Oil shocks and the macro-economy: a comparison across high oil price periods

Abstract

Oil prices are found to exert nonlinear effects on major advanced economies. Real output's reaction is much more visible during the periods of high oil prices of the mid-1970s and early 1980s. Oil prices have had inflationary consequences in several economies not only in those two periods, but also during the spike of 1990. There is also evidence that inflation has picked up in some economies as a result of the oil price hikes of 1999–2000, and even more recently in the case of the US.

Acknowledgements

We gratefully acknowledge discussions with Lieven Hermans and seminar participants at the European Central Bank. The views expressed in this article are those of the authors and do not necessarily reflect the position of the European Central Bank. The usual disclaimer applies.

Notes

¹ For a recent survey of the literature on the US, see Hamilton (2008). For evidence on countries outside the US, see Cuñado and Pérez de Gracia (2003) and Jiménez-Rodríguez and Sánchez (2005).

² In this regard, see e.g. Kilian (2005) who has warned against assuming on a priori grounds the exogeneity of oil prices, and Jiménez-Rodríguez and Sánchez (2005) who report block-exogeneity results in favour of bidirectional causality between oil prices and macro-economic variables.

³ We similarly construct a scaled and net oil price decrease variables, SOPD _t and NOPD _t , which are however never found to have statistically significant macroeconomic effects, except for the case of the NOPD _t for France.

⁴ We also considered the alternative ordering in which the real oil prices have a contemporaneous impact on output. The results were found not to change considerably from the baseline specification.

⁵ We opt to deflate oil prices with a US deflator, instead of converting them to domestic currency and then deflating by a domestic price index. The use of the US deflator, together with the inclusion of the real exchange rate in the VAR model, allows us to better decompose the direct impact from the oil shock and the indirect effects via transmission channels, including those involving the nominal exchange rate and domestic inflation. In addition, this facilitates comparability of the results by keeping the nature of the shock constant across countries.

⁶ These results are available from the authors upon request.

⁷ In practice, we detected only two such exceptional quarters, namely, 2001:1 and 2003:2.

⁸ Further to the present periodization based on the notion of ‘high oil prices’, it would be worth attempting to use a related one based on the definition of phases in which OPEC had considerable influence on world oil prices. We leave this for further research. The link between OPEC behaviour and oil prices has been studied in Böckem (2004), Bentzen (2007) and Dées et al. (2007).

⁹ We look at the AIC to select the best model for each economy.

¹⁰ HD reported in Figs 1 and 2 start in 1972:3. HD obtained from different conditional forecasts confirmed our substantive results.