Can you comment on how the typical ad elasticity of 0.1 - presumably short-term - compares to longer term effects like those studied in your time-series work with Professor DeKempe in the 1990s?

Correct. This elacity has no particular time dimension. They're not necessarily long run, so it's okay to call them short run, whether or not there is a translation into a long run effects. It really depends on two things. First of all is the magnitude of the short term effect. What I like to tell my students is if the short term effect is zero and the long term effect is twice that, which is a good approximation, two times zero is still zero. And so you can have a good short term effect. In many cases, to get a long-term effect, you can actually conservatively speaking, multiply that by approximately two, if you want to do it sort of in a quick and dirty way. But of course the much more data rich way is to take a look at specific data and take a look at, for example, on purchase results and repeat purchases, and then you get a real long-term effect.