Information asymmetries and principal-agent relations are crucial issues in network industries regulation. Therefore, besides incentive-based regulations, regulators developed tools dedicated to reveal hidden information and make “regulation smart”. In their very essence, these tools consist in benchmarking services and differ according to the use of benchmarking outputs. In this contribution, we consider performance measurement and sunshine regulation because of a discrepancy between literature and practices. Most of the empirical assessments conclude impacts of sunshine regulation on service performance are not significant while in policies the use of benchmarking increase. Instead of focussing on impacts, we look at the process of disclosure in sunshine regulation arguing that this process is subject to biases avoiding robust analysis of impacts. We assume there are three types of behaviour that cause these biases: opportunism, transaction costs minimisation and pro-social motivations. Our dataset combines 795 observations and results confirm impacts of opportunism and pro-social motivations while we find no persuasive evidence of the impacts of the complexity of indicators calculation.