Green when seen? No support for an effect of observability on environmental conservation in the laboratory: a registered report.

Understanding how humans navigate the tension between selfish and prosocial behaviour is central to addressing social dilemmas and several environmental issues. Many accounts predict that human prosociality would increase in the presence of observing individuals. Previous studies on this observability effect predominantly relied on artificial observability manipulations and low-cost measures of prosociality. In the present Registered Report, we used a recently validated laboratory procedure of repeated dilemmas to test whether the presence of actual observers affects costly prosocial behaviour in the domain of environmental conservation. When completing this dilemma task, participants repeatedly chose between minimizing the length of the laboratory session and minimising wasted energy from a bank of LED lights. Their choices were made either in private or in the presence of actual observers. Contrary to our expectation, we did not observe higher rates of energy-conserving behaviour when participants' choices were being observed. Manipulation and robustness checks indicate that this lack of a finding is unlikely to be owing to arbitrary methodological choices. In view of these findings, we argue that a more comprehensive analysis of situation- and behaviour-specific consequences might be necessary to predict how particular behaviours are affected by observability.