Modelling energy-economy interactions in small developing countries : a case study of Sri Lanka

Abstract

This report is addressed at modelling energy-economy interactions in small developing countries, those with populations less than 20 million or so and where neither the industrial or energy sectors are dominant. The overall objectives of the research were to learn more about how energy-economy interactions can be usefully modelled for policy purposes, to compare the pros and cons of alternative methods which have been used previously, and to test the feasibility of utilizing simple general equilibrium models by constructing an illustrative model for Sri Lanka.

Various approaches to energy policy analysis--project evaluation, technology assessment, energy sector assessment, macro simulation models, economy-wide optimization models, and computable general equilibrium models-- are surveyed and critically reviewed. A major deficiency of all but the latter two is their failure to account for the important two-way interactions between energy and the rest of the economy which are common in developing countries.

The latter models are general in scope and can include the important energy-economy relationships. Since the computable general equilibrium models are somewhat easier to formulate and solve, they seen most appropriate for the type of countries under consideration. These types of models can analyze a large number of interrelated issues such as: the impact of energy costs and prices on aggregate growth and its sectoral composition; the relationship between energy imports, investment rates, and the balance of payments; the scope for substitution between energy and other factors of substitution; and the effect of energy prices on income distribution and employment.

The Sri Lanka model is meant to illustrate how a simple computable general equilibrium model focussing on these issues can be built rather quickly in a situation with substantial data limitations. The model was constructed with data from existing sources, supplemented by some minimal econometric estimation, and was designed to run on a personal computer.

The model includes eleven sectors: (1) paddy and other annual agricultural crops; (2) tree crops; (3) industry; (4) transportation: (5) housing; (6) services; (7) refined petroleum products; (8) electricity; (9) non-competing imports; (10) crude oil: and (11) traditional fuels. Prices determine factor allocations, production, and final demands. Trade flows are adjusted to ensure that total supply equals total usage. For the tradable goods, prices are exogenous. Electricity prices also are set by government policy. The model calculates prices for transportation and housing which insure supply/demand equilibrium for these non-traded sectors. The model is "closed" by specifying a rule for relating aggregate investment and the balance of payments deficit to national income (GDP). In some cases, the trade deficit is fixed in terms of GDP, and in others aggregate investment is fixed as a share of national income.

Starting from a base year of 1983, the model simulates developments through 1989. Several alternative solutions are discussed to demonstrate how parametric changes can show the sensitivity of key variables to changes in prices, economic policy, and the external environment.