Joe Cortright of City Observatory recently looked at “the myth of revealed preference for [the] suburbs.” In it, he cites the work of Jonathan Levine, who is the author of a 2006 book called, Zoned Out: Regulation, Markets, and Choices in Transportation and Metropolitan Land-Use.

There’s an argument out there that, on average, people prefer the suburbs to urban neighborhoods because, well, more people in the US live in auto-oriented neighborhoods compared to urban ones. What Levine wanted to figure out was whether this was truly a result of consumer preference or simply a lack of urban neighborhoods – or a “shortage of cities” as Cortright calls it.

To do this Levine examined two cities with very different urban fabrics: Boston and Atlanta. The idea was to take a city with lots of urban neighborhoods (Boston) and compare it to one with relatively few (Atlanta). 

For his comparison, he classified all of the neighborhoods in both cities on a scale according to how urban they were. “A” meant very urban. And “E” meant sprawling/exurban. He then went out and interviewed residents, asking them about both the kind of housing they would ideally like to live in and how happy they were with their current housing. 

What Levine discovered, among other things, was that in Boston – where about half of all housing fell into the top 3 most urban categories – about 83% of people with a strong preference for urban neighborhoods were also living in one. Whereas in Atlanta, just 48% of people with a strong for urban neighborhoods were living in one.

Put differently, the study suggests that in cities with fewer urban neighborhoods, it is more difficult for people with a preference for that housing type to find and live in it, which makes intuitive sense. The spread between preference and reality widens, once again suggesting that this could be about supply rather than an issue of demand.

Anecdotally, I have seen this phenomenon play out here in Toronto. I often hear people talk about the neighborhoods that they would ideally like to live in, if only they could find a reasonably priced home. (Low supply leads to upward pressure on pricing.) How aligned would you say you are with your ideal level of urban-ness? 

For more on Levine’s work, head over to City Observatory.

So I was wrong. Amazon didn’t pick Toronto for HQ2. It instead picked Crystal City, Virginia (Washington) and Long Island City, NY (New York City). More on that, here, in the NY Times. Confession: My prognostication was at least partially about trying to create a self-fulfilling prophecy.

In any event, it’s interesting to consider the locations that they did pick – as well as the fact that they ended up picking multiple cities. This was not part of their RFP. Though, many have convincingly argued that this process was over before it even began. HQ2 was always going to end up on the east coast, near one of Bezos’ homes.

Nevertheless, urbanists such as Aaron Renn took the announcement as a direct repudiation of the American heartland. He believed that Amazon would be far more cost conscious in their decision making and ultimately elect for a lower cost locale in the middle of the country. Instead, the coastal hegemony won out. 

Joe Cortright of City Observatory correctly predicted that Amazon would, for a few reasons, parlay their HQ2 search into multiple smaller locations (HQ2, HQ3, and so on). One of the reasons for this is that it gives the company more leverage when it comes negotiating subsidies on a go-forward basis. If NYC doesn’t want our next round of hires, we’ll take them to Washington.

Looking at the locations, one of the first things I noticed is that both are just outside of their respective “downtowns” (across a body of water), as well as adjacent or on the way to an international airport. Crystal City is across the street from DCA and Long Island City is a 15 minute drive from LGA. Both are situated on top of higher order transit. Makes sense to me.

Now, who wants HQ4?

Joe Cortright of City Observatory recently looked at “the myth of revealed preference for [the] suburbs.” In it, he cites the work of Jonathan Levine, who is the author of a 2006 book called, Zoned Out: Regulation, Markets, and Choices in Transportation and Metropolitan Land-Use.

There’s an argument out there that, on average, people prefer the suburbs to urban neighborhoods because, well, more people in the US live in auto-oriented neighborhoods compared to urban ones. What Levine wanted to figure out was whether this was truly a result of consumer preference or simply a lack of urban neighborhoods – or a “shortage of cities” as Cortright calls it.

To do this Levine examined two cities with very different urban fabrics: Boston and Atlanta. The idea was to take a city with lots of urban neighborhoods (Boston) and compare it to one with relatively few (Atlanta). 

For his comparison, he classified all of the neighborhoods in both cities on a scale according to how urban they were. “A” meant very urban. And “E” meant sprawling/exurban. He then went out and interviewed residents, asking them about both the kind of housing they would ideally like to live in and how happy they were with their current housing. 

What Levine discovered, among other things, was that in Boston – where about half of all housing fell into the top 3 most urban categories – about 83% of people with a strong preference for urban neighborhoods were also living in one. Whereas in Atlanta, just 48% of people with a strong for urban neighborhoods were living in one.

Put differently, the study suggests that in cities with fewer urban neighborhoods, it is more difficult for people with a preference for that housing type to find and live in it, which makes intuitive sense. The spread between preference and reality widens, once again suggesting that this could be about supply rather than an issue of demand.

Anecdotally, I have seen this phenomenon play out here in Toronto. I often hear people talk about the neighborhoods that they would ideally like to live in, if only they could find a reasonably priced home. (Low supply leads to upward pressure on pricing.) How aligned would you say you are with your ideal level of urban-ness? 

For more on Levine’s work, head over to City Observatory.

So I was wrong. Amazon didn’t pick Toronto for HQ2. It instead picked Crystal City, Virginia (Washington) and Long Island City, NY (New York City). More on that, here, in the NY Times. Confession: My prognostication was at least partially about trying to create a self-fulfilling prophecy.

In any event, it’s interesting to consider the locations that they did pick – as well as the fact that they ended up picking multiple cities. This was not part of their RFP. Though, many have convincingly argued that this process was over before it even began. HQ2 was always going to end up on the east coast, near one of Bezos’ homes.

Nevertheless, urbanists such as Aaron Renn took the announcement as a direct repudiation of the American heartland. He believed that Amazon would be far more cost conscious in their decision making and ultimately elect for a lower cost locale in the middle of the country. Instead, the coastal hegemony won out. 

Joe Cortright of City Observatory correctly predicted that Amazon would, for a few reasons, parlay their HQ2 search into multiple smaller locations (HQ2, HQ3, and so on). One of the reasons for this is that it gives the company more leverage when it comes negotiating subsidies on a go-forward basis. If NYC doesn’t want our next round of hires, we’ll take them to Washington.

Looking at the locations, one of the first things I noticed is that both are just outside of their respective “downtowns” (across a body of water), as well as adjacent or on the way to an international airport. Crystal City is across the street from DCA and Long Island City is a 15 minute drive from LGA. Both are situated on top of higher order transit. Makes sense to me.

Now, who wants HQ4?