A fair pivotal mechanism for nonpecuniary public goods

The Clarke pivotal mechanism is inappropriate for nonpecuniary public goods, because the assumption of quasilinear utility is invalid, and because the mechanism gives disproportionate influence to wealthier voters. But by introducing a `stochastic' Clarke tax, we can convert any separable utility function into a quasilinear one. Also, by stratifying a large population by wealth, and applying different `weights' to the votes from different wealth-strata, we can ensure that the mechanism is `fair' in the sense that the voters in different strata all have equal influence (on average) over the outcome. These weights can be fine-tuned to their optimal values over time, by using the rich dataset generated by a series of large-population referenda. The result is a fair, strategy-proof implementation of weighted utilitarian social choice over nonpecuniary public goods.