This paper uses the input-output methodology known as structural decomposition analysis to discuss Namibian energy use. And the paper makes an additional contribution to the literature on structural decomposition analysis by showing that the hybrid units approach, which has frequently been used in other structural decomposition analyses and in other types of energy studies, is in fact unsuitable at least for this type of analysis.
Energy use in developing countries, and the effects of economic development on energy use, are issues of considerable importance. Economic growth might be expected to lead to higher energy use, both because increased production will require increased energy inputs and because higher income levels are likely to lead to increased household use of energy. On the other hand, higher income levels might also be expected to lead to greater consumption of services and immaterial goods, and with the improved education levels often associated with economic development it should also become easier to adapt already existing energy-efficient technologies from other countries. Which of these factors dominate is of considerable importance for future global energy needs.
The pattern, which was seen historically when the developed countries were industrialising, was one where energy intensities and energy use increased dramatically, together with economic growth, when industrialisation began, but levelled off as the economy matured (Proops, 1984). With increasing consumption of services and immaterial goods, the overall energy intensities of the economies declined, although total energy use continued to grow. After the oil crisis a new phenomenon known as “decoupling” has been observed: For the first time, economic growth has been taking place without energy use growing at the same time. The explanation for this is partly that the importance of non-energy intensive goods and services has increased, and partly that higher energy prices have encouraged the development of new energy-saving technologies.
Many previous studies of energy use in developing countries have concluded that they are in the early stages of a similar process. Park et al. (1993), for instance, in their overview of energy use in developing countries, state that “the link between energy and output growth appears to be still very strong in most developing countries”, that “the output effect (on commercial energy use) is several times greater than other effects in most cases”, that “in most (developing) countries, manufacturing energy consumption is likely to continue to soar” and that “improvements in manufacturing energy intensity may somewhat mitigate the effects of rapidly rising energy requirements, but may not be sufficient to curb their increase”. This fairly typical study concludes that some of the NICs may be entering a more mature economic phase and start resembling developed countries, but that in most developing countries economic growth will lead to higher energy requirements in the producing sectors.
However, the developing countries where energy use has been studied to date have almost exclusively been East Asian and South East Asian countries. Economic growth rates have been very high in many of these countries for extended periods of time, but this growth has largely been driven by increases in inputs and not by productivity improvements. Rather, factor productivities have increased very slowly for many production factors. This suggests that (perhaps for political and institutional reasons) economic agents in these countries have been slow to innovate generally, and not only in applying new energy-saving technologies. The finding that sectoral energy efficiencies have only increased slowly, leading to higher aggregate energy use when economic growth occurred, may thus be a product of specific circumstances in the countries studied rather than a characteristic of developing countries in general.
One important difference between the energy use histories of developed countries and developing countries is that developing countries potentially have access to alternative, energy-saving technologies that were not available when the developed countries were in the early stages of their own industrialisation. There is no obvious reason why agents in developing countries should be less adept than their counterparts in the developed world at utilising new energysaving technologies once these are available, if only the economic incentives and technical skill are in place. It is therefore somewhat worrisome that far-reaching conclusions about the energy use patterns of all developing countries are being drawn based on the experiences from Asian countries only, as countries with other development strategies might be following other patterns. This makes it important to analyse changes in energy use in other developing countries to see whether the trends in these countries are similar to those in Asian countries or not.
After independence in 1990, Namibian energy use has not been characterised by the high degree of regulations and subsidies which have marked the energy sectors in most Asian countries. Instead, some energy sources have been taxed, and cost-recovery policies have been in place for other energy sources, so changes in international energy prices have presumably had a more rapid impact on firms and households in Namibia than in many other developing countries. Since Namibia is not typical for developing countries in this regard, this means that Namibian energy use can provide an interesting comparison to the many developing countries where industrial energy use, and the industrial sector in general, have been subject to a greater extent of policy intervention.
The paper is structured as follows: The first section describes Namibian energy use. The subsequent section explains the input-output methodology known as structural decomposition analysis which is used in this study, followed by a section which discusses the data used. The results are given in the subsequent section, and the paper concludes with a discussion of these results. Apart from analysing energy use in a novel setting, the paper makes an additional contribution to the literature on structural decomposition analysis by showing that the hybrid units approach, which has frequently been used in other structural decomposition analyses and in other types of energy studies, is in fact unsuitable at least for this type of analysis.
