Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

15. The developing world: the new energy consumer

Anthony A. Churchill

1. Introduction

This paper is an attempt to look ahead at the energy production and consumption patterns of the developing world and the implications for both the countries themselves and the rest of the world.1 A period of three decades has been selected because anything less would not adequately reveal the underlying trends. The patterns of energy production and consumption are well established and represent a vast amount of capital; change will occur only with changes in basic parameters such as population, economic growth, and technology. The basic database used is that of the report by the Commission of the World Energy Council (WEC), Energy for Tomorrow's World (WEC Commission, 1993). This is the most recent summary of the global scene and its scenarios cover a sufficiently long period of time. The data and assumptions of the scenarios presented are all in line with other similar exercises by the International Energy Agency or the Intergovernmental Panel on Climate Change. I refer to this as the "consensus view."

Table 15.1 Energy demand to 2020: Four scenarios (billion tons of coal equivalent)








OECD 4.1 4.9 4.7 4.7 3.6
Central and Eastern Europe/CIS 1.7 2.0 2.4 1.8 1.5
Developing countries 2.9 10.3 8.0 6.8 6.1
World 8.8 17.2 16.0 13.4 11.3

Source: WEC Commission (1993).

2. Energy demand

Consensus views are not necessarily correct; for example, the consensus view on oil prices in the late 1970s looks slightly ridiculous from today's perspective. Table 15.1 summarizes the demand alternatives of the Commission report. Under the high-demand scenario (case A), the consumption of the developing world about doubles in the 30-year period. The alternatives presented differ in the assumptions about economic growth and the rate at which the energy intensity of output changes. The lowest-demand scenario (case C) is what would have to happen if greenhouse gas emissions were to be stabilized at 1990 levels. It is presented as an illustration of how difficult it would be to accomplish this objective. There are no realistic policies that would make this possible.

All of these scenarios imply a substantial shift in the way energy is used in the developing world. They imply a significant break between what is currently happening and what would have to happen if the demand of the developing world were to double only over 30 years as opposed to doubling every decade as at the present time.

Why should dramatic changes of these orders of magnitude take place? What are the sets of policy changes that would have to be put in place to make this work?

For the most part, these scenarios assume that increasing efficiency brought on by more market-oriented pricing and other policies will permit the continued growth of output with declining inputs of energy. But these are assumptions, not facts - even if the appropriate policy changes were to take place, we have no way of accurately estimating their impact on energy demand. In fact, as I argue below, it is equally plausible to assume that a better set of policies like those generally assumed in these scenarios will increase, not decrease, demand.

Even the high economic growth scenarios are pessimistic. If, on average, the developing world grows only at the rates assumed, the bulk of humanity will continue to live in wretched poverty 30 years from now. These assumptions do not fit the present aspirations of most of the developing world, and two countries alone, India and China, which account for a third of the world's population, are currently growing at rates that are multiples of those assumed. A doubling of the assumed growth rates would nearly double energy demand.

There are alternative ways of looking at the same facts and figures. Rather than focus on global aggregates, let us start with how the individual responds to changes in prices and incomes. If, through changes in policy, technology, or other measures, the production of a service becomes more efficient, it is equivalent to lowering the price of the service. If there is an improvement in the efficiency of the automobile engine so that it is now possible to drive more kilometres for a given fuel input, the price of vehicle kilometres has fallen. Similarly an improvement in the efficiency of the light bulb lowers the cost of the service. The consumer is purchasing the service, not the fuel, and in both cases will experience a decline in price of the service. What is his reaction? Does he purchase more or less? In other words, the amount consumed will vary with the price.

Most of those advocating increased efficiency and conservation as a way of decreasing the quantities consumed implicitly assume that energy services are price inelastic or that consumption does not change with a change in price. A perfectly inelastic demand curve is not supported by the evidence. Estimating price elasticities for energy and other services is notoriously difficult but most of the evidence points to some positive price elasticity - that declines in price will produce increases in the quantity consumed. Whether the increased quantity of the service consumed results in more or less fuel usage depends on the relationship between fuel costs and total service costs. In the United States, for example, improvements in the fuel efficiency of vehicles has not produced an aggregate decline in fuel consumption because of an increase in vehicle kilometres. In Japan, in particular, improvements in efficiency have been associated with rising consumption.

Another way of examining the same phenomena is in terms of expansion of markets. The efficiency of electric power production has continued to increase, for example, from less than 4 per cent heat efficiency at the turn of the century to over 50 per cent in today's combined-cycle plants. Real prices have fallen by a factor of at least five. In response to these changes in price, whole new markets have developed for electric power. In many cases it led to the substitution of a convenient and cheap form of energy for the sweat of the human brow.

In the developing world the potential for the expansion of energy services is enormous. The present inefficiencies of energy production and consumption result in high prices for the services, in effect rationing large segments of the population out of the market-place. Take, for example, electric power production. In almost all countries the price of electricity produced by the public sector is subsidized. But the subsidies introduce inefficiencies and in particular they lead to rationing because the subsidies are insufficient to produce the power that would be demanded at these prices. This forces those not lucky enough to be part of the system to seek more expensive alternatives. People using kerosene lamps or handpumps are paying multiples per unit of service compared with those connected to the grid. Even those connected to the grid are forced to take expensive measures to compensate for the poor quality of the service. In Indonesia or Nigeria, where nearly half the electric power production is in inefficient auto generation that is required to compensate for the poor quality of the public supply, the real cost of electricity is two to three times the price that would prevail under an efficient system of public supply. Imagine how much more electric power would be consumed in India if it were efficiently produced and distributed.

Probably even more important than the impact of efficiency on prices is the impact of increased incomes. The present per capita consumption levels of the developing world are between one-tenth and one-twentieth of those of today's industrial countries. In the developed world (although not yet in Japan), we have seen the per capita consumption levels stabilize or even begin to decline. Most of the assumptions behind the low levels of demand growth in the developing world assume a similar pattern, at least at the aggregate level, of a turn-down in the rate of increase in energy use relative to increases in income.

But is this a realistic assumption? Is it correct to take the changes in the developed world and apply them to the developing countries? In other words, what is the income elasticity of energy consumption? The evidence at hand is mixed and it is not an easy parameter to calculate with any certainty. At the aggregate level, most of the empirical studies come out at between 1.5 and 3.0. None shows this to be a declining figure. At a minimum, a per capita income increase of I per cent will generate a growth in energy consumption of 1.5 per cent. If one looks at both the planned and present construction of energy facilities in most of the developing world in the 1990s, the figures suggest that countries are working on the assumption that energy demand will grow at twice the rate of income. This is consistent with historical experience.

It is also useful to look at income elasticities from an individual or household perspective. At present, average per capita energy consumption is low in developing countries. The average Chinese, for example, uses less than 400 kilowatt hours (kWh) per capita - compared with the US average of nearly 10,000, or the Japanese of 5,000. In these higher-income countries, growth in GNP is increasingly concentrated in services and less energy-intensive activities. The developing world, in contrast, has yet to go through the stage of income growth in which demand for energy-intensive services is a large part of total consumption. Household demand for basic lighting, transportation, appliances, refrigeration, etc., is at relatively low levels. As income grows, demand for these energy-intensive services is likely to be quite income elastic.

One area of increasing demand that probably has been underestimated is transportation. The developing world is undergoing an unprecedented process of urbanization. In the next 30 years most developing countries will have shifted from being primarily rural to being primarily urban. This urbanization is part of the process of economic growth in which higher productivity in agriculture permits a greater degree of specialization in the production and trade in non-agricultural goods and services. Inevitably this will mean greater demand for the transport of goods, all of which are energy intensive. Add to this demand for personal mobility and you have a potentially explosive growth in the demand for energy related to transport needs.

The rapidly growing countries of East Asia are well into this process where the combination of rapid economic growth and rising incomes is producing double-digit rates of growth for transport services. All of today's developed countries have gone through this stage and there is no reason to expect a different experience in the developing world. Improvements in efficiency of transport will permit this increase in demand to take place at lower income levels.

Assuming modest income and price elasticities and a conservative rate of growth, it is possible to get increases in demand that are multiples of what is generally being projected, even with, or perhaps because of, substantial improvements in energy efficiency (Churchill, 1993: 6).

What are the quantitative implications? If energy services are income elastic and using the most conservative of these income elasticities (1.5), a growth rate in per capita income of 2 per cent would produce a growth in energy demand of 3 per cent. To this must be added the growth in population (assume an average of 1.6 per cent over this period), and we now have energy demand growing at 4.8 per cent. How do we take into account the price effects from improved efficiencies or technological progress? Casual observation on the part of many observers suggests that the developing countries are at least 25 per cent less efficient than the developed countries. Let us assume that by the end of 30 years they reach at least the present level of efficiency of the present developed world; that is, real prices will decline by at least 25 per cent. Again, let us make a conservative assumption that consumption will increase more or less in proportion to the decline in prices; that is, demand will be 25 per cent higher than it would have been in the absence of a price decline. We now have the growth of energy demand up to 6 per cent. This is more than five times greater than the growth rates of energy demand implicit in the WEC high scenario and in most other similar "high-case" scenarios, and it is based on using the low end of the estimates of per capita income elasticities.

Different assumptions about growth rates, income and price elasticities, technological progress, and policies can change these results. But it is difficult to see what plausible assumptions about any of these parameters could alter the basic outcome by a factor of five. One could assume declining income elasticities over time, but, given the low initial levels of energy consumption, it is difficult to see why much in the way of declines in this figure can be expected in this period. Present per capita income levels in the major developing countries are one-tenth to one-twentieth of those of the OECD average.

Policy changes can be important. Governments, for example, could choose to try and capture all of the efficiency gains through prices and to soak up the resulting consumer surpluses through the tax system. The practicality of doing so is open to question.

Is this rapid rate of growth in demand likely to be constrained by either supply or financing constraints? It is possible that specific countries will undertake policy and other measures that will restrain the growth in demand by slowing down the rate of economic growth.

But this is hardly a desirable objective. Given appropriate policies, there is no reason to assume that these increases in demand cannot be met within the existing constraints on physical supplies and financial resources.

3. Energy supply

Over the next 30 years there is no reason to assume shortages of the basic raw materials for energy production. The WEC scenarios and most others are consistent in this viewpoint. Particularly with improvements in efficiency and changes in technology, it is likely we face a world of relatively constant energy prices. This is consistent with the historical experience.

The two major economies in the developing world in terms of size are India and China. In both cases these are, and will remain coal-based economies. Both have large reserves of coal that go well beyond the needs of the next 30 years. There is also no reason to suppose this coal will be produced under conditions of rising costs; as has happened in the past, the combination of improved efficiency and more efficient technologies will permit the production of final energy services at prices that are not too different from today's price. It is interesting to note that the world prices of basic fuels into the energy transformation process have been relatively constant in real terms over the past 100 years.

Some concerns have been expressed about the availability of petroleum. Again, most of the supply projections indicate an adequate supply of petroleum at least for the next 30 years. Reserve estimates tend to be on the conservative side because they generally are based on today's technology, and at present costs of capital it makes little sense to spend resources to "prove" reserves for any extended period.

The issue may be more one of distribution. Undoubtedly the present distribution of these resources creates some concerns, but over the longer term it may not make all that much difference. In the production of electric power or other energy services resulting from the boiler use of petroleum, there are a multitude of substitutes, from nuclear to natural gas. It is only in the transportation area that petroleum products have a strong comparative advantage. But even here substitutes are developing. The cost of producing liquid fuels from natural gas has dropped by almost half in the past 20 years to the point where it is close to that of petroleum-based alternatives.

Substantial resources are being spent on the development of the electrically powered vehicle. Given present trends it is highly likely that a commercially feasible vehicle is within the horizon of the next 30 years. Once this is feasible, there are numerous substitutes for petroleum as a transport fuel.

4. Financial constraints

What about financial constraints? All energy transformation processes are capital intensive, and concerns have been raised about the availability of capital to finance the expansion of energy supplies, particularly in the developing world. The savings required to finance the supplies necessary to meet the type of demands forecast above all lie within historical parameters. Typically, in the energy-intensive stages of economic development, energy investments are about 20 per cent of total investment or between 3 and 5 per cent of GDP. These ratios are not expected to change and the type of demand growth forecast above is consistent with the historical pattern. In other words, economic growth and appropriate policies will generate the necessary resources.

The basic resource - the savings - will be available as part of the growth process. The challenge is in mobilizing those resources. Many countries have structured their energy enterprises so that they are both inefficient and subsidized. In these circumstances, only the government through its taxing powers is able to invest. Individuals will not knowingly invest their hard-earned savings in loss-making businesses. There also are limits on how much foreign savings can be attracted into the sector. At present most of these foreign savings are simply relying on the credit standing of the public sector. In any case it is not possible to finance a sector as large as the energy sector on the basis of foreign savings without running into overall balance-of-payment constraints. In order to mobilize the necessary domestic resources, a fundamental restructuring of many of the institutions in the energy sector is required. This is a process that is now under way in many developing countries.

5. Environmental issues

Most energy transformation processes produce wastes that, if improperly handled, can cause serious environmental consequences. Dealing with these wastes has become an increasingly important issue for all countries. Fortunately the technology exists for mitigating most of the serious problems. In most cases the costs of better environmental practices are a small part of total costs and are within the resources constraints of most developing countries.

In the case of the most serious environmental problem with respect to human health and welfare, that of particulates in the air, the costs are relatively minor and are easy to justify in terms of local costs and benefits. Others create more serious issues for cost-benefit analysis. Cross-boundary pollution from acid rain is one such problem. Emissions from Chinese coal plants, for example, impact on acid rain in Japan and other parts of South-East Asia. It is not clear that it is of benefit for China to spend the resources to eliminate this problem for its neighbours. On the other hand, it should be in the interest of these neighbours to provide some of the resources necessary to eliminate this problem. Given the rate of growth of new plant (some 1,000 MW per month) this will not be a trivial expense - approximately US$2 billion per year. This is not unaffordable, however, by the neighbouring countries.

In most cases the waste problems can be met through the direct application of appropriate technologies to eliminate the wastes rather than trying to achieve the same ends through improvements in efficiency. Figure 15.1 shows the trade-offs between reducing emissions through end-use and other efficiency efforts and removing them directly at source. In all cases, once one moves beyond the obvious measures to stop subsidizing inefficiency, it is cheaper to use existing technologies to remove the pollutants at source. Nuclear power, for example, is a cheaper alternative (assuming it can be operated safely) for removing CO2 than draconian policies to improve energy efficiency.

Fig. 15.1 The marginal costs of pollution abatement in electric power through energy efficiency and low-polluting technologies, with reference to developing countries

The one potential problem for which there is no immediate solution is that of greenhouse gas emissions. Over the coming decades almost all of the additional greenhouse gas emissions will be coming from the developing countries. India and China alone will account for about two-thirds of the increase. There is little alternative for these economies but to burn coal and neither one is likely to sacrifice economic growth in response to global concerns about the atmosphere. It is also unlikely that the rest of the world will be willing to contribute sufficient resources to these countries to offset the extra expenses that moving out of coal would entail. Shifting India and China from coal to nuclear power, assuming it were technically possible, would cost around US$15 billion per year over the next two decades.

Improved efficiency and conservation measures will have a limited impact. The 1992 World Development Report (World Bank, 1992) estimated that, under the best of assumptions regarding efficiency, savings of 20 per cent over the next 20 years were a possibility. These best of assumptions may not take place and, as has been suggested above, the cost reductions that are brought about by these efficiency measures may, in fact, increase demand.

If global warming proves to be a serious problem - there is still sufficient uncertainty about it to warrant some caution in spending too much in the way of resources to deal with it - then reliance will have to be placed on technologies as yet unavailable.

Is this an unreasonable expectation? Should we try to solve tomorrow's problems with today's technology? What are the potential alternatives that may become available in the next 30 years? To assume that technology will not change is an unreasonably conservative assumption that is contrary to the historical evidence. Technological change is continuing at a rapid pace. One has only to consider the world of 30 years ago to speculate on the potential magnitude of the changes ahead of us. In the production and consumption of energy there has been a steady improvement in overall efficiencies.

If we look ahead for the next 30 years, it would not be unreasonable to assume a continued or similar pace of development. A number of new technologies are close to commercial feasibility. Solar energy, whether in the form of photovoltaics or thermal, is a renewable technology that offers considerable potential. If costs continue to decline at even half the pace of the past 20 years, solar energy could well become competitive with fossil fuels in the next 20 years (see fig. 15.2). Using electric power to produce hydrogen or the direct production of electric power from fuel cells are all technologies that could prove commercially feasible in the near future. Continued improvements in existing fossil fuel technologies, particularly in using natural gas, are also well within the realm of possibility. Nuclear power is also a possible contender if costs and safety concerns can be met.

Fig 15.2 The cost of alternative means of generating electric power in high-insolation areas, 1970-2020 (Data after 1990 are predicted; future costs of fossil fuel and nuclear generation are uncertain, being affected by such factors as demand shifts, technological change, environmental concerns, and political conditions, which may act in opposite directions. Notes: a. Excluding storage costs. b. Including storage costs -on the basis of hybrid natural gas/solar schemes through 1990 and heat storage thereafter. c. Natural gas and coal)

Will these technologies be available to the developing countries? Inevitably they will be followers: only rich societies can afford to be at the margins of technological progress. Latecomers have the opportunity to capture the gains without bearing many of the risks and costs. In most developing countries the greatest benefits will come from using already tried and true technologies.

The key to the transfer of technology lies in the ability or absorbtive capacity of the recipient. Study after study shows that the greatest gains in productivity come not from the technologies, per se, but rather from the organizational and institutional changes that are required to utilize the technologies efficiently. If labour and management practices are unchanged, little may be gained from introducing new technologies. Unfortunately, the monopolistic structure of most energy enterprises, with their poor incentive systems, does not always permit the most benefits to flow from the introduction of new technologies. More competitive markets in which investors have a clear financial stake in the successful adoption of available technologies appear to be the direction in which countries will have to move if they are to take advantage of the substantial and growing volume of new technologies.

6. Conclusions

Within the next three decades the developing countries are set to become the world's major energy consumers. This growth is to be welcomed because it will be the result of economic growth, with its promise of a better life for some of the world's poorest populations. This growth will not be automatic; substantial institutional changes are required in the energy sector, particularly to improve efficiency and mobilize capital. Nor will it always be as environmentally benign as one would wish. The most serious problems with respect to human health and welfare are essentially local in nature and can be dealt with within the constraints of existing technology. Should global warming prove to be a serious problem, it will have to be dealt with using tomorrow's technology.


1. Much of the analysis developed for this paper is given in further detail in Churchill (1993).
2. There are many examples of this in World Bank (1989).


Churchill, A. A. 1993. Energy Demand and Supply in the Developing World 1990-2020: Three Decades of Explosive Growth. Annual Bank Conference of Development Economics. Washington, D.C.: World Bank, May.

WEC (World Energy Council) Commission. 1993. Energy for Tomorrow's World The Realities, the Real Options, and the Agenda for Achievement. London: Kogan Page.

World Bank. 1989. Technological Advance and Organizational Innovation in the Engineering Industry. Washington, D.C.: World Bank, Industry and Energy Department, Working Paper No. 4, Industry Series, March.

World Bank. 1992. World Development Report 1992. Oxford: Oxford University Press.

Contents - Previous - Next