Economic globalisation is the product of two distinct, but mutually reinforcing, forces: reductions in the cost of transport and communications; and liberalisation of barriers to movement of goods, services, capital and labour.

At the limit, a globalised economy would be one in which distance made no difference. If one were to focus on policy alone, a globalised economy would be one in which distance mattered, because transport and communications were costly, but national boundaries did not.

Such definitions demonstrate how far the world economy is from being globalised. Both distance and borders matter. That distance matters, even in the newest of new economies, is shown by the importance of clusters, such as Silicon Valley or the New York financial markets. That borders matter is shown by the ability of people working in the US to earn 10 times as much, in real terms, as Mexicans just across the border.

The best way to view globalisation is, in effect, as an ongoing process that is not new, has not progressed very far, and is far from irreversible.

First, globalisation is not new. Global commerce dates from the European voyages of discovery of the 15th and 16th centuries. Integration became more dynamic with the industrial revolution and reached an early peak in the period before the first world war. But integration fell back sharply under the impact of two global conflicts and the great depression. Thereupon a long and uneven recovery began. Over the past half century, global international integration has advanced substantially. The pace accelerated in the last two decades of the 20th century, with a worldwide movement towards liberalisation.

Second, notwithstanding the progressive opening of the world economy over a long historical time-scale, integration has not developed that far, either in relation to the mass of domestically-oriented economic activity or by historical standards.

Between 1950 and 1998, the volume of world production rose more than six-fold, while the volume of world merchandise exports rose over 19-fold. International trade has grown more quickly than output in virtually every year since 1950, while the ratio of world trade to world output has tripled. Even so, gross world trade, at $6,800bn in 1999, remained less than a quarter of world output.

Much more business is still conducted within countries than between them. Moreover, slightly over a half of merchandise trade is conducted within regions of the world rather than between them: in 1998, 31 per cent of world trade was within western Europe, 11 per cent was within Asia, and 7 per cent was within North America. Western European countries did 70 per cent of their trade with one another; for Asian countries, this proportion was 45 per cent; and, for the countries of North America, it was 38 per cent.

Notwithstanding the rapid growth of world trade over the past half-century, many countries are not much more open to trade today than before the first world war.

In current prices, the ratio of total trade to gross domestic product was higher for all members of the Group of Seven leading high-income countries in the mid-1990s than in 1910, except for Japan. But only in the US and Canada had the ratio more than doubled, from 11 to 24 per cent and from 30 to 71 per cent, respectively.

Capital market integration was also highly advanced in the late 19th century, before collapsing between the two world wars. The capital outflow from the UK averaged just under 5 per cent of GDP between 1870 and 1913 - far larger than the ratio for Japan, the world's largest capital exporter, today. The correlation between domestic investment and savings - a measure of self-sufficiency in capital - was lower between 1880 and 1910 than in any subsequent period.

What has changed is the composition of capital flows. Bonds dominated a century ago, but now stocks and bonds are of roughly equal importance.

Portfolio flows were more important than foreign direct investment before 1914, but FDI is now much more significant than portfolio flows. Before 1914, most of the capital outflow was long-term, while today there is ahuge volume of short-term activity, including currency trading. Finally, before 1914, FDI was largely undertaken by free-standing companies, particularly in mining and railways, while today multi-national companies predominate, with much investment in manufacturing and services.

While there has been substantial liberalisation of trade and capital flows in recent decades, the movement of people is far more restricted than in the 19th century. Almost 100m people migrated between continents during that century. In the 1890s, the inflow of people into the US was equal to 9 per cent of the initial population - equivalent to an inflow of 25m today. In the 1990s, the US was the only significant country with substantial immigration, but it was only equal to 4 per cent of its initial population over the decade.

Finally, the ups and downs of integration over the past century indicate that globalisation is far from irreversible. Technology has consistently lowered the costs of transport and communications, thereby making distance less important. But policy has not worked as consistently to make national borders less significant.

Thus, costs of transport and communications fell throughout the 19th and 20th centuries. The 19th century saw the arrival of steam ships and railways. The 20th saw air transport and containers. The first transatlantic cable was laid in 1866. By the turn of the century, the world was cabled. This reduced communication time from months to minutes. Then, in the 20th century, there arrived intercontinental telephony, radio, television, satellites, mobile phones and, finally, the internet.

Meanwhile, the world saw, over the past century, first a return to protectionism, the collapse of the gold standard, extensive controls over capital flows, the introduction of tight curbs on migration and the experiment of communist countries and most of the developing world with self-sufficiency.

Then, from the late 1940s, western Europe and North America began to liberalise trade and, later, capital flows. A few developing countries also began to rely on trade for economic development in the 1950s and 1960s. But it was only in the 1980s and 1990s, particularly after the fall of Soviet communism, that liberalisation of trade and capital movements became general.

It is, in short, a confluence of technological and policy changes that creates integration. Governments have chosen to integrate because experience has shown that self-sufficiency is a sure route to impoverishment. Yet, however difficult it may be, it is not impossible for a modern state to restrict international economic integration. Given the backlash against globalisation, such moves cannot even be ruled out.

Globalisation is an opportunity, not a threat. But it is also a choice, not a destiny. Whether the world seizes the opportunity for ever greater integration depends on political decisions, not technology alone.