Returns to Information and the survival of Broadway productions

Using total revenue as a measure of success in the market for Broadway productions, we confirm the existence of increasing returns to information using a rank-size regression. The resulting winner-take-all distribution of revenues across productions is explored via survival analysis, with the determinants of success being investigated using proportional hazards regression models. Data is compiled from Internet Broadway Database for 270 musicals, plays, and plays with music that opened between January 2011 and December 2018. Instantaneous rates of failure are found to be lower for revivals and musicals than for original productions and plays. The number of Tony Award wins likewise raises the probability of surviving beyond a given level of revenue. Interestingly, star power at a production's opening lowers the likelihood of closing while star power at the end of the run makes the production more likely to close.