This study experimentally examines the impact of the level of a minimum investment threshold to finance collective projects under uncertainty. The experimental design is based on a sequential game with imperfect information about the true state of the project, which can be "good" (generating benefits for investors) or "bad" (losing investors their investments). Following the theory of information cascades, we designed an experiment to analyze the effect of the threshold level for the financing of good projects. In concrete, we establish three treatments with different threshold levels: low, medium (optimal), and high. In this context, participants make sequential investment decisions, having both private information and the observed history of previous actions (public information).
The results show that the medium threshold significantly enhances the correct selection and funding of good projects by addressing key inefficiencies associated with both low and high thresholds. Specifically, the medium threshold reduces the proportion of incorrect down cascades compared with the high and low threshold, funds more good projects compared to a high threshold, and limits the number of bad projects funded relative to a low threshold. Although a low threshold increases the overall funding of projects, it hinders the aggregation of information prior to reaching the threshold, resulting in higher incorrect funding of bad projects. On the other hand, a high threshold often prevents good projects from being funded, as its strict requirements can only be met if a cascade up is formed early, that is, if in the initial turns it receives strong support from investing participants.
We find that the threshold level affects the funding of good projects both directly and indirectly. We consider whether pure down cascades have formed at the beginning of the financing round or otherwise whether a player has invested, as a mediating variable. Our treatment variable, the threshold level, significantly influences this mediator: a medium threshold reduces the formation of pure down cascades compared to both high and low thresholds, generating a positive indirect effect on the funding of good projects. While a medium threshold directly funds fewer good projects than a low threshold, this indirect effect partially offsets the difference. Moreover, compared to a high threshold, a medium threshold funds more good projects both directly and through the indirect effect of reducing pure down cascades.