Has geography ceased to matter in the age of globalisation, as many people believe? In this article John Kay shows that living standards and patterns of production are still largely determined by geographical factors.
We are told again and again that in the twenty-first century geography is no longer relevant to economics. And evidence of globalisation is all around us. McDonalds is everywhere; multinational companies ruthlessly source their products from the cheapest locations. The internet has established a single worldwide community of information.
And yet geography remains overwhelmingly important. The largest influence on the lifestyle you can expect is still your place of birth. Those born in Western Europe enjoy the highest material national standard of living achieved in human history. Most people in sub-Saharan Africa or the Indian sub-continent can expect a much shorter lifespan, in grinding poverty little different from the experience of their distant ancestors.
The only thing as important as where you are born is where you live. Immigrants to poor countries mostly enjoy living standards as high as they would in their country of origin. Otherwise they would not go. Immigrants to rich economies mostly have standards of living that are a blend of those of their birthplace and their adopted country. Germans in the United States have German or American incomes: Haitians in America earn more than they would in Haiti but less than the average American. Of course, the Germans are running Daimler-Chrysler while the Haitians are more often kitchen porters.
And globalisation does not mean that the location of production does not matter, only that it matters in a different way. Once most of what we consumed was produced near to where we live. That is no longer true, but the place of manufacture is not at all arbitrary. Cameras come from Japan, ties from Italy, software from the United States. And not because of the activities of any one firm in these industries. There are many Japanese camera makers, many Italian tie manufacturers, many US software businesses.
But the reasons these locations are associated with these particular activities are not at all obvious. The rationale of older patterns of international trade was clearer. Differences of climate and terrain explained why sugar came from Barbados not Belgium, oil from Saudi Arabia not Switzerland. But climate and terrain do not account for clusters of tie makers around Lake Como or software developers in Silicon Valley.
And the most striking feature of electronic communication is not how much geographical diffusion of activities it has produced, but how little. Market trading is now overwhelmingly screen based, and does not require that traders gather in a single location. But huge clusters of screens are sited within a few hundred yards of each other in densely packed and fabulously expensive areas of the City of London and lower Manhattan. And most received wisdom is that the financial services industry will become still more concentrated in a small number of centres. The attempt to consolidate European bourses is the product of the belief that regional market centres will not survive. Technology permits dispersion: but that dispersion is not happening.
Geography matters because even if the distribution of natural resources is no longer of overriding importance, the distribution of man-made resources is. Above all, the distribution of capital drives both personal standards of living and business location. Not just capital in the narrow sense of investment in plant and machinery, but human capital – investment in education – and social capital – the political and legal infrastructure and culture of personal relationships within which business is done. And the distribution of such capital today is as uneven as the distribution of natural resources was at the beginning of the industrial revolution.
Yet the distribution of capital, particularly physical capital, is not set by nature, but determined by people. With freedom of capital movement, its owners can locate capital where they will and they do not do so on nationalistic or patriotic grounds. Which is why differences in social capital are crucial. In a globalised world of freely moving capital and increasingly freely moving people, it is only social capital which remains tied to specific locations.
These differences in institutions are the main influence on differences in living standards. And differences in social institutions also explain the localisation of what would otherwise seem to be footloose production. Tacit skills and knowledge are developed where people exchange ideas with each other casually and daily. Flexible production relationships are based on personal relationships.
The information that can be handled by information technology is of a limited kind. Some of the data needed for financial services business can be found on screens. But much of it cannot. It needs not only body language, but the nuances that can be conveyed only between people who are, or have been, in the same room. The chat room is an impoverished form of human interaction, and e-mail an impoverished form of human communication.
So geography continues to matter, and because geography matters history matters too. It is not an accident that tie manufacture is centred in Italy, software production in the United States, and financial services are strong in England. These capabilities can be traced back to specific elements in the social and economic histories of the respective countries – the flourishing of design skills in Italy during the Renaissance, the early development of a large installed base of computers in the United States, the pivotal role of English merchants in the development of trading economies.
It is an accident that these activities are more specifically located near Lake Como, in Silicon Valley, and between Holborn and Aldgate. But such accidents of history are not easily reversed. As the evolution of London investment banking in the last decade has shown, the identity of the participants may change, but the location of the activity does not. So long as culture matters, history and geography will matter too.
963