energy-GDP relationship

From The New Palgrave Dictionary of Economics, Online Edition, 2015
Back to top


Energy use and GDP are positively correlated, although energy intensity has declined over time and is usually lower in richer countries. Numerous factors affect the energy intensity of economies, and energy efficiency is obviously one of the most important. However, the rebound effect might limit the possibilities for energy efficiency improvements to reduce energy intensity. Natural science suggests that energy is crucial to economic production but mainstream economic growth theory largely ignores the role of energy. Ecological economists and some economic historians argue that increasing energy supply has been a principal driver of growth. It is possible that historically energy scarcity imposed constraints on growth, but with the increased availability of modern energy sources energy’s importance as a driver of growth has declined. Empirical research on whether energy causes growth or vice versa is inconclusive, but meta-analysis finds that the role of energy prices is central to understanding the relationship.
Back to top


Back to top


Back to top


Figure 1 shows that energy use per capita increases with GDP per capita, so that richer countries typically use more energy per person than poorer countries. In fact, this relationship has been very stable over the last several decades, and the graph for 1971 looks very similar except that most countries were poorer and therefore used less energy. Van Benthem (2015) finds similarly that energy intensity – energy used per dollar of GDP – in today’s middle-income countries is similar to that in today’s developed countries when they were at the same income level.
The slope of the graph is a bit less than 1, so that a 1% increase in income per capita is associated with only a 0.7% increase in energy use per capita (Csereklyei et al., 2016). This means that energy intensity is on average lower in higher income countries, but that this relationship is not so strong (Figure 2).
Globally, total energy use has increased over time and energy intensity has decreased (Figure 3). This is mostly due to countries decreasing in energy intensity as they get richer, as the cross-sectional relationship in Figure 2 has been fairly stable over time. Energy intensity has, however, become more similar over countries over time, so that countries that were more energy intensive in the 1970s reduced their energy intensity by more than less energy-intensive countries on the whole and the least energy-intensive countries often increased in energy intensity. This convergence relationship is shown in Figure 4.
Though data are limited to fewer and fewer countries as we go back further in time, these relationships also appear to hold over the last two centuries – energy use has increased, energy intensity has declined globally and countries have converged in energy intensity (Csereklyei et al., 2016).
Of course, although energy intensity has declined, per capita energy use has increased over time, and when we also take population growth into account total energy use has risen strongly, though at a slower pace than total world economic output. Between 1971 and 2010 total world energy use increased by about 140% and GDP by 270%. Population increased by 80%.
This article next examines the factors that might lead to lower energy intensity with higher GDP and convergence in energy intensity, as seen in Figures 24. Then it reviews the literature on the theoretical relationship between energy and economic growth. Third, it reviews estimates of the elasticity of energy use with respect to GDP. The penultimate section looks at the empirical evidence on the question of whether changes in energy use cause changes in GDP or vice versa. Concluding remarks point to the main gaps in our knowledge.
Back to top

Factors affecting the linkage between energy and GDP

Back to top
We saw above that energy intensity is lower in richer countries and has declined globally over time. What are the reasons for this change in the ratio of energy to GDP? We can use a production frontier approach to examine the factors that could weaken or strengthen the linkage between energy use and economic activity over time. A general production frontier, assuming separability between inputs and outputs, is given by:

where the Qi are various outputs, such as manufactured goods and services, the Xj are various non-energy inputs, such as capital and labour, the Ek are different energy inputs, such as coal and oil, and the AXj and AEk are indices of the state of factor-augmenting technology. The relationship between energy and an aggregate of output such as GDP is then affected by:
  • substitution between energy and other inputs,
  • technological change,
  • shifts in the composition of the energy input, and
  • shifts in the composition of output.
We discuss each of these below. Also, shifts in the mix of the other inputs – for example to a more capital-intensive economy from a more labour-intensive economy – could affect the relationship between energy and output, but this issue has not been much discussed in the literature and so will not be pursued further here. An important factor offsetting the effects of technological change is the rebound effect, so we discuss this separately. Because all of these factors affect energy intensity, energy intensity is a poor proxy for energy efficiency, which is usually defined more narrowly (Ang, 2006; Stern, 2012b).
The theoretical rationale for considering energy as a factor of production is discussed later in this article. Of course, if energy is an input to production, then there is also a derived demand function for energy and the level of output or income is one of the factors that determine demand. Estimates of the income elasticity of energy are also discussed below.
Back to top
Substitutability of energy and capital
Koetse et al. (2008) conduct a meta-analysis of the (Morishima) elasticity of substitution (MES) between capital and energy for an increase in the price of energy. Their base case finds that the MES between energy and capital is 0.216, so that capital and energy are poor substitutes. The MES estimated using panel and cross-section data are greater: 0.592 and 0.848, respectively. It is likely that these larger values reflect long-run elasticities and the lower values short-run elasticities (Stern, 2012a). Relatively little research has looked at whether capital–energy substitution has driven part of the decline in energy intensity. Stern (2012b) found that capital deepening reduced energy intensity by 7% globally from 1971 to 2007. On the other hand, Wang (2011) found that capital accumulation was the main driver of reduced energy intensity in China.
Back to top
Energy efficiency and technological change
There are several ways of measuring changes in energy efficiency that take into account the shifts in other factors, such as the quantities of other inputs. The distance function approach measures the change in energy efficiency as the change in the minimum energy requirement to produce a given level of output holding all other inputs constant. Equation (2) compares the minimum energy requirements given the technologies in two different periods but the same levels of inputs and outputs:

where y is the vector of outputs and x the vector of non-energy inputs with subscripts indicating the periods and Ei( ) is a function indicating the minimum energy required in period i in order to achieve the given outputs given the level of inputs. The functions in Equation (2) can be estimated econometrically (e.g. Stern, 2012b) or non-parametrically.
A second approach is to use an index of energy augmenting technical change. Based on equation (1), the index of energy augmenting technical change can be constructed as:

where Si are the shares of each type of energy in the total cost of energy. Change over time in this index can be computed using an index method such as Divisia aggregation. The actual energy augmentation indices need to be estimated econometrically.
Bottom up, engineering-based measurements of energy efficiency represent a third approach. For example, Ayres et al. (2003) and Warr et al. (2010) estimate the useful work performed per joule of energy by various fuels and uses of energy. With some exceptions, the general trend over the 20th century in the USA, UK, Japan and Austria has been to greater energy efficiency measured in this way. Over shorter periods energy efficiency has declined in some countries, as it has in the long term for some fuels, especially food and feed.
Estimates of the trends in energy efficiency are mixed (Stern, 2011). The direction of change has not been constant and varies across different sectors of the economy. Judson et al. (1999) show that technical innovations tend to introduce more energy-using appliances to households and energy-saving techniques to industry. Stern (2012b) finds that energy efficiency improved from 1971 to 2007 in most developed economies, the former communist countries (including China) and in India. But there was no improvement or a reduction in energy efficiency in many developing economies. Globally, such technological change resulted in a 40% reduction in energy use over the period than would otherwise have been the case and so was the most important driver of reduced energy intensity.
When there is endogenous technological change, changes in prices may induce technological changes. Newell et al. (1999) provide some information on the degree to which energy price increases induce improvements in the energy efficiency of consumer products. For room air conditioners they found that only about one quarter of the gain in energy efficiency since 1973 was induced by higher energy prices. Another quarter was found to be due to raised government standards and labelling. For gas water heaters the induced improvements were close to one half of the total, although much less cost-reducing technical change occurred. Using US data, Popp (2002) similarly finds that increased energy prices have a significant though quantitatively small effect on the rate of patenting in the energy sector. Dechezleprêtre et al. (2011) broaden the analysis to cover patents from 84 national and international patent offices covering various climate mitigation technologies, including renewable energy and energy efficiency technologies. They find that until 1990 patenting in these fields closely followed oil prices. After 1990 patenting increased steadily, although oil prices remained stagnant until 2003. They argue that the increase in patenting since 1990 was driven by environmental policies as they occurred, especially in countries that ratified the Kyoto Treaty.
New energy-using technologies initially diffuse slowly due to high costs of production that are typically lowered radically by a fairly predictable process of learning by doing (Grübler et al., 1999). Diffusion tends to follow a logistic curve, with the speed of diffusion depending, among other things, on how well the innovation fits into the existing infrastructure. Energy-saving innovations such as LED light bulbs would be expected to diffuse rapidly once their price becomes competitive, while more radical innovations that require new support infrastructures diffuse much more slowly due to ‘network effects’.
Research also investigates the factors that affect the adoption of energy efficiency policies or energy efficiency technology (Matisoff, 2008; Fredriksson et al., 2004; Gillingham et al., 2009; Linares and Labandeira, 2010; Wei et al., 2009; Stern, 2012b). Differences in the adoption of energy efficiency technologies across countries and states, over time and among individuals might be optimal due to differences in endowments, preferences or the state of technology. But the rate of adoption may also be inefficient due to market failures and behavioural factors. Market failures include environmental externalities, information problems, liquidity constraints in capital markets, failures of innovation markets and principal–agent problems, such as between landlords and tenants (Gillingham et al., 2009; Linares and Labandeira, 2010). Fredriksson et al. (2004) find that the greater the corruptibility of policymakers the less stringent is energy policy, and that the greater lobby group coordination costs are the more stringent energy policy is. Matisoff (2008) finds that the most significant variable affecting the adoption of energy efficiency programs across US states is citizen ideology. A broad band of states from Florida to Idaho had not adopted any policies.
The rebound effect
Energy-saving innovations reduce the cost of providing energy services, such as heating, lighting and industrial power. This reduction in cost encourages consumers and firms to use more of the service. As a result, energy consumption usually does not decline by as much as the increase in energy efficiency implies. This difference between the improvement in energy efficiency and the reduction in energy consumption is known as the rebound effect. Rebound effects can be defined for energy-saving innovations in consumption and production. In both cases, the increase in energy use due to increased use of the energy service where an efficiency improvement has happened is called the direct rebound effect. For consumer use of energy, the estimated rebound effects are usually small: in the range of 10–30% (Greening et al., 2000; Sorrell et al., 2009). Roy (2000) argues that because high-quality energy use is still small in households in India, demand is very elastic, and thus rebound effects in the household sector in India and other developing countries can be expected to be larger than in developed economies. Fouquet (2014) confirms how the price elasticity of demand for energy services declines and how the direct rebound effect decreased as Britain developed.
In the case of energy efficiency improvements in industry, the rebound effect at the firm level could be large as the firm could greatly increase its sales as a result of reduced costs. However, under perfect competition for an industry supplying domestic demand it is much harder for the industry as a whole to expand output, so the direct rebound effect would be more limited. Rebound effects are likely to be larger for export industries that have more opportunity to expand production (Grepperud and Rasmussen, 2004; Allan et al., 2007; Linares and Labandeira, 2010).
As a result of the reduction in the cost of the energy service, consumers will demand less of substitute goods and more of complementary goods. These include other energy services. Firms will make similar changes in their demands for inputs. There will also be additional repercussions throughout the economy. Non-energy goods whose demand has increased require energy in their production. The fall in energy demand may lower the price of energy (Gillingham et al., 2013; Borenstein, 2015), increasing energy use again, and the efficiency improvement is a contribution to an increase in total factor productivity, which tends to increase capital accumulation and economic growth, which again results in greater energy usage (Saunders, 1992). These additional effects are called indirect rebound effects, though the latter two may be treated separately as ‘macro-level rebound effects’ (e.g. Howarth, 1997). Direct and indirect rebound effects together sum to the economy-wide rebound effect.
Estimates of the economy-wide rebound effect are few in number (e.g. Turner, 2009; Barker et al., 2009; Turner and Hanley, 2011) and vary widely (Stern, 2011; Saunders, 2013; Turner 2013). At the economy-wide level, ‘backfire’, where energy use increases as a result of an efficiency improvement, or even ‘super-conservation’, where the rebound is negative, are both theoretically possible (Saunders, 2008; Turner, 2009). It is usually assumed that the indirect rebound is positive and that the economy-wide rebound will be larger in the long run than in the short run (Saunders, 2008). Turner (2013) argues, instead, that because the energy used to produce a dollar’s worth of energy is higher than the embodied energy in most other goods, the effect of consumers shifting spending to goods other than energy will mean that the indirect rebound could be negative and the economy-wide rebound may also be negative in the long run. Borenstein (2015) presents further arguments for negative rebounds.
All evidence on the size of the economy-wide rebound effect to date depends on theory-driven models, which have limited empirical validation. Turner (2009) finds that, depending on the assumed values of the parameters in a simulation model, the rebound effect for the UK can range from negative to more than 100%. Barker et al. (2009) provide the only estimate of the global rebound effect, estimating the rebound from a set of IEA recommended energy efficiency policies at 50%.
However, these are rebounds in energy use rather than energy intensity. As the economy-wide rebound effect is largely due to an increase in output, the rebound effect probably has small effects on energy intensity.
Back to top
Energy quality and shifts in mix of energy inputs
In the course of economic development, countries’ fuel mix tends to evolve as the mix of energy sources used shifts to higher quality fuels (Burke, 2013). Energy quality is the relative economic usefulness per heat equivalent unit of different fuels and electricity. Fuels have a number of physical attributes that affect their relative qualities, including energy density (heat units per mass unit); power density (rate of heat units produced per unit or per unit time); ease of distribution; the need for a transfer medium; controllability (the ability to direct the position, direction and intensity of energy use); amenability to storage; safety; and environmental impacts (Berndt, 1978; Schurr, 1982; Cleveland et al., 2000; Stern, 2010). Some fuels, in particular electricity, require innovations to allow their use that must be embodied in capital equipment, which can transform the workplace entirely and change work processes, thus contributing to productivity gains (Schurr and Netschert, 1960; Toman and Jemelkova, 2003; Enflo et al., 2009).
In the least developed economies, as in today’s developed economies before the Industrial Revolution, the use of biomass and muscle power dominates. The evolution of the energy mix over the course of economic development and over history in the technologically leading countries depends on each country’s endowments of fossil energy and potential for renewables, such as hydroelectricity, but some regularities apply. The share of electricity in total energy use tends to rise. Low-income countries tend to generate electricity from hydropower and oil, while high-income countries have more diverse power sources, including nuclear power. Direct use of coal tends to rise and then fall over time and with income. Natural gas use has increased significantly in recent decades, mostly in more developed economies. Finally, electricity generated from solar and wind power is only now beginning to take off in more developed economies. Figure 5 illustrates this pattern for the USA.
Surprisingly, relatively few studies evaluate the role of the change in energy mix on energy intensity. Schurr and Netschert (1960) were among the first to recognise the economic importance of energy quality in understanding trends in energy and output. Noting that the composition of energy use has changed significantly over time, Schurr and Netschert argued that the general shift to higher quality fuels reduces the amount of energy required to produce a dollar’s worth of GDP. Berndt (1990) also noted the key role played by the shifting composition of energy use towards higher quality energy inputs.
Cleveland et al. (1984) and Kaufmann (1992, 2004) presented analyses that explain much of the decline in the US energy/GDP ratio in terms of structural shifts in the economy and shifts from lower to higher quality fuels. Kaufmann (2004) found that shifting away from coal use and, in particular, shifting towards the use of oil reduced energy intensity in the USA. This shift away from coal more than explained the decline energy intensity over the entire 1929–99 time period. Other studies find, however, a much larger role for technological change than for changes in the composition of energy in the reductions in energy intensity seen around the world. For example, Ma and Stern (2008) find that interfuel substitution had negligible effects on the decline in energy intensity in China between 1994 and 2003. Technological change reduced energy intensity by more than the actual reduction in energy intensity due to the intensity increasing effects of structural change. Stern (2012b) finds that between 1971 and 2007, changes in fuel mix within individual countries increased world energy use by 4%, while global energy intensity declined by 40%. Shifts in the distribution of economic activity towards countries with lower quality energy mixes, such as China and India, contributed further to increasing energy intensity globally.
Back to top
Shifts in the composition of output
Output mix also typically changes over the course of economic development. In the earlier phases of development there is a shift away from agriculture towards heavy industry, while in the later stages of development there is a shift from the more resource-intensive extractive and heavy industrial sectors towards services and lighter manufacturing. Different industries have different energy intensities. It is often argued that this will result in an increase in energy used per unit of output in the early stages of economic development and a reduction in energy used per unit output in the later stages of economic development (Stern, 2004).
However, there is reason to believe that the energy-saving effects of structural changes are overstated (Henriques and Kander, 2010). When the indirect energy use embodied in manufactured products and services is taken into account, the service and household sectors are more energy-intensive than they first appear. Service industries still need large energy and resource inputs. The service being sold may be intangible, but the office towers, shopping malls, warehouses, rental apartment complexes etc. where the activity is conducted are very tangible and energy is used in their construction, operation and maintenance. Furthermore, consumers use large amounts of energy and resources in commuting to work, going shopping etc.
Furthermore, on a global scale there may be limits to the extent to which developing countries can replicate the structural shift that has occurred in the developed economies, to the degree that this is due to outsourcing manufacturing overseas rather than simply from an expansion in service activities. However, the evidence shows that trade does not result in reductions in energy use and pollution in developed countries through the offshoring of pollution-intensive industries (Levinson, 2010, Aguayo and Gallagher, 2005; Kander and Lindmark, 2006). Additionally, if the service sector does require substantial material support, it is not clear whether the developed world can continue to shift in the direction of a growing service share of GDP indefinitely. In fact, as manufacturing prices have fallen relative to the prices of services, even the relative decline of manufacturing in developed countries is exaggerated when the relative sizes of the sectors are computed in current prices (Kander, 2005).
Kander (2002) and Stern (2012b) find a relatively small role for structural change in reducing energy intensity in Sweden (1800–2000) and the world (1971–2007), respectively. But, using a much finer disaggregation of industries, Sue Wing (2008) finds that structural change explained most of the decline in energy intensity in the USA (1958–2000), especially before 1980.
Back to top
The theory of energy in economic production and growth
Energy as a factor of production
Physical laws describe the operating constraints of economic systems (Boulding, 1966; Ayres and Kneese, 1969). Conservation of mass means that, to obtain a given production output, greater or equal quantities of materials must be used as inputs, and the production process results in residuals or waste (Ayres, 1969). Additionally, production requires energy to carry out work to convert materials into desired products and to transport raw materials, goods and people. The second law of thermodynamics (the entropy law) implies that energy cannot be reused and there are limits to how much energy efficiency can be improved. As a result, energy is always an essential factor of production (Stern, 1997) and continuous supplies of energy are needed to maintain existing levels of economic activity as well as to grow and develop the economy. Before being used in the production of goods and services, energy and matter must be captured from the environment, and energy must be invested in order to extract useful energy (Hall et al., 1986).
Back to top

The mainstream theory of growth

Despite these facts, the core mainstream economic growth models disregard energy or other resources. Aghion and Howitt’s (2008) textbook on economic growth does discuss growth and the environment, but only in a chapter near the end of the book. Acemoglu’s (2009) textbook does not cover the topic at all. There has been some analysis of the potential for resources to constrain growth in the journal literature, but it has mostly been contained within the sub-field of environmental and resource economics, and the main focus has been on the implications of non-renewable resources for economic growth.
Solow (1974) introduced non-renewable resources – which could represent fossil or nuclear fuels – into neoclassical growth models and showed that sustainability – or the ability of a nation to support a constant level of economic production indefinitely – is achievable under certain institutional and technical conditions. Assuming that there is no population growth or technological progress, Solow shows that technology must allow the use of natural resources and manufactured capital – machines and buildings – to be sufficiently responsive to changes in prices. As the price of natural resources relative to that of capital rises, capital is substituted for resources in production. In Solow’s (1974) model the elasticity of substitution is 1, as implied by the Cobb–Douglas production function. This means that resources are essential, but that a constant level of production could be maintained even with infinitesimally small resource inputs. An elasticity of substitution greater than unity means that resources are not essential and so achieving sustainability is much easier. These are all conditions concerning the technology available to society. But the institutional framework – for example, whether an economy is a free market economy or whether it follows a particular planning rule – is just as important. From an institutional perspective, sustainability can be achieved only if the welfare of future generations is given equal weight to that of the present generation. This implies that the discount rate used to aggregate costs and benefits over time must be zero.
If instead the economy is a free market economy with perfect competition, but has the same technology as Solow’s (1974) model, the resources are exhausted and consumption and social welfare eventually fall to zero (Stiglitz, 1974a). Dasgupta and Heal (1979) showed that with any constant discount rate the efficient growth path also leads to eventual depletion of the natural resource endowment and the collapse of the economy. Hartwick (1977, 1995) has shown that, if sustainability is technologically feasible, a constant level of consumption can be achieved by investing the rents from exhaustible resources in other forms of capital, which in turn can substitute for exhausted resources. It is difficult to apply this rule in practice, as the rents and capital must be valued at prices that are compatible with sustainability (Asheim, 1994; Asheim et al., 2003; Pezzey, 2004). Such prices are unknowable given that we have poor understanding of the costs of current environmental damage and resource depletion or of the future development of technology.
In addition to the substitution of capital for resources, technological change might permit continued growth or at least constant consumption in the face of a finite resource base. Stiglitz (1974b) showed that, when the elasticity of substitution between capital and resources is 1, exogenous technical progress will allow consumption to grow over time if the rate of technological change divided by the discount rate is greater than the output elasticity of resources. Technological change might enable sustainability, even with an elasticity of substitution of less than 1. Once again, technical feasibility does not guarantee sustainability. Depending on preferences for current versus future consumption, technological change might instead result in faster depletion of the resource (Smulders, 2005).
Back to top

The ecological economics approach

A prominent tradition in ecological economics, known as the biophysical economics approach (Hall et al., 1986), is based on thermodynamics (Georgescu-Roegen, 1971; Costanza, 1980; Cleveland et al., 1984; Hall et al., 1986, 2003; Ayres and Warr, 2005, 2009; Murphy and Hall, 2010). Ecological economists usually argue that substitution between capital and resources can only play a limited role in mitigating the scarcity of resources (Stern, 1997). Furthermore, some ecological economists downplay the role of technological change in productivity growth, arguing that growth is a result of either increased energy use or innovations allowing the more productive use of energy (Hall et al., 1986, 2003; Cleveland et al., 1984; Ayres and Warr, 2009). Therefore, in this view, increased energy use is the main or only cause of economic growth.
In this approach, value is derived from the action of energy that is directed by capital and labour. Energy flows into the economy from fossil fuels and the Sun. In some biophysical economic models, geological constraints fix the rate of energy extraction so that the flow rather than the stock can be considered as the primary input to production (Gever et al., 1986). Capital and labour are considered as intermediate inputs that are created and maintained by the primary input of energy and flows of matter. The level of the flows is computed in terms of the embodied energy use associated with them. Prices of goods should then ideally be determined by their embodied energy cost (Hannon, 1973) – a normative energy theory of value – or are seen as actually being correlated with energy cost (Costanza, 1980) – a positive energy theory of value (Common, 1995). This theory – like the Marxian paradigm – must then explain how labour, capital etc. end up receiving part of the surplus (Kaufmann, 1987; Burkett, 2003; Hornborg, 2014).
However, because the quality of resources and the level of technology do affect the amount of energy needed to produce goods and services, it is difficult to argue for a model where energy is the sole factor of production (Stern, 1999). For example, the quality of resources such as oil reservoirs is critical in determining the energy required to extract and process fuels. As an oil reservoir is depleted, the energy needed to extract oil increases. On the positive side, improved geophysical knowledge and techniques can increase the extent to which oil can be extracted for a given energy cost. Odum’s energy approach (Brown and Herendeen, 1996) and the framework developed by Costanza (1980) address the resource quality issue by including the solar and geological energy embodied in natural resource inputs in indicators of total embodied energy. An alternative approach is to measure material and energy inputs on the common basis of their exergy (Ayres et al., 1998; Ukidwe and Bakshi, 2007).
However, both approaches seem too reductionist. For example, other services provided by nature, such as nutrient recycling, the provision of clean air and water, pollination and the climate system, that make economic production – and life itself – possible should also then be accounted for. Models that allow a number of different factors of production while complying with the physical laws of the conservation of mass and thermodynamics to varying degrees were developed by Georgescu-Roegen (1971), Perrings (1987), and O’Connor (1993) among others. The ecological economics approach does not have to reduce to an energy-only model of the economy.
A key concept in biophysical economics is energy return on investment (EROI), which is the ratio of useful energy produced by an energy supply system to the amount of energy invested in extracting that energy. Lower quality energy resources have lower EROIs. Biophysical economists argue that the more energy that is required to extract energy, the less energy is available for other uses and the poorer an economy will be. In this view, the increase in EROI allowed by the switch from biomass to fossil fuels enabled the Industrial Revolution and the period of modern economic growth that followed it (Hall et al., 1986).
Thus, declining EROI would threaten not just growth but overall economic output and, therefore, sustainability. Murphy and Hall (2010) document EROI for many energy sources, arguing that it is declining over time despite the extensive innovation in the industry. Wind and direct solar energy have more favourable EROIs than biomass fuels, but worse than most fossil fuels. However, unlike fossil fuels, the EROI of these energy sources tends to improve over time due to innovation (Kubiszewski et al., 2010). Declining EROI could be mitigated by substituting other inputs for energy or by improving the efficiency with which energy is used. However, biophysical economists argue that both these processes have limits.
Substitution can occur within a category of similar production inputs – for example between different fuels – and between different categories of inputs – for example between energy and machines. There is also a distinction to be made between substitution at the micro level – for example within a single engineering process or at a single firm – and at the macro level – in the economy as a whole.
As shown in Figure 5 for the USA, the long-run pattern of energy use in industrial economies has been dominated by substitutions from wood and animal power to coal, oil, natural gas and primary electricity (Grübler et al., 2012). Meta-analysis of existing studies of interfuel substitution suggests that the long-run substitution possibilities at the level of the industrial sector as a whole are good. But there seems to be less substitutability at the macro-economic level (Stern, 2012a).
Ecological economists emphasise the importance of limits to inter-category substitution – in particular, the substitution of manufactured capital for resources including energy (Costanza and Daly, 1992). Thermodynamic limits on substitution can be approximated by a production function with an elasticity of substitution significantly below one (Stern, 1997). As discussed above, a meta-analysis of the existing empirical literature finds that the elasticity of substitution between capital and energy is less than 1 but much greater than 0 (Koetse et al., 2008).
In addition to this micro-economic limit to substitution, there may also be macroeconomic limits to substitution. The construction, operation and maintenance of tools, machines and factories require a flow of materials and energy. Similarly, the humans that direct manufactured capital consume energy and materials. Thus, producing more of the ‘substitute’ for energy – manufactured capital – requires more of the thing that it is supposed to substitute for. This again limits potential substitutability (Cleveland et al., 1984).
The mainstream economic argument that technological change can overcome limited substitutability would be more convincing if technological change were really something different from substitution. Changes in technology occur when new techniques are developed. However, these new techniques represent the substitution of knowledge for other factors of production. The knowledge is embodied in improved capital goods and more skilled workers and managers. But there are still thermodynamic restrictions on the extent to which energy and material flows can be reduced in this way. Although knowledge is non-rival in use, it must be used in conjunction with the other inputs, such as energy, and the productivity of knowledge is limited by the available quantities of those inputs.
Back to top

Synthesis: unified model of energy and growth

As a first step to integrating the ecological economic and mainstream approaches and explaining historical economic growth, Stern and Kander (2012) add an energy input that has low substitutability with capital and labour to Solow’s (1956) growth model. As discussed above, low substitutability between capital and energy is one of the key assumptions of ecological economists. Using 200 years of Swedish data Stern and Kander estimate that the elasticity of substitution between energy and the other two inputs is 0.65. This figure is similar to the other estimates of the elasticity also discussed in the previous section. Stern and Kander ignore the issue of whether the energy resource is non-renewable, as depletion of fossil fuels does not seem to have been a very important factor in constraining economic growth to date.
Assuming that the elasticity of substitution between energy and capital is less than 1 allows the share of energy in production costs to fall over time. When the elasticity of substitution is unity, cost shares must be constant in the long run. The cost share of energy has fallen in the long run in both Britain and Sweden, countries for which we have data from 1800 till the present (Figure 6). An elasticity of substitution of less than unity also allows us to distinguish between labour-augmenting innovations and energy-augmenting innovations, which again is not possible using a Cobb–Douglas production function.
The production function is given by:

Equation (4) embeds a Cobb–Douglas function of capital, K, and labour, L, LβK1β, in a constant elasticity of substitution production function of this combined input and energy, E, to produce gross output, Y. ϕ=(σ – 1)/σ, where σ is the elasticity of substitution between energy and the capital–labour aggregate. AL and AE are the augmentation indices of labour and energy, which can be interpreted as reflecting both changes in technology that augment the effective supply of the factor in question and changes in the quality of the respective factors. AEE and ALL are called effective energy and effective labour, respectively.
In Solow’s (1956) model, as long as there is technological change the economy can grow. In Stern and Kander’s model, depending on the availability of energy and the nature of technological change, energy can be either a constraint on growth or an enabler of growth. When effective energy, AEE, is very abundant, the model behaves very similarly to Solow’s original model and energy neither constrains nor drives growth. The more energy there is, the less important energy appears to be. But when effective energy is relatively scarce, the level of output depends on the level of energy supply and the level of energy-augmenting technology. Labour-augmenting technological change alone no longer results in economic growth.
Before the Industrial Revolution most energy was in the form of wood and animal and human muscle power – wind- and water-power contributed relatively little energy (Kander et al., 2014). The supply of this renewable energy was constrained by the availability of land, so energy was scarce (Wrigley, 2010). Therefore, as the data show (Maddison, 2001), until the Industrial Revolution, output per capita was generally low and economic growth was not sustained. Stern and Kander (2012) find that increases in energy use and energy-augmenting technological change were the main contributors to economic growth in the 19th and early 20th centuries, but in the second half of the 20th century labour-augmenting technological change became the main driver of growth in income per capita, as it is in the Solow growth model.
Back to top
The elasticity of energy with respect to GDP
How much does energy use increase with economic growth? Various studies have estimated by how much energy use tends to be higher as income increases without controlling for other factors, while other studies attempt to estimate the macro-level elasticity of demand for energy, controlling for energy prices and changes in technology. I summarise some recent econometric results. As mentioned in the introduction, Csereklyei et al. (2016) find that there has been a remarkably stable relationship between energy and GDP over the last four decades in a sample of 99 countries. The elasticity, which does not control for energy prices or technological change, is 0.7. Similarly, using panel data for middle-income countries, including today’s developed countries in earlier decades, Van Benthem (2015) finds an elasticity of 0.9 or 0.97 controlling for energy prices and time effects. He finds lower elasticities for higher and lower income bands. Similarly, Fouquet (2014) finds that energy income elasticities first rose and then fell over the course of economic development in Britain. Using a cointegration approach, Joyeux and Ripple (2011) estimate that the long-run income elasticity is 1.08 for OECD countries and 0.853 for 19 developing countries between 1973 and 2007. These estimates do not control for energy prices or time effects. Thus energy is a normal good and most estimates of the elasticity are between 0.5 and unity. However, there does not seem to be a consensus on whether the income elasticity declines or not with increasing income.
Back to top
Testing for causality between energy and GDP
Two methods for testing for causality among time series variables are Granger causality tests and cointegration analysis (Granger, 1969; Engle and Granger, 1987). Hendry and Juselius (2000) discuss the application of these methods to energy economics, where they have been applied extensively to test for causality and cointegration between energy, GDP and other variables from the late 1970s on (Kraft and Kraft, 1978; Ozturk, 2010). There are now hundreds of journal articles on this topic (Bruns et al., 2014).
Early studies relied on Granger causality tests on unrestricted vector autoregressions (VARs) in levels of the variables, while more recent studies use cointegration methods. A vector autoregression model consists of one regression equation for each variable of interest in a system. Each variable is regressed on lagged values of itself and all other variables in the system. If the coefficients of the lagged values of variable X in the equation for dependent variable Y are jointly statistically significant, then X is said to Granger cause Y. Cointegration analysis tests whether variables that have stochastic trends – their trend is a random walk – share a common trend. If so, then at least one variable must Granger cause the other.
Early studies also used bivariate models of energy and output, while more recent research tends to employ multivariate models. Ignoring other relevant variables can generate spurious causality findings. The most common additional variables used are capital and labour or energy prices. A third way to differentiate among models is whether energy is measured in standard heat units or whether a method is used to account for differences in quality among fuels.
The results of early studies that tested for Granger causality using a bivariate model were generally inconclusive (Stern, 1993). Stern tested for Granger causality in the USA in a multivariate setting using a vector autoregression (VAR) model of GDP, capital and labour inputs, and a Divisia index of quality-adjusted energy use in place of the usual heat equivalent of energy use. When both the multivariate approach and quality-adjusted energy index were employed, energy use was found to Granger cause GDP.
Yu and Jin (1992) conducted the first cointegration study of the energy–GDP relationship using the bivariate approach. Stern (2000) estimated a dynamic cointegration model for GDP, quality weighted energy, labour and capital. The analysis showed that there is a cointegrating relation between the four variables and, depending on the version of the model used, found that energy Granger causes GDP or that there is mutual causation between energy and GDP. Some subsequent research appeared to confirm these findings using other measures of energy quality (Warr and Ayres, 2010) or data for other countries (Oh and Lee, 2004; Ghali and El-Sakka, 2004) and panels of many countries (Lee and Chang, 2008; Lee et al., 2008).
Bruns et al. (2014) carry out a meta-analysis of 75 single-country Granger causality and cointegration studies comprising more than 500 tests of causality in each direction. They find that most seemingly statistically significant results in the literature are probably the result of statistical biases that occur in models that use short time series of data – ‘overfitting bias’ – or the result of the selection for publication of statistically significant results – ‘publication bias’. The most robust findings in the literature are that growth causes energy use when energy prices are controlled for in the underlying studies. Using a panel cointegration model of GDP, energy use and energy prices for 26 OECD countries (1978–2005), Costantini and Martini (2010) also find that in the long run GDP growth drives energy use and energy prices, though in the short run energy prices cause GDP and energy use, and energy use and GDP are mutually causative.
However, Bruns et al. (2014) find that studies that control for capital do not find a genuine effect of energy on growth or vice versa. But they had too small a number of studies that used quality-adjusted energy to test whether there was a genuine relationship between energy and growth when this measure of energy use was employed. So their findings do not necessarily contradict the previous research by Stern and others reviewed above.
Back to top
Gaps in knowledge
As this article has shown, the relationship between energy and GDP is one where there is remarkably little consensus, and large gaps in knowledge remain. The field of energy economics has expanded rapidly in the last decade, but much research is repetitive and adds little to existing knowledge (Smyth and Narayan, 2015). In particular, there is a very large literature using reduced form time series models to test for causality and cointegration between energy and output. But this literature is completely inconclusive, with equal numbers of studies finding causation in each direction (Bruns et al., 2014). Research in this area needs to be more closely based on testing potential mechanisms which link energy and output. However, researchers are only starting to build theoretical models of the role of energy in the economic growth process.
There is also a lack of consensus in research on the drivers of changes in energy intensity. In particular, energy intensity has risen in many developing countries. The reasons for this are little researched. There is also a lack of consensus on the size of the economy-wide rebound effect. Existing estimates are all derived from simulation models and range from negative rebound to backfire, where energy efficiency improvements actually increase rather than reduce energy use. Therefore there is little guidance on the potential for energy efficiency policies to actually conserve energy.
Research is also hampered by inadequate data. With the exception of traditional biomass, energy use data are normally of good quality. But data on prices is much more fragmentary. Most economic research is based on understanding the linkages between prices and quantities. So this is an important area where international comparable datasets could be very useful.
Back to top


Acemoglu, D. 2009. Introduction to Economic Growth. Princeton University Press, Princeton, NJ.

Aghion, P. and Howitt, P. 2008. The Economics of Growth. MIT Press, Cambridge, MA.

Aguayo, F. and Gallagher, K. P. 2005. Economic reform, energy, and development: the case of Mexican manufacturing. Energy Policy, 33: 829–37.

Allan, G., Hanley, N., McGregor, P., Swales, K. and Turner, K. 2007. The impact of increased efficiency in the industrial use of energy: a computable general equilibrium analysis for the United Kingdom. Energy Economics, 29: 779–98.

Ang, B. W. 2006. Monitoring changes in economy-wide energy efficiency: from energy- GDP ratio to composite efficiency index. Energy Policy, 34: 574–82.

Asheim, G. B. 1994. Net national product as an indicator of sustainability. Scandinavian Journal of Economics, 96: 257–265.

Asheim, G. B., Buchholz, W. and Withagen, C. 2003. The Hartwick rule: myths and facts. Environmental and Resource Economics, 25: 129–50.

Ayres, R. U. and Kneese, A. V. 1969. Production, consumption and externalities. American Economic Review, 59: 282–297.

Ayres, R. U. and Warr, B. 2005. Accounting for growth: the role of physical work. Structural Change and Economic Dynamics, 16: 181–209.

Ayres, R. U. and Warr, B. 2009. The Economic Growth Engine: How Energy and Work Drive Material Prosperity. Edward Elgar, Cheltenham.

Ayres, R. U., Ayres, L. W. and Martinás, K. 1998. Exergy, waste accounting, and life-cycle analysis. Energy, 23(5): 355–63.

Ayres, R. U., Ayres, L. W. and Warr, B. 2003. Exergy, power and work in the US economy, 1900–1998. Energy, 28: 219–73.

Barker, T., Dagoumas, A. and Rubin, J. 2009. The macroeconomic rebound effect and the world economy. Energy Efficiency, 2: 411–27.

Berndt, E. R. 1978. Aggregate energy, efficiency, and productivity measurement. Annual Review of Energy, 3: 225–73.

Berndt, E. R. 1990. Energy use, technical progress and productivity growth: a survey of economic issues. The Journal of Productivity Analysis, 2: 67–83.

Borenstein, S. 2015. A microeconomic framework for evaluating energy efficiency rebound and some implications. Energy Journal, 36(1): 1–21.

Boulding K. 1966. The economics of the coming spaceship Earth. In Environmental Quality in a Growing Economy (ed. H. Jarett). Johns Hopkins University Press, Baltimore.

Brown, M. T. and Herendeen, R. A. 1996. Embodied energy analysis and emergy analysis: a comparative view. Ecological Economics, 19: 219–36.

Bruns, S. B., Gross, C. and Stern, D. I. 2014. Is there really Granger causality between energy use and output? Energy Journal, 35(4): 101–34.

Burke, P. J. 2013. The national-level energy ladder and its carbon implications. Environment and Development Economics, 18(4): 484–503.

Burkett, P. 2003. The value problem in ecological economics: lessons from the physiocrats and Marx. Organization & Environment, 16(2): 137–67.

Cleveland, C. J., Costanza, R., Hall, C. A. S. and Kaufmann, R. K. 1984. Energy and the U.S. economy: a biophysical perspective. Science, 225: 890–7.

Cleveland, C. J., Kaufmann, R. K. and Stern, D. I. 2000. Aggregation and the role of energy in the economy. Ecological Economics, 32: 301–18.

Common, M. S. 1995. Sustainability and Policy: Limits to Economics. Cambridge University Press, Melbourne.

Costantini, V. and Martini, C. 2010. The causality between energy consumption and economic growth: a multi-sectoral analysis using non-stationary cointegrated panel data. Energy Economics, 32: 591–603.

Costanza, R. 1980. Embodied energy and economic valuation. Science, 210: 1219–24.

Costanza, R. and Daly, H. E. 1992. Natural capital and sustainable development. Conservation Biology, 6: 37–46.

Csereklyei Z., Rubio Varas, M. d. M. and Stern, D. I. 2016. Energy and economic growth: the stylized facts. Energy Journal, 37(2): 223–55.

Dasgupta, P. S. and Heal, G. M. 1979. Economic Theory and Exhaustible Resources. Cambridge University Press, Cambridge.

Dechezleprêtre, A., Glachant, M., Haščič, I., Johnstone, N. and Ménière, Y. 2011. Invention and transfer of climate change-mitigation technologies: a global analysis. Review of Environmental Economics and Policy, 5(1): 109–30.

Enflo, K., Kander, A. and Schön, L. 2009. Electrification and energy productivity. Ecological Economics, 68: 2808–17.

Engle, R. E. and Granger, C. W. J. 1987. Cointegration and error-correction: representation, estimation, and testing. Econometrica, 55: 251–76.

Fouquet, R. 2014. Long run demand for energy services: income and price elasticities over 200 years. Review of Environmental Economics and Policy, 8(2): 186–207.

Fredriksson, P. G., Vollebergh, H. R. J. and Dijkgraaf, E. 2004. Corruption and energy efficiency in OECD countries: theory and evidence. Journal of Environmental Economics and Management, 47: 207–31.

Gentvilaite, R., Kander, A. and Warde, P. 2015. The role of energy quality in shaping long-term energy intensity in Europe. Energies, 8(1): 133–53.

Georgescu-Roegen, N. 1971. The Entropy Law and the Economic Process. Cambridge, MA: Harvard University Press.

Gever, J., Kaufmann, R. K., Skole, D. and Vorosmarty, C. 1986. Beyond Oil: The Threat to Food and Fuel in the Coming Decades. Ballinger, Cambridge, MA.

Ghali, K. H. and El-Sakka, M. I. T. 2004. Energy use and output growth in Canada: a multivariate cointegration analysis. Energy Economics, 26: 225–38.

Gillingham, K., Newell, R. G. and Palmer, K. 2009. Energy efficiency economics and policy. Annual Review of Resource Economics, 1: 597–620.

Gillingham, K., Kotchen, M. J., Rapson, D. S. and Wagner, G. 2013. The rebound effect is overplayed. Nature, 493: 475–6.

Granger, C. W. J. 1969. Investigating causal relations by econometric models and cross-spectral methods. Econometrica, 37: 424–38.

Greening, L. A., Greene, D. L. and Difiglio, C. 2000. Energy efficiency and consumption – the rebound effect – a survey. Energy Policy, 28: 389–401.

Grepperud, S. and Rasmussen, I. 2004. A general equilibrium assessment of rebound effects. Energy Economics, 26: 261–82.

Grübler, A., Nakicenovic, N.and Victor, D. G. 1999. Dynamics of energy technologies and global change. Energy Policy, 27: 247–80.

Grübler, A., Johansson, T. B., Mundaca, L., Nakicenovic, N., Pachauri, S., Riahi, K.,Rogner, H.-H., and Strupeit, L. 2012. Energy primer. In: Global Energy Assessment – Toward a Sustainable Future, Chapter 1. Cambridge University Press, Cambridge, UK and New York, NY, USA and the International Institute for Applied Systems Analysis, Laxenburg, Austria.

Hall, C. A. S., Cleveland, C. J. and Kaufmann, R. K. 1986. Energy and Resource Quality: The Ecology of the Economic Process. Wiley Interscience, New York.

Hall, C. A. S., Tharakan, P., Hallock, J., Cleveland, C. J. and Jefferson, M. 2003. Hydrocarbons and the evolution of human culture. Nature, 426: 318–22.

Hannon, B. 1973. An energy standard of value. Annals of the American Academy, 410: 139–53.

Hartwick, J. M. 1977. Intergenerational equity and the investing of rents from exhaustible resources. American Economic Review, 66: 972–4.

Hartwick, J. M. 1995. Constant consumption paths in open economies with exhaustible resources. Review of International Economics, 3: 275–83.

Hendry, D. F. and Juselius, K. 2000. Explaining cointegration analysis: Part 1. Energy Journal, 21(1): 1–42.

Henriques, S. T. and Kander, A. 2010. The modest environmental relief resulting from the transition to a service economy. Ecological Economics, 70: 271–82.

Hornborg, A. 2014. Ecological economics, Marxism, and technological progress: Some explorations of the conceptual foundations of theories of ecologically unequal exchange. Ecological Economics, 105: 11–18.

Howarth, R. B. 1997. Energy efficiency and economic growth. Contemporary Economic Policy, 25: 1–9.

Joyeux, R. and Ripple, R. D. 2011. Energy consumption and real income: a panel cointegration multi-country study. Energy Journal, 32(2): 107–41.

Judson, R. A., Schmalensee, R. and Stoker, T. M. 1999. Economic development and the structure of demand for commercial energy. Energy Journal, 20(2): 29–57.

Kander, A. 2002. Economic Growth, Energy Consumption and CO2 Emissions in Sweden 1800–2000. Lund Studies in Economic History No. 19. Lund.

Kander, A. 2005. Baumol's disease and dematerialization of the economy. Ecological Economics, 55(1): 119–30.

Kander, A. and Lindmark, M. 2006. Foreign trade and declining pollution in Sweden: a decomposition analysis of long-term structural and technological effects. Energy Policy, 34(13): 1590–9.

Kander, A., Malanima, P., and Warde, P. 2014. Power to the People – Energy and Economic Transformation of Europe over Four Centuries. Princeton University Press, Princeton NJ.

Kaufmann, R. K. 1992. A biophysical analysis of the energy/real GDP ratio: implications for substitution and technical change. Ecological Economics, 6: 35–56.

Kaufmann, R. K. 1987. Biophysical and Marxist economics: learning from each other. Ecological Modelling, 38: 91–105.

Kaufmann, R. K. 2004. The mechanisms for autonomous energy efficiency increases: A cointegration analysis of the US energy/GDP ratio. Energy Journal, 25(1): 63–86.

Koetse, M. J., de Groot, H. L. F. and Florax, R. J. G. M. 2008. Capital–energy substitution and shifts in factor demand: a meta-analysis. Energy Economics, 30: 2236–51.

Kraft, J. and Kraft, A. 1978. On the relationship between energy and GNP. Journal of Energy and Development, 3: 401–3.

Kubiszewski, I., Cleveland, C. J. and Endres, P. K. 2010. Meta-analysis of net energy return for wind power systems. Renewable Energy, 35: 218–25.

Lee, C.-C. and Chang, C.-P. 2008. Energy consumption and economic growth in Asian economies: a more comprehensive analysis using panel data. Resource and Energy Economics, 30(1): 50–65.

Lee, C.-C., Chang, C.-P. and Chen, P.-F. 2008. Energy-income causality in OECD countries revisited: The key role of capital stock. Energy Economics, 30: 2359–73.

Levinson, A. 2010. Offshoring pollution: is the United States increasingly importing polluting goods? Review of Environmental Economics and Policy, 4(1): 63–83.

Linares, P. and Labandeira, X. 2010. Energy efficiency: economics and policy. Journal of Economic Surveys, 24(3): 583–92.

Ma, C. and Stern, D. I. 2008. China’s changing energy intensity trend: a decomposition analysis. Energy Economics, 30(3): 1037–53.

Matisoff, D. C. 2008. The adoption of state climate change policies and renewable portfolio standards: regional diffusion or internal determinants? Review of Policy Research, 25(6): 527–46.

Maddison, A. 2001. The World Economy: A Millennial Perspective, OECD, Paris.

Murphy D. J. and Hall, C. A. S. 2010. Year in review – EROI or energy return on (energy) invested. Annals of the New York Academy of Sciences, 1185: 102–18.

Newell, R. G., Jaffe, A. B. and Stavins, R. N. 1999. The induced innovation hypothesis and energy-saving technological change. Quarterly Journal of Economics, 114: 941–75.

O’Connor, M. P. 1993. Entropic irreversibility and uncontrolled technological change in the economy and environment. Journal of Evolutionary Economics, 34: 285–315.

Oh, W. and Lee, K. 2004. Causal relationship between energy consumption and GDP revisited: the case of Korea 1970–1999. Energy Economics, 26: 51–9.

Ozturk, I. 2010. A literature survey on energy–growth nexus. Energy Policy, 38: 340–9.

Perrings, C. A. 1987. Economy and Environment: A Theoretical Essay on the Interdependence of Economic and Environmental Systems. Cambridge University Press, Cambridge.

Pezzey, J. C V. 2004. Sustainability tests with amenities, and change in technology, trade and population. Journal of Environmental Economics and Management, 48: 613–31.

Popp, D. 2002. Induced innovation and energy prices. American Economic Review, 92: 160–80.

Roy, J. 2000. The rebound effect: some empirical evidence from India. Energy Policy, 28: 433–8.

Saunders, H. D. 1992. The Khazzoom–Brookes postulate and neoclassical growth. Energy Journal, 13(4): 131–48.

Saunders, H. D. 2008. Fuel conserving (and using) production functions. Energy Economics, 30: 2184–235.

Saunders, H. D. 2013. Historical evidence for energy efficiency rebound in 30 US sectors and a toolkit for rebound analysts. Technological Forecasting & Social Change, 80: 1317–30.

Schurr, S. 1982. Energy efficiency and productive efficiency: some thoughts based on American experience. Energy Journal, 3(3): 3–14.

Schurr, S. and Netschert, B. 1960. Energy and the American Economy, 1850–1975. Johns Hopkins University Press, Baltimore.

Smulders, S. 2005. Endogenous technical change, natural resources and growth. In: Scarcity and Growth in the New Millennium (eds. R. Ayres, D. Simpson and M. Toman). Resources for the Future, Washington, DC.

Smyth, R. and Narayan, P. K. 2015. Applied econometrics and implications for energy economics research. Energy Economics, 50: 351–8.

Solow, R. M. 1956. A contribution to the theory of economic growth. Quarterly Journal of Economics, 70: 65–94.

Solow, R. M. 1974. Intergenerational equity and exhaustible resources. Review of Economic Studies, 41(5): 29–46.

Sorrell, S., Dimitropoulos, J. and Sommerville, M. 2009. Empirical estimates of the direct rebound effect: a review. Energy Policy, 37: 1356–71.

Stern, D. I. 1993. Energy use and economic growth in the USA: a multivariate approach. Energy Economics, 15: 137–50.

Stern, D. I. 1997. Limits to substitution and irreversibility in production and consumption: a neoclassical interpretation of ecological economics. Ecological Economics, 21: 197–215.

Stern, D. I. 1999. Is energy cost an accurate indicator of natural resource quality? Ecological Economics, 31: 381–94.

Stern, D. I. 2000. A multivariate cointegration analysis of the role of energy in the U.S. macroeconomy. Energy Economics, 22: 267–83.

Stern, D. I. 2004. The rise and fall of the environmental Kuznets curve. World Development, 32(8): 1419–1439.

Stern, D. I. 2010. Energy quality. Ecological Economics, 69(7): 1471–8.

Stern, D. I. 2011. The role of energy in economic growth. Annals of the New York Academy of Sciences, 1219: 26–51.

Stern, D. I. 2012a. Interfuel substitution: a meta-analysis. Journal of Economic Surveys, 26: 307–31.

Stern, D. I. 2012b. Modeling international trends in energy efficiency. Energy Economics, 34: 2200–8.

Stern, D. I. and Kander, A. 2012. The role of energy in the industrial revolution and modern economic growth. Energy Journal, 33(3): 125–52.

Stiglitz, J. E. 1974a. Growth with exhaustible natural resources: the competitive economy. Review of Economic Studies, 41: 139–52.

Stiglitz, J. E. 1974b. Growth with exhaustible natural resources: efficient and optimal growth paths. Review of Economic Studies, 41: 123–38.

Sue Wing, I. 2008. Explaining the declining energy intensity of the U.S. economy. Resource and Energy Economics, 30: 21–49.

Toman, M. A. and Jemelkova, B. 2003. Energy and economic development: an assessment of the state of knowledge. Energy Journal, 24(4): 93–112.

Turner, K. 2009. Negative rebound and disinvestment effects in response to an improvement in energy efficiency in the UK economy. Energy Economics, 31: 648–66.

Turner, K. 2013. ‘Rebound’ effects from increased energy efficiency: a time to pause and reflect. Energy Journal, 34(4): 25–43.

Turner, K. and Hanley, N. 2011. Energy efficiency, rebound effects and the Environmental Kuznets Curve. Energy Economics, 33: 722–41.

Ukidwe, N. U. and Bakshi, B. R. 2007. Industrial and ecological cumulative exergy consumption of the United States via the 1997 input–output benchmark model. Energy, 32: 1560–92.

van Benthem, A. A. 2015. Energy leapfrogging. Journal of the Association of Environmental and Resource Economists, 2(1): 93–132.

Wang, C. 2011. Sources of energy productivity growth and its distribution dynamics in China. Resource and Energy Economics, 33: 279–92.

Warr, B. and Ayres, R. U. 2010. Evidence of causality between the quantity and quality of energy consumption and economic growth. Energy, 35: 1688–693.

Warr, B. Ayres, R. U., Eisenmenger, N., Krausmann, F. and Schandl, H. 2010. Energy use and economic development: a comparative analysis of useful work supply in Austria, Japan, the United Kingdom and the US during 100 years of economic growth. Ecological Economics, 69: 1904–17.

Wei, C., Ni, J. and Shen, M. 2009. Empirical analysis of provincial energy efficiency in China. China & World Economy, 17(5): 88–103.

Wrigley, E. Anthony. 2010. Energy and the English Industrial Revolution. Cambridge University Press, Cambridge.

Yu, E. S. H. and Jin, J. C. 1992. Cointegration tests of energy consumption, income, and employment. Resources and Energy, 14: 259–66.

Back to top

How to cite this article

Stern, David I. "energy-GDP relationship." The New Palgrave Dictionary of Economics. Online Edition. Eds. Steven N. Durlauf and Lawrence E. Blume. Palgrave Macmillan, 2015. The New Palgrave Dictionary of Economics Online. Palgrave Macmillan. 04 October 2015 <> doi:10.1057/9780230226203.3947

Download Citation:

as RIS | as text | as CSV | as BibTex