Robust estimates of the impact of broadcasting on match attendance in football

The paper employs data from 2,884 matches, of which 158 were televised, in the second tier of English football (currently known as The Football League Championship). It builds a model of the determinants of attendance that is designed to yield estimates of the proportionate changes in the size of crowds resulting from games being shown on either free-to-air or subscription based channels. The model has two innovatory features. First, it controls for the market size of home and away teams very precisely by including local population measures constructed from the application of GIS software and information on competition from other clubs. Second, it employs the Hausman-Taylor random effects estimator in order to take explicit account of the endogeneity of the television coverage variable and of other variables typically included in earlier studies based on ordinary least squares or fixed effects models of attendance. The Hausman-Taylor estimates of the impact of broadcasting are greater than those reported in such studies. In the case of free-to-air television, the negative impact is estimated as over 20 percent but for subscription television, which carried most of the transmissions, the negative effect was only of the order of 5 percent.