Individual versus group responses for non-market public goods

Stated preference methods have been used to value non-market public goods but questions remain about their incentive compatibility. A public goods referendum has proven to be incentive compatible for the more popular contingent valuation method. For attribute based choices, the choice experimental method values goods and services by attribute levels producing marginal values for the economic good. A referendum procedure has been proposed for this attribute based method but it is unclear how its use impacts participant choices. This research develops a controlled laboratory experiment eliciting individual and group responses for two non-market public goods using a referendum format for attribute based choice experiments. Results of the experiments find that a majority referendum rule induces individuals to form responses based upon group perceptions which may lead to non-truthful preference revelation.