Economic growth, inflation and oil shocks: are the 1970s coming back?

Abstract

This article analyses the relationship between oil price shocks and the macroeconomic evolution of the Group of Seven (G7) countries. Using the Qu and Perron (2007) methodology, we endogenously identify three breaks in the nonlinear relationship across our 1970 to 2008 sample. We compute long-term multipliers and find that the response of output and inflation to oil price shocks is greatest in the 1970s and progressively disappears until the late 1990s. In contrast to the previous literature, we observe that both effects reappear in the 2000s, especially on inflation. Nevertheless, the transmission of oil price shocks to the economy is weaker than in the 1970s, which means that oil price shocks have lost some of their explanatory power. Precisely identifying these effects is crucial for the design of adequate economic measures to control or smoothen them.

Keywords:

• oil shocks
• inflation
• production
• structural breaks

JEL Classification::

• C32
• E23
• E31
• Q43

Acknowledgements

We acknowledge the useful comments of an anonymous referee. We are also indebted to Pierre Perron and Zhongjun Qu, for their helpful suggestions. Financial support from Spanish Government CICYT projects ECO2008-03040 and ECO2009-13085 is recognized. The usual disclaimer applies.

Notes

¹ This relation has been reported by Bruno and Sachs (1982), Hamilton (1983, 1996, 2003, 2008), Mork (1989), Bohi (1989, 1991), Mork et al. (1994), Lee et al. (1995), Hooker (1996, 2002), Bernanke et al. (1997), Raymond and Rich (1997), Davis and Haltiwanger (2001), Barsky and Kilian (2002, 2004), Lee and Ni (2002), Hamilton and Herrera (2004), Baumeister and Peersman (2008), Blanchard and Gali (2008), Kilian (2008a, 2008b), and amongst others.

² Furthermore, there are several papers that model the effects of oil price shocks on the theoretical side (see, e. g. Rotemberg and Woodford (1996), Finn (2000) and Kilian (2009). More recently, there has been a renewed interest in Dynamic Stochastic General Equilibrium (DSGE) models to investigate the effects of energy price shocks (see, e.g. Nakov and Pescatori (2007), Blanchard and Galí (2008), Crucini et al. (2011)). Another stream of literature considers the asymmetries caused by oil prices shocks on the US business cycle (see, e.g. Hamilton (1996, 2003), Hooker (1996, 2002), Davis and Haltiwanger (2001) and Clements and Krolzig (2002)).

³ Less directly linked, De Gregorio et al. compute IR analysis and rolling bivariate functions to check the pass-through of oil prices to inflation.

⁴ See De Gregorio et al. (2007), Edelstein and Kilian (2007, 2009), Blanchard and Galí (2008) and Kapetanios and Tzavalis (2010), although the first one only studies the effects on inflation.

⁵ For more details, see Bernanke et al. (1997), De Gregorio et al. (2007), Nakov and Pescatori (2007), Blanchard and Gali (2008) and Kilian (2009).

⁶ For details, see Lütkepohl (2006).

⁷ Qu and Perron (2007).

⁸ However, in the case of Italy, given the location of the breaks from the global optimization with two breaks, we need to reduce the trimming to 0.15 in order to locate the third break properly.

⁹ Several calculations have been performed using the price of West Texas Intermediate (WTI) crude; nevertheless, the timing of the breaks does not fit the business cycle as well. Bentzen (2007) finds bidirectional causality among the three major oil prices (Brent, OPEC and WTI).

¹⁰ In this line, De Gregorio et al. (2007) find that, for a wide set of countries, there has been a reduction in the exchange rate pass-through of oil price to inflation.

¹¹ In the case of the UK, the CPI data have been extracted from the OECD's Main Economic Indicators.

¹² In Italy, the second period is longer than in the other countries, ending in 1995 (see Rossi and Toniolo (1996) for a historical economic review of the first two periods).

¹³ In most countries, this period finalizes at the end of the century-beginning of the new one, but in Canada it ends earlier, in 1996.

¹⁴ This lasted a little longer in the US due to the 9/11 terrorist attack but, nevertheless, the US economy reached its maximum mean growth on average in any of the periods.

¹⁵ Jiménez-Rodríguez and Sánchez (2009) consider this as a period of high oil prices, that is, over which the oil price lies above the average of the series since 1970.

¹⁶ In this recent period, speculation may have affected oil price swings even more than supply and demand changes. Oil is traded in the commodities futures market and it is held by speculative companies that consider it a shelter destiny. This futures market may have been responsible not only for the 2007 rises but also for the drastic drop in the price of oil during 2008, when it fell even more precipitously than it rose.

¹⁷ This could be related to the strong growth of the UK oil production and exports together with the decrease of imports during this period. In fact, the North Sea region began its production in 1975 and UK became a net exporter of oil in 1981, ending in 1984 when the increasing demand for oil made imports increase. Manning (1991) studies the relationship between retail petrol prices, excise duties and crude oil prices in the UK, finding asymmetries in the response only the first four months.

¹⁸ Unexpected results for the Japanese economy are also found in Jiménez-Rodríguez and Sánchez (2005) and Blanchard and Galí (2008).

¹⁹ Indeed, a glance at the Canadian series shows the asynchronous movements of these variables.

²⁰ A similar result was obtained by Blanchard and Gali (2008) and Kilian (2008a).