Modeling corner solutions with panel data: Application to the industrial energy demand in France

Abstract

This paper provides an empirical application of Lee and Pitt’s (1986) approach to the problem of corner solutions in the case of panel data. This model deals with corner solutions in a manner consistent with the firm behavior theory while controlling for unobserved heterogeneity. In this model, energy demand at industrial plant level is the result of a discrete choice of the type of the energy to be consumed and a continuous choice that defines the level of demand. The econometric model is, essentially, an endogenous switching regime model which requires the evaluation of multivariate probability integrals. We estimate the random effect model by maximum likelihood using a panel of industrial French plants from the paper and pulp industry. We calculate empirical price elasticities of energy demand from the model. We also study the effects on energy demand of an environmental policy aimed at reducing CO 2 emissions. Copyright Springer-Verlag 2004Energy demand system, zero expenditures, panel data, random effects model, C33, C34, D21, Q41,

Service Offshoring and White-Collar Employment

“…HWS implies that changes in the level of employment of a minor occupation do not a¤ect the marginal rate of technical substitution between any pair of occupations that belong to a di¤erent major group 22 . Under HWS, therefore, the relationship between occupations in the same group is unrestricted.…”

“…2 4 TSO has mainly been exploited in consumption theory to derive optimal demand functions for highly disaggregated goods [see Edgerton (1997) for an example]. In production theory, it has been applied to study optimal demand for di¤erent types of energy by Fuss (1977), Denny et al (1982), Woodland (1993), Chakir et al (2004). 2 5 Non labor inputs include, capital, energy and non-energy materials.…”

Estimating high-dimensional demand systems in the presence of many binding non-negativity constraints

“…Firstly, one may impose the parametric restrictions required to guarantee global concavity of the translog or other flexible functional forms (e.g. Chakir et al, 2004). However, since concavity of the cost or utility function is a sufficient, but not necessary condition, for coherency (van Soest and Kooreman, 1990), this approach is unduly restrictive, thereby destroying the very flexibility these functional forms seek to provide (Lau, 1978;Jorgenson and Fraumani, 1981;Gallant and Golub, 1984;Diewert and Wales, 1987;Terrell, 1996).…”

Essays in consumer theory

“…Lee and Pitt (1986) show how equations (2.39), (2.40), and (2.41) can be used for the derivation of the probability for each purchase regime and, hence, for the model's likelihood function. Specifically, Lee and Pitt (1986) (1986) are provided in Phaneuf (1999), andChakir, Bousquet, andLadoux (2004).…”

“…Specifically, Phaneuf (1999), employed a translog indirect utility function in order to estimate the demand for fishing in four sites of the Wisconsin Great Lakes region, while Chakir, Bousquet, and Ladoux (2004) used panel data and a translog cost function in order to estimate the industrial demand for electricity, gas and oil, in…”