On the effect of high energy prices on investment

Abstract

Empirical analyses of firm behaviour typically assume that there is a stable relationship between investment on the one hand and changes in the relative prices of inputs, output demand and other determinants on the other hand. However, because of the lumpy nature and irreversibility of investments and the presence of uncertainty about future economic developments, a specific percentage change in relative prices and output demand may not always lead to the same percentage change in capital stocks. That means that different regimes may exist in investment behaviour. We test whether such regimes exist using high-quality data on eight manufacturing industries in the Netherlands. Three different regimes can be identified that are characterized by differences in the relative input price levels and we find that if relative prices take on extreme values, the propensity to adjust the scale of production to changes in demand is very low.

Acknowledgement

We gratefully acknowledge the help of Hein Mannaerts and Henry van der Wiel (both CPB, Netherlands Bureau for Economic Policy Analysis) for providing us with data. We thank Bob Chirinko, Michael Funke, Shelby Gerking, Willem van Groenendaal, Simon Kuipers and the editor and referee of Applied Economics for useful comments on an earlier version of this article. All errors are ours.

Notes

¹Pawlina and Kort (2004) discuss the effects on investment of policies aimed at managing uncertainty.

²See for instance Shen and Wang (2005) and Shen (2005). The latter allows for a smooth transition between regimes. Bo et al. (2006) apply the Hansen panel data estimation procedure in a threshold uncertainty model of investment.

³Note that the specification of Equation 1 allows for two thresholds and hence is able to discern three regimes, but obviously the specification can easily be changed in order to allow for any number of thresholds.

⁴As capital and energy are often found to be complements, we use the wage rate as a deflator of the rental price of capital (see Kemfert, 1998).

⁵This price ratio is lagged one period based on statistical grounds.

⁶The coefficient of variation of the rate of growth of energy prices equals 10.7 while the coefficient of variation of the rate of growth of the wage rate is 0.7.