A review of Robert Gordon’s ‘Rise and Fall of American Growth’

| March 11, 2016

The economist Robert Gordon has recently written The Rise and Fall of American Growth, and its findings apply with some adjustments to Australia too. Ian McAuley says anyone who believes that changes in “productivity” and “efficiency” can be captured in a few simple metrics would do well to read at least a few chapters of Gordon’s work.

In 1956 the flight from Adelaide to Melbourne in a DC4 took 75 minutes gate-to-gate. Now in a Boeing 737 it takes 50 minutes. But add in an extra twenty minutes for security, and the net gain over sixty years is five minutes at best. And the DC4s were comfortable, with seating that would be considered business-class standard these days.

Of course we now have much easier reservation and payment systems, air travel is safer, and in real terms air fares are much less than in 1956.

If, however, we make a further sixty-year comparison, going back to 1896, Lawrence Hargrave was still experimenting with box kites and the Wright Brothers’ first flight was seven years into the future. What happened to civil aviation over the following two generations was truly amazing compared with anything we might have seen in our lifetimes.

As long as I remember, people have been talking about our living through a time of unprecedented change and marvellous technological progress. Anyone who thinks otherwise, or who questions the benefits of technological progress, risks at best the scorn of their children and at worst consignment to the Luddites’ wing of the nearest nursing home.

But is our pace of change really all that fast?

The economist Robert Gordon of Northwestern University uses air travel, as well as almost every other good or service we use in our daily lives, to trace the path of technological progress since 1870 – the dawn of what historians often call the “second industrial revolution”, and the start of reasonably reliable statistical data.

Over the period 1870 to 1940 people living in ‘developed’ countries (his focus is on the USA) enjoyed tremendous improvements in living standards. Electric lighting, telephones, electric appliances, refrigeration, clean water, sewerage systems and antibiotics are among the improvements he lists. He reminds us that when cars and trucks displaced horses, land previously devoted to growing horse fodder could be turned to food production, and that clean streets suddenly became a reality.

By comparison subsequent improvements, while welcome, have been incremental. Gordon does not dismiss the idea of a “third industrial revolution” associated with ICT, but he points out that compared with earlier periods of technological progress its benefits have been far less profound in improving our quality of life.

Gordon would argue that my smartphone, with its capacity to update the value of my share portfolio in real time and to tell me how far it is to a coffee shop is marvellous, but in terms of real benefits (“consumer surplus” in economists’ terms), its contribution to my wellbeing is trivial compared to the very first phone that connected my grandparents, living in outback Australia, to the rest of the world. My car with cruise control, anti-lock braking, GPS, airbags, a 15,000 km service interval and air-conditioning is certainly an improvement on my grandfather’s first car, a 1924 Hupmobile, but it would be hard to challenge the conclusion that my grandfather over his lifetime enjoyed much greater improvements in personal mobility than I have.

We can get a quick understanding of Gordon’s ideas from his 2013 TED Talks presentation “The death of innovation, the end of growth”.  The title itself summarises his ideas.

From a twelve-minute presentation, however, it’s too easy to see Gordon as yet another member of the chorus that since the time of Marx has been predicting the imminent collapse of capitalism. (Marx himself was circumspect about timing.)

The depth and detail of Gordon’s work is in his book The Rise and Fall of American Growth: The U.S. Standard of Living Since the Civil War (Princeton University Press 2016), with 760 pages of rich analysis, covering everything from details about power-to-weight measures for electric motors through to changes in life expectancy.

Cover of 'The Rise and Fall of American Growth' by Robert J. Gordon

He uses, but does not rely on, traditional measures such as GDP and inflation-adjusted time series to derive movements in the real prices of goods and services, but he goes into great depth to show the limitation of such metrics. Gordon is well-recognised for his work on index numbers, and those who have studied statistics at university will recall that deflators such as the CPI, when used to compare “real prices” over time, tend to understate improvements in standards. We can perhaps talk about the change in the real price of a banana purchased in 1924 and one purchased in 2016 (assuming there has been no genetic improvement), but when we come to a 1924 car and a 2016 car the comparison is next to meaningless.

Statistical agencies try to apply what is known as “hedonic adjustment” to products like cars and computers to cover quality improvement, but such adjustments are problematic and somewhat arbitrary. (My new iPad has a video editing app that I don’t want and will never use, but has the ABS, in compiling the CPI, assumed it is of some value to me?) Anyone who believes that changes in “productivity” and “efficiency” can be captured in a few simple metrics would do well to read at least a few chapters of Gordon’s work.

Gordon, like other researchers looking at long time series, finds an inverted “U” shape in official measures of productivity (as measured by output per hour). From 1870 to 1920 US productivity rose at 1.8 per cent a year, from 1920 to 1970 at 2.8 per cent a year (in spite of the depression and war), and since then at 1.6 per cent a year. These differences are significant.

But his main point is that, because of the enormity of the life-changing developments up to around 1970, those official metrics significantly understate the pace of improvement, and therefore understate the extent to which productivity has fallen since 1970.

Applying his work to Australia we would need to make some adjustments, because our living standards up to 1940 did not improve as quickly as America’s, and over the last 40 years we have enjoyed at least three commodity booms. Otherwise his analysis applies to us.

Many economists, brought up in the Pareto school which rejects any idea that we can compare what economists call “interpersonal utility”, may dismiss Gordon’s work. The Pareto purist makes the point that because I cannot get inside my grandfather’s mind, I cannot possibly make any comparison between his and my enjoyment of improved material living standards. In fact the Pareto purist would reject any generalisation that civil aviation has improved over the last sixty years, because there may be some people who value the comfortable seat of the DC4 more highly than all the other improvements in aviation.

Such nitpicking should rightly be confined to schools of philosophy rather than to practical economics, for they thwart any endeavour to make any aggregate measures of progress (or otherwise) in wellbeing, such as the Herald/Age – Lateral Economics index. Gordon is not bound by Pareto’s limitations, and his work makes a convincing case that in prosperous countries like the USA and Australia improvements in material wellbeing are well into the realm of diminishing returns.

His prognosis goes against the popular view that we’re on the cusp of another set of disruptive changes. At the turn of last century there was a lot in the technological pipeline – electricity, radio, internal combustion engines – but there is not so much now. Our much celebrated ICT revolution is simply about wringing out incremental improvements from the 1947 invention of the transistor. While there are advances in medical technology, they’re nothing in comparison with past improvements in public health and the overcoming of most forms of infection. Three dimensional printing will find new applications, but it’s hard to imagine they will be as transformational as Henry Ford’s production line.

Besides a much smaller technological pipeline, he identifies four “headwinds” to further improvements in material living standards – demography, education, fiscal constraints and inequality. On all four dimensions Australia shares the same problems as the USA, even if less intensely. Our population is ageing, but we still have strong immigration. Our education standards are slipping in comparison with other ‘developed’ countries, but we don’t have the entrenched problems of the American school system which relies on local taxes to fund school education. More importantly neither country is making adequate public investment in education. Our public debt is lower than America’s, but we share with Americans an aversion to public expenditure, even though well-directed public expenditure is important for a country’s economic health. And our inequality, while not as bad as America’s, is growing strongly.

It’s on inequality that he makes his strongest point. Since 1973, all the gains in real income in the USA have gone to the ten percent of the population with the highest income, while the other 90 percent have gone backwards. He makes the point that should be obvious that there is little point in economic growth if nine in ten people enjoy no benefit from it, or have to be compensated with transfer payments. The Pareto economist may say this doesn’t matter: just so long as some people benefit we can celebrate economic progress. Gordon’s prescriptions, involving high minimum wages, progressive taxation and public investment in education is a strong challenge to the Pareto orthodoxy that has been so dominant in American and Australian public policy.

Some commentators on Gordon’s work, such as the reviewer for The Economist, believe he downplays the possible impact of ICT on the physical economy. They claim his idea that the “third industrial revolution” is past its peak is premature. Indeed Gordon’s own research shows long lags between invention and economy-wide transformational adaptations. (His work is reminiscent of the work of the Soviet economist Nikolai Kondratiev who found half-century long waves of technology-based economic growth in capitalist economies.)

My own comment is that he tends to disregard innovations that emerge away from the world of scientific and technology breakthroughs. The ISO shipping container is one such example that gets little coverage in his work. Its development and adoption relied on cooperation between worldwide standards and regulatory agencies. The humble shipping container should remind us of the importance of international cooperation in areas such as combatting global warming, dealing with terrorism and avoiding costly arms races. A retreat to economic and security isolationism would be at least as strong as the other headwinds he identifies.