Fra arkivet: Avindustrialisering er bra

Deindustrialisation is often perceived as evidence of economic decline. On the contrary, it is a natural consequence of economic progress.
ECONOMIC FOCUS (The Economist, 26.april 1997)

Phrases such as "hollowing out" and "deindustrialisation" continue to strike fear into the hearts of many workers, manufacturers and politicians in rich economies. And yet the reason for this phenomenon is widely misconstrued. The relative decline in manufacturing jobs in rich countries has coincided with a rapid increase in manufactured exports from countries such as China and Brazil, prompting many rich-country voters to assume that these upstarts are stealing their jobs, and that the solution lies in job subsidies and trade barriers. Really?

The trend is clear; its causes less so. Two explanations are commonly offered for deindustrialisation. One is that, as they become richer, consumers want to buy relatively fewer manufactured goods and relatively more services. The other pays less heed to a shift in consumer preferences and more to the alleged migration of manufacturing jobs from rich countries to poorer ones. But in a careful new analysis of the numbers*, the IMF suggests this week that neither explanation captures what is really going on. A bigger cause of change by far is that productivity is growing much faster in manufacturing than in services.

Manufacturing's share of GDP measured in current prices has fallen, suggesting that deindustrialisation may indeed reflect a shift in spending from goods to services. But the rise in the nominal value of services in GDP reflects a rise in the relative price of services. In constant prices the share of manufacturing output turns out to have remained broadly stable over the past three decades in the rich economies as a whole.

The decline in manufacturing jobs mainly reflects sectoral differences in productivity growth. In 1960-94, service-sector output and manufacturing output grew at roughly the same pace in rich economies. Productivity in manufacturing, however, rose more than twice as fast as in services. This naturally shifted employment from the more productive manufacturing sector, where fewer workers are needed to produce a given increase in output, to service industries where more workers are needed. Productivity improvements in farming caused the same thing to happen in agriculture over the past century. Having made up 50% of all American jobs in 1860 farming now employs only 3%.

As a country gets richer, it is inevitable that a smaller proportion of workers will be needed by manufacturing. This trend, which began in America and then spread to Europe and Japan, is now visible in the Asian tigers. The share of manufacturing jobs has been falling in Hong Kong since the 1970s, and in Singapore, South Korea and Taiwan since the 1980s.

Deindustrialisation causes problems in economies unable to absorb the workers released by manufacturing. But those who would tackle this by subsidies or trade barriers are missing the point. As manufacturing continues to shrink in an economy, overall growth will increasingly depend on boosting productivity in services. Policy should therefore focus on removing obstacles (such as trade barriers and regulation) to such productivity growth, and creating a labour market in which workers can move freely from factory employment to services. Protection and subsidies push just the wrong way.

*World Economic Outlook, April 1997. This subject is treated in more detail in "Deindustrialization: Causes and Implications", by Robert Rowthorn and Ramana Ramaswamy, IMF Working Paper, April 1997.

Se også Polecon om landbruk.




La verket dø: Færre og mer lønnsom fabrikker, færre og mer produktive arbeidsplasser.