Lump together and like it: Cities and growth

…Yet new research published by the World Bank in its annual flagship World Development Report* suggests that pessimism over the future of huge cities is wildly overdone. The bank argues that third-world cities grow so big and so fast precisely because they generate vast economic advantages, and that these gains may be increasing. Slowing urbanisation down, or pushing it towards places not linked with world markets, is costly and futile, the bank says. At a time of contagion and bail-outs, the research also reaffirms the unfashionable view that the basic facts of geography—where people live and work, how they get around—matter as much as financial and fiscal policies. (The award of this year’s Nobel prize for economics to Paul Krugman of Princeton University for his work on the location of economic activity was another reflection of that view.)

The bank’s research yields lots of new insights. It argues, for example, that the share of humanity that lives in cities is slightly lower than most people think. The bank drew up a fresh index to get around the knotty problem of defining “urban”; this new measure puts the world’s city-dwelling population at about 47% in 2000. In fact—as Indermit Gill, who oversaw the report, acknowledges—it is impossible to pinpoint the proportion: the urban slice of humanity may be anywhere between 45% and 55%, depending on how you count. The report’s main point is that, whatever their exact dimensions, the Gotham Cities of the poor world should not be written off as a disaster simply on grounds that they are too big, too chaotic, too polluted and too unequal.

It is true that they are unprecedented in size. Mexico City, Mumbai, São Paulo and Shanghai each have over 15m people, whereas Paris and London, after their surge in the 19th century, had less than half that. The average population of the world’s 100 largest cities now exceeds 6m. In 1900, it was only 700,000.

 But relative to the size of countries’ populations, the current growth is far from unusual. Between 1985 and 2005 the urban share of the population of developing countries rose by eight percentage points. Between 1880 and 1900, the bank says, the urban share in then-industrialising Europe and America went up by about the same amount. Over time, cities as disparate as Santiago (in Chile), Seoul, Lisbon and São Paulo have all followed strikingly similar paths, rising fast until they made up about a quarter of their countries’ populations, then stabilising when the country’s income hit about $5,000 per person (see chart). This path roughly tracks the transition of a country from an agricultural to an industrial base. Many countries are undergoing that sort of transition now, and therefore urbanisation is accelerating. But history suggests it will not go on rising at this rate for ever.

History also suggests that the income gaps that worry governments will narrow. …

…So one answer to the question—why are third-world cities so big?—is that they are not in relative terms all that large. But another answer, suggests the World Bank, is that they are big because they do an economic job that is becoming more, not less, important.

Cities are products of trade. Market towns trade crops. Second cities produce and trade manufactured goods. Metropolises design, make and sell everything, especially services. Over the past 50 years world trade has expanded hugely, especially in services, and giant cities have thrived correspondingly. …

…What has made such growth possible, argues the bank, is cheap transport. Falling transport costs in the 18th and 19th centuries enabled Britain and Portugal to trade wool and port (as the political economist David Ricardo memorably pointed out). Cheap transport in the past 25 years has produced a second sort of trade revolution. Countries now sell each other not final products like port but intermediate ones such as recording heads for hard drives. That has been made possible by an extraordinary fragmentation of production: every step in the production line is broken down. Parts are made separately, then shipped for assembly. …

…These prescriptions have something in common. For poor countries, the key to development is to link up flagging and fast-growing regions. To do that, governments often overemphasise policies targeted on particular places. In practice, there are more powerful instruments of integration than “spatially targeted” efforts—eg, land markets that unify all places, or infrastructure that connects some places to others. Growth, says the bank, is inevitably lumpy. Governments must learn to like it.