Amazon—the world’s largest online retailer and fourth largest company as measured by market capitalization—is soliciting North American regions in a competitive selection process for a second corporate headquarters, or HQ2.
In the world of economic development—of which business attraction is a core function—bringing an Amazon headquarters to town is a huge deal, likely the most significant corporate decision in recent memory.
Simply, very few companies have the potential to reshape American cities. But Amazon is probably one of them. It employs tens of thousands of workers in Seattle and claims an economic impact of $38 billion between 2010 and 2016. CEO Jeff Bezos has intimated that HQ2 will “be a full equal to our Seattle headquarters.” Specifically, the Amazon announcement projects as many as 50,000 new jobs with average compensation exceeding $100,000 per year in the new site.
And not surprisingly, the speculation of where Amazon will go began immediately after the announcement. We examined data to see which metro areas in North America will be top contenders for HQ2.
Before diving in, it is important to note that the specific function of this second headquarters is still a bit unclear. Preliminary signs from the “request for proposals” (RFP) suggest that HQ2 aims to be just that, a second command and control post that could share in the global management and product development done in Seattle. But much remains unknown, which makes this an admittedly speculative exercise.
What we do know is that Amazon is only seeking bids from metro areas with more than one million people. This immediately narrows the list to a little more than 50 U.S. metropolitan areas. We include Montreal, Toronto, and Vancouver in Canada as well.
Talent is a second major consideration. Innovation is a key distinguisher of Amazon from competitors, and that depends on top technical and managerial talent. The RFP notes that a “highly educated labor pool is critical” and that the site must be able to attract and retain “technical talent.” In this regard, a region must be able to offer a highly educated labor force, and that labor force must be large and deep enough to not be overwhelmed by the volume of Amazon’s worker needs. Therefore, we ranked metros on a “talent pool” index based on four simple metrics: the total metro population with a bachelor’s degree, the share of metro population with a bachelor’s degree, the metro area’s total number of STEM jobs, and the share of STEM jobs in the local labor market.
Seattle ranks 9th of the 55 metro areas on this composite “talent pool” index. Near the top of the list are the nation’s largest and most educated cities: New York, Washington, San Jose, Boston, San Francisco, Los Angeles, and Chicago. But the top half also includes medium-sized cities like Austin, Detroit, Hartford, Minneapolis-Saint Paul, and Raleigh. Smaller metros with more limited technical worker pools like Las Vegas, Orlando, Riverside fall near the bottom of the list, as do many Southern metro areas including Miami, Memphis, and New Orleans. Given how important human capital is to Amazon, it is likely that they favor the 22 metro areas that rank as above average metros on this combined index.
A second set of factors involve land and transportation access, specifically whether a city actually has a physical site that can accommodate up to 8 million square feet of development and provides public transportation. These public transit requirements likely eliminate Hartford, and probably Raleigh, which do not have rail systems (whether bus rapid transit is sufficient for Amazon is an additional question mark). Therefore, we end up with 20 North American metro areas that all have rail transit systems, albeit of varying coverage and quality, and international airports with the baseline connectivity necessary to meet Amazon’s demands of daily direct flights to the Bay Area, New York, Seattle, and Washington, DC. We list those below.
Of course, markets like Atlanta, Dallas, Los Angeles, New York, and Washington, D.C. will rank higher on aviation connectivity—especially to other international cities—than Austin, Baltimore, Montreal, and Vancouver. If Amazon wants HQ2 to have immediate global connectivity, then it will favor the largest international air hubs. But domestic direct flights are a local characteristic that cities can work with the airlines to establish pretty easily, especially if it means landing a major whale like Amazon. In this regard, beyond connecting to the four major U.S. cities listed above, domestic aviation capacity may not be a hugely decisive factor within these top 20 markets.
Land may be a binding constraint for some places, but there is limited data on which to rank metros on land availability. The reality is that there are few development-ready sites of this size that have good public transit connectivity. East coast metro areas like New York, Boston, and Washington have the most comprehensive rail transit systems but may struggle to find sites of the right size. Available land may also constrain bids from San Francisco and San Jose, although they would be logical technology hubs. Meanwhile, non-coastal metro areas like Austin, Dallas, Denver, and Houston have more land, but less developed transit systems. No clear answer here.
Amazon also indicates they want a high quality of life and a cultural community fit. These are admittedly subjective concepts, but it seems like Amazon favors sites that resemble urban Seattle—complete with a diversity of housing options and residents, walkability, and transit and entertainment options. In other words, a relatively urban neighborhood (although Amazon is accepting suburban site applications). This may prove difficult for sprawling metro contenders in the Sunbelt.
Tax incentives will play a role, to be sure, but only after the list has been narrowed significantly to metro areas that meet these fundamental economic and cultural requirements.
A final consideration is Amazon’s strategic direction. Jeff Bezos moved to Seattle to found an online bookseller because of its deep pool of engineers and access to a large book wholesaler in Oregon. Are there similar metro intangibles that Amazon will seek out? For instance, Atlanta’s legacy of large companies with major logistics operations such as Coca-Cola, Home Depot, and UPS has created a unique specialization in supply chain technology that may be attractive to Amazon if they are creating a new hub for managing their logistics system. Amazon’s major push into movies could favor Los Angeles as a headquarter site. Minneapolis-Saint Paul, home of Best Buy and Target, have a deep pool of employees in retail management, an Amazon focus. Meanwhile, Toronto has been ramping up investments in artificial intelligence research and development, a key technological priority for the firm. Completely importing expertise is difficult, even for Amazon, so they will likely aim to build on existing metro strengths.
In sum, it is relatively easy to build the list of 20 metro areas in the United States and Canada that can offer table stakes. But narrowing within that list will be subject to how Amazon weights these factors against one another and how big the local tax incentive packages get, which leaves this highly speculative.
Of course, that isn’t stopping predictions, which you can find from various journalists and experts here, here, and here.
For what it’s worth, I think Amazon will ultimately make this decision based on where they can get a quality technical workforce at scale, especially one that has a concentration in a key area of expansion for the firm. Regions with research universities with good business schools and computer science departments will be a logical fit, and they will likely want a site with some baseline density and vitality.
New York and the Bay Area offer very large technical labor pools but also a very high cost of living, which will likely exclude them. The draw of Boston’s labor and university base is strong, as evidenced by General Electric’s recent arrival based on those factors. Atlanta is intriguing: its sprawling physical development may be disqualifying, but the city provides a combination of a deep white-collar labor pool, supply chain technology capabilities, Georgia Tech, and a relatively low cost of living. Toronto is Canada’s strongest contender. Going abroad would be politically controversial but Toronto would offer a diverse, cosmopolitan and educated labor force, the University of Toronto’s globally relevant computer science and business school, and a hedge against U.S. political risk. Of course, Jeff Bezos recently bought a house in Washington, DC., so that would provide a convenient east coast location for perhaps the world’s most valuable CEO.
Cities constantly compete for firms, but the scale of Amazon means that the scale of this competition will be historic. Its outcome will give us a notable signal as to the continent’s ever-shifting economic geography.