A robust nding of repeated public goods experiments is that high initialcontribution rates sharply decline towards the end. This paper reports onan exploratory experiment designed to discover whether such a decline is simply triggered by the usual experimental practice of publicly informing...