The Energy Connection

Despite growing interest in the interrelationship of energy and economic growth and its emergence into the front rank of policy concerns within industrialized countries, there is a striking absence of solid information on its history and on the more tenuous question of its prospects for the future.

Some years ago, we published at Resources for the Future a detailed quantitative study of the interrelationships between energy and economic growth in the United States over a long period. In that study we found that, between the latter half of the nineteenth century and the first decade of the twentieth century, energy consumption in the United States increased at a faster rate than GNP, while following the end of World War I the growth in energy consumption had generally been at a slower rate than the growth in GNP. Note that these findings do not support the widely held belief that energy and GNP have grown at essentially the same rate in advanced industrial economies. Not only did they not grow at the same rate, but their comparative rates show divergence in different directions depending upon the particular period of U.S. economic history being covered.

This set of statistical findings was subjected to a deeper probing in order to explain the divergent trends for the different time periods. We found that a major explanatory factor for the rise in energy relative to GNP during the earlier period, and its decline relative to GNP in the later period, was to be found in the changing structure of output of the U.S. economy. Whereas the period prior to World War I was one in which the development of heavy industry was the dominant element in national economic growth, the period following the 1920s was one in which lighter manufacturing and the broad service component of national output were growing rapidly. The heavier energy consumption relative to national output was thus associated with the stronger influence of heavy industry, while the comparative decline in energy relative to national product was associated with the less energy-intensive service sector and light manufacturing.

In addition to the change in the composition of the national output, another major influence during the period of declining energy consumption relative to GNP was the change in the composition of energy output. Of particular importance was the far more rapid growth of electricity than of other elements within the energy total. From 1920 to 1955, electricity grew by a factor five times greater than that of all other energy.

Two subtle points illustrated by the impact of electrification reveal major aspects of energy and economic relationships that are frequently overlooked. They are, however, essential to a proper understanding of the role of energy in economic growth.

One point concerns the question of thermal efficiency as compared with the economic efficiency of energy use. There is a school of thought gaining much attention these days which is concerned with what is called "net energy use." In the net energy approach, major attention is paid to energy balances: how much energy is put in compared with how much useful energy is obtained. This is in many cases a useful exercise, but it can convey a partial and, therefore, misleading impression. Thus, in net energy terms, electricity might be regarded as an undesirable energy form because it requires several Btus of fossil fuel to produce 1 Btu of electricity. However, in economic efficiency terms, which we would argue is the decisive factor to consider, electricity has been a very desirable energy form, because its unique characteristics have permitted the performance of tasks in altogether different ways than if fuel had to be used directly as a source of energy.

The impact of electrification in industrial processes is the clearest case in point. A significant but not well-recognized aspect of electrification was in its effect on the overall productive efficiency of the economy, particularly in the manufacturing sector. An historical examination of the organization of production within manufacturing shows that the growth of electrification permitted the layout and organization of productive processes within the factory enormous importance in the growth of manufacturing productivity and thereby in the productivity of the total economy.

Beyond its key part in industrial processes, electricity also has been of unique importance in communications, automatic controls of various kinds, and in the performance of numerous household tasks, to cite but a few examples. The essentially different applications made possible by electricity have greatly multiplied the efficiency with which labor and capital are employed, and have, thereby, enhanced the overall productivity of economic processes. Thermal efficiency considerations notwithstanding, the economic efficiency of electrification has been of outstanding significance.

The second point, related to the first, is that because of electrification not only was the overall productivity of economic processes enhanced, resulting in the greater output of goods and services per unit of labor and capital employed, but, more subtly, so was the productivity of energy use enhanced, resulting in a decline in the amount of raw energy required per unit of national output. In this sense, despite the heat losses involved in its generation, electricity has also enhanced the productivity of energy use, as measured by the ratio of GNP to energy consumption. To repeat, it has done this not because it is a thermally efficient energy form, but because it is an economically efficient energy form.

This is not to suggest that greater thermal efficiency in the generation of electricity has not been significant in the historical record. Indeed, there have been vast improvements in the efficiency with which electricity has been converted from fossil fuels. And there are undoubtedly opportunities for achieving still greater thermal efficiency in the future. Nor do we mean to suggest that electricity is an appropriate energy form for all types of use—comfort heating may frequently represent an undesirable application. However, in historical perspective electricity has left its mark despite, not because of, its net energy characteristics.

In an analogous fashion, it may be reasoned that the utilization of the internal combustion engine, powered by liquid fuels, permitted the substantial mechanization of agriculture which played so great a part in the rising productivity of the American economy. Similarly, the growth of truck transportation made possible the movement of industry away from sites dictated by the location of railroad facilities or waterways. Again, it is worth noting that it is not through its thermal efficiency characteristics, but in its broader economic impacts, that the internal combustion engine has left its imprint.

Thus, the change in the composition of energy output toward the more flexible forms of electricity and gasoline made possible shifts in production techniques and locations within industry, agriculture, and transportation which greatly enhanced the growth of national output and productivity. In enhancing the efficiency with which labor and capital are employed, the changes in the composition of energy output have also enhanced the efficiency with which energy itself has been employed as a factor of production; that is, the relationship between GNP and the consumption of energy has been characterized by a persistent trend of declining energy consumption per unit of national output throughout a large part of this century.

What this past course of development has signified for the growth of jobs is a point worth considering a bit more explicitly. The interrelationship of energy consumption and employment in the process of economic growth displays certain distinct and intuitive features in a macroeconomic and long-term historical context. In a disaggregated and shorter-term perspective, the picture is more elusive. The self-evident proposition that labor productivity is sensitive to the availability of energy-using capital equipment as a complementary productive input shows up in the close historical movement of energy per worker and output per worker. Of course, the paths of these two series are not proportionate. Indeed, the more rapid rise in output per worker than in energy per worker after World War I points to the innovative characteristics of industrial mechanization and electrification already commented upon in connection with the declining energy/output ratio which dates from around this time.
But one cannot draw the inference from these long-term trends that any deceleration in energy use, leaving out the obvious reduction stemming from increased efficiency, necessarily jeopardizes the prospects for future productivity growth. At least 40 percent of U.S. energy consumption represents as much the "proceeds" of income growth (which persons deploy on such things as passenger transportation and household fuels and power) as it does the springboard for growth through its role in the productive process. In other words, no more than 60 percent of yearly energy use goes to the business sector—industry, freight transportation, agriculture, and commercial enterprises.

For the same reason, despite the similarity of movement in energy consumption and employment in the aggregate and over the long term, it seems to us questionable to assert a rigid linkage between these two factors, at least not without probing underlying elements in the relationship. A given shortfall in gasoline supplies, whatever its siege-economy characteristics, is not job-threatening to the extent that the same deficiency in industrial fuel availability would be to the assembly line worker.

We cite these instances to make the point that an understanding of the relationship among energy, economic growth, and employment requires a deep probing in relation to both the structure of the economy and the structure of energy output. It is in these changes in structure that one begins to see more clearly the two-way relationship between energy and economic growth: energy consumption as it is affected by the composition (and not just the level) of national output, and the level of efficiency of national output as it is affected by changes in the structure of energy output. Both are of critical importance to a proper appreciation of energy's relationship to economic growth.

Of what value are these findings for the United States in assessing the future? One factor that has been highlighted as an explanation of the past experience is that of structural change. Now, if as many believe, the economy of the United States and other advanced countries will in the future be shifting ever more heavily in the direction of services, there is some reason, using history as our teacher, for believing that energy relative to GNP will continue to be on a declining trend and, perhaps, at an accelerating rate of decline. This would be the energy counterpart of what has been called the "postindustrial economy."

However, a word of caution is called for. Services are a highly heterogeneous category of activities, and some of them may turn out to be quite energy-intensive. Consider, for example, leisure time activities, which will in the future account for increasingly larger percentages of the personal services which consumers will demand. It is not unusual in the United States for people to travel great distances by plane or automobile to have a skiing weekend, or to engage in other types of leisure activity which require substantial travel. This is obviously a high energy-intensity form of services. There is also a growing trend toward second homes, perhaps the future counterpart of the earlier phenomenon of second and third cars. The construction of such homes, and the travel required to go from the city residence to the weekend residence, may both turn out to be comparatively high in energy intensity. We mention these types of developments only to make the point that one should not too easily fall into the trap of believing that the growth of nonindustrial activities in the future will necessarily be associated with lower intensities of energy use.

One bit of insight into unfolding energy-consumption patterns is provided by a picture of the energy-using characteristics of different income groups. If the share of household budgets devoted to energy fell as families entered higher-income classes, one would be tempted to suggest that, whatever else occurred, rising affluence implied a less-than-proportionate increase in energy demand. In fact, it has been estimated that energy consumption does rise pretty much in line with income as households pass through successively higher income brackets. True, relative utilization of direct energy—motor gasoline, residential fuels, and power—drop off as income grows. But this is apparently more than offset by the rise in indirect energy which is embodied in purchases of nonenergy goods and services. An unanswerable question, of course, is what will happen to the spending patterns of those currently in the high-income brackets. That aspect of future behavior involves pure conjecture.

Another factor that needs to be stressed in interpreting the historical record is that the long period examined in the RFF research was one in which energy prices were falling relative to the prices of other productive factors. Thus, price movements served to favor the substitution of energy for other factors of production. In the future, however, energy prices are expected to be rising relative to the prices of other factors. It may, therefore, be correct to expect that the use of energy will decline relative to that of other factors of production. Clearly, the whole question of energy demand and its response to price change, both up and down, is one which requires intensive investigation.

It is also reasonable to expect that the dampening effects of rising energy prices on the growth of demand will be abetted by advances that will be achieved in energy-use technology. There are undoubtedly a number of energy-saving developments, not necessarily easy to foresee in detail, which will flow from technological response to higher prices of energy and to institutional and other reforms designed to achieve conservation. Just as declining energy prices in the past encouraged the growth of energy-using machines and appliances, so price rises in the future should result in the development and use of energy saving equipment of various kinds.

A better understanding of the combined effects on the growth of future energy consumption of structural changes in the economy, of relative price increases in energy, and of changes in energy-use technology in response to rising energy prices is sorely needed if we are to achieve a dependable basis for projecting energy needs relative to future economic growth. Unfortunately, the historical record for any particular country may be quite inadequate to this task. Consequently, international comparisons of the use of energy should help in assessing consumption to differences in energy prices, and also to differences in economic structure and in the energy-using technologies employed.

An ongoing effort of this kind at Resources for the Future, which is nearing completion in draft form, points to a complex set of factors determining comparative relationships between energy use and economic activities in different countries. A detailed report on this study will be made at a later date.

There is a range of other considerations which will need to be taken into account in assessing the energy-economic growth relation for the future. These arise in the sphere of public policy and social attitudes, and do not readily lend themselves to quantitative study.

There appears to be a growing tendency in public policy to want to mandate energy conservation, or to encourage it through the use of various types of incentives. The lasting impact of such regulatory and institutional factors in altering the relationship between energy and economic growth is difficult to foresee.

In addition, it is necessary to consider the matter of how changes in social attitudes may affect energy-economic growth relationships. This subject generally goes under the name of "changing lifestyles." It is difficult to give description of what this general heading covers because the term means different things to different people.

However, woven into all such considerations is the belief that, in the future, society can (or will) move in direction of wanting a simpler basket of goods and services than has evolved in the high-consumption economies of the industrialized Western World.

There is an additional perspective worth noting. Energy developments have in the past been a dynamic influence in the economic growth and development of industrial economies. Many of the fundamental features of contemporary Western society have had their origins, at least in part, in developments in the field of energy supply technology. One need only think of the emergence of coal in the nineteenth century (and earlier) as a replacement for fuelwood, to recognize the significance of this one energy supply shift for the growth of industrialization in Europe and the United States. For it was the transition from wood to coal that made possible the unimpeded growth of the iron and steel industry and the rapid expansion of railroad transportation. The subsequent development of electrification in the twentieth century, as already noted, has, in similar fashion, been of strategic importance in supporting the growth of productivity, which has led to higher living standards in the industrialized countries. It has permitted the introduction of new technologies into the home in the form of electric lighting and various electric appliances which have fundamentally altered the comfort and convenience of living for the mass of the population. The more recent growth of liquid fuels has brought with it the use of automobiles, which has fundamentally altered the patterns of life of ordinary people by giving them far greater mobility in all aspects of their lives.

It is not necessary to present a complete catalog of all of the basic socioeconomic changes that have depended upon, and indeed been produced by, developments in the field of energy-supply technology. The point is that there is a dynamic element that we overlook at our own risk if we try to freeze energy-supply technology by adopting the cry that is increasingly being heard, "this far and no farther." To be sure, numerous effects which have been produced are negative rather than beneficial in nature. There is air and water pollution, a blighted landscape, and deleterious effects of various kinds which are by-products of energy development. Many of these social costs can be substantially reduced through improved technologies and through institutional and regulatory reforms of various kinds. It is essential that we bend every effort to achieve environmentally benign energy technologies, but it is essential also that we not lose sight of the benefits in terms of improved living standards, decreased drudgery at work and at home, and greater mobility of all kinds—not just physical but also societal—which are the result of the dynamic thrust which energy technology has given to economic life in the industrialized world.