Models for Energy Policy

Jean-Baptiste Lesourd, Jacques Percebois, Francois Valette

• Publisher: Taylor & Francis
• Year: 1995
• Print ISBN: 9780415129756
• eBook ISBN: 9781134793204
• Edition: 1
• Topic: Business
• Subtopic: Business generale
• Index: Business

Energy policy is a key area in each of the world's economies. The oil shocks of the 1970's emphasised how important energy had become. In recent years a growing awareness of environmental issues has had a major impact on perceptions of energy use as growing numbers of people express concern at the relationship between energy and the greenhouse effect, acid rain and the depletion of the ozone layer. All of this has created a demand for more, and better, models of energy use.

Part I
New models and new modelling methodologies

Chapter 1
Methodological advances in energy modelling: 1970–90

James M.Griffin

1.I
INTRODUCTION

The 1970s were unique in that a micro-economic issue, energy economics, suddenly captivated the policy agenda. Not since the 1930s and the Great Depression, had the emphasis of policy makers fallen so squarely on economics. Generous energy research funding fuelled a virtual explosion of energy-related research, attracting a widely diverse group of researchers. Besides varying greatly in equality, the resulting literature varied greatly in breadth, reflecting the different perceptions of what the major policy questions were. Major topic areas included theoretical models of exhaustible resources, cartel theory applications to OPEC, environmental economics, as well as an assortment of energy supply/demand models designed to address some subset of a continually changing menu of policy questions.

Just as oil prices reached their peak in the early 1980s and declined markedly for the remainder of the decade, so too did the energy policy agenda and research funds. Whether measured by the frequency of articles published on energy related issues, NSF funding for energy research, or the energy economics staff at Exxon or British Petroleum, evidence of a sharply reduced research effort is uniform.

Perhaps it is time to look back over the last 20 years and ask, ‘What have we learned?’ The purpose of this chapter is to recount three major methodological advances which have profoundly affected energy modelling and have generated positive externalities for applied research in general. No attempt will be made to survey this voluminous literature or to summarize empirical findings. We begin by summarizing briefly in Section 1.II the status of energy modelling circa 1970. Section 1.III recounts three major methodological advances of the 1970s and 1980s. Section 1.IV recapitulates the major advances and offers my suggestions for a work agenda.

1.II
THE STATE OF ENERGY MODELLING CIRCA 1970

1.II.1
Energy demand

Energy demand research at Resources for the Future (e.g. Schurr et al. 1960) typified energy modelling. An implicit assumption underlying energy demand modelling was that energy was tied to output through a Leontief technology as follows:

According to equation (1), output (Q) was determined by a Leontief relationship depending on value added (VA(K,L)),¹ energy (E), and materials (M) inputs. The technical coefficients γ₁, γ₂ and γ₃ measure the Leontief recipe for the dollar value of any given input necessary to produce a dollars worth of output. Taking GNP as a proxy for output, energy demand was tied to GNP through a technically determined coefficient ( γ₂) as follows:

Thus energy demand forecasts depended entirely on GNP growth. Refinements of this methodology attempted to explain changes in γ₂, linking it to changes in the technical efficiencies governing autos, railways, refrigerators and so forth.

1.II.2
Supply of individual fuels

Having determined aggregate energy consumption in this manner, researchers then proceeded to allocate BTUs among the four primary fuels, hydroelectric, coal, natural gas and oil based on the supply of these fuels. Typically, hydroelectric and coal consumption were set equal to projected hydro and coal production. For example, coal production was viewed as constrained by existing mining capacity and the technical feasibility of adding capacity over a given planning horizon. Natural gas consumption was usually treated as constrained by pipeline capacity constraints and/or reserve levels, leaving oil to serve as the residual. By allowing oil to serve as the residual, all fuels were implicitly treated as perfect substitutes, despite the fact that the BTU equivalent prices of the four fuels were seldom equal.

1.II.3
Price determination

In this system, prices were largely irrelevant both in the determination of aggregate energy demand or the supply/demand functions for individual fuels. Lacking a demand specification for individual fuels, price determination was handled in an ad hoc manner. The long-run supply schedule for coal was assumed perfectly elastic, given wages, so that coal prices could be determined as a markup over mining costs. Natural gas prices and crude oil prices in the US were administratively determined, obviating any need to model these prices. Because of the perceived unimportance of prices, the analysis tended to take on an engineering orientation with a focus on technical factors affecting both supply and demand.

1.II.4
Important caveats

The above description of energy modelling circa 1970 is a stereotypical view based on modelling work at Resources for the Future, the US Bureau of Mines, and the leading oil companies. The problem with stereotypes is that they can be grossly unfair to certain individuals and organizations. Important academic work lead many industry practitioners to begin questioning the prevailing methodology. In a provocative paper in the Journal of Industrial Economics in 1967, F.G.Adams and his student, Peter Miovic, collected energy consumption data for 11 European (including six EEC) countries and showed that the energy/GNP ratio varied significantly over time because of the differing thermal efficiencies of fuels. The transition in Europe from coal with very low thermal efficiencies to petroleum with high thermal efficiencies caused the energy/ GNP ratio to fall markedly. Adams and Miovic argued for adjusting fuels for their differing thermal efficiencies. In another development, looking at the discovery of oil reserves, Franklin Fisher (1964) found that oil prices impacted drilling rates, discovery rates and average discovery size. Fisher’s model offered a clearcut econometric alternative to the prevailing trend line analysis. On the question of interfuel substitution, the role of relative fuel prices as a determinant of demand was illustrated nicely in a famous paper by Balestra and Nerlove (1966). The Balestra/Nerlove paper was important not only for its findings of interfuel substitution between electricity and natural gas but also for its pioneering use of a panel data set.

Besides these key academic papers, linear programming techniques made major inroads in refinery scheduling and expansion planning in the 1960s. Even though these models were non-econometric and engineering based, they provided a complete description of the production surface. Moreover, these optimization models were driven by prices—either cost minimization or profit maximization. Refinery optimization models played an important role in optimizing crude choice, i...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Figures
• Tables
• Contributors
• Acknowledgements
• Introduction
• Part I: New models and new modelling methodologies
• Part II: Application to particular energy policy problems