ECONOMIC FOCUS
Birds of a feather

If EMU encourages greater specialisation in particular industries, economic management will be more difficult
 
More about
The Euro
 
Resources
Search
archive
THE economic case for the euro rests largely on the idea that it will hasten economic integration. This in turn will promote competition, leading to higher productivity and incomes. The case against rests largely on the idea that “asymmetric shocks”—disturbances that affect some countries much more than others—will be harder to deal with in the absence of national monetary policies. But what if it turns out that the more successful the euro is in promoting integration, the more susceptible the single-currency area becomes to asymmetric shocks?

At first sight, this seems unlikely. In a more integrated Europe, surely, countries will be more alike. Maybe not. There are reasons for thinking that integration across a region makes sub-regions more specialised—that is, less alike. If this turns out to be true for Europe, the gains from integration in good times may be great, but the risks in bad times will be bigger too.

Economic theory suggests several connections between integration and regional specialisation. In standard “neoclassical” growth theory, integration makes countries more similar: once capital and labour are free to move, the mix in which the two are used tends to converge internationally. Traditional trade theory complements this finding. It argues that a country’s incomes and industrial structure reflect its endowment of labour, capital and other resources. As countries become richer, the accumulation of physical and human capital comes to outweigh resources such as land and climate: endowments become more similar, and so do patterns of output.

More modern theories emphasise different links. “New” growth theory is interested in externalities. The return to a unit of extra investment may be higher in a country that already has a lot of capital: the existing stock confers an external benefit on the new addition. If this effect is powerful, integration may promote divergence rather than convergence: investment will move all the faster to areas that already have lots of capital. “New” trade theory complements this finding (just as traditional trade theory sat happily alongside traditional growth theory). Externalities in production will cause high-technology, high-wage industries to form clusters. The more integrated these areas are through trade, one with another, the greater the tendency for clustering to take place. Again, integration makes economies less alike.
 

At your service
The evidence on all this is fiercely contested. There is no consensus on whether the new theories of trade and growth are closer to being true than are the current versions of their predecessors. So far as specialisation is concerned, it is not even clear exactly what the term means, let alone how it should be measured. A lot depends on the level of magnification—that is, on whether you compare broad sectors (manufacturing, farming) or sub-sectors, or even particular industries.

It would be hard to deny, though, that the economy of the United States is more regionally specialised than the euro zone. Big industries do seem to have clustered in certain areas: classically, car making in Detroit, software in Silicon Valley, movies in Hollywood. Europe has clusters here and there, but they are less pronounced. In America, asymmetric shocks have often pushed regional economic cycles out of synch, but fiscal transfers and labour mobility have eased the adjustment. Europe may find that integration causes sharper regional cycles—but since its central budget is small and its labour mobility low, its ability to cope may be poorer.

A recent study by the OECD reports some further evidence. First, it looks at a measure of output divergence (which asks, in effect, what proportion of each country’s output would need to switch sectors for the country’s pattern of GDP to conform to the euro-area average). The results suggest that specialisation has increased since 1980 in most of the countries examined. (It also suggests that Italy is the most “European” economy, and Belgium the least.) But the findings are inconclusive. Using a similar measure, but looking at 60 sub-national regions within the euro area rather than at countries, and using employment in place of output, the researchers find the opposite: specialisation has been falling, albeit gently, since the mid-1980s.

It is possible, indeed likely, that “new” and “neoclassical” effects are going on at the same time. A paper cited by the OECD shows that specialisation in America has traced a bell-curve over time. Regional specialisation increased during the 19th and early 20th centuries as manufacturing, prone, in the “new” view, to form clusters, gained ground over agriculture. Later, this trend reversed and specialisation fell—notwithstanding the fact that America is today more specialised than Europe.

What happened? “Neoclassical” factors—trade and comparative advantage—caused a long-term shift in industrial structure: the economy moved from manufacturing to services. Many services have to be produced where they are consumed, which means less regional specialisation. This was enough to outweigh any tendency within manufacturing for specialisation to increase. (Note, however, that specialisation even within American manufacturing has declined in the second half of this century: the tendency to cluster may be weaker than before.)

Europe can draw comfort from this. Its shift from manufacturing to services lags far behind America’s. Integration, other things equal, will speed it up. This should mitigate any tendency towards clustering. For now at least, service economies look more alike than manufacturing economies, and should prove less susceptible to shocking developments.

 
Paul Krugman first drew attention to this question in his book “Geography and Trade” (1991). The OECD study is “EMU: Facts, Challenges and Policies” (1999). The study on America is “Economic Integration and Convergence: US Regions, 1840-1987” by Sukkoo Kim, NBER Working Paper 6335 (www.nber.org).