Complements of the house: Estimating demand-side linkages between residential water and electricity

Past studies have estimated residential demand for water and electricity in isolation, but these goods are often used as joint inputs in household production activities. As such, separately estimating electricity and water demand may lead to biased demand parameter estimates. If prices are positively correlated and goods are complements, ignoring cross-price effects will exaggerate own-price elasticity estimates, leading to inaccurate revenue and conservation forecasts. Moreover, understanding the water-electricity demand relationship will allow for synergistic conservation strategies. We propose a joint estimation procedure using 3-Stage-Least-Squares-Fixed-Effects (3SLS-FE) to highlight linkages between water and electricity and conclude that water demand, in particular, appears less own-price elastic when cross-prices are included in the demand system. Results from our study region suggest that water and electricity are gross complements (with an average cross-price elasticity of approximately - 0.1). A simple simulation is included to highlight how omitting cross-price elasticities may lead to inaccurate forecasting and suboptimal decisions.