Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences


    The emission targets agreed in Paris require a radical reduction of material extraction, use and disposal. The core claim of this article is that a radical dematerialization can only be part and parcel of degrowth. Given that capitalist economies are designed to grow, this raises the question of whether, and under what circumstances, the inevitable ‘degrowth’ can become socially sustainable. Three economic policies are discussed in this direction: work-sharing, green taxes and public money.

    This article is part of the themed issue ‘Material demand reduction’.

    1. Introduction

    Nation states committed in Paris to significant reductions in carbon emissions. Reducing carbon emissions requires reducing material use [1]. This dematerialization is ‘radical’, first because of its unprecedented scale, and second, because of the significant social, cultural and political changes it will entail [2]. The thesis of this paper is that radical dematerialization is not compatible with economic growth. Less material use will, in all likelihood, come only with less output, that is with a smaller economy. Vice versa, policies of radical dematerialization will slow down the economy. Capitalist economies, however, become unstable when they do not grow [3]. Economic slowdowns often have negative social effects. The question is then whether, and how, could we manage, or even prosper, without growth [3,4]. A main obstacle to strict policies for dematerialization is their presumed negative effect on growth. This rests on the assumption that a lack of growth is always disastrous. If the absence of growth is manageable, then this makes it easier to implement policies that drastically reduce material demand.

    Section 2 outlines the field of ecological economics (EE) and its understanding of the economic process. This provides the theoretical and empirical backbone for the claims made in this article. Section 3 argues that it is unlikely that material use will be reduced if the economy grows. Section 4 explains why the efficiency and demand policies espoused by Allwood et al. [2] are unlikely to be compatible with economic growth. Section 5 presents three economic policies in the direction of sustainable degrowth and §6 concludes.

    2. Ecological economics, an economics where matter matters

    In 1989, economists and natural scientists that were dissatisfied with the treatment of environmental problems by mainstream economics founded the society for EE [5]. EE aspires to rethink economics from an ecological point of view [6]. It is a diverse field of research reclaiming the classical economic approach, in which land and resources were essential factors of production. Five basic insights from EE are presented below:

    • (1) The economy is a process of material transformation driven by human and non-human work.

    • (2) Efficiency improvements lead to more, not less, resource use.

    • (3) Factors of production—resources, labour and capital—are complements, not substitutes.

    • (4) Services embody significant amounts of energy and materials.

    • (5) Externalities are cost-shifting successes.

    (a) The economic process is a material process driven by work

    The economy is a complex process that converts raw materials (and energy) into useful goods and final services. The conversion and final use of energy and materials inevitably produces waste. Nicholas Georgescu-Roegen (G-R) [7] proposed a theory of the economic process based on the principles of thermodynamics. G-R's work is the departure point for EE, though there was a long lineage of ecological thinking in economics before [8]. From a thermodynamic point of view, what the economy does is process materials and ‘convert high quality (low entropy) raw materials into goods and services, while disposing of, and dissipating, large and growing quantities of high entropy materials and energy waste (that is waste heat)’ ([9], 139). All economic goods have a material origin—hence G-R's phrase that ‘matter matters’ [9]. But what moves the economy?

    The engine of the economy is work [9]. This view dates back to classical political economy, for which labour was the central analytical category. Production is the outcome of work. Humans work to transform raw materials into useful products. This is what ‘adds value’ to unprocessed raw materials [10]. But humans do not work alone. They often use the work of draft animals, appropriate the work of non-human species such as bees, and, crucially, exploit the vast amounts of work supplied from fossil fuels for free. Fossil fuels are solar energy captured by photosynthesis, accumulated over millions of years below the ground. To appropriate this free work from nature, humans have to put work and energy. But the free work they appropriate from fossil fuels is much more than the work they put in.

    An important indicator in this respect is the Energy Return on Energy Investment (EROI). This is the ratio of the amount of energy produced to the amount of energy used to produce it [11]. To drill oil, you use electricity; to build a photovoltaic park, you have to power schools that train engineers and machines that extract the minerals that go into the panels [11,12]. Fossil fuels are a stock [7], a reservoir of ‘bottled photosynthesis’. The Sun's energy may be vast in comparison, but it is a diffuse flow. Energy and land have to be spent to concentrate the Sun's power [7]. A lot of energy has to be spent to capture and store solar power. Hence its EROI is likely to remain lower than that of fossil fuels [11,12]. The appropriation of net energy though is crucial for economic growth.

    Economic growth is a process whereby increasing amounts of human and non-human work process raw materials and energy, faster and faster. Civilizations before the advent of capitalism experienced very slow rates of economic growth. What was it that capitalism did differently? The major innovation was that surplus work from humans and fossil fuels was concentrated in the hands of a class of people—‘capitalists’, the owners of the means of production—who did not just store this surplus, or ceremoniously waste it for prestige, as elites and Royal courts did before, but put it back into production [13]. Capitalists invested in machines, which made it possible to appropriate more work from humans and nature, creating ever-more surplus. A substantial portion of surplus was directed to science, research and development that led to the invention of new machines, and the appropriation of more work from nature, further fuelling this cycle of growth.

    Empirical facts confirm this theory. There is a strong correlation between the input of useful energy—that is energy left after subtracting the energy lost in its production and conversion—and economic output [14]. Economists accounting for growth attribute it mostly to ‘total factor productivity’—a name given to all factors other than capital and labour that affect growth—a shorthand for ‘technological progress’. But as Ayres & Warr [9] show total factor productivity is strongly associated with the amount of energy used and the efficiency of conversion of energy inputs into useful work. Technological progress is therefore progress in using more energy, more efficiently.

    If energy is a ‘cause’ of economic growth, as these studies suggest that it is, and if fossil fuels are a unique source of high EROI, then decoupling growth from fossil fuels is less likely than conventionally believed. If the economic process is the application of energy and human work on the transformation of materials, then economic growth is also likely to be strongly coupled with material use (see §3). Couldn't, however, economic growth and resource use be decoupled by using resources more efficiently?

    (b) Efficiency and scale

    Energy and materials can indeed be used more efficiently. This is precisely what new technologies do. With the invention of the steam engine, for example, a given quantity of coal produced much more useful work than before. This, however, led to an increase, not a decrease in the use of coal, as noted at the time by economist Stanley Jevons. The ‘Jevons paradox’ is the name given to this counterintuitive—from a resource conservation standpoint—outcome, whereby a more efficient use of resources leads to more, not less resource use [15]. A simple economic explanation is that a resource used more efficiently costs less as a result. The demand for it and consumption increase (rebound), compensating for the savings from the more efficient use. Does this always have to be the case? Can't the rebound be less than the savings?

    In principle yes, but in practice one finds that the total use of resources increases as a result of technological improvements that increase resource efficiency. The same is the case with improvements in labour productivity. Increases in labour productivity have not led to masses of unemployed people—the increase in total production as a result of increased labour productivity led to the employment of more, not fewer, workers [15]. To understand why this happens, one has to look at the underlying growth dynamic of capitalism. As more surplus is extracted from a given amount of labour or a given stock of resources, this surplus is invested into more production, which in turn requires more resources and more labour. Efficiency brings growth, which annuls the savings from efficiency. It is unlikely that resource use will be avoided this way, unless a concrete limit is imposed upon the maximum allowable scale of resource use—or in the case of labour, the maximum amount of hours workers can work [16].

    Couldn't resources however be spared by ‘substituting’ them in the production process? Instead of using more materials or energy, couldn't we use more capital or more labour?

    (c) The complementarity of the factors of production

    Capital machinery is not produced from thin air—it has also a material origin. Machines are the product of human work and energy and they are made with raw materials. To produce machines, energy and materials are used. Substituting energy and materials with machines in production will not necessarily save energy or resources. What about labour though?

    If we were to substitute energy and resources with human labour, this would spare resources, but would not grow the economy. Growth is the result of substituting labour with energy and adding to the work of humans the free work of fossil fuels. This is what is behind rises in ‘labour productivity’. What appears as an increase in productivity is, partly, a substitution of the work done by human muscles and brains with the work done by fossil fuels [17]. Fossil fuels did not simply substitute the ‘horse power’ of horses—they brought millions of new, and cheaper, ‘horses’ into production. Reversing this and making humans again do more of the work presently done by energy may increase employment, but would make the economy less productive, reducing output.

    Certain resources may also not be substitutable. Without getting into details, standard economic theory assumes perfect substitution between the factors of production (capital, labour, resources), and unlimited possibilities of substitution between specific resources or materials (say fossil fuels substituting energy from biomass). There are two problems with this.

    Firstly, mutual and full substitutability means that goods and services could in principle be produced by any of the three factors alone, without the use of the other two. This cannot be right. To fly, a plane needs both pilots and kerosene. The factors of production are complements, not substitutes. Machines require workers to operate them, and labour requires tools to be productive [9]. For certain economic processes, there is a minimum requirement of each factor without which the process cannot take place.

    Secondly, it is reasonable to expect that as resources become scarce and their relative price increases, investment will go into research and development of substitutes. As fossil fuels get more expensive, for example, other sources of energy may emerge as substitutes. Whereas this might have been true in the past and for certain resources, for example, substituting fuel wood with fossil fuels, the possibility of substitution cannot be asserted a priori independently of concrete situations. There might or there might not be a substitute for fossil fuels or certain scarce materials. This question is one of engineering, not of economics. One cannot assume that there will always be substitutes. And even if there are substitutes, one cannot assume that these substitutes will be able to maintain the same scale of the economy or rates of growth. If fossil fuels, for example, are substituted by lower EROI renewables, this will reduce the total amount of useful work provided by energy sources and dampen the scale and growth of the economy.

    A final theoretical possibility under which the economy could dematerialize and grow is through structural change. If a growing share of output is provided by services that use few materials and these services substitute output from more materially intensive goods, then it might be possible to combine growth with dematerialization.1 But do services use fewer materials and energy?

    (d) Embodied energy and materials

    Products and services embody not only the materials and energy directly found in them, but also the materials and energy spent in the various stages of their production. Howard Odum proposed the concept of ‘Emergy’, with an m. This is the energy embedded in products and services—the energy that went into their production. A fish, for example, embodies all the solar energy that went into making the phytoplankton that it ate [19]. A service worker embodies all the energy consumed for feeding, clothing, educating and moving her around.

    William Rees [20], a biologist turned ecological economist, put in practice this idea with the concept of ecological footprint. Recent research on consumption-based emissions and material flows, including indicators such as water or material footprint [21,22], is based on this insight: behind any final product or service, there is a long chain of resource and energy use, conversion and waste. In a metaphorical sense, all this is ‘embodied’ in the final product and marks its ‘footprint’.

    The next section shows that the material footprint of nations is strongly correlated with the size of their economy. The data show that there is no absolute dematerialization to date—a tentative confirmation of the EE theory presented in this section.

    (e) Cost-shifting

    The environmental costs of economic activities are often considered as ‘externalities’, costs that are not accounted for within existing markets. This view presupposes the possibility of an all-encompassing market where nothing will be external to it. In reality, ‘cost-shifting’ is a pervasive feature of market activity, where businesses try to increase their profits by shifting their costs to others, or to society as a whole, when and if they can [23]. Externalities are therefore cost-shifting ‘successes’, rather than market failures [23].

    Ever since the colonies, growth is fuelled by the extraction of surpluses in the imperial centres by exploiting people and environments elsewhere [24]. If such costs were internalized, probably there would not have been growth to begin with. An ‘internalization’ of ecological or carbon costs today would possibly bring growth to an end, especially if accumulated debts from the past were included. This is why carbon or other environmental taxes are resisted by the powerful economic interests that profit from the externalization of carbon or environmental costs. As Raj Patel [25] vividly puts it, the social and ecological cost of a McDonald burger is in the order of 200$. McDonalds would not have a business if it was to internalize its costs.

    3. Why dematerialization is unlikely to be compatible with economic growth

    Is the use of metals and other materials bound to increase because of growth, or not? This is an argument dating back to the publication of the Limits to growth report [26] and the response of economists to it [27]. The Club of Rome claimed that economic growth increases material use and will eventually exhaust mineral deposits unless slowed down. Solow [27] responded that as prices of scarce metals increase, and technologies and substitutes are developed, the material intensity of the economy will decline (that is the kilograms of a metal used per dollar of economic product). Eventually, this should lead to a decline of resource use.

    Efficiency (or intensity), however, is not the same with scale. We might use a resource more efficiently, and still end up using more of it. Data on the aggregate use of materials by the global economy show that there is only a minor reduction of the intensity of use (or, put differently, a minor improvement of material efficiency or productivity). Yet overall material use has more than doubled since 1980 (figure 1).2

    Figure 1.

    Figure 1. Trends in global material extraction, GDP and material intensity (kg/$) 1980–2013 (data from the WU Global material flows database and the World Bank; indexed values, 1980 values equal 100).

    Some authors see signs of a forthcoming dematerialization. Recent data show stagnation of material consumption in some Western economies such as the UK [28]. Could this be a point of ‘peak stuff’, a natural peaking of material consumption as economies reach a mature stage, after which material use declines? ([29], see also [30]). If that were the case, growth and dematerialization would be compatible, at least in the long-run after developing countries had developed sufficiently to reach their ‘peak-stuff’.

    As Gutowski et al. [30], however note, data on domestic material consumption does not account for the material embedded in imported consumer goods. The true material use of a national economy does not include only the raw materials that it extracts, imports or consumes within its borders. It includes also the materials embedded in the finished goods or services that it imports. The material footprint of developed nations, that is the total amount of materials used to produce the goods and services that they consume, increases hand in hand with the size of their economies (figure 2). There is no sign of peak-stuff if one looks at material footprint instead of domestic material consumption (the actual materials consumed, plus imports minis exports). Despite substantial deindustrialization and de-agrarianization, the material demands of the so-called ‘service economies’ continue to grow.

    Figure 2.

    Figure 2. Relative changes in OECD material footprint, domestic material consumption, and GDP from 1990 to 2008. GDP is calculated as purchasing power parity using 2005 US Dollars as the base year. Values are plotted as ΔX = (Xt2 − Xt1)/Xt1; t1 = 1990. Figure from Wiedmann et al. [22].

    Take California: the information sector accounts for 8% of the total state product, 21% together with business and professional services that include computer systems and design [31]. Agriculture's share of the economy is down to 2%. Yet, the state's water footprint grew almost 40% from 1992 to 2010 [21].

    Research on individual metals confirms a similar story to that of aggregated material flows. The Jevons' paradox is confirmed for 57 different materials for all of which there is no evidence of dematerialization. Increases in consumption and production outpace savings from technological improvements [32]. And at the sites where metals are extracted, conflicts intensify because of the negative environmental and social consequences of extraction [33].

    The separation in the trends of domestic material consumption and material footprint in high-income economies is not evidence that they are doing something better. It is a by-product of the globalization of the economy. Industrializing economies produce the consumer goods of service economies [30]. (The idea that one day all economies could graduate to be service economies with saturated material consumption raises the question who would produce then their industrial goods?) This is a systemic pattern, revealed at the global scale, the only appropriate scale to study a globalized economy with a global division of labour. At this scale, the prediction of EE is confirmed: material extraction and consumption grow as the economy grows.

    Cross-country comparisons confirm the same picture. A GDP growth (degrowth) of 1% leads to a 0.6% growth (degrowth) of material footprint [22]. Same for carbon: a 1% increase (decrease) in GDP leads to about 0.5–0.7% increase (decrease) in carbon emissions [34]. These are strong, statistically significant effects, as significant as one can hope to find in econometrics. One may well claim that the causal relationship between GDP and resource use will finally change in the future with structural changes in the economy, or the advent of new materials and technologies. But on this basis, no econometric study could ever be a basis for policy since we can always hope that with sufficient will we can change established causal relations.

    In conclusion, the more (less) an economy grows, the more (less) materials it uses. The next section argues that the reverse holds true too. Radical dematerialization is likely to slow down the economy.

    4. Dematerialization would slow down the economy

    There are four reasons why a radical material reduction is unlikely to be compatible with economic growth.

    First, material efficiency at the factory level will lead to profits that will then be reinvested for growth in other parts of the economy. Efficiency fuels growth and increases material use. Only a limit to the scale of the economy—and the total amount of raw materials that can be extracted and consumed—ensures that efficiency reduces material use [35]. But if the rate of economic growth depends on the use of resources, then such limits will also limit growth.

    Second, using labour instead of new materials can reduce material use. Say for example, that instead of knocking down and replacing commercial buildings after 40 years, we did a major retrofit. This would require more labour, but a fraction of the material input. But isn't this what most pre-growth societies did? Reversing to a more labour-intensive economy will increase the number of jobs; but is unlikely to produce profits and lead to growth. Labour productivity will go down (see also [30]). A retrofit economy is possible, but its houses and cars may look more like Cuba's than California's.

    Likewise, a transition to a solar and wind economy is desirable and necessary. Given, however, that the energy return on energy investment of renewable energies is much lower than fossil fuels [36], decarbonization is unlikely to be compatible with a global economy at its current size—and much less with a global economy converging to the output of a US economy growing 2 or 3% per year.

    Investments in renewable energies create jobs. In the short-term, they may act as a stimulus for the economy. But this does not mean that in the long term it is possible to sustain growth with diminished energy surpluses, substituting energy with human labour. Efficiency improvements, demand reductions, and material and energy source substitutions are all necessary for a sustainable future. But they are not compatible with more growth.

    Third, consider a scenario where people fix their own phones or their cars instead of buying new ones. The circulation of new goods in the economy would slow down. The unpaid, domestic portion of the economy will grow, but market economy will shrink. The surpluses that drive growth will diminish, and therefore growth will decline. An economy that grows 2% per year has to double the speed with which it churns goods and services every 35 years. There must always be demand to satisfy the bigger and faster production. Production requires consumption. Planned obsolescence and an ever-faster turnover of goods are necessary if growth rates of 2–3% per year are to be sustained. Taking consumption out of the market will inevitably slow down production.

    Fourth, dematerialization and decarbonization require significant investments, for example for new infrastructures. These investments will ‘crowd out’ conventional investments that typically go towards improving labour productivity. This is also likely to dampen economic growth [37].

    A retrofitting, renewable economy is desirable environmentally. People may under certain conditions come to desire and find meaningful a shift to more manual work. They may embrace their reduced material or energy purchasing power in the knowledge that the alternative—growth in material and energy use followed by climate change—would be worse. Many people already find meaning in a more frugal lifestyle, and enjoy relational instead of material goods [38]. Such ‘preferences’ may be strengthened if basic needs are secured, and wealth distributed equitably [39]. The question is then how and under what conditions this may happen. ‘Sustainable degrowth’ refers to such a scenario of increasing wellbeing in a context of declining output [40,41].

    Assuming that the EE diagnosis is correct, a decline of economic output is inevitable. The question is whether we will follow a ‘prosperous way down’ [19]; or grow more and collapse after crossing planetary boundaries. Degrowth research investigates under what social and political conditions the inevitable downscaling can become prosperous and not catastrophic [42,43]. This includes policies that make degrowth stable, reducing the dependence of wellbeing upon growth [3,42].

    The next section focuses on three economic policies that can act as levers of change in a possible transition: work-sharing, green taxes and public money. These policies do not directly target material use. Yet they are important from a radical dematerialization perspective. They facilitate a downscaling of the economy that will reduce material use. And they secure wellbeing in conditions of economic contraction—a likely outcome of strict dematerialization or decarbonization policies.

    5. Policies for sustainable degrowth

    Without growth, unemployment increases because rising productivity reduces the amount of labour necessary. Work-sharing refers to a reduction of the hours of paid work without loss in income. Fewer working hours per person leave more jobs for everyone to share. If people work less, the cost of labour may rise. The reduced output might reduce then the additional number of jobs. But overall a reduction of working hours has a positive effect on employment [44].

    Work-sharing redistributes the gains from productivity. Instead of producing profits for capitalists, productivity serves to liberate time for workers [45]. Lange [46] shows that work-sharing is a fundamental condition for a stable, zero-growth path in both neoclassical and Keynesian growth models. Econometric studies suggest that reduced working hours reduce carbon emissions and environmental pressures and increase wellbeing [4750].

    There is a limit, however, to how far work can be reduced. If the use of fossil fuels is to be limited in the future, there will be less ‘energy slaves’ doing work for free. Humans will have to do more of this work—if we want to work less, then we will have to suffice also with less consumption [51,52]. In a carbon- and oil-constrained world, automation and ‘robotization’ are not inevitable outcomes.

    A risk of work-sharing is also that resource use can rebound if the liberated time is devoted to resource-intensive leisure. A taxation policy that dis-incentivizes resource-intensive consumption is a vital complement to work-sharing [44].

    Taxation should shift from taxing work, which is a ‘good’, to taxing resource use and environmental damage, which are ‘bads’ [16]. A carbon or resource tax (green tax) could gradually substitute income tax, without increasing overall tax revenue. The green tax should apply not only to domestically produced goods, but also to the resources and carbon embedded in imports.

    A problem is that poor people devote a larger share of their income to resource and energy use. The rich save a comparatively larger share of their income. In an income tax, the rich pay more; with a green tax, the poor may end up paying more. To alleviate this, the revenue from the green tax could be used to eliminate mostly the income tax of low (or middle) income groups.

    Green taxes are not proposed only by degrowth advocates. The difference is that from a degrowth perspective, there is a recognition that strict green taxes at the scale necessary to reduce resource and fossil fuel use, will have in all likelihood a negative effect on the economy. Work-sharing (and other policies) are then meant to deal with the social consequences of lack of growth.

    Without growth, inequalities may also rise if the rate of return to capital is higher than the rate of growth—a larger share of national income will concentrate this way in the hands of capitalists compared to workers [53]. Jackson & Victor [54], however, argue that less growth may suppress the return to capital, and redistributive policies reduce inequalities. A maximum wage (income ceiling) could act as a cap on permissible levels of inequality [55]. A return to the progressive taxes of the 1950s and 1960s, with very high taxes for high-income brackets, combined with inheritance taxes or a global wealth tax could also reduce inequalities [53].

    Is degrowth, however, compatible with the way the monetary system works? Currently private banks create money and lend it with an interest. Repaying interests creates an imperative for the economy to grow [56]. This leads some authors to conclude that states should take back the control of the creation of money from private banks (‘public money’, see [57]).

    Money now enters the economy as debt. When a bank issues a loan, new money is created. Before the financial crisis money and debt grew out of proportion with the real economic activity. While private banks can only issue money as debt through loans, the state, under a public money policy, could issue money free of debt to pay for public needs. For example, states could issue money to finance a minimum universal basic income [55] or to subsidize green investments [57]. States would reclaim seigniorage (the difference between the nominal value of money and the cost of producing it), and would not borrow from private banks to finance public expenditures. This will give leverage to states to stimulate desired forms of consumption (e.g. green) and suppress others (e.g. environmentally damaging).

    Private money creation, resource and territorial expansion, and the constant push for reducing labour costs and requirements are central features of capitalist economies. This has led some to question whether a capitalist economy without accumulation and growth is possible (see [58] versus [59]). Capitalist economies are characterized by relentless competition. This creates a ‘grow or perish’ imperative for individual firms, which translates into a growth imperative for the economy as a whole. It is not impossible to imagine in theory a capitalist system confined within limits, many firms perishing, and a few growing [58]. But the experience from periods of stagnation and recession suggests that a capitalist system without growth is socially unsustainable. A shrinking and limited pie intensifies distributive conflict between capitalists and workers. Growth defuses social conflict and this is why it is essential for the stability of capitalism [27].

    Where most analysts agree is that if reforms like the ones discussed above were to be implemented, the resulting economies would look nothing like the capitalist ones of today [3,46,60]. Systems where wage labour is diminished, private banks are redundant—constrained at best to the role of intermediary lenders—and resource and energy use are low and do not expand, would be very different. To see such ‘reforms’ happen, a radical change of political and economic power would be necessary [61]. Dittmer [62], for example, in a critical review of public money concludes that implementing the policy would require ‘a tremendous reconfiguration of power relations between states and finance capital’.

    The above policies apply principally to ‘overdeveloped economies’ [16]—economies whose scale could not be sustained by planetary ecosystems, if it were to become generalized. OECD countries fall under this category. What about the very poor countries though whose basic needs still go unmet? Gutowski et al. [30] argue that there is no historical precedent of a country developing without industrializing, and there is no industrialization without increased resource use. Satisfactory levels of development, as measured by the human development index, can be achieved at lower levels of resource use than those of the richest economies today. But generalizing even these decreased levels of resource use will overshoot planetary boundaries [30,63,64].

    This creates an impossibility theorem. If greening and dematerializing growth is not possible, and if development without growth and materialization are impossible, then in the long-term sustained development is impossible. There is some hope, however, insofar as the human development index includes GDP. It is almost tautological to claim that human development requires GDP growth. The important question instead is: are there different understandings of wellbeing and human development that could be materially sustainable? The work of Gough [39] on defining a set of basic needs that could be sustainably met moves in this direction.

    6. Conclusion

    The take-home message of this article is that decarbonization and dematerialization are incompatible with economic growth; and that strict policies for decarbonization and dematerialization will have a negative effect on growth. This requires thinking how to manage without growth. Three policies were indicatively discussed: sharing work, taxing resources and energy instead of work, and creation of money by the state, redirecting it to a green transition.

    All this is a tall order. Given the current political climate, opting for a voluntary ‘prosperous way down’ is extremely unlikely. The pursuit of growth has the features of a ‘collective action’ tragedy. Without sufficient international governance, each nation is bound to pursue growth in the search for economic and military advantage in a globalized world. The end result is likely to be ruin for all.

    Understandably, some prefer to imagine against historical experience, that it is possible to decouple resource use from economic growth, and to fuel growth with renewable energy. The technological achievements during the last 200 years or so of capitalism fuel this hope. Yet previous technological achievements were energy and resource-intensive, accelerating the transformation of matter and the fresh occupation of new territories. Capitalism has been extremely good at relentlessly and violently expanding; there are few signs that it can be as good at peacefully contracting.

    It might well be the case that ‘the politically acceptable is ecologically disastrous while the ecologically necessary is politically impossible’ [65]. Scientists, however, should not limit the pursuit of truth within the confines of the politically acceptable. This article provided theoretical and empirical evidence for such an inconvenient truth—further growth is not ecologically sustainable.

    Competing interests

    I declare I have no competing interests.


    This research was supported from the ‘María de Maeztu’ Unit of Excellence (MDM-2015-0552) grant from the Spanish Ministry of Economy and Competitiveness (MINECO) and the SINALECO grant no. CSO2014-54513-R SINALECO.


    1 Note that dematerialization would last only as long as there are possibilities of substitution. When the transition has been completed, any further growth will increase material use, no matter how dematerialized the economy is. This is the basic insight behind Ward et al.'s [18] claim that decoupling GDP growth from environmental impact is impossible. Assuming continuous substitutions and efficiency improvements, at some point the physical limits of efficiency improvements will be reached, and after that any further growth will be unavoidably coupled to environmental impact.

    2 A question also is to what extent GDP growth might be inflated from financial services—20% of GDP in some advanced economies—and private or public debt.

    One contribution of 19 to a theme issue ‘Material demand reduction’.

    Published by the Royal Society. All rights reserved.