Determinants of electricity demand elasticity: A global mixed-effects meta-regression analysis

Abstract

Electricity demand elasticity is a fundamental parameter in energy economics because it informs electricity pricing, demand-side management, energy efficiency programmes, and long-term power system planning. However, reported elasticity estimates vary considerably across empirical studies owing to differences in research design, econometric methods, data characteristics, and study contexts. While previous meta-analyses have primarily focused on estimating average elasticity values, limited attention has been given to explaining the sources of heterogeneity underlying these estimates. This study addresses this gap by examining the methodological and contextual determinants of electricity demand elasticity using a global mixed-effects meta-regression framework. A database comprising 2,354 price and income elasticity estimates extracted from 113 empirical studies was constructed following the PRISMA 2020 guidelines. Separate mixed-effects meta-regression models were estimated for short-run and long-run price elasticity as well as short-run and long-run income elasticity. Moderator variables representing study characteristics, econometric specifications, functional forms, estimation methods, data structures, publication characteristics, and geographical contexts were incorporated to explain between-study heterogeneity. The findings demonstrate that heterogeneity in reported electricity demand elasticity is systematic rather than random. Methodological characteristics exert stronger influences on price elasticity estimates, whereas contextual factors play a greater role in explaining income elasticity. Furthermore, the determinants of elasticity differ across short-run and long-run behavioural responses, highlighting the importance of distinguishing adjustment horizons when synthesising empirical evidence. These findings contribute to the electricity demand literature by providing a unified framework for understanding heterogeneity in reported elasticity estimates and offer practical guidance for researchers and policymakers when selecting elasticity estimates for forecasting, electricity pricing, and energy policy evaluation.