The principal findings of this experiment are: It is somewhat time-consuming but not otherwise difficult to employ the demand-revealing process in groups of 10 to 60. The ability of the demand-revealing process to take account of intensities of preferences is appreciated by many of the persons who have used the process, but the departure from one-man-one-vote is strongly resisted by others. The outcomes reached by demand-revealing differed from majority rule in about one-tenth of the cases. The improvement in efficiency associated with these reversals of majority rule was 2.25 percent of the net benefit of making the efficient decisions. The Clarke taxes, which would decline in percentage terms as group size increased, were 3.04 percent of the net value of efficient decisions for the decisions studied. Copyright Martinus Nijhoff Publishers 1983
