Natural gas demand at the utility level: An application of dynamic elasticities

Previous studies provide strong evidence that energy demand elasticities vary across regions and states, arguing in favor of conducting energy demand studies at the smallest unit of observation for which good quality data are readily available, that is the utility level. We use monthly data from the residential sector of Xcel Energy's service territory in Colorado for the period January 1994 to September 2006. Based on a very general Autoregressive Distributed Lag model this paper uses a new approach to simulate the dynamic behavior of natural gas demand and obtain dynamic elasticities. Knowing consumers' response on a unit time basis enables one to answer a number of questions, such as, the length of time needed to reach demand stability. Responses to price and income were found to be much lower-even in the long run-than has been commonly suggested in the literature. Interestingly, we find that the long run equilibrium is reached relatively quickly, around 18 months after a change in price or income has occurred, while the literature implies a much longer period for complete adjustments to take place.

Research highlights► Dynamic elasticities are estimated by simulating the dynamic behavior of NG demand. ► LR equilibrium is reached around 18 months following a change in price or income. ► NG demand elasticities are lower than what is generally suggested in the literature.