The USA’s geography is very distinct, from an air-transport point of view. First of all, it’s a big country, so air had a major role to play, not just on the trunk routes. But most importantly, the population is distributed around the edges (more or less). Years ago when I had time on my hands, I drew a detailed map of where the US population was, and a very high proportion (I forget how much but I guess it was 80% or even more) was contained in a belt that went most of the way around the country.
It didn’t take a genius to notice that the hubs are distributed around the inner edge of this population belt, not just ATL and DFW but at the time I did this exercise there was also MSP, MKE, CVG, DAY, PHL, PIT, CLT, IAH. While not absolutely in the same mould, CHI, DEN, SLC and STL all had shorter hauls behind and longer hauls beyond. Plus the new hubs that American was trying to set up at Nashville, Raleigh/Durham and San Jose (remember them?).
I did a similar exercise for Europe, and the outcome was almost exactly the opposite. Something like 75-80% of the population of ‘Western’ Europe (one didn’t take much notice of the East in those days) was in the area roughly bounded by Manchester – Hannover – Munich – Marseille. Again, most of the hubs (LHR, FRA, ZRH, CDG) were situated on the edge of this area, which meant that they were strongly directional – LON and PAR had good behind connections for transatlantic routes but not for Eastern routes, the opposite was the case for FRA and ZRH. AMS had most of Western Europe as its catchment for transatlantics and the UK market behind for Eastern routes.
I wonder just how much of a hub LHR is these days, with so many of BA’s spokes, both long and short haul, having moved to LGW. How good a market is longhaul/longhaul? I did see some figures not so long ago stating that the biggest individual LH/LH flow through LHR was Bombay-Toronto, which would explain why Emirates is so desperate to expand into the Canadian market – and dangerous for BA if they succeed.