The introduction of performance measures in the public sector is largely upheld by a strong believe in the positive effects of naming and shaming on organizational performance. That is, publicizing performance data will by itself have a reputational effect on failing organizations which thus will tend to improve subsequent performance. Citizens are often the ones expected to react on reputational damage with exit and voice options. However, little is do we know about if they in fact do that. Here the effect of naming and shaming is approached in a setting of a benchmarking initiative on gender equality for Danish public organizations between 2003 and 2005. The initiative aimed at measuring intra-organizational gender policy efforts, goals, and actual gender representation at the leadership level. A novel regression discontinuity design is proposed for estimating various causal effects of the initiative. These include the effect of being named and shamed in a national comparison and through comparison with reference organizations based on organizational size. The analysis shows that the naming and shaming efforts were at best ignored. The initiative did not significantly improve performance or cause any unintended gaming effects. The preliminary results point to that naming and shaming had a very limited effect on subsequent organizational behavior and performance.
