Competitive attention, superstars and the long tail

We propose a model of competitive attention based on two key premises: i) People have limited information processing capacities and ii) consideration sets are formed according to relative salience. The equilibrium predictions we obtain can help to understand, and connect, diverse empirical phenomena, such as the Paradox of Choices, the Power Law dispersions of key market data (sales, profits, online clicks,...), the relation between advertising expenditures and market shares, the evolution of market inequality, or why evidence favoring a "Long Tail" effect is mixed at best.