Head office clustering in the mega-regions

Economists and geographers have long recognized the social and economic importance of cities, metropolitan areas, and other “sub-national” divisions. However, some are now realizing that a new and broader perspective on urban agglomeration helps us to understand productivity potential, employment growth, consumption patterns and innovation activity. In the 1950s, economic geographer Jean Gottmann first described a massive new regional formation he called “megalopolis”, dubbing the great conurbation along the Northeastern seaboard “Bos-Wash”.

More recently, Richard Florida and colleagues identified some 40 cross-metropolitan mega-regions around the globe – 12 in the US and Canada, alone – using world light emissions data. Florida’s research indicates that these mega-regions increasingly represent the peaks of the global economy, generating some two-thirds of the world’s economic activity and nine in ten of its innovations.

A new MPI analysis shows that corporate head offices in the US and Canada are overwhelmingly concentrated in the core cities of great mega-regions: 85% of the 1,100 largest companies we analyzed.i But these head offices are also strikingly specialized by mega-region. We know that firms tend to cluster at the local or regional level in order to take advantage of location-specific assets such as proximity to shared customer and supplier markets, pools of talented labour, and to share in knowledge spillovers between firms and institutions. As a result, regional economies tend to become specialized. Using a composite index created by multiplying a measure of industry concentration (i.e., the percentage of an industry’s head offices found in a specific mega-region) by regional specialization (i.e., the percentage of head offices in a mega-region in a specific industry), we observe a similar phenomenon at the mega-region level as well.ii The most important clusters are indicated on the map below, along with a few key companies from each.

What we found is that for a few key industries, one or two mega-regions tend to dominate. In manufacturing, Chi-Pitts and, to a lesser extent, Char-lanta have substantial clusters, collectively accounting for almost half the corporate head offices in these sectors (100 of 211 total). Manufacturing accounts for 36% of all head offices in Chi-Pitts and 28% in Char-lanta. The next two most significant manufacturing mega-regions, Bos-Wash and Tor-Buff-chester account for a total of 56 head offices.

Energy-related head offices are clustered in Hou-Orleans, extending along the Gulf Coast of Texas, Louisiana and Mississippi with 43 head offices – roughly 4.5 times what would be expected from its population – and a 67% share of corporate presence in the region.

High-technology head offices are concentrated in the Nor-Cal mega-region, anchored in the San Francisco Bay Area which has more than half of all high-tech head offices, more than four times the number one would expect. Conversely, only two large high-technology firms are headquartered in Hou-Orleans, and two energy-related firms in No-Cal.

Finance/insurance head offices are somewhat less concentrated, with three significant centers. Bos-Wash accounts for 17% of all head offices in this sector, Tor-Buff-chester has a significant concentration with two times the number of finance/insurance head offices that would be expected and a 25% share of corporate presence in the region, followed by Char-lanta with 1.3 times the number expected and a 16% share.

The implications of this research are significant: clustering, and its consequences, may in fact occur on a broader scale than previously understood. For example, those metropolitan and surrounding regions that represent the heart of the North American finance and insurance industries have both benefited and, more recently, suffered as a result of dramatic market changes. Yet, the broader economic fallout from the financial crisis may pose less of an economic threat to regions like Bos-Wash which is buoyed by strength in high-technology and media and entertainment, sectors which have thus far proven more resilient during the current downturn, and Tor-Buff-chester with its more conservative (and thus solvent) banking sector.

For other regions of the US and Canada, corporate clustering at this scale may create challenges.

Chi-Pitts has long benefited as the preeminent location for North America’s largest and most venerable manufacturing firms – the Big Three automakers, Procter and Gamble, US Steel and others. However, with the globalization of ownership and declining competitiveness of US and Canadian firms in many of these sectors, the challenges that currently face small factory towns in the Midwest may begin to spread more broadly to the corporate centres of Illinois, Ohio and Michigan. Rather than a Midwest mega-region with multiple economic hubs, Chi-Pitts may increasingly come to rely on the diverse and vibrant core area of Chicago. Similarly, the corporate orientation of Hou-Orleans may prove to be a liability for the Gulf Coast as a whole, as their fortunes are tied to the increasingly volatile price fluctuations of a single commodity, petroleum.

In our future research we will attempt to build a deeper understanding of industrial clustering in the mega-regions and its implications for their economic development and regional integration.

i) Analysis based on the largest companies in the United States and Canada – the 2009 Fortune 1000 supplemented by the 107 Financial Post Canadian companies that would qualify for that list based on revenues.

ii) Full table of results is available.

Download this Insight (PDF)

The Martin Prosperity Institute at the University of Toronto’s Rotman School of Management is the world’s leading think-tank on the role of sub-national factors—location, place and city-regions—in global economic prosperity. Led by Director Richard Florida, we take an integrated view of prosperity, looking beyond economic measures to include the importance of quality of place and the development of people’s creative potential.