Yo, Philly! We crunched the numbers and we’re not Amazon’s prime location
Amazon has created a competition among cities worthy of reality TV, and dozens of regions are vying for its hand.
At stake: a massive second corporate headquarters for one of the largest publicly traded companies on Wall Street, filled with 50,000 full-time employees making more than $100,000 a year on average. When complete, the HQ2, as Amazon is calling it, could exceed eight million square feet and $5 billion in capital investments over 15 or so years.
We’ve built a tool to compare the competitors based on Amazon's stated criteria. You can select different criteria to rank the suitors, playing the role of the Amazon selection committee.
It's a big prize for cities; in Seattle, where its first headquarters has grown to 8.1 million square feet, Amazon estimates every dollar it invested generated an additional $1.4 for the city's broader economy. Now the company is dangling the prize of a second headquarters, seeking proposals from cities whose metropolitan regions have more than one million people.
The nation has 52 such regions, excluding Seattle. In its guidance for these cities, Amazon spells out several factors it says are important.
No one area ranks highly in all of Amazon's criteria, so much depends on the specific location in the region and the specific pitch a region makes.
"The perfect city just frankly doesn't exist," said Ram Mudambi, a professor of strategy at Temple University's Fox School of Business. "There is no such place that has perfect physical infrastructure and perfect human resources and perfect policies all together in one place."
If Amazon cared equally about all the criteria, in our analysis Philly would come in at No. 12, and Los Angeles, Phoenix, and Tampa would be the best bets.
But Amazon probably doesn't care equally about all the criteria, Mudambi said. Still, using his ranking of factors, Philly also still sits at No. 12, with Phoenix, San Diego, and St. Louis rising to the top.
We also asked David Reibstein, a marketing professor at the University of Pennsylvania's Wharton School, what he thought Amazon's preferences would be among its stated criteria. By his rankings and our calculations, Philly drops to No. 19, with Tampa, Pittsburgh, and San Diego at the top.
Here are the factors we included in the tool, along with information on why each is important, what data we used, and where Philly ranks.
Tax burden and stable business environment
This is a major consideration for Amazon, Mudambi and Reibstein said, and both said the company would rank it as top among its preferences. In our system, we have assigned each city the same ranking because taxes essentially are a wash for two reasons.
First, Amazon cares more about its specific incentives than the general environment, Mudambi and Reibstein said, so a fair comparison would be of each region's proposals, not the general tax structure.
Second, the competition Amazon has created gives it the leverage to negotiate down to essentially no taxes, they said.
"It probably will not be a determining factor, even though it's very, very important," Reibstein said. "And the reason is, I think Philadelphia will say, 'We'll cut your real estate taxes in half," and then Denver says, 'Well, we'll take your real estate taxes down to zero,' and Philadelphia will respond and say, 'OK, we'll take your taxes down to zero.'"
Reibstein compared the situation to cities' negotiations with sports teams, where cities hope for economic benefits outside of taxes.
"It'll just be negotiated away such that it doesn't differ significantly across cities," Reibstein said.
Educated labor force
"A highly educated labor pool is critical and a strong university system is required," according to Amazon's request for proposals. Reibstein and Mudambi said this would be a top priority for the company. You can have all the physical requirements, Mudambi said, but it is the people who make a city work.
"If you have really, really creative people, you can overcome all kinds of weaknesses in the hard infrastructure that surround you," Mudambi said, describing cities as akin to Dr. Frankenstein's creation: "You can have all the parts, you can have the arms and legs and the heart and everything, but it's not alive. The people are what make it live."
We ranked regions according to the number of adults 25 and older who have a bachelor's degree or higher, according to the Census Bureau's 2016 American Community Survey estimates. We used a sheer number instead of a rate because Amazon doesn't need a large percentage of people to be educated — it needs a large number of educated people, period. This means that having a large population helps, but this isn't a direct measure of population: Houston ranks fifth by population but ninth by education; Riverside, Calif., is 13th by population but 21st by education.
With more than 900,000 adults with bachelor's degrees, Philly ranks seventh on this measure.
A related measure is the number of current college students, an indicator both of the higher education system in a region and the number of people interested in pursuing college, whether or not they finish.
We ranked regions by the total number of undergraduate students, according to the 2015 American Community Survey five-year estimates.
Philly is seventh among the regions by population, but moves to sixth by number of students because of the large number of colleges in the region. The Miami-Fort Lauderdale-Palm Beach area, which is eighth by population, jumps up to fifth for number of college students.
Amazon's headquarters will be a massive operation, and its tens of thousands of staffers need easy access to work and travel around the region. The company asks proposals to include information on transit and transportation options, traffic congestion data during rush hour, and information on bike lanes and pedestrian access.
We used the 2015 Urban Mobility Scorecard from the Texas A&M Transportation Institute, drawing on traffic speed data and highway performance data from around the country, to rank cities.
Philly is 34th on this metric. The top cities are Raleigh, N.C., Richmond, Va., and Birmingham, Ala.
By another metric, the average travel time to work regardless of mode of travel, Philadelphia ranks 42nd of the cities, with its average commute time of 29.8 minutes.
"It's a business decision," Reibstein said, and Amazon has tens of thousands of employees it'll be paying.
We used Census Bureau County Business Pattern survey data from 2015 to estimate the average pay for workers in three general sectors: information, which includes software publishing, web search, and data processing; professional, scientific, and technical services, which includes legal services, accounting, engineering, and computer services, and management of companies and enterprises, which includes non-governmental firms that control or oversee other companies.
Philly ranks 46th on this measure, because those workers make an average of $100,333.35 here. Compare that to the top three cheapest cities for these sectors: Riverside, Calif. ($58,234.27), Tucson, Ariz. ($62,007.26), and Buffalo, N.Y. ($62,485.53).
The most expensive city is San Jose, Calif., at $191,675.78 average pay, followed by San Francisco ($148,635.14), and Houston ($115,720.19).
Cost of living
Amazon needs to attract employees, Reibstein said, and cost of living also affects what employees should be paid. The lower the cost of living, the better.
We compared median monthly housing costs for occupied housing units, whether rented or owned, as estimated by the 2016 American Community Survey one-year estimates.
Philly ranks 36th on this measure, with residents paying a median of $1,217 in monthly housing. As Philadelphians like to point out, that's cheaper than New York, which ranks 46th with its $1,557 cost.
The most expensive regions are San Jose ($2,214), San Francisco ($1,945), and Washington ($1,735). The cheapest are Pittsburgh ($814), Buffalo ($820), and Birmingham ($845).
This is where Philly shines. Amazon mentions several times in its guidance that "personnel travel and logistics needs … are critically important." The company includes access to major highway arteries and "an international airport with daily direct flights to Seattle, New York, San Francisco/Bay Area, and Washington, D.C."
We searched flight schedules for a Monday in October to see which cities had at least two direct or nonstop flights to one of those four cities, or other easy access. (Baltimore, for example, has access via a short drive to D.C.)
For each city, we came up with a number of cities one can easily get to via airport or highway; 23 cities, including Philadelphia, have access to all four areas. Oklahoma City has no direct access to any of the four areas, and Tucson has direct access only to San Francisco. The remaining cities have access to either two or three of the cities Amazon names.
Amazon says its headquarters "requires a compatible cultural and community environment for its long-term success," which it says "includes the presence and support of a diverse population," among other factors. Reibstein and Mudambi both rank this factor very low on Amazon's preferences: "It is on their list because it is important," Reibstein said, "but all the items above are economic issues."
Similarly, Mudambi said diversity usually does not weigh heavily for corporations searching for locations.
For this, we used the Census Bureau's 2016 American Community Survey estimates and ranked cities by the size of their non-white resident population.
Mudambi and Reibstein disagree on the importance of crime. Mudambi said he believes Amazon would see it as the least important of the criteria, since crime is highly localized in a given region. Reibstein said crime would fall around the middle of the preferences because of perception — the company needs to attract people to the city, which means public perception of safety is important.
We ranked cities by violent and property crime, using data submitted by police departments to the FBI. Our ranking includes only the major city in the region, and not suburbs or smaller cities. (So it includes Philadelphia, but not Camden or Wilmington; New York City, but not Newark.) We used Norfolk, Va., data for the Virginia Beach/Norfolk/Newport News region and Las Vegas metro data for that region. Portland data is from 2014 and Raleigh data is from 2013.
On this measure, Philly comes in at No. 18. New York, San Diego, and San Jose have the lowest crime rates; Salt Lake City, St. Louis, and Birgimingham have the highest. Salt Lake City has a middling violent crime rate, but its property crime rate is much higher than the other cities' and about three times the property crime rate of Philly.