Standard Life Investments

Weekly Economic Briefing

US

Solving the productivity puzzle

Back

We like to think of ours as age of extraordinary technological progress. New and cheaper ways of servicing customers and intermediating production via digitalisation, the advent of the smartphone, and the rapid progress of robotics and artificial intelligence, all appear to be disrupting business models, driving prices down and changing the way we live our lives. The transformation in the way we do business and buy goods and services is also readily apparent on the world’s stock exchanges where Apple, Google, Microsoft, Amazon, Alibaba and Tencent are all among the 10 largest companies in the world by market capitalisation. Yet it is much harder to see in the official statistics, which continue to show labour and total factor productivity growth rates lagging well behind their long-term averages (see Chart 2). What explains this paradox?

Productivity growth in the doldrums The rising cost of new ideas

One theory is that statistical agencies have failed to keep up with these changes and are thus underestimating the increases in real investment, production and the consumption they have generated. Prime examples of the types of activity that may not be adequately captured in GDP include: the explosion of short-term rentals via peer-to-peer websites such Airbnb; the rapid growth of self-employed transportation workers in the so-called gig economy; the blurring of the line between consumption and production as the internet and smartphones have made it easier for individuals to do things like book holidays in their leisure time rather than pay for the service in the market economy; and the ‘free’ provision of apps, internet searches and social media platforms. Moreover, even when statisticians properly estimate the nominal value of such activities, identifying how much of it should be attributed to the volume and how much to the price is immensely difficult, especially when the quality of such goods and services is changing rapidly. This disconnect between what the economy ‘is’ and what we can ‘measure’ at a particular point in time may explain why productivity growth tends to be revised up in the decades after its first release.

However, for all the conceptual challenges of accounting for these types of activity in the national accounts, the empirical evidence suggests that mismeasurement is likely to account for no more than half of the productivity puzzle.  Some of it is quantitatively too small to matter. And some is non-market activity that raises welfare more than it lifts market activity, much as the advent of the radio and television did in the 20th century. In addition, some of the progress we are seeing may only be skin deep. While innovation at the technological frontier is progressing at around its long-term average, the gains are being captured by a smaller proportion of ‘winner-takes-all’ firms, with the diffusion of technological progress through the firm structure now very weak. Moreover, in contrast to the perception that disruptive technologies have become much more pervasive, many industries are becoming less competitive and contestable, with the average age of firms increasing in recent decades. And if that were not enough, the most recent evidence shows that while the frontier is expanding at a healthy pace, the cost of that progress is increasing even more quickly (see Chart 3), reducing the efficiency of extracting new ideas. The upshot is that it will take more of a statistician’s sleight of hand to restore productivity growth to its long-term average.

Jeremy Lawson, Chief Economist, Standard Life Investments