Abstract
Many empirical exercises estimating demand functions, whether in energy economics or other fields, are concerned with estimating dynamic effects of price and income changes over time. This paper first reviews a number of commonly used dynamic demand specifications to highlight the implausible a priori restrictions that they place on short and long-run elasticities. Such problems are easily avoided by adopting a general-to-specific modeling methodology. Second, it discusses functional forms and estimation issues for getting point estimates and associated standard errors for both short and long-run elasticities—key information that is missing from many published studies. Third, our proposed approach is illustrated using a dataset on Minnesota residential electricity demand.
Highlights
Many empirical exercises estimating demand and supply functions are concerned with estimating dynamic effects of price and income changes over time
The LR elasticities and their standard errors are read directly from the coefficients in the error correction term of the Error Correction Model (ECM), while the SR elasticities and their standard errors are read from the coefficients of the first difference terms in the ECM
Using annual data for the years 1953 to 1990, they find that their regression variables are all I(1) and cointegrated. They use both an ECM and an autoregressive distributed lag (ADL) model to estimate short-run and long-run demand elasticities for coal in China, but they restrict all variables to have the same number of lags and they go up to a maximum of three lags only
Summary
Many empirical exercises estimating demand and supply functions are concerned with estimating dynamic effects of price and income changes over time. Researchers are typically interested in estimating both short-run (SR) and long-run (LR) elasticities, along with their standard errors. It follows immediately from the discussion of the ADL(1,0-1,0-1,01) model above that this specification contains the implausible LR/SR ratio restriction: kβp dqg(q,p,∞) ≡ dp = βp g(q,y,0). The practical implication of the foregoing calculations bears repeating: Estimate an ADL model using a standard lag selection criterion (e.g., the Schwarz or Akaike information criterion) but be sure to allow for at least two terms involving p, ps, and y; the number of lags of the dependent variable (q) does not matter. To our knowledge no energy demand studies have yet used these techniques
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
