The Food-Energy-Climate Change Trilemma: Toward a Socio-Economic Analysis

Introduction

The principal purpose of this paper is to develop understanding of the interaction between socio-economic and bio-physical systems by focusing on a particularly significant interaction complex. The concept of a ‘food-energy-climate change trilemma’ identifies a strong interdependency between the growth in food demand, the decline and increasing insecurity of fossil energy resources, and the increase in anthropogenic climate change arising from land conversion and use for both food and energy. The trilemma represents an unprecedented challenge to continued economic and social sustainability. Yet the trilemma takes varied forms and dynamics in different regions, with different pathways of growth and development, in the context of varying resource endowments, and differential access or potential access to food and energy, of whatever kind.

As a concept, the trilemma first emerged within the natural and environmental sciences (Tilman et al., 2009), highlighting the evolving dynamics between various kinds of resource finitude (fossil energy, land, water, etc.) and the generation of greenhouse gases from shifting patterns of utilization of these resources in economic development. Central to the trilemma concept, as discussed further below, is the complex interaction between different kinds of resource demand and use, on the one hand, and different kinds of finite resource constraint, on the other, and their combined consequences for climate change. Much of this type of analysis operates in terms of global aggregations (e.g. of oil demand, CO₂ emissions, food supply and demand, etc.), and thus talks generically of ‘anthropogenic’ climate change. Understanding these evolving and shifting dynamics – especially resulting in the increasing importance of land use, land use change, and food production – presents an ongoing challenge for natural and environmental science, and at the same time a quite distinctive challenge for social sciences. The core argument of this paper is that analysis of trilemma dynamics in terms of sociogenesis arising from the interaction between diverse political economies and their variously finite natural environmental resources requires significant new theoretical development. Different political economies generate different trilemma challenges as a consequence of their distinctive polity-economy-nature interactions.

The paper first reviews a sample of theoretical approaches from classical political economy, innovation studies, political ecology, economic geography, and economic sociology, arguing in particular for the openness of a neo-Polanyian approach to meet the challenge. The second section outlines the natural science account of the trilemma. Using a neo-Polanyian approach, the third section will analyse the case of Brazil as exemplary of socio-economic trilemma variations and their uneven and emergent historical character. The conclusion draws together the elements that point to the nature of the theoretical challenge to the social sciences, emphasizing that the paper seeks to identify the dimensions of the challenge rather than provide a fully developed theoretical framework adequate to it.

The Theoretical Challenge to Social Science

Reading a 1986 paper of the economist Nikolas Kaldor entitled ‘Limits on Growth’, one is struck by two things: first, he points to the long-standing pre-occupation of political economy to understand the dynamics of interaction between economic growth and finite global resources, between socio-economic and biophysical systems; second, he was still intellectually trapped in a ‘pre-climate change, pre-peak oil’ epoch. The classical political economists of the time of the industrial revolution, Malthus and especially Ricardo, had enunciated a Law of Diminishing Returns with respect to land as a finite resource. In their view, growth would inevitably find a limit as a consequence of the ever increasing economic resources, including notably labour, required to produce food and other products from ever more marginal land. Applied initially to food and land, Kaldor argues that the same principle can equally operate for all other finite resources, such as fossil fuels, metals, and water, which also may draw labour and technology more and more into the primary sector to the detriment of growth in the secondary (industry and services) sector.

Yet, in a period only shortly following Hubbert's prediction of an American, rather than global, peak oil (Defeyes, 2001), this classic historical conception of a dynamic of diminishing returns lacks recognition of the ‘energy cliff’ that confronts industrial capitalism for the first time in its historical experience. The Law of Diminishing Returns conceptualization does not visualize a finite resource disappearing on use: landmass does not shrink progressively to zero as more and more of it is settled and cultivated. The classic political economy fails to distinguish clearly between finite amounts of a resource available for use (such as land area) and finite amounts of resources being used up (such as oil), and the range of intermediate kinds of finitude (such as land temporarily or permanently exhausted through use or pollution). The Law of Diminishing Returns operates within a fairly undifferentiated concept of resource finitude, although tilting, from its origins, to the ‘availability for use’ end of the spectrum. Moreover, with respect to energy resources, until the present epoch, the Law scarcely made an appearance, either in reality or in political economic thought. For, in previous energy transitions, coal energy had been adopted and developed without an exhaustion of wood-, ¹ charcoal- and peat-based energy. Likewise, oil and nuclear power were additions to the existing range of energy sources.

From the perspective of this paper, moreover, the Kaldor review of the Law of Diminishing Returns also conspicuously lacks any spatial dimension. Notably, there is no discussion of where finitudes of land use resulted in effects on economic growth or, more generally, of how different geographies of resource finitudes (types of fossil energy, water, etc.) relate to different trajectories of development and/or sustainability crises. However, recent accounts of the ‘great divergence’ between regions of the world from the inception of the industrial revolution argue that, in addition to the accessibility of relatively cheap coal, especially in England (Allen, 2009), Europe and other thriving regions of the world such as the Yangtze river basin, were increasingly faced by crises of land availability – finite resources of land limiting growth (Pomeranz, 2000). Only the new colonization from European countries to the ‘New World’ broke out of that resource constraint, so providing European populations with huge new resources of calories (sugar) and clothing (cotton) as well as the new slave labour forces to cultivate them. ²

The industrial revolution and its accompanying urbanization depended in significant measure on the slave plantations of the New World, and subsequent forced labour regimes stretching into the early decades of the 20th century. It was the unrecognized Great Escape from the Malthusian and Ricardian Law of Diminishing Returns, fixated as this was on implicitly national territorial finitudes of land and labour resources. From 1700 to 1890 the area of cultivated land grew by 466 per cent, historically by far the most rapid expansion, and predominantly outside Europe and China. North America witnessed an expansion of 6,666 per cent over the same period – a rather unreal figure given the low initial base (Meyer and Turner, 1992).

This escape from national finitudes of land resources can now also be understood for its significance for anthropogenic climate change. The conversion of uncultivated to cultivated land, and then the uses of energy, fertilizers and simply the emissions from tillage agronomies releasing the stored carbon in the top soil surface, are now recognized to be major sources of anthropogenic greenhouse gases (Houghton, 2003, 2008). At present, anthropogenic GHG from land conversion and agriculture is 2½ times greater than that from total global transport and its oil consumption (World Resources Institute, 2005). So, at the inception of the industrial revolution, there was a particular dynamic between socio-economic and bio-physical systems, a specific historical and spatial interaction between land resource constraints and energy consumption. The satanic mills burnt coal, certainly, but they also processed cotton and were worked by labour forces in part dependent on the calorific energy of sugar, entailing a rapid expansion of cultivated land in the New World as an additional major source of anthropogenic climate change.

This historical interaction between socio-economic and bio-physical environments can only now be seen as a distinctive ‘food-energy-climate change’ constellation, spatially centred around England and the New World sugar and cotton economies. The constellation was an historically emergent phenomenon, with a specific spatial and temporal scale. The ways in which classic political economy – here represented by Kaldor – conceived the interaction between political economies and finitudes of resource in terms of a generic capitalist economy and despatialized finite resources limits the understanding of those interactions.

Although historically overlapping, yet with no evidence of their ‘speaking’ to each other, the more renowned Limits to Growth (Meadows et al., 2004 [1972]) operated with a very different perspective on the interaction between economic systems and the planet earth, yet one which was in its own way equally despatialized. Indeed, both sides of the debate between the Limits to Growth and innovation studies perspectives (Freeman, 1984, 1996; Cole et al., 1973) share the characteristic of theorizing in universal global terms. The ‘limits to growth’ perspective aggregates total greenhouse gas emissions and projects their growth into the future, on the assumption of projected exponential growth in demand for energy, food, minerals, and global resources of all kinds. It suggests that ‘exponential growth has been a dominant behaviour of the human socioeconomic system since the industrial revolution’. And, of course, greenhouse gases as such do aggregate, and have an aggregate impact on the planet, although, as argued below, not as a consequence of a single ‘human socioeconomic system’: the anthropogen. A counter-argument from innovation studies suggests that the growth of scientific and technological knowledge lacks any discernible limits, suggesting an equally global but ‘infinite’ knowledge resource, potentially enabling, for example, the emergence of renewable energy or sustainable agriculture tapping – again for example – the also relatively infinite resource of solar energy (Freeman, 1984, 1996; Cole et al., 1973). From this standpoint, scientific technical fixes arising from inexhaustible knowledge resources are projected as a global potential for a portfolio of ‘stabilization wedges’ (e.g. wind farms, solar energy, nuclear power using thorium, sustainable intensification of agriculture, etc.) (Pacala and Socolow, 2004; Royal Society, 2009; UK Government Office for Science, 2011). In contrast to the classic political economy treatment of an abstract economy in interaction with spatially unspecified resource availability, the much more empirically driven ‘limits to growth’ view and its innovation studies counterpoint are each propounding an interaction between a global economic system (as characterized by statistical techniques of aggregation) and the earth, and especially its atmosphere, as a global whole.

Transition theory has a much more circumspect view of the capacity for technological innovation and growth of scientific knowledge to transcend limits to growth in its distinctive approach to transitions to sustainability. The approach stresses different levels of economic organization and the interdependency of socio-technical systems, reinforcing the idea of path dependencies constraining radical technical change. Obstacles to radical and rapid decarbonizing of economies can be illustrated by the interlocking of the petro-chemical complex, fossil fuel transport and power generation, its physical infrastructures, and, not least, the associated taxation regimes, as analysed by Unruh (2000; Geels, 2002, 2004). The multi-level perspective (MLP), where different innovation and change processes occur at the ‘ground-level’ of niches, through socio-technical regimes, to the overall landscape of a socio-economy, articulates the complexity of major socio-technological transitions. In developing the MLP approach, historically comparative examples have been drawn from a range of economic fields (land transport, sail to steam ships, sewerage, continuous flow mass production systems in factories, etc.) to provide examples of different types of transition (transformation of landscapes, reconfiguration, technological substitution, and re-alignment and de-alignment) (Geels and Schot, 2007). And finally, more recently, unlike earlier work, issues of economy-nature interactions have been introduced, suggesting that climate change constitutes a novel environment for innovation, requiring some adaptation of the theoretical framework. Transitions to sustainability are contrasted with earlier cases. Notably, the state and public authorities as key agents of innovation achieve much greater recognition than previous accounts, which were characterized by more bottom-up niche novelty and innovation (Geels, 2010). This theoretical shift has further been amplified by the analysis of transitions to renewable energy, where the state is attributed a much greater role (Smith et al., 2005).

However, a lack of comparative analysis of political economies of capitalism – let alone of non-capitalist economies contemporary and historical – is a feature of transition theory. In that sense, the MLP as an analytical framework is presented as a general model, applicable to any political economy in any spatial context. It is striking, for example, that the analysis of the transition from cesspits to sewerage systems in the Netherlands paid so little attention to economy-nature interactions (Geels, 2006). A comparison with a similar transition in the United Kingdom suggests that a major sustainability crisis, the ‘Great Stink’ of London, was a significant and decisive disruptive event in a transition process. The crisis arose from an interaction between particular socio-technical regimes, a polity, and a specific natural environment (Harvey, 2012; Halliday, 2001). Likewise, Evans's analysis of the transition of water and sewerage provision in Hamburg emphasizes the distinctive spatial and historical configuration of a ‘free city’, largely independent of the Prussian state. Hamburg, with its mercantile elite, was a unique point of confluence of migrating people, international trade, rivers and water-borne disease (Evans, 1987). The societal and spatial decontextualization of the MLP representation into three levels, and an implicit assumption of its general applicability, filters out analysis of politico-economic variation and spatio-environmental context in the cases it analyses. So explanation of major variations in transition trajectories and fossil carbon path dependencies are difficult to encompass in MLP representation as it stands, especially given the significance of spatially located resource finitudes and their use in inducing climate change.

Unlike these theoretical limitations in addressing resource finitudes and climate change arising from abstract and unspatialized economies, the aggregate anthropogen, and decontextualized MLP transitions, some political ecology, resources geography and recent economic geography approaches take these issues as central. In part derived from Marxist tenets of O'Connor's (1998) second contradiction (the intensifying tendency of capitalism to undermine the reproduction of the necessary environmental and resource conditions for continued accumulation), and Marxist theories of neoliberalism (Harvey, 2005), a strong theoretical strand takes neoliberal attempts at marketization of the environment as the focus of this perspective on political-economy interactions with natural environments. Supportive evidence of this approach is found in the formation of markets for renewable energy from previously uncommodified natural resources, the use of carbon trading mechanisms favouring large corporations for CO₂ emissions reduction (Castree, 2009), and the erosion of common pool resources (Prudham, 2007; Mansfield, 2004; Swyngedouw, 2004, 2007). The most strident version of this view of the ‘neoliberalization of nature’ assumes a single hegemonic form of global capitalism, both ideological and institutional (McCarthy and Prudham, 2004; Heynon et al., 2007). Rather than examining interactions between different political economies and their environments, whether in terms of access to, and use of, finite resources or GHG emissions, this deployment of the concept of neoliberalization emphasizes general worldwide processes such as deregulation, appropriation by dispossession, and marketization.

However, others have stressed both the variety of neoliberalizations (Peck and Tickell, 2002) and the disjunctures between neoliberalist ideology and its instantiations, emphasizing the contradictions between the ideology of spontaneous emergent self-regulating markets and strong state intervention and regulation, often with the state as architect and governor of markets (Mansfield, 2004; Bakker, 2003, 2005). In a recent review of the literature on the neoliberalization of nature, Castree in particular has pointed to definitional weaknesses, but above all to a methodological failure of using singular exemplary cases as instances of general neoliberalizing processes, rather than a comparative and historical method necessary for understanding the kind of variation indicated by Peck and Tickell (Castree, 2010). More recently, Peck (2013) has advocated a Polanyian turn to economic geography, especially the later ‘anthropological’ Polanyi, precisely because of its comparative and historical approach.

The neo-Polanyian ‘instituted economic process’ (IEP) approach adopted here takes historical and spatial variation as central to all its analysis, and to an as yet limited extent has addressed some polity-economy-nature interactions (Harvey, 2007). Building on Polanyi's later work (Polanyi, 1957), economies are analysed in terms of relational configurations of processes of production, distribution, appropriation and consumption. The analytical approach explores how economies – the configurations – are instituted and transformed in space and time, and how spatio-temporal scales of configurational organization are formed, rather than found, within a pre-given multi-level ontology. Economies of food (exemplified by the tomato) (Harvey et al., 2002), knowledge (Harvey and McMeekin, 2007), water (Harvey, 2012), and energy (Harvey and Pilgrim, 2012) have been analysed for their variation and transformation in space and time. In Exploring the Tomato, contrasts were drawn between the Northern European glasshouse and the Southern European polytunnel regimes, related to the availability and use of, respectively, fossil fuel or solar energy. The particular competitive advantage of the Dutch use of their North Sea natural gas resource for heating glasshouses was seen as contributing to the demise of the Guernsey tomato. Moreover, by contrasting the UK supermarket with the Dutch export-oriented small producer economies of tomatoes, these political economy/natural environment interactions pointed to sources of variation within given, similar environmental settings, notably the distinctive forms of hybridization and innovation resulting in Tesco tomato varieties or the Dutch ‘Greenery Tomato’: different hybrids in different socio-economic spaces. A similar type of analysis of political-economy/environment interactions was developed in relation to European, Brazilian, and American biofuels. Continental European biodiesel, with its use of rapeseed as an already established temperate zone crop as its feedstock, interlocked with a tax regime that had established diesel powertrains as the dominant light vehicle type in the previous decade. The European biofuel trajectory contrasted with the Brazilian tropical resource of sugarcane bioethanol, or the USA's lock-in to maize as a feedstock. Different political economies in interaction with different environmental resources and dominant agricultures resulted in different transitions to renewable transport energies.

Further, Polanyi's central concept of the ‘shifting place of the economy in society’ has been expanded by considering different modes of instituting economic configurations, especially the changing relation between polity and economy manifest in responses to peak oil and climate change. Whether for biofuels or wind farms, solar or nuclear energy, nation-states have engaged in highly interventionist, often long-term strategic reconfigurations of economies of energy, entailing new industrial divisions of labour, operating in politically instituted markets. In this respect, diverse political modes of instituting economies of energy have been widely commented on, attracting the somewhat anachronistic label of ‘climate Keynesianism’, for example, in the case of carbon offset markets (Newell and Patterson, 2010). This perspective of varied modes of economic transformation and reconfiguration, with different polity-economy dynamisms, has parallels with Block's conceptualization of the ‘hidden developmental state’, a characterization of strong but unspoken state interventionism in the US economy, which departs sharply from its paradigm reputation as the homeland of neoliberal, free market economic organization (Block, 2007, 2008; Block and Evans, 2005). Moreover, as economic reconfigurations responding to peak oil, food security, or climate change are witnessing new forms of political intervention, so new spaces open up for social and political movements to shape these reconfigurations – notably in the case of NGOs with respect to European renewable transport energy (Pilgrim and Harvey, 2010; Harvey and Pilgrim, 2012).

Yet, although the IEP approach has addressed polity-economy-nature variation, involving dynamics of interaction with natural environmental settings and access to resources, it has done so only to a limited extent. Notably, it lacks analysis of the kind of feedback loop interactions involved in the food-energy-climate change trilemma. It is yet to address variation in sociogenic (as against anthropogenic) climate change. The theoretical challenge is to take the state-of-the-art natural scientific understanding of these complexities – how land use change generates greenhouse gases, what are the biophysical consequences of shifts from fossil energy to bioenergy, what are the diverse energy inputs involved in food production, etc. – in order to develop an analysis of the varied politico-economy/nature interactions generating varied trilemma consequences in different regions of the world. So, rather than attempting to construct a parallel social science ‘materiality’ (Bakker and Bridge, 2006), it is to current state-of-the-art natural and environmental science (including its controversies) of the trilemma that we first turn, before looking to Brazil as an example of a particular trajectory of trilemma evolution.

The Emergence of the Abstract, Natural Science, Concept of the Trilemma

Although scientifically appreciated for many years, the significance of land-use change and agriculture as sources of greenhouse gases came into full focus with the controversies surrounding the competition for land between biofuels and food (Gallagher, 2008). At the height of the food and oil price spikes of 2007–8, Searchinger and Fargione (Searchinger et al., 2008; Fargione et al., 2008) published papers in Science arguing that the carbon footprint of biofuels was far greater than previously calculated and, far from being benevolent for climate change, biofuels were a cure worse than the disease (Doornbosch and Steenblick, 2007). Simplifying the argument, the production of biofuels, particularly from food crops such as maize in the USA, raised the price of food and stimulated agricultural production elsewhere, thus leading to the conversion of more uncultivated to cultivated land. This is termed Indirect Land-Use Change (ILUC), and ILUC was suggested as a significant source of greenhouse gases, which should be included in the carbon footprint of biofuels. The argument has been contested from many angles, and it could just as easily be argued that high oil prices stimulated the production of alternative biofuels, which then led to ILUC, so increasing the carbon footprint of oil (Harvey and Pilgrim, 2011). Nonetheless, once land-use change fell under the floodlights of controversy, attention to the much wider issue of land-use change, especially direct land-use change for food to feed the growing global population, achieved much greater prominence, and led to the articulation of the natural science concept of the ‘food-energy-climate change trilemma’ (Tilman et al., 2009).

In Figure 1, the interactions between socio-economic and bio-physical systems are displayed as an undifferentiated global phenomenon, aggregating all socio-economic activity. As an exercise, it is worthwhile running through the natural/environmental science account of the dynamics, in simplified form. OPEN IN VIEWER

From the top of the diagram, assumptions are made that there will be both an increase in demand for food, with rising standards of nutrition, and a world population growing from 6.5 billion to a projected plateau of 9 billion (IAASTD, 2009; Evans, 2009; Royal Society, 2009; Pretty, 2008), and an increase in demand for energy. Increased demand for food directly increases demand for productive land, and for increased productivity of agriculture. Increased demand for energy, and in particular transport energy, with the transport fleets of China and India either overtaking or having just overtaken that of the USA, will increase demand for oil. Further, plastics derived from oil and other key industrial materials also place increased demands on oil. Yet, oil being a finite resource, with many suggesting that global peak oil has already occurred (Aleklett et al., 2010), or will shortly (IEA, 2013), increasing pressure is placed on finding substitutes for oil, of which biomass is a prime candidate, not only from biofuels but also from biomaterials and industrial biotechnology. Failure to find substitutes for oil, in the view of the International Energy Authority, is likely to result in an economic depression which would make the current economic crisis appear no more than a dimple on the way (IEA, 2011–12). Oil price-induced depression affects the poorest regions far more extremely than the rich, in part from impacts on food prices. Given especially that agriculture is dependent on considerable energy inputs, the choice is not energy or food. Moreover, less intensive agriculture would mean more land for the same amount of food, and it has been argued that the use of chemical fertilizers, in spite of their carbon footprint, has consequently less environmental impact than less intensive agriculture (Burney et al., 2010). But finding biomass alternatives to fossil fuels also increases pressure on land cultivation and conversion. So the pressure for land comes from a double pincer movement, and risks to climate change and biodiversity increase. This then constitutes the trilemma: more food results in more climate change. Burning fossil fuels leads to more climate change. Failure to find alternatives to oil results in unending economic depression. Biomass alternatives increase demand for land-use change, leading to more climate change. More climate change results in less land availability and less productivity of much existing agricultural land, and so undermines food supply. And, indeed, results in major disruptions to life on the planet as we know it.

Trilemma Interactions and Reactions: The Exemplary Case of Brazil

The natural scientific account of the trilemma, necessary for understanding complex biophysical dynamics, assumes generic, unspecified human socio-economic activity. In this section, a preliminary attempt will be made to develop an IEP approach to understanding sociogenic varieties of trilemma generation, the emergence and progression of the trilemma, or, as importantly, the transitions or transformations responding to its varied sustainability crises. Brazil is taken as an exemplary case to explore these interactions: the sources of variation in the dynamics between regions, and the spatially and temporally uneven development of, and responses to, the different horns of the dilemma (energy, food, climate change). To anticipate, Brazil reacted similarly to the USA to the first oil shocks of the 1970s, but then pursued a very different strategic response. And, partly as a consequence of its biofuels strategy, it has been in the forefront of controversies and responses to developing sustainable agriculture regulation. In so doing, for example, its Zero Deforestation Policy (however effective or not in practice) is symbolic of a political redefinition of the finitude of land, by ruling in and out the availability of different types of land for agricultural or other human use. Finitudes are politically modified, thus changing the goal posts of limits to growth.

To illustrate the significance of this social scientific focus of trilemma dynamics, a good starting point is the contrast in the actual carbon footprint of Brazil compared with other global economies, as portrayed in the UN Environment Programme's analysis of a prospective failure to adequately respond to the challenge to reduce that footprint. The two figures below, taken from The Emissions Gap Report 2012 (UNEP, 2012), represent, first, the distinctive carbon footprint signatures of different countries and regions; and, second, the better-known summary of the relative per capita carbon footprints of different countries and regions.

Figure 2 demonstrates the starkest contrast between South Korea, Canada, Australia and the USA, at one extreme, and Brazil at the other. As a medium per capita greenhouse gas emitting economy (Figure 3), and one of the most dynamic economies of the BRIC group, Brazil displays a remarkably distinctive carbon footprint. The combined total of emissions from energy conversion, industrial use, and transport is below 30 per cent of their total emissions. In all countries left of Mexico in Figure 2, this energy-industry-transport subtotal is above 80 per cent of total emissions. Conversely, in the case of Brazil, over 60 per cent of its emissions arise from agriculture, especially deforestation and peat destruction. Yet Brazil is not notably less industrialized, energy-consuming, or lacking in transport systems than many of the comparator cases. Rather, Brazil has been engaged in a particular trajectory of political economy-environment interactions, a particular dynamic of development, over several decades. To take two key components of this distinctive trajectory, Brazil has continued to invest heavily both in hydroelectric power and the development of biofuels as an alternative to fossil fuels for transport, discussed more fully below. As a consequence, in 2005, Brazil had the highest level of renewable energy of any world economy – indeed, at 29.1 per cent, almost three times the world average of 11.4 per cent (DIEESE, 2007). OPEN IN VIEWER

OPEN IN VIEWER

Figure 3.

Per capita greenhouse gas emissions for 1990, 2005, 2010, and 2020 for G20 countries. Source: Olivier et al., 2012.

Clearly, both biofuels and hydroelectric power have a land-use related carbon footprint, whether in terms of deforestation or cultivation, so contributing to the contrast with another BRIC economy, Russia, where flaring from gas and oil production combines with the fossil fuel extraction to give that economy its distinctive carbon signature. At the same time, however, Brazil has become a major, and advanced, agricultural export economy: the world's leading exporter of poultry, red meat, coffee and sugar, the second largest exporter of soy beans, soy meal and soy oil, and the third largest exporter of corn (Wilkinson, 2009; FAOSTAT, 2012). In terms of the emergent trilemma, Brazil is contributing to meeting growing global demand for food, and this is manifested in its pattern of GHG emissions from food production and deforestation. The emerging trilemma dynamics of Brazil contrasts with other regions and nations of the world, while interlinked through trade with those regions, with their distinctive dynamics. However, as with other ‘exported’ carbon footprints, these cannot straightforwardly be re-attributed to consumer (importing) nations. As will be seen below, the key issue is how food is produced, and where, in the producer country. On the one hand, much depends on the forms of agriculture and land conversion, with potentially huge variations in carbon footprint (e.g. slash-and-burn versus recovery of degraded land). On the other, Brazil's natural endowments of high rainfall and sub-tropical solar energy in principle permit lower carbon footprints for producing food and biomass than temperate zone regions, so potentially contributing to a global realignment of agriculture as a response to climate change.

In analysing the emergent character of the Brazilian trilemma dynamics, different phases are distinguishable, with different horns of the trilemma developing unevenly: energy security, land finitude, renewable and environmental concerns, first for biofuels, then for food. Energy security and the issue of finite resource constraints was the first ‘horn’ to markedly shift the configuration of the Brazilian economy. There has not been, and will not be, any simple physical ‘peak oil’ created by technologically available means of extracting finite physical resources (Cavallo, 2005; Bridge and Wood, 2010; Bridge and Le Billon, 2013; Sorrel et al., 2009). The 1970s oil price spikes occurred just as the USA passed its Hubbert's peak of domestic conventional oil production, leading to progressively increased dependency on Middle East oil, and its vulnerability to political shocks. These price shocks were hence powerful events which conditioned political perspectives to the geopolitics of oil as a finite, spatially concentrated resource in the decades that followed.

Finitudes, Transformations and Transitions

The Brazilian example represents an exploratory attempt to develop a neo-Polanyian IEP analysis of the socio-economic and bio-physical environment interactions central to trilemma dynamics. Challenges of ‘peak oil’, food production, land as a resource, and climate change vary importantly from one economy or region to another. To make the very banal but increasingly significant point, it matters greatly how much and where carbon fossil (conventional and unconventional) or land resources are located. There is simply more solar energy available in Brazil than in the temperate North, and that opens possibilities for variation in political-economic trajectories. Moreover, these finitudes are not just physical quanta located in geographical space. As we have seen with land – the original focus of the Law of Diminishing Returns – how much land is available for use has been redefined in certain political spaces by sustainability regulations responding, at least in part, to climate change risks. The 2010 catastrophe of the Deepwater Horizon oil well in the Mexican gulf, or similar events in polar regions, may likewise set political and economic limits to what oil resources may be available to be ‘used-up’. The contrary dynamic of some innovation pathways (fracking, horizontal drilling) opening up resources previously deemed unexploitable may themselves be subject to regulatory limitations of varying rigour in different political contexts. Although the International Energy Agency has significantly revised its view of US energy supply for trilemma dynamics (IEA, 2011–12, 2013), the critical issue is the US political priority of energy security over climate change in determining whether and how much of this resource is exploited. So, again, the critical interaction is between the economy and environmental resources within a political space.

Illustrated by the example of Brazil, a similar argument also applies to food security, and the major trilemma challenge of feeding a global population expected to plateau at approximately 9 billion. Although in purely bio-physical terms, the energy inputs required to produce a given quantity of food are much lower in sub-tropical rain-fed zones (reducing climate change impacts), how and where such food is grown, what lands are converted to what uses, becomes critically important, and increasingly subject to politically strategic responses to climate change. The finitude of land in terms of availability for use is becoming – however tardily and inadequately – politically redefined in terms of what use, what agronomies, rather than use in the abstract. The choice between slash-and-burn and sustainable intensification – to counterpose two extremes – is a political as much as a technological one, as is evident already from the political regulations in different regions for the use of GM technologies. Certain lands are available for certain uses (GM soya, cotton, corn, etc.) in some regions, not in others.

The argument is not that finitudes are politically constructed, let alone discursively construed. There are natural ‘givens’ open to some political circumscription in varied triangular relations of polity-economy-nature in different material environments. The argument in this paper is nonetheless that this conceptualization of resource finitudes significantly changes perspectives on limits to growth by placing them within the varied dynamics of these triangular relations as they develop in different regions and nations across the globe. The dynamics are not uniform and global, however much the planet and its atmosphere are shared by the world's peoples. Although gases and world demand for oil may aggregate, it is critically important to disaggregate the dynamics that generate them.

In closing, the neo-Polanyian approach illustrated here is far from fully developed in relation to the theoretical challenge for understanding interactions between political economies and natural environments. To emphasize this incompleteness, it is useful to point to some key missing dimensions yet to be addressed. First, the analysis needs to address issues of resource depletion in a more fundamentally comparative way, in terms of the located finitudes of resources and the significance of the material rootedness of economies in their natural environments. Second, in abandoning anthropogenic in favour of sociogenic climate change, the IEP framework needs to address the developing dynamisms which produce the markedly varied carbon signatures of different nations and regions. The very different political challenges for climate change mitigation in Brazil, the US, Europe, China, India, Africa … are embedded in these contrasting polity-economy-nature relational spaces and dynamisms. The analysis needs to be complemented by natural scientific analysis of the impacts of these different trajectories as they develop. Third, the Brazilian case is marked by a radical political change of regime, and of politics within a democratic regime. The different political institutions and politics in the USA, Europe, and China, for example, themselves may condition and constrain radical transitions to sustainability, in terms of lock-in or entrapment, in ways that have not been analysed here. Fourth, emergent trilemmas bring to the forefront the nexus of interconnections between food, land, energy and climate change, and the interdependent and antagonistic forces of sustainability transformations and doomed path dependencies. The choice is not to feed the 9 billion or develop renewable energy, to mitigate climate change or develop new directions of economic development. To explore these developments, analysis needs to be undertaken at the geopolitical level, incorporating international trade and supranational political regulation and strategic responses to trilemma dynamics. And this list of further dimensions is certainly itself incomplete.