Are US gasoline price adjustments asymmetric?

Abstract

We use the LSE-Hendry general to specific approach to analyse if US gasoline price adjustments are asymmetric with respect to changes in crude oil prices. Furthermore, we modify some weaknesses in the earlier works by Borenstein et al. (1997) and Bachmeier and Griffin (2003) and shows that if the price adjustment equations are properly specified and estimated, alternative specifications and temporal aggregation of data do not affect the results. Monthly US data are used to show that alternative specifications give equally good results and there is no asymmetry in the US gasoline price adjustments.

Acknowledgements

We are grateful to Professor Lance Bachmeier of the Kansas State University and Bill Horvath of the US National Energy Information Center for valuable advice on the US data. We take responsibility for the errors in data and analysis.

Notes

¹ Price changes were never zero in our sample.

² Interestingly in the BG ECM-EG equation the point estimate of the downward adjustment coefficient was also more than the upward adjustment coefficient.

³ It may be of interest to point out that Rao and Rao (2005) used both the ECM-EG and GETS based gasoline price adjustment equations for Fiji. Both equations, with quarterly data, gave similar estimates of the price adjustment coefficients and imply that gasoline price adjustments in Fiji are highly asymmetric. This finding is not surprising because Fiji is a small developing country and, unlike in the developed countries, regulation policies are difficult to implement and police in the developing countries.