From Baghdad to London: the Dynamics of Urban Growth in Europe and the Arab World

In 800 A.D., Baghdad, then the heart of the Abbasid Caliphate, was a thriving, cultured metropolis while Western Europe remained a stagnant economic backwater. Yet over the next thousand years an extraordinary reversal of fortune took place: the Christian West experienced a dramatic flowering of trade, wealth and population, while the Arab world entered long-term decline. The urban population of the Arab world increased by a half between 800 and 1800 A.D. - but the urban population of the Christian West grew by more than twenty-fold.

Economists have debated for many decades about what caused growth in the West to take off - was it the 'Great Discoveries' of the sixteenth century when explorers opened up new trade routes to China and India? Did it depend on the distinctive political, cultural and economic institutions of the Christian states? Or was it instead connected to more permanent factors, such as geography?

A new paper CEPR Research Associate Jan Luiten van Zanden, Maarten Bosker and Eltjo Buringh tries to test these various theories in more detail by examining the rise and fall of individual cities throughout the two regions - the Arab world and the Christian (mainly Catholic) west. In order to carry out an analysis at this level, the authors construct a new database, including every city which had more than 10,000 inhabitants at some point during the millennium they study. For Europe, they update a database originally published by Bairoch et al. in 1998; but for the Arab world they use a variety of sources, and in some cases estimate cities' population size on the basis of their area, and the number of mosques and hammans that were built to serve the inhabitants.

Simply looking at the broad trends in the database already identifies a number of differences between the development of cities in the Arab world and Europe. There tend to be far more cities in Europe for example - on average four to the same land-area. Arab urban populations tend to be more concentrated in the largest city in a particular state, rather than spread between a number of evenly sized conurbations.

In order to carry out a more formal examination of these ideas, the authors use a series of regression analyses to measure the impact of different factors on the population growth of cities through the centuries. They include a large number of potential factors in these regressions in order to try and capture the effects of economic and political institutions, and geographical location.

One classic dichotomy that the authors find helpful, developed by the sociologist Max Weber, is between 'consumer cities,' and 'producer cities'. Consumer cities sit at the heart of a state, offering services such as administration and protection in exchange for taxes and land rents. Their success is intimately linked to that of the state within which they exist - and they tend to crowd out other cities nearby.

Producer cities, meanwhile, exist through making and trading goods and materials with their hinterland and with other cities in their vicinity. Because they have their own economic basis, the fortunes of these producer cities need not be bound together with those of a great state or empire. The growth of other nearby cities can increase rather than diminish their economic success.

One key criterion for deciding which of these categories cities fall into, therefore, is the strength of feedback effects between them: if having other large cities nearby appears to promote growth it is likely that these are producer cities, trading and interacting freely with each other.

In order to measure these feedback effects, the authors calculate what they call the cities' 'foreign urban potential,' in effect capturing the possibilities for trade. The 'FUP' for each city is a sum of the size of all the other cities, weighted by the transport costs of getting goods there (differentiated by sea, road and river transport costs).

As well as this calculation of FUP, the authors include in their regressions terms for whether or not the city is the capital or the seat of a bishop or archbishop, and whether it has a university, to help reflect the extent to which institutions may be important. They also add terms for a city's geographical location, including whether it is on the sea; on a Roman road; or on the route of a camel caravan.

When they carry out regressions using all these terms, the authors are able to identify very different patterns of growth in the Arab and Latin West over the period in question. They find that there are strong positive feedback effects between Arab cities and between the cities in the Christian area, but not in between the two. In fact, Christian cities that are near large Arab cities tend, on average, to be smaller than would otherwise be expected, and vice versa. This suggests that there are strong economic linkages between cities in each of the two areas - but that transactions across the cultural divide are difficult, supporting the idea that the separate cultures and the distinct economic institutions to which they gave rise, were important.

To examine this idea a bit further the authors carry out another set of regressions, this time calculating the FUP of each city for different centuries instead of using the same measure throughout. This allows them to understand changes in the pattern of development over the millennium.

During the Middle Ages, in the early years of the period in question up to 1300, the authors find that the positive feedback effects between Muslim cities are actually very strong, suggesting that they were part of a powerful economic and trading system and could thus profit from each other's success. Meanwhile, in Europe, cities remain isolated and there is little positive interaction between them.

Later, however, the position is reversed: after the Abbasid caliphate collapses, perhaps partly as a result of the incursions made by the raiding Mongols, the economic system in the Arab world appears to disintegrate and it no longer helps a particular city to have successful neighbours nearby.

In the Latin West, meanwhile, from about 1300, positive feedbacks began to develop. This suggests that, despite the absence of an Empire, by uniting the region into a single political entity, economic institutions began to emerge from the bottom up. These were able to bypass the barriers created by different languages and particular political institutions, and allow cities to profit from each other's success thus helping to drive a powerful wave of urbanisation.

Importantly, the authors point out that their analysis shows these positive interactions to have been present before the 'Great Discoveries' of the 1500s - so the economic success of Europe cannot simply have been a result of the discovery of the sea routes to Asia and the Americas. The disintegration of the Arab economic system which had been united by the great Caliphates was already well underway before that point.

However, proximity to the sea does become important in determining a city's success after 1500 in the West; while it remains unimportant for Arab cities which instead are more reliant on closeness to caravan routes.

As far as political institutions are concerned, the authors find that cities tended to fare better when they were given some independence, as began to happen in France and Italy from about 1100 and had spread to more than a third of cities in Europe by 1500 - though not at all in the Arab world. Although political institutions appear to be significant, including them in the regression does not wipe out the separate importance of integration with other cities through economic links.

By repeating their regression analysis with different factors in this way, the authors build up two separate and distinct patterns of development in the pair of regions they study. In the Arab East, after the collapse of the Caliphate, which had united a large geographical area, cities became disparate and disconnected. They wisely switched to using camel transport instead of sticking to the roads left by the Romans, as exhibited in the importance of proximity to camel routes in determining Muslim cities' success; but they stuck with them, even as sea routes became cheaper and more efficient.

While the once-mighty cities of Baghdad, Damascus and others declined, Europe was developing a separate urban system founded on merchant guilds, and cities with some measure of political independence. Although this was no more efficient than the system in place in the Arab world around 800 - the feedbacks and linkages between cities were no stronger in Europe than they had been in the Muslim world centuries earlier - the Christian system proved more robust, surviving without an overarching territorial empire to nurture it. For the consumer city of Baghdad, meanwhile, once the Caliphate collapsed there was little else to sustain its economic power.

Using a comprehensive database covering hundreds of cities over an ambitious thousand year period, Bosker et al. uncover a complex picture of the divergent growth paths of the two regions they study. They show that Weber's distinction, between producer and consumer cities, remains useful in understanding how a once-mighty metropolis can become a mere backwater once it no longer has an empire to support it; while smaller, less remarkable cities can thrive by trading successfully with their neighbours. This appears to be what Europe's Christian cities succeeded in doing from about 1300, long before the 'Great Discoveries' of the New World.

However, while these economic linkages and positive feedbacks between cities were crucially important, they did not stretch across the religious divide. Having even a successful, rapidly expanding Muslim city on the doorstep did not help its Christian neighbours, and vice versa - so culture, too, appears to have been important in separating winners and losers. In fact, Bosker et al. show that any theory which suggests that geography, or culture, or political institutions are the key to successful economic development, is likely to be much too simple: all play a part, but perhaps most important is how well cities manage to profit from each other's success.

DP 6833 From Baghdad to London: the Dynamics of Urban Growth in Europe and the Arab World, 800-1800

Maarten Bosker, Eltjo Buringh and Jan Luiten van Zanden