Where did we go right? The effect of theatre choice on Broadway success

Draft of manuscript available here. Abstract: Broadway theater is a billion dollar-per-year industry, with the lion’s share of profits coming from well-known, long-running shows and properties, such as The Phantom of the Opera, The Lion King, and Wicked. While it is almost every Broadway producer’s dream to produce a show that attracts audiences for years, the vast majority of shows close after only a few weeks or months, usually failing to recoup their capitalization costs, which can range from anywhere between 3 to 75 million dollars. Consequently, producers are interested in identifying factors that contribute to a show’s success or failure. However, trends are challenging to study, as there are no easily downloadable datasets available online for public use that contain information about Broadway shows. We provide a web scraping algorithm that organizes data from the Internet Broadway Database for all plays and musicals produced on Broadway for the past 37 years (n=1,334). We then fit a logistic regression model where we define success for a production as running for at least 300 performances. Our analysis is focused on the effect of which of the 41 Broadway theatres a production calls its home on success. Despite small sample sizes for some individual theatres, we find that even when controlling for type of show (play vs. musical) and number of Tony nominations accrued, productions that take place at the Belasco Theatre tend to have shorter runs, whereas productions at the Helen Hayes Threatre and the Palace Theatre tend to be more successful.