Long before records, film, or radio, live theatrical genres like Vaudeville knit together diverse audiences in communities across the United States. Centrally organized through a set of informationally-oriented offices in New York and Chicago, a handful of large-scale corporate circuits moved tens of thousands of performers between thousands of venues for a weekly audience of millions, bridging the gap between local mores and mainstream tastes. Such systems formed the basis for the American entertainment industry—an innovative business sector that would explode into global prominence during the 1920s.
While scholarship has long pointed towards the importance of Vaudeville for the emergence of American mass culture, the sheer scale of the genre’s operations has made analyzing its systemic functioning difficult. As a result, we still struggle to understand the basic labor patterns that defined the industry—how many performers were employed? How long did they stay on the stage? How did their careers progress over time?
Computational methods offer a way forward. Using a large language model, it was possible to extract touring data from Variety magazine’s “Bills Next Week” column, constructing a database of over 200,000 georeferenced Vaudeville performances. Analyzed at scale, the results shed new light on the workers who built American mass culture, revealing the extent to which the industry was defined, on one side, by rapid turnover among employees, and on the other, by a limited number of constantly performing “troupers.” Examining this previously unstudied cohort through mapping, data analysis, and archival close-reading, it became possible to reconstruct how syndicates strategically moved performers between markets, using metropolitan reputations to pull in profits from smaller communities. Demonstrating the importance of digital methods for recovering the political economy of culture, this paper also develops a set of techniques applicable to a far wider array of business historical questions.