Back to
English Articles
Publié le
Lundi 10 Août 2015
Questions about technical progress in the United States and capital accumulation and resource allocation elsewhere should not be confused.
What to Make of Productivity Paradoxes

Technical innovations transform our everyday lives at a seemingly breathless pace. Pundits wonder when human labor will soon be made redundant by robots. Yet productivity growth is slowing down almost everywhere. According to the latest report by the Conference Board, a business research group, overall efficiency in production – a measure of technical progress that economists call total factor productivity – grew at 1.3 percent per year globally between 1999 and 2006, but its pace slowed down to a meager 0.3 percent from 2007 to 2014. Put differently, efficiency in the use of worldwide capital and labor was only 2 percent higher in 2014 than seven years before, whereas it increased by 9.4 percent over the course of the seven preceding years.

These are very important numbers, because total factor productivity is what ultimately determines at what pace income per capita is bound to grow. A developing country can certainly increase its growth potential by piling up more and more capital per worker. But the limit to such a capital accumulation strategy is that it cannot permanently sustain high growth. The more capital there is already in the country, the less effective is the additional investment. Relying on this strategy to foster growth would require devoting an ever-larger share of GDP to investment rather than to current consumption. After the capital stock per worker has stopped growing, efficiency, knowledge and innovation become the only sources of per capita growth.

The Conference Board says that in the early 2000, about half of China's growth came from capital accumulation and the other half from all the other factors, including additions to the labor force, better educational attainment and total factor productivity. But in recent years about 90 percent of growth came from capital accumulation, while total factor productivity reached a near standstill. This is not sustainable development. As forcefully argued by Yu Yongding of the Chinese Academy of Social Sciences and others, the share of investment in GDP is already about 50 percent and it is not imaginable to increase it further. Future growth must rely more on new sources.

The question then becomes, why has total factor productivity growth slowed? Most of the discussion on this question concentrates on the U.S. case. Careful research by John Fernald identifies the reasons why American performance has been disappointing over the last 10 years. First, capital per worker does not increase anymore, which is quite natural for a mature economy. Second, the quality of the labor force measured by the average educational attainment is not progressing fast anymore; in fact, it set to stagnate as the cohorts born in the late 1980s are not better educated than those born 10 years before. Third, the great technology wave that lifted U.S. productivity growth between 1995 and 2005 petered out before the financial crisis erupted, and there are no signs it is being succeeded by a new wave. Lastly, the financial crisis of 2008 has left scars, although unlike in Japan in the 1990s, they do not seem to be very deep.

Of these factors, the most discussed one is the third. Techno-pessimists such as Robert Gordon of Northwestern University argue that the potential of information technologies has been overestimated in comparison to the great inventions of the past, such as the steam engine, running water and electricity, while techno-optimists such as Erik Brynjolfsson and Andrew McAfee of MIT emphasize the transformational potential of artificial intelligence. Both sides claim the evidence is on their side, but the truth is that the hard data support pessimism while anecdotal evidence suggests that the optimists are right. Some observers also point out that the impact on GDP of new digital innovations is likely to be underestimated because the exchange of service for data is a modern form of barter that statisticians are at pains measuring.

What is true of the United States does not necessarily apply to the other economies, however. America is the frontier economy of the world, the place where most radical innovations emanate from. With a few exceptions, most advanced economies lag behind in the adoption of these innovations. This simple fact has a profound implication: a slowdown on the frontier must impact the United States much more than other, less advanced countries, where there is room for continued catching up simply through adopting innovations already in use in America.

[...] Lire la suite sur le site de Caixin Online


Jean Pisani-Ferry
Anciens auteurs de France Stratégie
Tous nos travaux sur  :