Determinants of per game average annual attendance for NBA teams

I explore which characteristics of NBA teams are significant determinants of average annual attendance by team for the 1990-91 to 2014-15 seasons. I use regression analysis to investigate which characteristics determine demand for attendance. My panel data includes independent variables measuring each team’s success, its playing style, level of “star power,” local price of NFL ticket prices, home city population, real income per person, and the local unemployment rate. My paper employs a two-stage regression model to address the endogeneity of ticket prices. The data used in the model spans twenty-five of the most recent NBA seasons. This is a longer range of data and includes more recent data than previously studied. In addition, a city, or in some cases a regional, consumer price index is employed to convert monetary figures into real terms. In past research, only a national consumer price index was used. My paper also considers variables not used in past research. For example, I consider the total number of all-stars on a team to account for “star-power,” and the unemployment rate of each team’s home city. By determining which variables are significant and the direction and magnitude of their impact on a team’s annual attendance, the results may be used to predict future NBA attendance figures, analyze how different teams can improve attendance and revenues, and identify what other cities are the best candidates for attracting new NBA teams. Keywords: Instrumental Variables, National Basketball Association, Attendanc