That was my first thought, too, but that's not totally ignored. It all hinges on this part: "...when compared to census tracts that were scheduled to receive a

That was my first thought, too, but that's not totally ignored. It all hinges on this part: "...when compared to census tracts that were scheduled to receive a Starbucks but did not do so..." So the sampled population is the neighborhoods that were scheduled to receive a Starbucks, and lack prior other coffeeshops. Which means we're basically only looking a newly gentrifying neighborhoods (that Starbucks has calculated are worth opening something in, but that's probably mostly redundant with "newly gentrifying"). This severely limits the scope of applicability, but is fine in terms of the statistics. The big question is: why did some locations not receive the Starbucks they were "scheduled" to receive? Because that is a potentially huge confounding factor. If it's because Starbucks corporate said "hey, we only want to open up 2000 new locations this year, let's roll dice to see which ones get it", then it's all good. If (as seems more likely) it's more like "let's start the process for all candidates, but drop the ones that are slowest to get through the permitting process or hardest to get a loan/insurance/workers/supplies/building materials for", well then -- might those affect the ability to start up a company just a wee little bit? I'm torn. It's an interesting enough result that I'd almost like to go to the extreme measure of doing the "actually read the paper" thing that I've heard rumored some people do. And yet, my cynicism says my doubts are probably correct, so I don't want to bother with the effort to find out if they aren't.