Something missed in the analysis of Hotellings law is the difference between positive and negative competition. In a positive sense neither food cart or shop a

Something missed in the analysis of Hotellings law is the difference between positive and negative competition. In a positive sense neither food cart or shop are any better than the other. However intense negative competition continues to happen and if one food cart op coughs on the food or calls in sick, the reliability of the market is boosted by having the other cart a step away. Likewise, under normal circumstances there is little reason to select one shop over the other, until one shop has a rude clerk, or a slow clerk, or a sick clerk, and suddenly reliability and alternatives is very important. If you're in a non-Hotelling market, you win by doing something revolutionarily better than everyone else. In a Hotelling market, you win by perfect execution of a commodity. I guess a dotcom analogy would be '07 smartphones were a non-Hotelling market when the iphone was introduced, and the android device market is clearly a Hotelling market where they're all about the same but the phone mfgr that screws up the least (hardware reliability, specs, bloatware, whatever) is the winner. The transition between those markets sometimes really trips up businesses and customers.