The concept is that it is fundamentally wrong to smear the population of NY across the total area of the US. You make a point to back an argument - the argument might be correct but your point is not.
> Frankfurt-Hahn or Paris Beauvais
Note that they are close to dense population areas Frankfurt and Paris. Not part of average density areas.
I am just saying I have seen the pattern occur repeatedly where "population density" is used to attempt to prop an argument: density is averaging and is often meaningless.
Hmm. So you don't disagree that cheap semirural airports with a population base and short flights are important parts of the model, but you don't think they are related to population density? I don't see how those could be unrelated.
Perhaps a simpler way to convince you of the importance of population density to today's LCC model: Take a look at a map of where Ryanair flies. Under the 58th parallel, on the entire European continent, there is only one place over 250km from a Ryanair airport! Look at all the tiny airports, Limoges or Valladolid or Brive or Osijek or anywhere really. Ryanair flies to 250 airports and they do it with hundreds of 737s operating like clockwork, one short flight to another, not a tiny regional plane to be seen. Meanwhile, Delta flies to a similar number of airports in the US (in the 200s) but the majority of those are with small regional jets which are way less cost-effective than 737s; if you're an LCC, that means that you can forget about flying there. Obviously Delta is flying all those regional jets not for fun, but because they can't fill a 737, because the local population is not large enough to support one. I don't see how population density is not a core part of the issue here.
I think transport in the US might be a factor, though. While Frankfurt-Hahn and Beauvais are particularly extreme examples (they're further from the notional place than most budget airports) you can still get to the the relevant city relatively cheaply (if not particularly _quickly_, in the case of those two; about two hours). In many parts of the US, I suspect you'd really have no option but to hire a car, which instantly wipes out any price advantage of the budget airline.
I think a better measure would be an average of averages of adjacent populations.
Define a node as a population center (say NYC) and then edges as cities (Boston, Philly, DC, Pittsburgh) with the weight being the population of the respective city. Average the populations for each mode and then average those averages.
The idea of this number is it would represent the average population of points of interest.
> Frankfurt-Hahn or Paris Beauvais
Note that they are close to dense population areas Frankfurt and Paris. Not part of average density areas.
I am just saying I have seen the pattern occur repeatedly where "population density" is used to attempt to prop an argument: density is averaging and is often meaningless.