The Wall Street Journal

ACCORDING to their bent, journalists and politicians like either to enthuse or to rant about "globalization". Some hail the greater ease with which capital and goods can move between countries as a cause of prosperity. Others rail against it, saying that it will impoverish workers in rich countries.

Economic historians, however, are a far less excitable bunch. Globalization, they point out, is nothing new. The half-century or so before the first world war was similar to modern times, in that national economies became more closely linked. Declining transport costs made it cheaper to send goods across oceans. Migrants left Europe for the Americas and Australia in huge numbers. Investment poured from the Old World into the New.

In a paper published last year Jeffrey Williamson of Harvard University stressed another similarity: the convergence of wages among many countries now in the OECD.* Since the 1950s, the amount by which Americans' wages exceed Europeans' has shrunk markedly. In the same way, in the second half of the 19th century, European wage rates caught up on American ones (see chart), although the trend changed after about 1895. Within Europe, some smaller countries closed the gap with Britain, then the continent's leader. In 1870 Swedes' wages were about half of Britons'; by the first world war, they were ahead.

The interwar years, however, were quite different. Not only did countries retreat behind tariff walls and curb immigration, thus reversing international economic integration, but, says Mr. William-son, the convergence of wage rates ceased and, in the late 1930s, was reversed.

Is there a link: does economic integration cause wage rates to converge? In principle, there could be other causes of convergence, such as improved education in lower-wage countries. In another recent paper, in which he draws on research jointly undertaken with other economic historians, Mr. Williamson argues that, at least in the late 19th century, economic integration was the main cause.**

In part, integration happened through increased international trade--as a result, in particular, of a fall in transport costs. According to standard economic theory, this should have had two effects. First, the prices of internationally traded goods should have become more alike in different countries. Second, countries with relatively large amounts of land and little labor (such as America and Australia) should have specialized in producing agricultural goods; more crowded countries (i.e., Europe) should have produced more labor-intensive goods.

This in turn should have increased the demand for relatively abundant factors of production in each country, raising their prices, and reduced the use of relatively scarce factors, lowering theirs. Thus the ratio of wages to land rents in America and Australia should have fallen; in Europe it should have risen. A study by Kevin O'Rourke of University College, Dublin, Alan Taylor of Northwestern University and Mr. Williamson finds that this indeed happened.**

The authors estimate that between 1870 and 1913 the ratio of wages to rents in America fell by half, while Australia's ratio dropped by three-quarters; Britain's and Sweden's more than doubled, while Ireland's more than quintupled. Furthermore, the change in wage-rent ratios was smaller in European countries, such as France and Germany, in which protection remained high.

Standard theory explains much of the convergence of wages and rents between Britain and America, where the gap between national commodity prices narrowed greatly: between 1869-71 and 1911-13, for instance, the amount by which grain prices in Chicago exceeded those in Liverpool fell from 60% to 15%. But the theory is less useful for explaining, say, Sweden's catch-up on Britain: commodity prices in the two countries converged only modestly, perhaps because Sweden increased tariffs in the 1880s.

Go west, young man

Emigration to America, which made labor scarcer at home, explains much of the relative increase in Swedish wages. The same goes for Ireland, where a huge wave of emigration began after the country's famine in the 1840s. With no emigration after 1870, says Mr. Williamson, urban wage rates would have been lower by one-third in Ireland and one-eighth in Sweden in 1910.

By making labor more abundant, migration also held down wages in receiving countries. Had there been no immigration into America after 1870 and had it had no effect on American investment, reckons Mr. Williamson, real wages in American cities would have been 34% higher than they actually were in 1910. But, he says, it is nonsense to assume that the capital stock did not respond. Immigration increased the return on investment; capital chased labor across the Atlantic. The result? Without immigration, wages would still have been higher in 1910--but only by 9%.

That said, Messrs O'Rourke, Taylor and Williamson point out that the type of investment reinforced the effects of migration on wages and rents. In Europe, investment tended to economize on scarce land; in America, the emphasis was on saving labor.

Nonetheless, the prediction of standard theory--that global economic integration may hurt owners of a country's relatively scarce resources (American labor, European land)--is borne out by the evidence of the late 19th century. Today, rich countries' relatively scarce resource is low- skilled workers: there are many of them, but relatively fewer than in emerging nations. They may lose from globalization. Mr. Williamson muses that in the early 20th century such fears contributed to barriers to migration and trade, with dire results. Might it happen again?

* "The Evolution of Global Labor Markets since 1830: Background Evidence and Hypotheses". Explorations in Economic History. 1995

** "Globalization, Convergence, and History". Journal of Economic History. 1996

** "Factor Price Convergence in the Late Nineteenth Century". International Economic Review. 1996

Friday, April 20, 1996

Modern times, old trends