious to a lot of these innovations, information technology innovations, and therefore, there’s a lot of room to install these innovations and hopefully have a lot greater productivity in those two sectors. There certainly is a lot of venture capital interest in both those sectors viewing them as unplowed ground for that IT revolution. Isn’t there the potential for fairly dramatic increases in productivity?
There’s opportunity for progress, but think of the typical hospital with nurses, the kinds of things that nurses and orderlies do for patients. There’s room perhaps for robots to push patient beds through the hallways to go from the hospital room to, say, get a CAT scan or an X-ray and bring the patient back, replacing human labor.
But we’re a long way from having robots that are capable of doing virtually anything that a human nurse is capable of doing, much less a doctor. Granted that laser machinery has helped surgeons become more precise and much of the medical technology involves better devices that work along with the human beings, far from replacing them.
You say you’re making a 25-year forecast. How confident are you that what you’re saying will be true of the next 25 years, or will be equally true over the next 50 years?
Well, that’s why I forecast with 25 years. I think, after all, 25 years is… Just looking back at how long ago 25 years ago, that was 1991, and much of what we’re enjoying now could have been forecast in 1991. I don’t forecast beyond 25 years because I’m fully humble enough to realize that that’s a long enough period going out 40 or 50 years for things to be invented that we haven’t even dreamed of. But I’m talking about the things that we have dreamed of that are far, far, far from being rolled out into the real-world economy. And I see the arrival of robots, 3D printing, artificial intelligence and autonomous vehicles is taking a very long time to occur.
The proponents of driverless cars, for instance, talk about an improvement in safety and a reduction in crashes. That might indeed be possible when all the cars are autonomous, but we’ve got a fleet of motor vehicles of 200 million-plus existing motor vehicles that require humans to drive them. They’re not all going to be scrapped overnight. It’s going to take decades for the fleet of motor vehicles in the country to transform. And many people may not want to have autonomous vehicles which by their very nature, with all of their sensing devices, are going to have to cost more. Some people will prefer to drive themselves. I certainly would.
The economics team over at Goldman Sachs has written a lot about this issue. Here’s what Goldman wrote:
Profit margins have risen to record levels. Inflation has mostly surprised on the downside. Overall equity prices have surged and technology stocks have performed even better than the broader market. It was just the opposite when the productivity slowdown began in the 1970s. Maybe metrics devised for a wheat-and-steel economy of physical commodities are poorly suited for one experiencing rapid growth in software and digital content. Maybe there’s a systemic understatement of productivity and GDP growth.
And they conclude, “one should be skeptical of confidence pronouncements that the standard of living is growing much more slowly now than in the past.” Might we just be missing the digital economy because we can’t measure it properly?
There have always been parts of our economy that have been left out of the measurements. And I mentioned before running water, the conquest of infant mortality, the transformation of cities when motorcars replaced horses and we no longer had to remove horse manure from the streets and sidewalks.
And the thought that we are missing the benefits of smartphones is very familiar because we missed the benefits of – in our official measurements of the invention of e-commerce, the invention of free encyclopedias, the invention of electronic catalogs that tell us not just what’s available but what’s in stock and what’s out of stock. So I see the omission of benefits of modern society not being included in GDP as a phenomenon that is not just true of the last eight years of the smartphone, not just true of the last 25 years of e-commerce and search engines, but going all the way back to the beginning of the 20th century and the end of the 19th century.
And I would venture to suggest that when we talk about infant mortality, running water, waste disposal, the curing of infectious diseases, all those things that happened in the first part of the 20th century, that the value to human beings of those things that were omitted from GDP back then are greater than the value of those things like Skype and free Internet telephoning that have taken place recently.
To quote one Silicon Valley venture capitalist, what is the value or potential of four billion people being able to have in their hand a small sheet of glass with full access to the accumulated knowledge of human civilization? There would seem to be a lot of potential there.
Well, when we talk about four billion people, we’re expanding the topic, going beyond the history of the United States and talking about the enormous transformations taking place in China and India and Southeast Asia and Latin America as countries begin to enter the modern world.
But we would benefit certainly by having hundreds of millions of brains come online and think up who knows what.
Well, I think the inventions are going to take place, and the question is: who benefits from them? Will there be Chinese entrepreneurs and Chinese companies or will they be American or European entrepreneurs and companies? The fact that the Chinese have come on and done much of the manufacturing for us is behind a number of very deep and difficult social problems in the United States.
So, we’ve got economic growth and catching up in the rest of the world, but globalization has brought with it, along with automation, a hollowing out of US manufacturing. Manufacturing employment is down to barely 10% of the population. We have Rust Belt cities throughout the Midwest and the Northwest that are shadows of their former selves. We’ve heard a lot recently about Flint, Michigan, with its population falling by half as its manufacturing base departed. So globalization is not just an unalloyed benefit.
Before we started our chat, I went onto Twitter and asked some of the folks there if they had questions for you. One was: Have Brynjolfsson and McAfee, referring to Erik Brynjolfsson, Andrew McAfee, authors of “Race with the Machines,” a more of a techno-optimist book, have they apologized to you? Obviously, this person finds your arguments so correct that they’re wondering if they have been persuaded at all by your argument.
Well, just a few weeks ago, I was with Erik Brynjolfsson in a back-and-forth discussion for public television in which he said he agreed with 90% of what I’ve been saying. And I think the forecast of the future, we differ in emphasis rather than in substance. After all, I’m not saying that there’s going to be no innovation. My 1.2% productivity growth allows plenty of room for innovations to occur. In fact, the economy is running behind – at slower speed than I forecast. Productivity growth in the last five to six years has only been about half a point per year.
So Erik and Andy are enthusiastic about inventions that I see coming just as they do. I just don’t see them replacing human labor to the extent that they worry about it. We, after all, have managed to get our economy’s unemployment rate back down from 10 to 5%. Five percent used to be thought to be about as low as it can go. And the big problem we have is not that the machines are eliminating all of the jobs. There are plenty of jobs – 5.4 million job openings at present in a given month.
But the problem is that those jobs are low paying, their wages are rising very slowly, and many of the jobs are for menial tasks in the service economy which have proven to be very difficult for robots to replace.
Toward the end of the book, you list a number of potential policy ideas to boost productivity growth. But it seems to me that what you’re most strongly arguing for is a lot more redistribution. If you have a lot of people who are experiencing stagnant incomes, the obvious thing is to redistribute from those who are not experiencing stagnant incomes. Is redistribution the great challenge over the next 25 years? Or is it trying to tweak productivity and get somewhat faster GDP growth?
Well, you have to go at both. I distinguish between equality of opportunity and equality of outcomes. To make the outcomes better for those who are in the bottom half of the income distribution, we need to make a number of improvements in our educational system, starting with much more widespread publicly financed pre-school that goes down below ages four and age three for at least the poverty population, which is not getting the enrichment they need at home. We need to have preschool that goes back even at ages below three.
And moving up the ladder to elementary and secondary education, in our country, unlike others, we have that level of education financed by local property taxes, so rich suburbs are able to provide their children with lavish facilities and opportunities that don’t exist for poor children in central cities. We need to switch to financing education at the level of states rather than individual localities.
For college, where we have enormous tuition inflation and huge amounts of college debt, we need at least to transform that debt so that it’s repaid contingent on people’s income, so-called income-continent loans so that if you suffer a spell of unemployment or if you decide to go into a low-paying social service kind of occupation, you don’t have to repay the full amount of your loan. All of those improvements would very gradually help to overcome the barriers to higher productivity that much of the population faces.
We’re in the middle of this presidential election and there’s a lot of claims about what policy can do for the economy, for economic growth, for improving the lot of the middle class — claims that if we had the right tax policy or immigration policy or public investment policy, we can at least grow as fast in the future as we have in the past, if not considerably faster. What expectations should our politicians be setting for Americans?
Well, we know that cutting taxes on the rich has been achieved in the Reagan administration and in the second Bush administration and it didn’t do anything to boost growth. The underpinnings of growth are the things that I talked about before, the opportunities for inventions and the opportunities to diffuse those inventions throughout society.
The Republican candidates almost uniformly are recommending very drastic reductions in tax rates paid at the top; in some cases going to an outright flat tax, in some cases dropping the top tax rate from 45% to 25% or 20%. We had much faster growth in the first two decades of the postwar era, back in the ’50s and ’60s when the top tax rates were much higher than they are today.
So I think there’s room to raise government revenue by having a special surtax or incomes above a million and another tax bracket for incomes above 10 million. Those people will continue to do what they’re doing even if we take a little bit more of it and redistribute it into society in such forms as publicly financed preschool.
Even if we were to do your entire list of reforms dealing with inequality, demographics, hours worked and everything, we’re still not talking a dramatically different outcome than sort of your baseline forecast.
That’s right. And the forces that are raising incomes in the top 1% are very strong. The forces that are holding down wages for the middle and the bottom are very strong. They’re due to the fact that so many of the remaining jobs are service occupations that do not require much training and do not differentiate workers from each other.
We do have labor shortages in some areas, but many of them don’t require a lot of education. We have a shortage of long-distance truck drivers. We have a shortage of some kinds of construction workers. Solving that would be best handled by coming closer to the Germans’ system of apprenticeships. Most American parents are expecting their children to go to college. And for some of them, they may have a better future if they go into blue collar, skilled work. We need to facilitate that through apprenticeships that are paid for by a combination of government and private business.
Comments are closed.
1789 Massachusetts Avenue, NW, Washington, DC 20036
© 2018 American Enterprise Institute