Not only has America’s productivity wonder survived its first recession; it has positively thrived. Output per man-hour in the non-farm business sector rose at an annual rate of 8.6% in the first quarter of this year, its fastest growth in 19 years. Quarterly figures are volatile, yet the year-on-year growth in productivity was also impressive, at 4.2%. This bodes well for America’s future economic growth—but not necessarily for company profits, or for share prices.
Commentators cheered the latest evidence of rapid productivity gains, hoping that it might promise fatter profits ahead. That America’s productivity continued to rise last year, in contrast to previous recessions, seems to confirm that an increase has taken place in trend productivity growth. Still, the latest numbers overstate the underlying trend.
First, the growth in output, and hence productivity, was inflated in the first quarter by a big swing in inventories. Productivity often surges in the first year of a recovery after recession, as firms produce more without needing to hire extra workers. Productivity rose by 4-5% in the first year following both the 1981-82 and the 1990-91 recessions. Firms have actually continued to cut jobs this year, lifting the unemployment rate in April to an eight-year high of 6%. Today’s best guess is that trend productivity growth is around 2-2.5%. That is less than the 3-4% claimed at the height of the new-economy bubble; but still well above the 1.4% average over the two decades to 1995.
A second, more fundamental quibble is that, although profits will certainly rebound this year, as firms continue to trim their costs and revenues rise, in the longer term faster productivity growth does not automatically mean faster profits growth. A new study by Stephen King, chief economist at the HSBC bank, concludes that workers and consumers have received the lion’s share of the productivity gains of the revolution in information technology (IT). Companies have received relatively little reward for their risk-taking.
In the late 1990s it was widely assumed that faster productivity growth would mean higher profits (so justifying higher share prices). Over the previous half-century a strong positive relationship had indeed held between productivity and profits. In the 1990s that relationship broke down. Despite a surge in productivity, national-accounts profits (as opposed to profits reported by companies, a less accurate measure) fell between 1997 and 2000, even before the economy dipped into recession. At the end of 2000 the profits of America’s non-financial firms were no higher in real terms than in 1994, implying a big fall in their share of GDP.
Mr King argues that workers (who are, naturally, also consumers) were virtually the sole beneficiaries of the new economy, in the shape of faster real wage growth. This was partly thanks to a fall in the prices of IT goods that they bought. More important, the same IT that spurred productivity also increased competition more widely across industries, from airlines and banking to insurance and cars, squeezing prices and profits. Information technology reduces barriers to entry, and makes it easier for consumers to compare prices.
What is more, globalisation, itself spurred by information technology, has further trimmed the pricing power of firms. HSBC finds that, in most economies, the correlation between domestic inflation and domestic unit-labour costs has declined over the past 40 years; the correlation between domestic inflation and average OECD inflation has risen. In most countries in the 1990s domestic inflation was more closely correlated with OECD inflation than it was with domestic costs.
The dismal performance of profits should not surprise. As the IMF’s World Economic Outlook last October pointed out, productivity gains from previous technological revolutions, from railways and textiles to electricity and the car, have gone largely to consumers. Each time, a decline in the prices of goods and services has given a big boost to real incomes. Consumers gained from cheaper travel or clothes, but profits disappointed. The difference this time is that new technology has increased competition and squeezed profit margins across the whole economy. None of this lessens the overall benefit of faster productivity growth. But it does lead to some interesting conclusions:
• The profit expectations built into share prices are unrealistic. Even if productivity growth remains robust, long-term profits growth is likely to be weaker than expected, making shares overvalued.
• A lower return on equities means that consumers will need to save more. In the late 1990s consumers spent more and saved less, in the expectation that capital gains would finance their future pensions. If future returns are lower, consumers will need to tighten their belts.
• The dollar might well fall. The expectation of higher returns on capital in America has attracted foreign money and made it easy to finance America’s big current-account deficit in recent years. If actual returns are disappointing, foreign investors will dump the dollar.
• A lower than expected return on capital could also lead to a more prolonged slump in IT investment. That firms have not benefited from their massive investments in the late 1990s raises questions about their capital spending over the coming years—and, in turn, over the sustainability of faster productivity growth.
Perhaps two-fifths of the acceleration in productivity growth between the first and second halves of the 1990s is explained by companies’ increased spending on IT equipment rather than by higher total factor productivity (the efficiency with which both capital and labour are used). Spending on IT has since fallen sharply from a peak in 2000. If it fails to return to its earlier, clipping pace—because firms can see no pay-off—this could dampen future productivity growth.
The good news is that companies still have plenty of scope to boost productivity by reorganising their businesses to use information technology more efficiently, which could yet boost growth in total factor productivity. That theory might soon be put to the test.