Robert Gordon’s Rise and Fall of American Growth offers a vast narrative, encompassing some 150 years of us economic history since 1870, with prospective views up to 2040 or so.footnote1 The conclusion is that low per capita growth is here to stay, with dire consequences for worsening inequalities in income distribution and other headwinds. It demands to be taken seriously. It should be noted at the outset that the very long-run view that Gordon provides may serve to underestimate the impact of the 2008 financial crisis—indeed, finance is nearly absent from the book. However, the trend decline in productivity growth that it detects from the late 1970s is concomitant with what has been labelled ‘the financialization of the firm’, a transformation in corporate governance that is itself linked to financial liberalization and globalization. Furthermore, the long-run slowdown in productivity since the late 1970s or early 1980s is not specific to the us; it plagues all large advanced economies.

This poses a question for Gordon, because his core argument is that technological inventions are the source of growth; other factors of production contribute as inputs, induced by the pace of innovation. It is clear that Gordon does not buy the neo-classical production function. Nor does he agree with the ‘secular stagnation’ hypothesis, which he sees as just a prolongation of the standard Keynesian argument about demand deficiency. Since the output gap in the us has been closed, low potential growth is a supply phenomenon; however, if one rejects the neo-classical notion of the independence of the long run and the short run, the interaction between demand and supply is much more important. Gordon would agree with this. His richly detailed descriptions of innovatory techniques place great emphasis on the transformations of lifestyles and social relationships that they bring about. The America of 1870 was largely rural and got by with candlelight, horse-power and the steam engine. Within a few decades, in his account, electrification, the internal combustion engine, wireless transmission, the telephone and the phonograph had already begun to revolutionize the social existence of a young, fast-growing population. By 1950 further innovations—electric-powered machine tools, assembly-line production, aviation, white goods, air-conditioning, inter-state highways, petro-chemicals, supermarkets, television—had altered it again.

However, Gordon is inclined to view this as a one-way process from invention to social transformation. The institutional framework that gives innovations social acceptance and enables their diffusion tends to be underestimated. His book also raises the question of American ‘uniqueness’, since there have been other success stories; Germany and Japan also saw dramatic growth, starting around 1870. Another issue is the role of finance—and the subsequent drastic change in corporate management, from the 1980s onwards—in dampening the productive investment rate, apart from a short period in the second half of the 1990s. Last but not least, integrating the environment into the economy, ignored in the book, raises two further questions: could ecology be the source of a fifth industrial revolution? And if sustainable growth is to be taken seriously, is gdp the proper indicator of development?

The greater part of The Rise and Fall of American Growth is the work of a historian. It provides a narrative explanation of the ways in which a vast range of innovations changed the fabric of American society, nurturing the growth process. The main message is that innovation was a strong and sustained impetus for growth, starting after the Civil War. Indeed, in this account wars have generally been beneficial events for the us, triggering long-lasting surges in productivity. After World War I, growth in total factor productivity (tfp) accelerated in the 1920s and 30s, despite the Great Depression; World War II launched what would be the apex of us manufacturing’s domination, over three decades. All in all, there were fifty years of robust productivity growth; Gordon calls them the ‘productivity miracle’. Then, from the 1970s, the sun began to set, only interrupted by the brief revival of the Information and Communications Technology (ict) revolution in the late 1990s. Abstracting this respite, tfp growth decelerated over three decades, down to the present. Is this trend going to continue—and if so, why? The third part of Rise and Fall tries to answer these hard questions.

Gordon makes no predictions for the distant future, but he argues that low growth rates will provide our horizon for the next generation. However, he is keen to assert that the exhaustion of innovation is not the cause. Rather, the explanation for low growth lies in four ‘headwinds’: demographics, inequality, education and government debt. Here the reader grows frustrated: there is no question that these are impediments to growth. But the four headwinds are treated as exogenous phenomena, with very little discussion about why they have developed from the 1980s on. Demographics can to some extent be taken as exogenous. But why has inequality taken root at precisely this stage? Why has elementary and high-school education deteriorated? Why is public debt a problem, but not total debt? After all, private debt was the spur for the massive wave of financial speculation that led to the crisis. But why has private debt increased so much, for so long?

Lacking here, clearly, is an analysis of the institutional mechanisms of macroeconomic regulation and how they have gone astray—and, with them, feedback on productivity. Rise and Fall does hint at these relationships. Gordon invokes the decline in unionization to explain the decoupling of median real-wage growth from productivity gains. But this does not go far enough. Why did the decline in unionization take place? Gordon does not examine whether it was one of the consequences of the sea change in corporate governance, from a managerial to a shareholder-value model. Nor does he explore the impact of these new incentives on rising indebtedness or on the decoupling of financial strategies from real productive-capital accumulation. As for education: the consequences of deteriorating schools are clear in Gordon’s account—the ict revolution has not had the hoped-for impact on productivity, because it has not been accompanied by the necessary increase in workers’ skills and improvement in organizational processes. But what are the causes of their deterioration in the first place, despite the technological innovations being already at hand?

If we consider the most recent period, the oecd has estimated the decline in us potential growth between 2003–07 and 2013–17 at 0.7 points. The largest share of the decline (0.3 points) is due to the reduced contribution of capital stock; lower contributions from labour and tfp account for 0.2 points each. The future of r&d spending—the matrix for tfp growth—remains uncertain, as its recovery from the financial crisis has been slower than before. How do Gordon’s long-run processes of growth and innovation compare to other views? Angus Maddison made it his life’s work to elaborate and study the data on the very long-run development of capitalism. He acknowledged that, from 1820 to 2000, worldwide growth had been very unequal. There have been only two leaders: the uk, from 1820 to 1913, and the us, from World War I to the present. However, the acceleration of gdp per capita growth started at around the same time in Germany and Japan as in the us. In all three cases, the initial condition was a sweeping political event that unified the country: the Civil War in the us, the creation of the German Empire after the 1870 victory against France and the eradication of feudalism by the Meiji Revolution in Japan.