Consider a policy proposal to require each of the 100 households in an economy to rent a
solar panel that costs $400 per year, net of the value of the energy provided to the user. Suppose each panel would prevent $600 worth of harm from pollution in the economy each year. The pollution is uniformly distributed, so each of the 100 households incurs 1/100 × $600 = $6 worth of the harm that could be avoided by each panel yearly. Although society's $600 annual benefit from each panel exceeds the $400 annual cost, each household only internalizes $6 worth of the environmental benefit—far less than the rental cost of a panel. So the privately
optimal decision is to not rent a panel. To reach the socially optimal decision, residents could vote on the policy proposal. If enacted, the policy would cost each household $400 per year. The total damage each household would avoid each year if the policy were enacted—the household's annual benefit from policy enactment—would be 100 x $6 = $600. So the voting mechanism causes each household to internalize the entire $600 yearly benefit to society of purchasing a panel, and the incentive is for households to vote in favor of the socially optimal policy. Suppose instead that each panel would prevent only $300 worth of harm from pollution in the economy each year, again spread uniformly among 100 homes. In that case, it would not be socially optimal for residents to purchase panels, because the $400 annual cost would exceed the $300 annual benefit. Again, a vote would yield the socially optimal solution: If the policy were implemented, each resident would avoid its 1/100 x $300 = $3 share of the harm from each of 100 panels yearly, but this $300 benefit would fall below the $400 annual cost of a panel, so each resident would vote against the requirement and collectively the community would achieve the socially optimal outcome. == References ==