Is Giffen behaviour compatible with residential demand for cooking gas and kerosene? Evidence from a state in Nigeria

摘要

Purpose - This paper aims to examine the compatibility of Giffen behaviour with residential demand for kerosene and cooking gas. Design/methodology/approach - In total, 600 questionnaires were administered on selected households in Ondo State while 485 were retrieved. Both ordinary least square and instrumental variables (FVs) were estimated, while, the IV estimated result was preferred. Findings - The result showed that Giffen behaviour is compatible with the demand for kerosene in Ondo State, but not for cooking gas. As regard to other factors, prices of the alternatives to kerosene and cooking gas have positive but insignificant impact on the demand for the respective products. Age of the household has a positive significant impact on the demand for kerosene and cooking gas. Household in which the heads has tertiary education demand for kerosene and cooking gas more than those without any form of education. Larger households consume more of both commodities than smaller households. Research limitations/implications - Based on these findings, the authors recommend that government should continue to subsidize either the production or consumption of household kerosene. Practical implications - Consumers should not mind the initial expenditure in purchasing cylinder for cooking gas as subsequent expenditure would be lower than that of kerosene. Social implications - Regulators should brace to ensure that kerosene and cooking gas be made available at government-regulated prices, particularly by checkmating the activities of the "black-marketers." Originality/value - Two outstanding knowledge gaps that this paper filled are in the novelty of this paper regarding the application of Giffen behaviour to kerosene and cooking gas. Second, previous studies did not account for the potential endogeneity problem that is inherent in the joint demand for kerosene and cooking gas. This paper took care of this by estimating the model using IVs.