ABSTRACT
Research question
Sport management scholars often draw on institutional theory to examine the adoption of new approaches among sport organizations, but understanding of variation in innovation adoption remains limited. Using a relational perspective to better account for the complexity of the institutional environment, we examine the following questions in the context of esports: how do organizations’ ties to different entities across the field affect the adoption of innovations in sport? And how are these relationships shaped by organizational status?
Research methods
Our sample includes 1274 universities and colleges in the U.S. We used event history analysis to investigate the influence of ties to field-level interest associations (i.e. athletic associations) and local-level peer networks (i.e. athletic conferences, state community settings) on organizational adoption of collegiate varsity esports programs. We also examined the moderating role of organizational status (i.e. university/college ranking). With respect to model selection, we used the Cox proportional hazard model.
Results and findings
Results show that interest associations and local peer networks have a significant effect on the establishment of varsity esports programs. Additionally, organizational status moderates the influence of different relational ties (i.e. interest associations and peers) on esports adoption.
Implications
: This study contributes to the institutional literature on innovation adoption in sport by revealing how relational pluralism and organizational status affect the adoption of new sport programs.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 The U.S. is divided up into eight geographical regions, including Atlantic, Central, Great Lakes, Mideast, Midwest, New England, South/Southeast, and West.
2 Constrained by data availability issues, we were not able to collect longitudinal data on the number of sports programs for all NCAA and NAIA schools. The size of athletic programs is treated as a fixed variable in this study.
3 We ran additional analyses using (1) the exponential proportional hazards model with time as one of the predictors and (2) the Weibull proportional hazards model. Our results across these three models are consistent.
4 We also conducted a robustness check by including all three interaction terms in one saturated model. The results are largely consistent with the outcomes from main models (Model 3–5 in ). The estimator on the interaction term between interest association and organizational status is still significant at the .05 significance level in the saturated model. The estimator on the interaction between geographic proximity and organizational status becomes insignificant at the .10 level in the saturated model.