Anticipating the honeymoon: Event study estimation of new stadium effects in Major League Baseball using the imputation method

Abstract

This paper estimates the impact of new stadiums on attendance and revenues of Major League Baseball teams between 1970 and 2019. Recent studies reveal that two‐way fixed effects (TWFE) models may produce biased estimates, proposing an “imputation” method instead. This paper uses the imputation method to generate a counterfactual estimate, based on untreated observations: the average treatment effect on the treated (ATT) equals the difference between actual and counterfactual estimates. The analysis shows that there were significant anticipation effects associated with new stadiums, up to three seasons before opening. It suggests previous estimates may significantly understate new stadium revenue gains.

Estimating the Fiscal Impact of Stadium Developments: Evaluating a Pro Forma Model

“…Commissioned studies often fail to account for the "novelty" or "honeymoon" effect, which is the shortterm boost in consumer demand that occurs with new stadiums. First, hypothesized and identified 50 years ago by Noll (1974), numerous studies identify an initial attendance shock from opening a new ballpark that diminishes back to its normal level over a decade (Clapp and Hakes 2005;Coates and Humphreys 2005;Depken 2006), and the effect continues to be observed in modern baseball stadiums (Bradbury 2023;Szymanski 2023). These studies use a variety of empirical methods and stadium samples, which supports the robustness of the novelty effect phenomenon.…”

Public policy toward professional sports stadiums: A review

“…The premature replacement of functional sports facilities is incentivized by the "novelty effect"the temporary boost in attendance and revenue that new venues experience during their first few years of operation, which diminishes rapidly within a decade of opening as facility novelty wanes. The phenomenon was first identified by Noll (1974) and has since been documented in numerous empirical studies of second-wave (Clapp & Hakes, 2005;Coates & Humphreys, 2005;Leadley & Zygmont, 2006) and third-wave (Bradbury, 2019(Bradbury, , 2023aSzymanski, 2023) venues.…”