In order to examine the role of non-economically measurable factors into the models of voluntary provision of public goods, Experimental Economics showed two systematic findings: (a) in early rounds of game, individual contributions are surprisingly high and increasing, but they tend to decrease progressively for all the agents and to reach a level of cooperation close to zero and (b) complete free riding is never observed. In explaining these evidences, experimental economists have pointed out the importance of factors such as kindness, confusion and strategic cooperation and most laboratory experiments have used clever designs to control for particular factors in order to isolate the effects of others. However, attempts to disentangle and measure the relevant variables of voluntary cooperation are partially and mainly realised comparing pairs of variables, which leaves one of the three hypotheses unexplained. In particular, there are at least two aspects left to be explained: (1) which has more influence in determining the level of contributions: kindness or confusion? (2) What is the role of strategic cooperation? Earlier experiments appear not able to include kindness, confusion and strategic cooperation in the same model and it seems to be very important to separate strategic cooperation from other altruistic motivations. In this work I aim at providing evidence on the way that kindness, confusion and strategic cooperation all affect contributions and to discuss new findings in the measurement and in the disentanglement of these variables made by using a new experimental design with the Voluntary Contribution Mechanism (VCM). I will adopt the same approach as Andreoni (1995), subtracting one component in each treatment and leaving other components as reasonable (and measurable) explanations for cooperation.
