Working Paper: CEPR ID: DP17200
Authors: Michael Cahana; Natalia Fabra; Mar Reguant; Jingyuan Wang
Abstract: We study the distributional impacts of real-time pricing (RTP) in the Spanish electricity market, where RTP was rolled out as the default tariff for a large share of residential customers. We complement aggregate patterns of distributional effects with a novel method for inferring individual households' income using zip code income distributions. We identify three channels for the distributional impacts of RTP: consumption profiles, appliance ownership, and locations. The first channel makes the switch from monthly to hourly prices progressive since high income households consume disproportionately more at peak times when real-time prices are higher. However, in the Spanish context, the other two channels make the switch from annual to monthly prices regressive. In particular, since low income households tend to have more electric heating, they benefit from the price insurance provided by time-invariant prices during winter, when prices are higher and more volatile. Given that price differences are greater across months than within months, the regressive effect dominates in our application. Using counterfactual experiments, we find that RTP makes low income households particularly vulnerable to adverse price shocks during winter. In the future, the wider adoption of enabling technologies (including storage and demand response devices) by the high income groups might worsen the distributional impacts of RTP. Our findings should allow to design an equitable real-time pricing system while retaining its efficiency properties.
Keywords: dynamic pricing; electricity; distributional effects; generalized method of moments; clustering
JEL Codes: L94; H23; C33
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| high-income households (R20) | increase in consumption during peak times (L97) |
| low-income households (R20) | benefit from time-invariant prices during winter (L97) |
| larger price differences across months (E30) | regressive effects (H23) |
| adoption of enabling technologies by high-income groups (O14) | exacerbate distributional impacts of RTP (D39) |
| RTP (R19) | increase in bills of low-income households (H53) |
| RTP (R19) | regressive effects (H23) |