Working Paper: NBER ID: w18037
Authors: Severin Borenstein
Abstract: While time-varying retail electricity pricing is very popular with economists, that support is not matched among regulators and consumers. Many papers have been written estimating and extolling the societal benefits of time-varying rates -- especially dynamic rates that change on a day's notice or less. Yet, such tariffs have been almost completely absent in the residential sector. In this paper, I present a potential approach to implementing an opt-in dynamic pricing plan that would be equitable to both customers who choose the rate and to those who choose to remain on a default flat-rate tariff. The approach bases the dynamic and the flat rate on the same underlying cost structure, and minimizes cross-subsidies between the two groups. I study the potential distributional impact of such a tariff structure using hourly consumption data for stratified random samples of customers from California's two largest utilities. I find that low-income households would, on average, see almost no change in their bills, while low-consumption households would see their bills decline somewhat and high-consumption households would see their bills rise. I also show that the opt-in approach is unlikely to increase the flat rate charged to other customers by more than a few percentage points. I then discuss the most common approach to implementing dynamic electricity pricing -- critical-peak pricing -- and suggest how it might be designed to more accurately match retail price spikes with periods of true supply shortages. Finally, I study the incentive problems created by an alternative program in growing use that pays customers to reduce their consumption on peak usage days.
Keywords: No keywords provided
JEL Codes: L51; L94
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| opt-in dynamic pricing (D49) | customer bills (L84) |
| dynamic pricing (D49) | stability of flat rates (E64) |
| dynamic pricing (D49) | electricity generation costs (L94) |
| program design (C88) | customer behavior (D16) |
| customer consumption patterns (D12) | financial outcomes (G39) |