A socially efficient water tariff under the English optional metering scheme

We design a socially-efficient water tariff in the institutional context of England, where water metering is largely optional and non-metered households are levied proportional to the rateable value (RV) of their property. Within this context, it is theoretically demonstrated that: the larger the RV, the more likely the household to opt for metering; and the larger the RV, the smaller the Demand Effect of Metering (DEM; the fall in water consumption resulting from metering). These two hypotheses are confirmed with econometric analyses using datasets provided by a water company operating in East Anglia, England. The results signify an adverse-selection problem: wealthier households are more likely to opt for metering, yet they are expected to exhibit a smaller DEM once a meter is installed. In order to overcome this, we propose a two-part tariff for metered households consisting of: a variable charge levied proportional to water consumption at a uniform price; and a progressive standing charge to place a heavier burden on wealthier households. The latter component has a potentially major role in attaining social efficiency of metering, by encouraging poorer households to install meters whilst discouraging wealthier ones. The optimal two-part tariff is determined empirically