Only if the introduction of other ride-sharing services happened at the same time - I have no idea of the timeline in NY, but the author send to be using the sudden introduction as a natural experiment.

Of course, the results are probably applicable to other ride-sharing schemes.