This paper analyzes the effect of energy prices on energy efficiency, separately accounting for operational and investment choices in different sectors. For this purpose, capital stock is characterised by vintages with different intensities of energy use, calculated as a function of exogenously-evolving technology availability and energy prices. Our model separately accounts for substitution between inputs to for production (labour, energy and materials), and the potential for more efficient

more »

... of these inputs by choosing more efficient technologies at the time of investment. The model is estimated for 23 OECD countries across four sectors, and their respective prices for final energy consumption over the period 1990-2005. Vintage representation of capital stock significantly improves the explanatory value of the model at the sector level. Our results imply that rising energy costs result in substantial decline in energy use in the long-run. Abstract This paper analyzes the e¤ect of energy prices on energy e¢ ciency, separately accounting for operational and investment choices in di¤erent sectors. For this purpose, capital stock is characterised by vintages with di¤erent intensities of energy use, calculated as a function of exogenously-evolving technology availability and energy prices. Our model separately accounts for substitution between inputs for production (labour, energy and materials), and the potential for more e¢ cient use of these inputs by choosing more e¢ cient technologies at the time of investment. The model is estimated for 23 OECD countries across four sectors, and their respective prices for ...nal energy consumption over the period 1990-2005. Vintage representation of capital stock signi...cantly improves the explanatory value of the model at the sector level. Our results imply that rising energy costs result in substantial decline in energy use in the long-run. 1 capital. Energy consumption would then depend on the utilization and e¢ ciency characteristics of the stock of equipment. Such an elaborate model could then be simulated to describe the adaptation of the capital stock to energy price shocks. But given the absence of capital stock data needed to re ‡ect the adjustment of the capital stock of energy using equipment, econometricians estimate reduced form single demand equations featuring a distributed lag on price to capture the adaptation of the capital stock." This paper attempts to address this limitation in modelling the e¤ect of energy prices on energy use. Our econometric model explicitly incorporates the capital stock, and separately accounts for operational and investment choices in di¤erent sectors. Speci...cally, we expand the traditional estimation of energy, materials, and labour responses to input price changes by including vintages for the capital stock. Each vintage of the capital stock has its own energy e¢ ciency, which is a function of input prices at the time of investment, and exogenous technological change. In our vintage capital model, a rational cost-minimizing ...rm chooses both the optimal input quantities and the e¢ ciency of new capital stock. The model therefore accounts separately for the ‡exibility of substitution between input factors to production (labour, energy and materials), and the potential for more e¢ cient use of these inputs by choosing more e¢ cient technologies at the time of investment. In doing so, our model allows for adaptation of the capital stock to energy price shocks. 1 Our analysis is based on a new panel dataset, which covers 23 OECD countries and four sectors (agriculture, commerce, manufacturing, and transport) between 1990 and 2005. Compared to earlier studies, our analysis relies on more accurate energy prices in di¤erent sectors and countries based on the end-use fuel prices and sector-speci...c energy mix. As a result, this study is among the few to analyze the e¤ect of energy prices from a cross-country, cross-sector perspective. 2 We estimate the vintage capital model using a translog cost function approach suggested by Berndt and Wood (1975) . However, our cost-share equations are non-linear in factor prices because of the composite e¤ect of input substitution and changes in the e¢ ciency of capital stock. This introduces additional complexity for the estimation of the relevant parameters of the model, and provides a better explanation of energy demand at the sector 1 An alternative econometric approach, which allows for adaptation of capital stock to energy prices is the quasi-...xed input demand model (Berndt, Morrison, and Watkins 1981; Pindyck and Rotemberg 1983; Popp 2001; and Sue Wing 2008). This approach is based on entirely di¤erent assumptions about the nature of the adaptation of the capital stock, and should be treated complementary to our model. For comparison of these approaches (and defense of the vintage capital approach), see Atkeson and Kehoe (1999) . 2 A large number of studies have considered the e¤ect of energy prices from a cross-country withinsector (typically, manufacturing and residential sectors) perspective. The only econometric study known to authors based on cross-country time-series data disaggregated by sector activity and fuel type, which uses theoretically appropriate measures of income and price is Pesaran, Smith, and Akiyama (1998) 2