Reputation and Competence in Publicly Funded Science: Estimating the Effects on Research Group Productivity