Green et al. (1998) [Green, D., Jacowitz, K.E., Kahneman, D., McFadden, D., 1998. Referendum contigent valuation, anchoring, and willingness to pay for public goods. Resource and Energy Economics 20 (2), 85-116] show theoretically that stated preference questions about public services can be framed in such ways that if the subjects accept the frame the payoffmaximizing behavior will be to answer truthfully. One key element of such a theoretically incentivecompatible framing is that the (hypothetical) decision rule specified in the survey instrument is understood to be a majority rule rather than the efficiency rule typically used in costbenefit analysis. We conducted field experiments in Germany and Switzerland to test if a referendum framing as suggested by Green et al. is effective in reducing strategic misrepresentation in a contingent valuation setting. We did not find the expected effects of the framing treatments on stated willingness to pay or on individuals' (stated) beliefs about the social choice context. The results do not support hopes that a theoretically incentive compatible framing could be purposefully used to invoke the specific beliefs about the linkage between responses and policy implementation that would make stated preference questions incentive compatible.