This is, of course, a disabling lacuna for understanding world cities as nodes in a global space of flows as famously envisaged by Manuel Castells. Public statistics provide much information on the attributes of cities but very little on the flows between cities. And it is precisely in this respect that our knowledge has been found to be wanting. Cities are the crossroads of society: they exist through their connections. These are important features of world cities but we can never understand any city in terms of just its stock of functions and practices. Most definitions of world cities have relied upon such factors as the number of corporate headquarters, the size of capital markets, and the quantity of international institutions located within a city. And it is at this step in the argument that the certainties give way to many unanswered questions. 2 Thus the point is not whether London is a major world city but how it operates as a world city. 1 And, of course, London is one of Saskia Sassen’s famous trinity of ‘global cities’. In a recent review of the world cities literature London was one of only four cities cited by all sources.

London relations with global cities