I watched a TV movie last week about a bird-flu pandemic. Starts in China, whole neighborhoods closed off, food runs out, Penn Station giant hospital, millions of deaths, vaccine, baby on the way, ambiguous ending. Not bad, but not good, either. My mind kept drifting to the spread of The Shangri-La Diet.
It is like a contagious disease in the sense that it can be spread person-to-person: Person A reads the book, tries its weight-loss methods, loses weight or at least appetite, tells Person B about it, Person B reads the book, and so on. It is unlike a contagious disease in the sense that it can be spread by media — newspaper article, radio interview, etc. After spiking to #2 on amazon.com after an interview on the Dennis Prager show, the amazon.com rank has drifted down to about #100.
A contagious disease spreads faster and slower: faster when people are close together (subway), slower when they are apart (sleeping). But with a weight-loss idea the variations of rate of transmission are much larger. At one extreme (slow), someone tries it and then three weeks later someone else notices her weight loss and asks about its source and (if I’m lucky) decides to try it — one person “infected” in three weeks. At the other extreme (fast) is something on national TV, where hundreds of thousands of people become “infected” in minutes.
To get some idea of how the idea is spreading, I have been tracking the growth of the forums at sethroberts.net. Here are graphs of the total number of posts, topics, and members (upper graph) and the rates of their growth (lower graph).
There has been little publicity since the Prager interview about two weeks ago, so the recent growth level of about 5% per day presumably reflects word of mouth/blog. The interesting fact is that the growth appears steady at about 5%.
Berkeley Public Library Watch: The Shangri-La Diet, 4 holds on 1 copy. The Omnivore’s Dilemma by Michael Pollan, 110 holds on 7 copies. Website Watch: Distinct hosts served at sethroberts.net, latest 24-hour period: 948. One week ago: 875. Distinct hosts served is close to the number of different visitors.
There are a number of modeling and predictive tools available in applied economics for this sort of thing.