To suggest that reputation is a component of merit is a category error. Reputation isn't part of merit; it doesn't cause merit. Merit may cause reputation. You

To suggest that reputation is a component of merit is a category error. Reputation isn't part of merit; it doesn't cause merit. Merit may cause reputation. You probably mean reputation has some positive correlation with merit, so that reputation can be used as a predictor of merit. And of course, you're right. However, I wasn't addressing how consumers can ascertain merit, but whether merit has similar increasing chacteristics to market success, as described in the article. That is, if someone starts with a little extra merit than someone else, does it also compound as they practice? I wanted to exclude competitive factors. For example, Gladwell's Outliers claims that Canadian hockey players have birthdays biased to part of the year, and that this was because the slight age difference makes child players more successful against their sub-year younger opponents in the same age-group. They get more practice and confidence, and get selected for extra training. But, this competitive aspect gives an artificial advantage to players who are slightly relatively better. And also, this advantage is not present just once, but every time they play, for years. BTW both the points you and I made regarding the merit-market relationship are mentioned in the literature posted by gwern https://news.ycombinator.com/item?id=17052074 https://news.ycombinator.com/item?id=17052074 (last link): > in part because people use the popularity of products as a signal of quality—a phenomenon that is sometimes referred to as “social” or “observational learning” (Hedström 1998)—and in part because people may benefit from coordinating their choices; that is, listening to, reading, and watching the same things as others (Adler 1985). - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3785310/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3785310/