It's been a while since I had my probability course based on Sheldon Ross' book "Probability Models", and while I never went into econometrics, I was very interested in the queuing theory section. I thought about this problem, and figured it has a simple answer, but don't recall the formulations well enough to work it:
In the United States and Europe, I observe that when people ride lift/elevator, they push the button for the direction they want to go. If an elevator stops on the floor where they're waiting, but is headed in the opposite direction, they do not embark. This is sound, as a) you want to ride the shortest path and b) the risk that the elevator that does arrive is overburdened is offset by the possibility that another elevator may arrive earlier or soon after with fewer intermediate stops to your destination. Call this the "shortest ride" approach.
However, in some African and Arab countries I've traveled to, I observed that people tend to push both the up and down arrows and embark on the first elevator that arrives, staying with it even if traveling in the wrong direction, until eventually it lands on the floor they want. Call this the "soonest on" approach.
Can we show (with a minimal number of assumptions) that, given other people ride soonest on, riding "shortest path" is on average faster?