Professor Sterman is presently Director of the MIT System Dynamics Group at the Sloan School of Management.
A system dynamics model is developed for the analysis. The model provides a general disequilibrium representation of the major linkages between the energy sector and the economy. GNP, consumption, investment, wages, and prices, and other major energy and economic aggregates are determined endogenously. Though the model generates the macroeconomic dynamics of the economy, it is based on an explicitly behavioral theory of economic decision-making at the microeconomic level of individuals and firms.
Results show a substantial potential for the first-order effects of depletion (rising capital requirements for energy production, rising real energy prices) to be amplified by feedback mechanisms in the economy, worsening economic performance. Delays in substitution and the development of new energy sources further exacerbate the intermediate term impact of depletion. In particular:
The United States finds itself thrust into a transition. Dependent on nonrenewable oil and gas for about three-fouths of its energy, the nation must find a way to do without these rapidly depleting fuels. Such transitions have occurred before. Figure 1.1 shows the energy transitions of the industrial era in the United States. The two previous transitions each required about sixty years, providing a useful time frame for consideration of the current transition. But unlike the shift from wood to coal and from coal to oil, the transition to a post-petroleum era comes on us unwillingly, through depletion. Unlike the previous transitions shown in the figure, there is no readily available replacement for dwindling oil and gas reserves. Unlike the previous transitions, the nation is dependent on unprecedented amounts of energy (per capita energy use is about 10 times greater now than a century ago while total energy use expanded nearly 35 times in the same period). Unlike the previous transitions, the current transition will be marked by rising real energy prices. Never again will the nation enjoy energy as abundant, inexpensive, and environmentally benign. [3]
Since 1973 there has been growing awareness that the energy transition will be more difficult, time consuming, and expensive than anticipated. There are already many signs of economic stress. During the 1970s, economic growth faltered from the 3.7%/year rate of the 1950s and 1960s to 2.7%/year. The nation experienced the two deepest recessions since the Great Depression, high unemployment, large trade deficits, slackened productivity growth, and the most severe peacetime inflation in U.S. history. While not all the nation's economic woes can be traced to energy, the impact of energy on the economic health of the nation is undeniable. The unemployment, factory shutdowns, hardship, and inconvenience caused by the OPEC embargo of 1973, natural gas shortages of 1976, coal strike of 1978, and gasoline shortages of 1979 all suggest the importance of energy in a modern industrial economy. But energy also affects the economy in more subtle ways: energy prices outpaced inflation for most of the decade, raising the real price of energy and adding to inflationary pressures; growing capital requirements for energy production threaten investment in other sectors of the economy; the costs of producing synthetic fuels and other alternative sources rise as OPEC prices rise; and high OPEC prices transfer income and wealth from oil consuming nations to oil producing nations.
Despite the signs of stress and the acknowledged urgency of the problem there is little agreement on the nature, strength, and relative importance of the myriad interconnections between energy and the economy. A framework is needed to integrate in a consistent and realistic manner the dynamic effects of energy depletion and rising energy costs on economic growth, inflation, and the standard of living.
This study seeks to develop such a framework. The framework consists of a system dynamics model of the national economy. The model is designed to provide a vehicle for understanding the role of energy in the economy and for evaluating the macroeconomic consequences of energy policies. Based on an explicit causal theory of economic behavior at the level of individuals and firms, the model endogenously generates the major energy and economic aggregates including GNP, consumption, investment, real and nominal wages and prices, the rate of inflation, interest rates, and energy production, imports and prices. Because of the model's detailed behavioral representation of the physical and decision-making structure of the various sectors of the economy, policy initiatives such as price controls, tax credits, and subsidies for energy production can be tested realistically in a macroeconomic context.
As a case study, the model is used to gauge the potential for depletion of conventional energy resources to influence economic growth and inflation. The analysis isolates and explores the structural mechanisms and disequilibrium dynamics responsible for the macroeconomic effects. Special consideration is given to the potential for the first-order effects of depletion (declining energy production, rising energy prices, growing capital requirements for energy production) to be amplified by feedback mechanisms in the economy and for delays in substitution and the development of new energy sources to worsen the intermediate-term impact of depletion [4].
The next chapter establishes the need for an integrating framework by examining the models currently available for energy-economic analysis. The discussion of the assumptions made in existing models is used to develop criteria for dynamic modeling of energy-economy interactions. Chapter 3 describes the structure of the model in non-technical terms. Chapter 4 examines the ability of the model to replicate the historical behavior of the energy-economy system, presents the analysis of the effects of depletion on economic growth and the real economy, and explores the sensitivity of the real effects to major uncertainties. Chapter 5 discusses the effects of depletion on inflation. In Chapter 6, the use of the model for policy analysis is demonstrated through detailed examination of a large excise tax on energy coupled with compensating reductions in income taxes. Finally, Chapter 7 summarizes the conclusions, offers caveats where necessary on the limitations of the model, and closes with a discussion of possible extensions to the model. The appendices provide a detailed description of the model equations, the data sources, and instructions for reproducing the simulations presented below. ...
(under construction)
The purpose of this study is to demonstrate the importance of the interactions between the energy sector and the economy at large in studies and policy planning related to the energy transition. Focusing on the journey rather than the destination, the study stresses the adjustment path and disequilibrium dynamics likely to arise as the economy moves through the energy transition. In contrast to many studies of energy-economy interactions which emphasize equilibrium, this study suggests the road to an economy freed from dependence on nonrenewable energy sources is likely to be quite long and rocky, even when a number of idealized and optimistic assumptions are made to cushion the ride.
The feedbacks between energy and the economy and the delays in the adjustment of the economy to the changes wrought by depletion appear to be crucial determinants of both economic behavior and the evolution of the energy sector itself during the transition - a period likely to persist well into the next century. Consequently, neither energy planning nor economic policy can be conducted in isolation from the other, or without consideration of the disequalibrium dynamics of the transition and the long-term (ten to fifty years) repercussions of decisions made in response to short-term pressures.
The substantive findings supporting these conclusions are summarized below, followed by methodological implications for research into energy-economy interactions. Finally, the limitations of the present study are discussed and important extensions identified.
First, as depletion continues to raise the price of conventional energy, energy demand will eventually shift to unconventional sources. As demand for conventional energy drops, the desired workforce and capital stock in the conventional energy sector will drop. Because of long delays in the adjustment of capital to its desired level (both construction delays and the long life of capital itself), and to a lesser extent, delays in the migration of labor between geographical regions and industries, there is transient but significant under-utilization of capital and unemployment of labor in the conventional energy sector as energy demand shifts to unconventional sources. The under-utilization of productive resources occurs just when the unconventional energy sector is most in need of those resources to build up its own capacity, forcing extra capital and labor to be diverted from the other sectors of the economy and thus worsening the first-order effects of depletion.
Second, conventional energy production depends on inputs of energy, both directly and through the energy embodied in the capital stock of the industry. Depletion increases the physical inputs of labor, capital, and energy required to produce energy, thus reducing the net energy yield of exploration and development. More energy is required to produce each barrel, creating a powerful vicious cycle of higher exploration costs, higher energy prices, and still higher exploration costs even when the potential for substitution and conservation in the energy industry is assumed to be high. The positive feedback loop or multiplier effect substantially boosts conventional energy prices above the level indicated by depletion alone.
First, it takes time for firms and consumers to react to an increase in real energy prices. Social surveys suggest people are suspicious of energy price increases and reluctant to accept the permanence of higher energy prices, delaying decisions to improve energy efficiency. Further delays are introduced by research, development, retraining, and retooling needed before suppliers can begin to produce the efficient products demanded by their customers. Institutional delays in, for example, revising building codes to accommodate heavily insulated or solar homes further delay the adjustment of efficiency.
Second, and by far the largest source of the delay, is the long life of energy-consuming capital itself. The average physical life of industrial plant and equipment is on the order of two decades, while that of housing is closer to a century. Though retrofits can improve the efficiency of existing capital and housing somewhat, the actual energy intensity of the economy will lag the efficiency of the newest structures by ten to thirty years. Changes in settlement patterns, transportation networks and lifestyles may require even longer. Thus, even if the long-run potential for substitution and conservation is large, short-run flexibility is limited, increasing the magnitude of the economic impacts of the energy transition and worsening the vulnerability of the economy to shocks.
Third, the economic stresses caused by depletion itself can delay the adjustment of the economy to higher energy prices, further worsening those very stresses. Depletion reduces the resources available to finance retrofits, research and development that can improve efficiency, and investment in efficient plant, equipment, and housing. As a result, the adjustment of the economy to higher energy prices is prolonged, energy demand is maintained, and depletion is worsened, contributing to further price increases and hence to even more economic stress.
More important, the price of unconventional energy itself overshoots its initial (and final equilibrium) value. Unconventional sources of energy represent the backstop sources which, according to accepted theory, should provide a cap or backstop on energy prices. Despite extremely optimistic, idealized assumptions concerning unconventional energy (no depletion effects, a flat long-run supply curve, no environmental constraints, no non-energy resource limits, and a constant construction period), the price of unconventional energy substantially and persistently overshoots its initial value under a wide range of assumptions. The overshoot directly delays the energy transition by over a decade by forcing continued reliance on conventional energy resources, and substantially worsens economic performance during the transition by accelerating depletion and forcing the average price of energy to overshoot its equilibrium level.
Rapid growth and long lead times lead to chronic cash flow problems for unconventional energy industries. Despite heavy reliance on external financing, investment and hence capacity are constrained below the rate required to fill orders, and as a result:
The fifth source of escalation is due to the energy-intensive nature of unconventional energy technologies. The lower the net energy yield of an energy technology, the stronger the cycle of energy price increases, rising production costs, and further price increases. Unlike the other sources of overshoot, the dependence of unconventional energy production on inputs of energy permanently raises production costs.
Though energy price increases alone cannot cause sustained inflation, they can cause the symptoms of economic stress traditionally fought with monetary expansion, thus indirectly leading to sustained inflation. However, allowing the money supply to expand in an attempt to ease the stresses caused by depletion does not appear to be effective. If, at one extreme, firms, workers, and consumers respond rapidly to increases in the money supply, then monetary accommodation will have little impact on real activity, and will primarily add inflation to the other symptoms of stress. If, at the other extreme, firms, workers, and consumers are slow to adjust inflationary expectations and hence prices and wages, there may be a short-run improvement in economic performance (along with inflation). But if so, and as a direct consequence, energy demand will increase, accelerating depletion and worsening the stress on the unconventional energy industry, maintaining import dependence and worsening the long term performance of the economy.
After more than a decade of high and accelerating inflation, the economy is rapidly becoming fully indexed to inflation and the adjustment of wages and prices to the available money supply is occurring faster and faster. As a result, monetary policy probably offers little leverage for mitigating the impact of higher energy prices on the standard of living.
The behavioral insensitivity of the model is the result of 'compensating feedback,' the interactions of the highly interdependent network of (predominently negative) feedback loops that make up a complex system. For example, increasing the assumed lifetime of undiscovered conventional energy in conventional 1980 by thirty years extended production of conventional energy by just ten to twenty years, and had virtually no effect on energy imports. Because higher resource availability and slower depletion lead to lower energy prices, higher demand, and faster economic growth, the additional resources are consumed faster than expected. Higher resource availability also slows the increase in conventional energy prices, delaying the emergence of unconventional sources and further speeding depletion. Compensating feedback and the resulting insensitivity of behavior modes to environmental changes (such as changes in the numerical value of assumptions) is a general property of complex systems, and implies confidence can be placed in the robustness of the modes of behavior identified by the model despite the uncertainty surrounding major assumptions.
The major limitations of the model in its present form fall into two categories: insufficiently broad model boundary and insufficiently detailed and endogenous representation of monetary and fiscal policy. In the present version, there is a highly abbreviated model of the financial system and the channels of monetary policy. Explicit inclusion of the monetary base and money supply process including an endogenous representation of Federal Reserve policies would improve the model's ability to evaluate inflation. Similarly, a more detailed representation of fiscal policy, including endogenous government spending, tax rates, deficits, and transfers would enhance the model's ability to respond realistically to the economic stresses caused by depletion. The effects of energy imports and OPEC's investment and consumption policies on exchange rates could be included.
Another profitable extension to the current model would add explicit installation costs for retrofit, entailing each sector undertaking retrofits acquire 'retrofit services' either from the goods sector or, preferably, from a separate 'retrofit services sector.' Provision of retrofit services would depend on inputs of capital, labor, and energy, and would thus be subject to startup dynamics and interdependencies similar to those facing the unconventional energy sector (though presumably involving lower costs and shorter lead times).
The model could be disaggregated to represent different energy types. While detailed representation of a large variety of primary energy sources and end users is not necessary to gauge the macroeconomic effects of the energy transition, disaggregation of electric and non-electric energy would be useful due to the large differences in the costs, efficiencies, and capital stocks required to utilize electricity and other fuels.
The second class of limitations are potentially more important. The exclusion of non-energy resources and environmental constraints may exclude other important interdependencies and dynamics that could magnify the economic impacts of depletion. In particular, the current model includes a highly idealized and optimistic representation of unconventional energy production. No depletion effects or resource constraints are assumed, long-run costs are stable, no environmental costs or constraints are included, and no constraints on non-energy resources are assumed. Though these assumptions were justified in the present analysis to highlight more fundamental interactions, none of them are true. Many of the technologies touted as backstops depend on nonrenewable resources (synfuels from coal and oil shale, nuclear power, heavy oils, and tar sands). The costs of many of these technologies will rise on the margin as the highest grade ores, thickest coal seams, and sites nearest transportation networks are exploited. These technologies have massive water requirements and will compete for water in regions already beset by shortages and dependent on nonrenewable aquifiers.
The analysis of the energy transition can be integrated with other important long-term modes of economic behavior. Two such modes are particularly important: the Kondratiev or long wave, and the lifecycle of economic development.
Research on the System Dynamics National Model has led to development of an endogenous theory of Kondratiev or economic long waves, 45 to 60 year fluctuations of output and employment revolving around the overexpansion and collapse of the capital-producing industries of the economy. Though not essential in the genesis of the long wave, innovation and technology appear to be strongly influenced by the long wave. Each long-wave expansion appears to be organized around a particular ensemble of technologies. The long-wave expansion that culminated in the Great Depression was characterized by dependence on railroads, steam power, and the telegraph; the post-war boom has been characterized by automobiles and airplanes, internal combustion and electricity, and radio and television, each of which largely replaced it predecessors over the same period despite wide variations in their date of invention. Innovations tend to be 'bunched' together by long waves because
Midway into a capital expansion, opportunities for applying new inventions that require new types of capital become poor. The nation is already committed to a particular mix of technologies and the environment greatly favors improvement innovations over basic innovations. During a long-wave downturn, basic innovation opportunities gradually improve, as old capital embodying the technologies of the preceding buildup depreciates. Near the trough of the wave, there are great opportunities for creating new capital embodying radical new technologies. The old capital base is obsolescent, bureaucracies that thwarted basic innovation have weakened, and traditional methods are no longer sacrosanct.The energy transitions in the United States have also coincided with the peaks in innovation that occur during the troughs of the long wave (Figure 7.1)
A growing body of evidence suggests the economy is currently entering the decline phase of the long wave, raising questions about the interactions of depletion and the current energy transition with long-wave dynamics. The model should be extended to examine the energy transition in the context of the long wave.
Finally, it should be recognized that depletion and the energy transition are one dimension of a larger mode - the life cycle of economic development, a centuries-long process that traces the evolution of an economy through a pre-growth period, economic takeoff, rapid growth, and finally social and physical limits to growth. Energy represents just one - albeit an important one - of the physical constraints to growth.
The present study has pointed out the importance of the mutual dependence of the sectors of the economy on energy, and suggests the inclusion of non-energy resources and environmental constraints could, by introducing still more interdependencies, substantially worsen the economic impacts of depletion. Examining the energy transition at the level of detail presented here but in the context of other physical constraints to growth such as soil fertility, water supply, pollution, and other renewable and nonrenewable resources would help develop an integrated framework for evaluating policies to encourage a smooth transition from material growth to equilibrium in the United States.
Copied when spare time allowed between March 2000 and present.
April 19,2004
http://world.std.com/~awolpert/gtr370.html