Amazon is going to choose a city for its second headquarters based on the availability and sustainability of labor. There’s no getting around the labor play when you plan to hire 50,000 workers. Yet the final decision for Amazon, or any firm making a site selection decision, is not just about absolute workforce numbers.
Labor sustainability is as important as labor availability. A tech-savvy workforce is a must. And tapping into a strong flow of millennials, if not members of Generation Z, is vital, too.
We’ve combined high-level elements of labor data metrics, and other data variables, into an Amazon Talent Index to rank the 52 U.S. metros that have populations of at least one million to qualify. The cities at the top—San Jose, San Francisco, Austin, New York, and Dallas—are a mixed bag: some likely have no shot (i.e., the Bay Area), and some are strong options. While many qualifying cities have the tech-saturated workforces to land Amazon’s HQ2, the right decision will be based on comparing markets at a micro level.
In this piece, we outline a few innovative data components—some of which go beyond our macro Amazon Talent Index—that we recommend presenting in your HQ2 proposal. (And pretty much every major city has said it will respond to what must be the most-viewed RFP in history.)
Bay Area, Austin, New York Lead Amazon Index
Our index includes the major labor and community-flavored data points that Amazon requests. This includes total jobs, job growth, and concentration in the five occupational categories listed in its RFP: executive and management, legal, administrative, accounting, and engineering—with a focus on software development engineers.
We’ve also included some of the community variables that Amazon mentioned, including a well-educated labor pool (we used bachelor’s-and-above educational attainment for that), cost of living, and share of millennials.
Our index is tilted to tech talent and specifically software engineers, so it’s no surprise to see San Jose and San Francisco as 1-2. It would be a shock if Amazon selected the Bay Area because of the competition for talent, the cost of living, and tight housing supply, but there is no better place for existing tech talent.
Most of the top 15 metros in the Amazon Talent Index, in fact, have been heavily debated by economic development junkies on Twitter and in articles on which city Amazon should or will pick.
We’re not saying all these metros have the infrastructure and culture that Amazon is looking for, which can dramatically impact this ranking and bring Amazon’s analysis down to the top three cities versus a top 10. (No city, as Lyman Stone argued well, is a perfect fit.) But from a labor perspective—a tech labor perspective above all—it’s hard to beat this group.
Note: The Amazon Talent Index is a composite z-score index that ranks metros with over one million population by two categories. The first is labor-focused: 2016 jobs, 2011-2016 job change, and 2016 concentration of tech/IT occupations, software occupations, and management, accounting, legal, and administrative occupations. The second is community-focused: millennial population share and growth, bachelor’s-and-above educational attainment share and growth, and C2ER cost of living.
Other Relevant Data for Your HQ2 Proposal
Beyond what we’ve provided for this macro index, there are other relevant data that we suggest economic development organizations consider when drilling down into the information that will make your community stand out to Amazon. Really, they are the same things that most companies are interested in: a strong labor force, reasonable cost of labor, the right mix of skills, a robust educational pipeline for that talent, quality of life, and more.
What follows are innovative spins on some of the above metrics.
The Skills Gap: Comparing the Supply and Demand of Skills
When conducting micro analysis, we recommend looking at the skills embedded in your workforce to see how they compare to the skills that Amazon has sought in Seattle, its main headquarters. This will give you a sense of how closely your workforce matches with what Amazon (and you can do this for other companies, too) is looking for.
In Seattle, 64% of Amazon job postings during the last year were for IT-related workers, so we know they are going to emphasize tech skills. (We also know this from the RFP itself.)
We took the 50 most-listed skills in Amazon’s Seattle job ads for software and IT positions, from Java to C++, and cross-referenced them with the skills that show up for those same positions in major metros using Emsi’s resumé/social profile database. This allows us to get two perspectives on skills—those Amazon has requested (demand) and those that workers list on their resumés and social profiles (supply).
Among the top 10 metros in our Amazon Talent Index, San Francisco has the highest occurrence of the top skills, with 33 of the top 50. Austin, Denver, and Boston are next, at 29. There’s not much separation among the top 10; Raleigh has the lower number, at 24.
|MSA||Skills in Common to Amazon's Top Skills in Seattle|
|San Francisco-Oakland-Hayward, CA||33|
|Austin-Round Rock, TX||29|
|New York-Newark-Jersey City, NY-NJ-PA||28|
|Dallas-Fort Worth-Arlington, TX||28|
|Los Angeles-Long Beach-Anaheim, CA||28|
|San Jose-Sunnyvale-Santa Clara, CA||27|
|Atlanta-Sandy Springs-Roswell, GA||27|
Interested in seeing this for your area? Contact us for your metro or region’s skills gap.
Workforce Availability and Sustainability
It’s one thing to have a healthy number of workers in the core occupations that Amazon is targeting. It’s another thing to be able to indicate that you have the workers available in a broader skill set, and that your city’s workforce is sustainable.
Last month, Emsi introduced a workforce availability report that allows you to visualize the available workforce at a given level of compensation. Amazon listed an average salary of $100,000 for the jobs it will fill in its second HQ. Using roughly that comp level ($48 per hour), we estimated how many software engineers it could expect to be available in Austin and the other top cities in our index. We did this by estimating the number of developers at each point along the wage curve and finding workers with compatible skills.
Austin has nearly 8,900 software engineers (software developers, applications, and software developers, systems software) who could be available at $48 per hour. Yet if Amazon paid $60 an hour, a reasonable expectation for software talent, it could tap into 14,000 software engineers, or 60% of the total software development pool in Austin.
In Austin, the number of available workers with a similar skillset (i.e., they work in occupation at 95 and above in our compatibility index) balloons to over 17,000 at $48 per hour.
What about the availability of software engineers among other top contenders and tech metros? As the below table shows, Austin has the smallest available labor pool among the top seven in our index (and Seattle, which we included for comparison). Denver is the next smallest. In comparison, the same $48-an-hour wage would capture 42% of the software engineers in New York (37,507), an immense labor pool that explains in part why it’s positioned near the top of our index.
|Exact Match||Exact Match % of Total||Similar Skillset||Similar Skillset % of Total|
|San Jose-Sunnyvale-Santa Clara, CA||12,930||19%||19,915||22%|
|San Francisco-Oakland-Hayward, CA||13,391||24%||24,014||30%|
|Austin-Round Rock, TX||8,874||46%||17,371||56%|
|New York-Newark-Jersey City, NY-NJ-PA||37,507||42%||61,661||45%|
|Dallas-Fort Worth-Arlington, TX||14,982||41%||32,853||50%|
For labor sustainability, Emsi has conducted studies based on the churn and age of the workforce, as well as available workers, to determine how soon communities will run out of labor at a certain wage. For more on this, contact us.
Amazon is going to locate in a metro with a strong university system. Whether that’s University of Texas in Austin or MIT in Boston, most of the major players are well known. But what’s less well known is the output of software or management talent that those universities, and other colleges, are producing.
This is where data from the National Center for Education Statistics comes in handy. And where comparing that output to estimated demand, either with job postings or annual openings that look at new and replacement jobs, can be valuable.
Here are the metros that rank the highest for output of computer science graduates the last three years. Note that other CS-related grads can come from coding bootcamps that aren’t captured with this data. Warrensburg, Missouri, shows up highly because of the emergence of University of Central Missouri’s computer science program.
Talent Attraction and Retention
Validating that you have an educational pipeline only takes you so far. A natural, and fair, question after showing your educational pipeline is, how many of these graduating students is your city retaining? And further, are you attracting other skilled talent?
We took the same resumé/social profile data that we used in the above supply and demand of skills section and mapped where graduates of local institutions in the top 10 metros wind up. The expansive New York MSA retains 40% of graduates from local colleges and universities. Austin keeps 30% of its graduates, while 11% go to Houston and 9% go to Dallas.
Another way to get at some of this is by looking at Census or IRS migration data.
There are only so many regions that have the technical clout to satisfy Amazon’s intense demand. But if the multitude of analyses on where Amazon should go are proof of anything, it’s that most regions on the list have a story to tell. And that story is easier to tell with compelling, and sometimes unique, data like we showed in this piece.
For help on preparing data-driven, micro, and multivariate analysis that Amazon is expecting in response to its proposal, please contact us for a consultation on how we can position your city as the right city for the next Amazon HQ.