Doesn't it apply to taxis? There weren't as many taxis as ubers, so they were more expensive and harder to get and took longer to show up. If this was a purely

Doesn't it apply to taxis? There weren't as many taxis as ubers, so they were more expensive and harder to get and took longer to show up. If this was a purely artificial supply constraint (medallian monopolies and abuse, say), then that would be one path to Uber profitability. If this was just a flexibility problem - where taxi driving full time as a requirement for drivers meant you couldn't hit peak demand without an unsustainable amount of idle time too - Uber could have a path there. If it was a sort of "step change" demand function, where if you make it trivially available and cheap the growth in demand will be enough to offset the lower per-ride revenue, there's another potential path there. But it's been a long time now for really none of those theories to pan out.