A structural time-series model is estimated to investigate the relation between competitive balance, measured by the actual-to-idealised standard deviation ratio, and average match attendance in the Australian Football League from 1945 to 2005. The unobserved components approach allows the data to be modelled in ways new to the literature on this topic. A seemingly unrelated time-series version shows much of the explanatory power of the data to be in the irregular (fast-moving) component. An OLS regression produces robust goodness-of-fit and diagnostic results, and the coefficient estimates produce inferences in contrast to those of Schmidt and Berri (2001) , with persistent shocks.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados