Sunday, July 13, 2014
Neil Irwin (at Economix, July 11) tell us the latest from Uber and Lyft (still unable to link from you-know-where but source is easy to find). Two things bear repeating. First, one cannot recognize enough the simple fact that sellers learn about demand (elasticities) via trial-and-error pricing (as Uber is apparently doing) if we let them -- if prohibitions against "gouging" are relaxed. I have no idea whether the companies hire econometricians to estimate elasticities but I suspect that trial-and-error is better. This type of essential learning is full-time work. Second, while the old man-vs-machine question has given way to humans-vs-computing, we see that complementarity is more accurate. How do they integrate? Not,how do they compete? It is now hand-held apps plus automobiles. Along these lines, it is not a question about "the end of cities" (George Gilder, 1995, and many others) or not -- now that we are smart-phone-tethered-communicating-continuously-at-near-zero-marginal-cost. Rather how will cities change? The complements-rather-than-substitutes questions points to complements. Cities will not disappear (no signs of that since 1995) but they will continue to spread out -- as always and perhaps a little more and a little faster.
Posted by Peter Gordon at 7/13/2014 12:56:00 PM