Human existence changed irreversibly after the invention of indoor plumbing and the municipal supply of water and sewage. The advent of electricity also changed life as we know it, and so did automobiles, the telephone, penicillin, pasteurization, the polio vaccine, and much more. In his book, The Rise and Fall of American Growth; The US Standard of Living Since the Civil War (Princeton University Press. 2016), Robert Gordon distills many of these innovations, and presents a lucid history of their economic impact on living standards in the United States over the last century and half.

This book has worked its way into common conversation. There is merit to that attention, as many readers will love the first eighty percent of this book. However, I expect most will not like the last twenty percent. The latter is almost a different book, as it covers contemporary events, policy debates, and unresolvable questions. The common reader may not appreciate how it has sparked professional disagreement. Today’s column provides a window into that disagreement.

Innovation and GDP

The marriage of ‘innovation’ and ‘history’ and ‘economic’ might sound like an enormous agenda, and it is. It also might sound like too much for one book to discuss, but Gordon organizes copious details about history into very readable chunks, and the book moves forward at an engaging pace. To his credit, Gordon also displays vociferous self-awareness about what standard economic measurement can do and cannot do well. He displays the standard government statistics, but does not stop there. He draws widely from other data and quotes, and he tells a rich historical story.

One proposition motivates Gordon’s inquiry: Innovations only come along once in the history of a country’s growth, so each unique innovation changes Gross Domestic Product (GDP) only once. GDP is a measure of the total flow of economic activity in the domestic economy in a given year. Gordon is a master of GDP measurement, and, accordingly, he sets before himself a tall task: to trace the links between improvement in many different innovations and measured growth in GDP.

Gordon doggedly pursues this agenda through a ‘greatest hits’ of post-civil war and twentieth century innovation – electricity, telephony, rail shipping, sanitation, the automobile, mass market medicine, housing, and television. The book delightfully wanders into unobvious territory too, devoting many pages to, for example, the rise of frozen food, the use of time-saving devices in households, the importance of air conditioning for the south, and the spread of catalogs into rural America.

No book can do everything, and that poses a problem for the scope of Gordon’s agenda. In practice the book says much more about the consequences of innovation for growth and less about its origins. There also cannot be a clear boundary between the ‘greatest hits’ and the innovations not on Gordon’s radar. Most of the included topics are delightful, but some of the missing discussions are puzzling. It is an inescapable tension.

Gordon’s treatment of historical communications technology illustrate this tension. Gordon has a great discussion on the invention of the telephone and its spread. Yet, he does not delve deeply into innovation at the Bell System except with token stories about the invention of the transistor. Despite its regulatory obstinacy, the Bell System could be innovative. It deployed the electro-mechanical switch and the digital switch, for example, which were major technical achievements that linked the country. Did Gordon not devote time to the topic because he did not think it mattered, or because he did not look into it?

Later he acknowledges the contribution of cell phones and smart phones, though, for my taste, I would have liked more on the value of mobility. Does he think mobility is of minor importance to living standards or are its effects simply unmeasured? Telephone booths became obsolete, and a large fraction of the landline business for homes went away. Sure, another game app won’t change the world, but the platform has taken over leisure time, altered many business processes, and its consequences are far from finished. Isn’t that a big deal for living standards? How would a reader know? Gordon is silent.

More recent times

In the last twenty percent of the book Gordon turns attention to contemporary events. What a record: the US experienced slow growth in the 1970s and 80s until the Internet boom between 1995 and 2004. It was slow after that, largely due to a financial meltdown. He foresees more slow growth, a forecast that has generated attention from the punditry.

A diligent reader of this book will notice a change in this final section. The writer who so celebrated historical innovations transforms into someone else. Gordon has upbeat observations to make about airline deregulation, the PC, the CT scanner, and electronic commerce, but his mood becomes mostly dismal. He does not foresee many major innovations on the horizon, and he illustrates the argument with extended discussions about medicine and information technology.

The tension about scope arises again. Consider Gordon’s discussion of the Internet. After noting the value of mobile telephony and the PC, he remains skeptical of exaggerated claims for IT (fine with me). He cares for nothing less than an important innovation. Taken on his own terms, Gordon sets a high bar: he wants innovation to simultaneously change consumption, alter the allocation of leisure time, and upend standard business processes across the entire economy.

Given his outlook, Gordon should love the rise of the commercial Internet, which did all that for a sustained period, and recently. Gordon lauds some of it, such as Amazon’s reach, but then displays no serious appreciation for how much business processes changed, nor how it supported the expansion of world trade. It is simply impossible to imagine how large scale outsourcing and offshoring – as we see today – could take place without digital support (or, relatedly, how hundreds of millions of Asian workers would have taken a step above abject poverty). He then provides a few reasons why any major innovative information technology is unlikely to arise tomorrow or contribute to large economic growth.

Call me unconvinced. Aside from being a killjoy, the discussion is incomplete and unbalanced. Gordon poses a question that requires a thorough and definitive analysis of why the market for information technology does or does not contain the capacity to renew itself. A thorough analysis must grapple with the many potential pathways through which radical innovation emerges today.

To say it blithely, the history of technology commercialization teaches one to make economic forecasts with severe caution, cognizant of the numerous possible ways market processes leads to the accumulation of advance. General purpose technical advance often appears unexpectedly. Yet, Gordon seems to ignore this lesson from the first eighty percent of his book, and writes as if he can forecast today with certainty, and in under ten thousand words. Seemingly unaware of the inconsistency, Gordon pronounces his skepticism with confidence.

The discussion about medical technology contains the same features. Gordon highlights some important issues, but the discussion is incomplete without a survey of advances in many areas of medicine, such as cancer treatments, medical devices, and bio-engineering. I expect any person involved in the development of precision medicine today to simply throw this book at the wall. More to the point, assessing the capacity of the medical system to yield radical improvements in living standards requires a nuanced and thorough discussion. Not here.

By the final chapter the ebullient economic historian disappears, replaced by a downbeat macroeconomic forecaster. The last chapter enumerates a number of challenging societal and economic factors that make growth difficult – which Gordon calls headwinds. These headwinds are the pessimistic fare of macroeconomic policy debate, such as inequality in consumption and underfunded entitlements. Gordon argues that these headwinds will overwhelm the impact of any but major innovation.

The dismal outlook rests on the forecast that few major innovations will appear on the horizon. If a reader regards this forecast as unsupported, then this conclusion seems hasty and unwarranted.

Hard question, challenging answers

I respect Robert Gordon for what he achieved in this book. Gordon has assembled an enormous amount of historical evidence.

He also framed a provocative argument about contemporary experience. If prior generations picked the low-hanging fruit, do modern innovators face a thornier and costlier set of remaining innovative challenges? What changes would raise the likelihood of major innovations from the US university system, US corporate labs, the Department of Defense, and Silicon Valley? Is GDP the best way to measure technical progress in the modern era?

I think most readers will take issue with how Gordon mixes the historical and contemporary. Today’s headwinds do not look any worse than yesterday’s, which, after all, included primitive scientific instrumentation, a couple bank panics, a few minor and major world wars, and the great depression, not to mention the fires that burnt down the cities of Chicago and San Francisco, to name just a few challenges. If prior generations overcame their headwinds, why can’t the present generation?

The pessimism also stands uneasily against the unique status of the US on the global stage. It is still the global mecca for innovators today, and governments around the globe still admire the US innovative engine.

More bluntly, Gordon’s closing pessimism stands at an uneasy place next to many of his uplifting historical anecdotes. I wish Gordon would lighten up, and cut the present generation of innovators some slack.

Copyright held by IEEE. To view the printed essay, click here.

Tagged with →  

2 Responses to Economic Growth from Technical Advance

  1. richardhserlin says:

    “One proposition motivates Gordon’s inquiry: Innovations only come along once in the history of a country’s growth, so each unique innovation changes Gross Domestic Product (GDP) only once.”

    This is to an extent, and in some ways, true, and to an extent, and in some ways, not true. Consider electricity. It led to assembly line production “one time”, but the assembly lines kept getting better for over a century and running. And it much later led to ultra powerful computers which would otherwise not be possible, which continue to improve. The full implementation of an innovation, or even the realization of most of the potential benefit, can be a process taking centuries or more.

    I’ll have a lot more to say on Gordon as time permits, even if the super-fast blogging window closes.

Leave a Reply

%d bloggers like this: