What is the information value of energy efficiency certificates in buildings?

I study the information value of energy efficiency certificates. By using data on repeatedly observed buildings, I separate the rent premium for certified energy efficiency from the premium for readily observable energy efficiency. The buildings were observed before, in-between and after two consecutive law changes that first made certification compulsory and then introduced fines for non-compliers. The strategy allows to control for time-fixed effects of the buildings and for changes in energy efficiency premia over time. I find a precisely measured, but economically and statistically insignificant effect of certification. Supplementary analysis suggests that consumers do value energy efficiency per se, and that energy cost savings translate into higher rents 1-by-1. Further, in a simple theoretical framework, I study the channels through which certification of buildings affects energy consumption. One implication from theory is that compulsory certification is unlikely to be welfare-increasing, even if certificates carry additional information. Given the empirical results of this paper, it is almost certainly welfare-decreasing.