Are football managers as efficient as coaches? Performance analysis with ex ante and ex post inputs in the Premier league

There is a controversy on sport performance literature about what type of inputs might explain more deeply the performance of sports clubs (inputs specification controversy). By one side, several papers have analysed sports teams' performance using the match-related statistics or wages as inputs, well-known as ex post inputs. By other side, some authors have criticized the use of these ex post inputs, and recommend the use of ex ante inputs, as the market value of the players. We have analysed the performance of football teams estimating technical efficiency with three different inputs specification. The methodologies employed were data envelopment analysis (DEA) and a bootstrapped DEA. Our sample is composed by English Premier League football clubs, during three seasons (2012/13-2014/15). The DEA results indicate that the correlation between the three models is positive and significant. The DEA-bootstrapped results help to restate the robustness of the estimations and endorsed the inputs choices. The correlations of the estimations with market value and match-related statistics are the most striking (90% and 94%, DEA and bootstrapped DEA), which indicate that the existent discussion related to the use of match-related statistics as input is unjustified, because it does not affect significantly the efficiency estimations.