Global City - Studies - GaWC Study

GaWC Study

The first attempt to define, categorize and rank global cities using relational data was made in 1998 by Jon Beaverstock, Richard G. Smith and Peter J. Taylor, who all worked at the time at Loughborough University in the United Kingdom. Further by academic researchers like Ronald Daus in Germany (Free University of Berlin), who focused on «extra-European cities, predominantly in the Southern hemisphere». Together, Beaverstock, Smith and Taylor established the Globalization and World Cities Research Network. A roster of world cities was outlined in the GaWC Research Bulletin 5 and ranked cities based on their connectivity through four "advanced producer services": accountancy, advertising, banking/finance, and law. The GaWC inventory identifies three levels of global cities and several sub-ranks.

The 2004 rankings acknowledged several new indicators while continuing to rank city economics more heavily than political or cultural factors. The 2008 roster, similar to the 1998 version, is sorted into categories of "Alpha" world cities (with four sub-categories), "Beta" world cities (three sub-categories), "Gamma" world cities (three sub-categories) and additional cities with "High sufficiency" or "Sufficiency" presence. The following is a general guide to the rankings:

• Alpha++ cities are New York City and London, which are vastly more integrated with the global economy than any other city.
• Alpha+ cities complement London and New York by filling advanced service niches for the global economy.
• Alpha & Alpha- cities are cities that link major economic regions into the world economy.
• Beta level cities are cities that link moderate economic regions into the world economy.
• Gamma level cities are cities that link smaller economic regions into the world economy.
• Sufficiency level cities are cities that have a sufficient degree of services so as to not be obviously dependent on world cities.

The 2010 roster of leading world cities and "sufficient" cities is as follows: