climate change – Institute for Interdisciplinary Research into the Anthropocene

Homeland’, borders, and business-as-usual

Framing the environmental crisis

It has long been accepted amongst various communities of academics that both political ideas and discourses matter in framing political issues, rendering actors and phenomena visible or invisible, and shaping political outcomes.1 A pertinent example of this is the phrase ‘Anthropocene’ – used to denote a new geological era in which human activity has significant impacts on planetary ecosystems – but which is itself contestable for the phenomena it captures and elides. Some have put forward the alternative term of ‘capitalocene’ to reflect the understanding that the primary driving force of ecological change in this era is not human activity per se, but the capitalist systems which continue to drive resource extraction, greenhouse gas emissions, and rising inequalities.2

“The far right discourse on the ecological crisis has historically been to deny its existence”

The ecological crisis is subject to a series of political discourses which each imperfectly capture the complex myriad of social, economic, and technological dynamics that are degrading planetary ecosystems. These discourses shape the public understanding of the environmental crisis and the appropriate strategies for its resolution, with each discourse purveyed by distinctive but evolving political factions and social forces.3^,4

The far right discourse on the ecological crisis has historically been to deny its existence.5^,6 This denial has taken many forms, but most commonly the science of ecological degradation has been disavowed and this has been matched by the refusal to accept any national responsibility for addressing the unfolding global ecological catastrophe. Customarily, the scientific evidence has been pronounced as a conspiracy designed to benefit ‘globalist elites’ or a plot to undermine national sovereignty through the ratification of multilateral agreements. This has served to bolster resistance to effective environmental policies.

However, this environmental discourse is no longer as central to the far right movement as it was in the 2000s and 2010s. Increasingly, climate science is tacitly accepted, but the finger of blame is being disingenuously pointed towards the far right’s traditional enemies.

The shifting environmental discourses of the European far right

As environmental issues have risen up the political agenda (becoming salient to younger voters in particular), far right parties have seemingly shifted away from denialism of the science. This shift has not led to a recognition of the need for a just economic transformation or, indeed, any political action commensurate to the scale and character of the environmental crisis. Instead, the increasing (albeit belated) recognition of environmental issues (primarily those which exist within national borders) has been fused with an anti-immigration agenda to create a new invidious framing of environmental politics. The emerging discourse, which we have conceptualised as ‘ecobordering’ elsewhere,7 is characterised by climate nationalism and seeks to depict immigration (of which migration from the Global South is made hyper-visible) as a threat to local and national environments.

This discourse takes two primary forms. First, it aims to politicise the environmental impacts of ‘mass immigration’ from the Global South, while depoliticising the impacts of ‘natives’. This includes linking ‘mass immigration’ with rising demand for natural resources and local environmental problems such as the pollution resulting from greater traffic and consumption. Immigration, it is suggested, is to blame for such problems, which were not issues of concern for local areas prior to multiculturalism.

At the same time, this narrative stokes fears that mass immigration will lead to population growth amongst non-white communities which will exacerbate these local environmental issues further and deplete finite natural resources, in what could be termed ‘racialised Malthusianism’. This was particularly exhibited by the British National Party (BNP),8 the National Rally,9 the Swiss People’s Party,10 Vlaams Belang,11 and Alternative for Deutschland.12 The Swiss People’s Party repeatedly claimed that it is the bulwark against “the greatest environmental killer, overpopulation… by urging people to limit immigration”,13 while the British National Party adopted the same Malthusian logic that it “is the ONLY party to recognise that overpopulation – whose primary driver is immigration, as revealed by the government’s own figures – is the cause of the destruction of our environment”.14

“The depiction of Global South migrants is juxtaposed with the depiction of ‘natives’ as responsible stewards of their ‘homeland’”

The second form this discourse takes is the depiction of Global South migrants as environmental hazards, with no personal aptitude for managing natural resources due to a lack of belonging to, or lack of financial or emotion investment in, local areas. This has been most strongly exhibited by far right parties such as Golden Dawn,15 the National Rally,16 the BNP,17 the Swiss People’s Party,18 and Vox.19 This has included the disparagement and scapegoating of migrants in numerous ways, such as littering, causing forest fires, the inhumane treatment of animals, and the destruction of ‘indigenous wildlife’ amongst other environmental offences.

“The purported threat posed by immigration and migrants… seeks to vindicate the notion that border policies are key forms of statecraft for the protection of the environment”

The lack of belonging is key to understanding this portrayal; as Le Pen explicitly put it: “environmentalism [is] the natural child of patriotism, because it’s the natural child of rootedness… if you’re a nomad, you’re not an environmentalist… Those who are nomadic… do not care about the environment; they have no homeland”.20 The depiction of Global South migrants is juxtaposed with the depiction of ‘natives’ as responsible stewards of their ‘homeland’ and adept stewards of their ‘little platoons’ (to invoke the eco-fascist and Burkean logics which this framing draws upon). This typically entails glorifying the historic stewardship of pastoral national citizens (such as farmers21 or foresters22) and the proclaiming the sound management of domestic natural resources by ‘natives’23 over the ‘homeland’.24^,25 The National Front and Golden Dawn have even established wings of their movements called ‘New Ecology’26 and ‘Green Wing’27 designed to protect “family, nature and race”28 and “the cradle of our race”29 respectively.

Both of these discursive traits have since been identified more recently in Marine Le Pen’s recent presidential campaign in which she obtained 41.5 per cent of the vote. Dubbed ‘patriotic ecology’ by her followers, the fallacious depictions of culprits and saviours in the environmental crisis have become normalised in French politics to the extent that they are echoed by rival conservative politicians.

The purported threat posed by immigration and migrants to previously ‘pure’ and ‘sustainable’ spaces of European nature seeks to vindicate the notion that border policies are key forms of statecraft for the protection of the environment. As a senior figure in Marine Le Pen’s National Rally, Jordan Bardella, declared in 2019: “borders are the environment’s greatest ally… it is through them that we will save the planet”.30

A shift away from climate denialism, but at what cost?

The potential electoral potency of fusing border securitisation and climate issues – however fallaciously – underlines the importance of recognising and challenging these discourses. Should the ascendant far right in Europe gain any further power, or have further influence on traditionally conservative political parties, this discourse could more forcefully shape public understandings of the environmental crisis and the strategies for its resolution in the future.

“To ignore the root causes of the ecological crisis at this juncture would be catastrophic for the natural world”

This would be catastrophic on two fronts. On the one hand, the discourse prescribes a form of statecraft centred on border security rather than systemic economic transformation, which represents an apocryphal programme of environmental protection. It does so by focusing narrowly on ‘national’ nature (peripheralising global issues) and obscuring the material economic drivers of ecological degradation (such as the heavily polluting energy and aviation industries, for which Global North populations are primarily culpable). To ignore the root causes of the ecological crisis at this juncture would be catastrophic for the natural world, but that is precisely what this political framing inculcates.

Just as importantly, ecobordering seeks to inflict further structural violence on those who those exploited at the peripheries of the global economy. The nationalistic framing emerges at a time when immigration is rising because of climate change, and the discourse thus seeks to diagnose the symptoms of ecological degradation as the causes of it. There is already evidence that the rise of the far right strengthens political resistance to climate migration,31 and this framing serves to justify this resistance from an environmental perspective. At a global scale, these framings threaten to rationalise a de facto climate apartheid; with Global North populations and elites in the Global South enjoying the spoils of an environmentally deleterious global economy, while poorer Global South populations become confined to increasingly uninhabitable areas facing escalating risks of climate shocks and deteriorating health conditions.

The meaning and practical implications of climate justice will become an increasingly hot topic in the Anthropocene. Challenging the depictions of culprits and saviours purveyed by far right figures is only an initial step to preventing injustices mounting further.32 Recognising the historical constitution of the global economy and the inequalities and vulnerabilities resulting from it underlines the injustices of far right framings and the need for progressive actors to advance more transformative approaches.33 Progressive responses to the rise of the far right in the Anthropocene requires formulating and advancing notions of a just transition which accounts for the movement of people affected by climate change as well as other less privileged groupings in society.34 This will require far more progressive forms of statecraft which are a world away from those advocated in the framings of the far right.

Biographies

Dan Bailey is a senior lecturer in international political economy at Manchester Metropolitan University. His is interested in the evolving and complex interactions between the global economy, climate change, the objectives and strategies of political institutions, and the ideas and discourses that shape public understandings of the ecological crisis and sustainability transitions. He has authored a series of academic publications and policy reports on topics relating to these interactions.
Joe Turner is a lecturer in international politics at the University of York. His interdisciplinary examines how border regimes in post-imperial states like Britain are structured by imperial and colonial histories and hierarchies of human value. He recently published the book Migration Studies and Colonialism with Lucy Mayblin.

Uncategorized

Three Scenarios for the Future of Education in the Anthropocene

April 12, 2020 Updated:January 7, 2026 16 Mins Read

By Kathleen Kesson

We have entered the Anthropocene — a new era in geological history — a phase of planetary development in which human impacts on the Earth may cause or have caused irreversible damage. We are witness to “the great acceleration” in which geothermal, biological, ecological, and atmospheric changes threaten to bring about irreparable changes in the planetary ecosystem, and by extension, our social and economic systems. Every day brings news of wildfires, drought, floods, conflicts, hurricanes, locusts, extinctions, and the latest, a Coronavirus pandemic, which has managed to shut down many of the global systems we rely on for survival.

Humans (GR: ánthrōpos) have been blamed for the tragic despoliation of our Earth. It is not humans in general, however, but a specific human civilization that has driven the processes of resource extraction, labor exploitation, capital accumulation, and what we can only call “ecocide.” While historically, empires have come and gone and laid waste in countless ways to people and planet, the current modern era of industrialization/capitalism, paralleling a centuries-long narrative of conquest, genocide, plunder, slave labor, and economic imperialism has created the conditions of this new age that some scholars suggest we more rightly call the “Capitalocene” (see Moore, 2016).

Given the climate and other ecological crises, the rise of authoritarian/totalitarian governments, and the general breakdown of multiple systems, there is an urgent need to create new, nimble configurations of communities, ecologies, and learning centres to respond to the uncertain and rapidly changing environment. The education (not necessarily “schooling”) of young people is at the heart of the future; it is only through education that a “new human” might emerge, capable of enacting the mindset and behaviors that might create a livable world. Education alone, however, absent substantial changes in culture, thinking and behavior, is incapable of bringing about the fundamental changes necessary to survival.

I offer here three scenarios for the future of education, each of them tied to various components of a dominant governing ideology. Each Scenario is accompanied by structuring metaphors as well as a dominant “binding quality.” The notion of a binding quality comes to us from an ancient Indic episteme; it is said that consciousness and matter operate in three fundamental modes: sattva (sentient), rajah (mutative), and tamah (static), collectively known as gunas in Sanskrit. Understanding the gunas is a complex philosophical matter; I use them here metaphorically, to describe the predominant energy of each Scenario. I have drawn largely on the comprehensive projections of P.R. Sarkar (1992; 1999) for the vision of the future portrayed in Scenario 3, though it must be said that the various components of this vision are emerging from multifarious directions and under different appellations at the present time.

Futures thinking is an uncertain art. It is likely that the future of humanity will include dimensions of each Scenario; in fact, the present moment contains all of them, though Scenario 2 dominates because of the globalization of the economy and hegemonic forms of culture. I believe, however, that the survivability of humanity is dependent on learning the lessons of the multiple current crises we face, and figuring out how to navigate through complexity, chaos and the general breakdown of systems to facilitate the self-organized, positive evolutionary outcomes highlighted in Scenario 3.

An important caveat: When considering the “Big Picture,” generalizations are unavoidable. These scenarios are mapped in very broad strokes, and we must remember that the map is not the territory. Details, diversities, exceptions, and contradictions certainly need to be taken into consideration.

Scenario 1

Regression/Devolution

I start with the grimmest of the forecasts, in order to disabuse us of the modernist notion that history is an inevitable trajectory of progress, of increasing individual freedom and rights, of economic growth, constantly improved standards of living, and the capacity of positivist reason and logical thinking to solve all human problems. As in the aftermath of the Roman Empire or perhaps more vividly, in modern dystopian films, societies can deteriorate rather swiftly.

In European history, the years between 500-1250 AD are usually considered the “Dark Ages.” After the fall of the Roman Empire, and due to many factors including ineffective leadership, economic failures, internal struggles for power, external invasions, and yes — climate change — the western territories of the Roman Empire entered a long period of decline. Historians disagree on many of the details, though there is a general consensus that it was a period of breakdown and change of the social and economic infrastructures. Schools were closed, and illiteracy spread. Travel and trade were restricted, epidemics wiped out huge populations, and conflict was prevalent.

While our modern era may seem to have little to do with the European Medieval period, it’s altogether possible that we (at least in the “West”) are living through the deterioration of an empire begun in the European colonial period and culminating in late capitalism and the economic imperialism that is an essential component of the globalized economy. This world-historical empire has been engaged in endless wars throughout its reign, has deep internal fractures and multiple external pressures, not least from other empires. Most important, as noted above, the bio-systems upon which life depends, and upon which so much of its wealth was created, are deteriorating.

In times of collective stress such as the current pandemic, it is tempting to withdraw, to retreat from the forward flow of life and pull into individual and social cocoons, burrow into the past. That tendency is currently exacerbated by the pandemic related strictures to isolate, to distance ourselves from the social world. Should these tendencies persist after the disease is brought under control, we could see a “devolution.” In such a regressive move, we are likely to see rising xenophobia, racism, religious prejudice, sexism, strong borders, and ever-increasing economic inequality.

Scenarios and metaphors	Worldview/Philosophy	Power	Social/economic organization	Ecologicalperspective	Knowledge	Education Institutions	Spirituality
Regression/Devolution Binding quality: Tamah (static) Contraction, decay, degeneration, ignorance, death and inertia.	Pre-Humanist submersion in forces thought to be beyond human control. Recycling of medieval ontologies and philosophies. People concerned with their own immediate land, clan, family and social group.	Power/over-exerted through superstition and propagation of false ideas; patriarchal structures control behavior, social life, and education.	Provincial, feudal, mostly dispersed rural populations. Centralization of (weak) control in urban centres. Subsistence economy for the masses; wealth flows upward—vast inequalities.	Nature as a force to be feared. Attempts to exert dominion over nature. The exploitation of natural resources benefits the few.	Past knowledge valued over experimental, new knowledge. Knowledge distribution restricted as a form of social control.	Knowledge production concentrated in centres of power.Private teachers/schools for the wealthy. Survival skills adequate for the general population.	Traditional/orthodox/dogmatic; power centralized in the clergy.Metaphysical beliefs grounded in irrationality and superstition—emphasis on domination and control of thought.

Scenario 2

Status quo/Business as usual

Thinking optimistically, we’re unlikely to sink into the miasma of Medieval Europe, but young people who have not lived through a Depression, or an epidemic, or a war on their own territory cannot be blamed for fearing that this is the “end of the world as we know it.” This pandemic, however, and the economic dislocations, the social isolation, the fear and uncertainty that it has brought, while perhaps not the apocalypse much fear, may be a harbinger of the future. It is human nature to want to “get back to normal” following a crisis of great magnitude, to restore a sense of equilibrium and stability. But what if “normal” forms of social, economic, and ecological behaviors are themselves at the root of the crisis? Astute observers of our current modernist trajectories, including a majority of the scientific community, warn us that we are now living through a transition period, which, depending on collective decisions we make in this next decade, have the potential to transform the conditions of life as we know it on Planet Earth, and not for the better. If we continue the rate of petroleum extraction, fossil fuel burning, deforestation, unrestrained consumption, pollution, and so much more, it is clear that humanity is in for a century of increasingly deadly wildfires, droughts, floods, ocean acidification, pandemics, rising sea levels, and massive extinctions on a scale heretofore unimagined. If current power relations persist, and we do not affect a deep reordering of our economic system, power structures, worldview and ways of thinking, if we merely tinker with existing conditions while hoping to achieve what could only be a “false equilibrium,” elites will prosper while our life systems continue to degrade and masses of people suffer. The kind of thinking that has created the multi-faceted crises we face is unlikely to help us solve them, but humans may not, in this Scenario, demonstrate the will or the capacity to radically transform their thinking and their behaviors, or challenge the existing power structure.

Scenarios and metaphors	Worldview	Power	Social/economic organization	Ecologicalperspective	Knowledge	Education Institutions	Spirituality
Status quo/ Business as usual Binding quality:Rajah (mutative) Pulsation, change, growth, movement, restlessness and activity.	Secular. Mainstream rejection of spirituality based on widespread materialistic worldview. Man is seen as the pinnacle of creation. Humanistic emphasis on individualism, independence, personal autonomy.	Power/over-exerted through economic domination and hegemonic media; Power/with only mythology of democratic capitalism. Dramatic concentration of wealth; oligarchical rule.	Increasing inequalities. The illusion of a relatively prosperous (if shrinking) “middle class” sustains myths of growth and progress.	Humans are seen as separate from nature (dualism). Nature understood as a resource to be exploited for profit.	Conventional, hierarchically organized. Positivist thinking dominates. Scientific and technological advances are double-edged (i.e. air travel creates mobility + air pollution, greenhouse gases and rapid spread of disease). Sifting and sorting mechanisms maintain inequities of race, ethnicity, gender, and social class.	Increasing concentration of influence over standards and curriculum in the interest of global economic competition. Higher education commodified, fewer young people have access. Western forms of education spread globally, resulting in loss of languages, local cultures and epistemes.	Mostly secular. Fundamentalisms operate at the fringe, often with major impacts on systems (re 9/11). Commodified “new age” practices amongst middle classes are oriented towards individual well-being.

Scenario 3

Evolution/Revolution

The current crisis has brought into sharp relief the injustice and unsustainability of socio-economic systems that value profits over human needs and the well-being of the planet. It is possible that this moment in time could signal the “great awakening,” the tipping point that pushes us into creative new ways of thinking about what it means to be human and how we should live our lives. What if the present moment were a space of “liminality” — a moment between what has been and what will be? A space between the ‘what was’ and the ‘next.’ A space of transition, a season of waiting, during which we collectively question where we have been and where we are going. A space in which we reconceptualize the entire edifice – the mental and the material structures that have brought us to the current crossroads in our evolution.

https://spotterdavey.files.wordpress.com/2011/09/ccl_bg_en.jpg

In Scenario 3, we find the courage to design and implement new economic structures that serve the welfare of the whole of humanity, not just the elite few. We begin to understand our essential embeddedness in nature and explore how to cultivate relations of harmony and reciprocity with the “more-than-human-others” with whom we share the planet. And perhaps most important, we overcome the false notion that matter and spirit occupy independent realms, separated by an impassable abyss. We begin to understand that the purpose of life is not the mere accumulation of material goods, or the acquisition of political power, or even the development of a brilliant intellect, but the unification of body, mind and spirit in the quest for spiritual enlightenment.

Unlike the “tinkering” referred to in Scenario 2, Scenario 3 represents a radical paradigm shift, an evolutionary transformation of consciousness, values, and human behavior. Education has a core role to play in that it is young people who will carry the present into the future. A philosophy of Neohumanism (Sarkar, 1999), in which we reconsider the fundamentals — the nature of human beings, the nature of knowing, what we value, and how we are to live — asks us to rethink the purposes of education. Rather than educate so that a tiny sliver of people rise to the top of the global income chain, a Neohumanist education would prepare all people for the art of living well on a fragile and sacred planet. It would emphasize not just academic achievement and high test scores, but shift the focus to fostering compassion, community, empathy, imagination, insight, friendship, creativity, communication, justice, practicality, pleasure, courage, humor, wisdom, introspection, transcendence, ethics, service, and the ability to live well within the carrying capacity of our ecosystems. It would tear down the walls that have separated school and community and invite local and intergenerational knowledge and traditional ways of knowing into conversation with modern empirical science and technological know-how. Importantly, Neohumanism would welcome our inner lives into education and foster multiple epistemologies (embodied knowing, intuitional knowing, narrative knowing, aesthetic knowing, mythic knowing). Adults and young people together would plant gardens and reinvigorate forests, clean up our waterways, and regenerate the soil. We would “rewild” our children and ourselves so that we might begin to understand the vital part we all play in a living web of interconnection, a web that encompasses not just humans, but the eight million other species with whom we share the planet. Only with such an educational process might we “elevate humanism to universalism, the cult of love for all created beings of this universe” (Sarkar, 1999, p. 7).

Scenarios and metaphors	Worldview	Power	Social/ economic organization	Ecological	Knowledge	Education Institutions	Spirituality
Evolution/Revolution Binding quality: Sattva (sentient) Awareness, purity, happiness, sensitivity, expansion and lightness.	Human life an integrated whole encompassing the material and spiritual worlds. Neohumanism: the liberation of the intellect and the expansion of mind. Emphasis on interdependence of all species. Resilient local cultures, universal, inclusive outlooks.	Power/with radical democracy, people organized to resist domination. Co-creation of new systems that serve the whole. Gender partnership, full inclusion. Moral leadership based in service replaces corruption and self-interest. Cooperative global governance regulates international affairs.	Progressive Utilization Theory (PROUT) — Social equality fostered through worker’s cooperatives, caps on wealth accumulation, food sovereignty, the gift and sharing economy, the rights of all people for a decent job, housing, food, health care and education, and the protection of biodiversity and natural habitats. (see Sarkar, 1992).	Deep connection and sense of interrelatedness of all species; humans learn to live in balance with the ecosystem and practice reciprocity. All living beings accorded moral standing and rights.	Integration of modern science/technology and ancient wisdom and indigenous perspectives. Epistemological pluralism. Elimination of dogma.Knowledge balanced between introversial and extroversial.	Schools take on new role as centers of resource, connections, healing, community building, mentorship. Self-organizing learning groups form around real life problems and issues. Eco-versities. Decolonizing pedagogies.	Transformative, new understanding of human potential and the cosmic dimensions of individual life. Pragmatism and contemplative practice exist in mutual harmony (subjective approach/objective adjustment); intuition and rationality complement each other.

Scenario 3 is not a pipe dream. In this present crisis, multitudes of people are acting selflessly to care for others and serve the greater good. Heroic health workers are struggling to mitigate suffering without adequate resources. Teachers are working to reinvent schooling so that children might stay connected to their peers and engaged in learning. Regular folk creating mutual aid societies, ensuring that those who are sick, disabled, or elderly are not forgotten. In many places, small organic farms are beginning to supply much of the local food. Young people are inclined towards egalitarian socio-economic formations, and they are willing to challenge the status quo and struggle for the future of their planet. People the world over are awakening to spiritual wisdom. We are making the road by walking.

The world right now is in a state of chaos – a “far-from-equilibrium” state. Chaos is unpredictable and destabilizing, and small inputs can have huge effects, illustrated by the compelling image of the fluttering wings of the butterfly in the Amazon, causing a cyclone in China.

But chaos theory also teaches us that systems re-organize, often in surprising new ways. A far-from-equilibrium state is a liminal space; liminality is described by one author as “the sacred space where the old world is able to fall apart, and a bigger world is revealed.” (Rohr, 1999). Will we find the courage to allow this dissolution, in order to make way for the world we hope to create? Or will we eagerly seek the status quo, business as usual, or worse, regress into barbarism? I believe that we are in the thick of what may come to be understood as the “great transition” – the death of an old era and the birth of the new. Such a birth is not accomplished painlessly, but with extraordinary labor. Those of us who share the values of Scenario 3, who hold a Neohumanist vision of human potential and a social vision of a just, ecological and joyful Earth home (PROUT) share a responsibility to be midwives to this birth. Systems demand that we evolve and adapt. The butterfly effect reminds us that small actions can have big impacts. Our small collective actions, mindfully taken, could have important collective impacts, so let us proceed with Scenario 3 as consciously and compassionately as we can.

About the Author

Kathleen Kesson is Professor Emerita, LIU-Brooklyn, and is the former Director of the John Dewey Project on Progressive Education at the University of Vermont and Director of Education at Goddard College. She currently lives in Barre, Vermont and is actively engaged in the work to make Vermont schools more equitable, sustainable, and joyful. Her latest book is Unschooling in Paradise. You can read other writings by her as well as an excerpt from this book at https://www.kathleenkesson.com

References:

Moore, J. (2016). Anthropocene or capital scene? Nature, history, and the crisis of capitalism. Oakland, CA: PM Press.

Rohr, R. (1999). Everything belongs: The gift of contemplative prayer— The Crossroad Publishing Company.

Sarkar, P.R. (1992). Proutist Economics: Discourses on economic liberation. Kolkata, India: Ananda Marga Pracaraka Samgha

Sarkar, P.R. (1999). The liberation of intellect: Neo-Humanism. 4th edition. Ananda Nagar; Kolkata: Ananda Marga Publications.

Uncategorized

Impacts of agrisolar co-location on the food–energy–water nexus and economic security

Nature Sustainability volume 8, pages 702–713 (2025)

Abstract

Understanding how solar PV installations affect the landscape and its critical resources is crucial to achieve sustainable net-zero energy production. To enhance this understanding, we investigate the consequences of converting agricultural fields to solar photovoltaic installations, which we refer to as ‘agrisolar’ co-location. We present a food, energy, water and economic impact analysis of agricultural output offset by agrisolar co-location for 925 arrays (2.53 GW_p covering 3,930 ha) spanning the California Central Valley. We find that agrisolar co-location displaces food production but increases economic security and water sustainability for farmers. Given the unprecedented pace of solar PV expansion globally, these results highlight the need for a deeper understanding of the multifaceted outcomes of agricultural and solar PV co-location decisions.

Main

Climate change threatens our finite food, energy and water (FEW) resources. To address these threats by transitioning towards net-zero carbon emissions energy systems, new energy installations should be designed while considering effects on the complete FEW nexus. The rapid expansion of solar photovoltaic (PV) electricity generation is a key part of the solution that will need to grow more than tenfold in the United States (US) by 2050 to meet net-zero goals¹. However, solar PV expansion presents threats to agricultural production due to its land-use intensity and potential in croplands². A considerable portion of ground-mounted solar PV facilities in the US are installed in agricultural settings^3,4,5. Yet regions with high solar breakthrough, such as the California Central Valley (CCV), are often among the most valuable and productive agricultural land in the US^3,5,6. It is not yet clear how the current solar PV landscape affects agricultural security, much less under 2050 net-zero expansion. Here we quantify both the agricultural offsets of solar PV land-use change and the decision-making processes behind these transitions for existing solar PV arrays in agriculture.

Competition between solar PV and agricultural land uses has led to various co-location methods where installations are sited, designed and managed to optimize landscape productivity across a wide range of ecological and anthropogenic services⁷. This approach differs from conventional solar PV deployment, which is often installed and managed primarily for electricity output and reduced maintenance⁷. Emerging concepts such as techno-ecological synergies (TES)⁸ and more recently, ecovoltaics⁷, encompass a wide range of co-location strategies enabling renewable energy installations to serve multiple productive ecosystem services. Agricultural production and solar PV can be laterally integrated (agrisolar co-location)⁹ or directly share land and photons via vertical integration (agrivoltaic co-location)^10,11.

Agrivoltaic co-location involves the direct integration of solar and agriculture (crops or grazing) or ecosystem services (pollinator habitat, native vegetation) within the boundaries of solar infrastructure¹¹. The earliest technical standardization, originating from Germany, specifies that this can occur under or between system rows, but not adjacent to, while agricultural yield losses are reduced to less than one-third of reference (without solar PV) yields¹⁰. Effective agrivoltaic management can improve agricultural yield, microclimate regulation, soil moisture retention, nutrient cycling and farmer profitability, while enhancing public acceptance^12,13,14,15. Thus, agrivoltaic co-location can address the agricultural competition concerns created by solar PV expansion.

The term agrisolar is more broadly defined (modified from SolarPower Europe⁹), as the integration and co-management of solar photovoltaics, agriculture and ecosystem services within agroenergy landscapes, explicitly considering the trade-offs and co-benefits of agricultural, environmental and socio-economic objectives. Thus defined, agrisolar practices align with TES and ecovoltaic principles and encompass both coincident (‘agrivoltaic co-location’) and adjacent co-location where agricultural land is replaced (hereafter ‘agrisolar co-location’)^11,16. However, replacing agricultural land with solar PV (‘adjacent agrisolar’) without implementing agrivoltaic management has historically been considered conventional solar and thus excluded from co-location research because agricultural production is ceased on site¹⁰. There is some evidence, however, that converting portions of agricultural fields to solar PV in water-stressed regions can also provide water and economic benefits that enhance agricultural security despite food production losses^17,18. Adjacent agrisolar replacement appears to be the dominant practice, with recent work showing that there have been relatively few documented agrivoltaic installations compared to total solar PV deployment in agriculture in the CCV^5,19. Because agrisolar practices are understudied relative to literature on other forms of co-location^14,20, there is a need to assess regional resource outcomes for most existing solar PV installations and consequences for lost food production without agrivoltaic management. Conceptual examples of solar PV co-location are shown in Fig. 1.

**Fig. 1: Conceptual diagram of trade-offs and co-benefits with agrisolar, agrivoltaic and ecovoltaic co-location.**

We argue that by enhancing water, energy and economic security, transitioning farm fields to solar PV installations can be considered adjacent agrisolar management in water-stressed regions. Here security is the capacity of a farmer to maintain or improve their financial well-being, operational resilience and access to essential resources, such as water and energy, while preserving the integrity and future of their agricultural practices. We assess the FEW security effects of these agrisolar PV installations across the CCV through 2018 and estimate the economic potential of those arrays throughout a 25-year operational-phase lifespan. We compute landowner cash flow including net energy metering (NEM) for commercial-scale PV installations and land leases for larger utility-scale arrays. All resource and economic effects are referenced to a counterfactual business-as-usual scenario with no solar PV installation, assuming continued agricultural production and operation on the same plot of land. The purpose of this analysis is to evaluate the lifespan FEW and economic impacts of existing agrisolar arrays in the CCV. Rather than projecting future installations or policies, we report on the existing agrisolar placement, design and policy practices to inform future practices on a per-hectare basis, tailored to regional needs. We also highlight the need for, and opportunities within, additional research into agrisolar practices.

Results

Commercial- and utility-scale agrisolar arrays in CCV

We assembled a comprehensive dataset of agriculturally co-located solar PV installations within the CCV through 2018. We identified 925 solar PV arrays installed between 2008 and 2018, with an estimated capacity of 2,524 MW_p on 3,930 ha of recently converted agricultural land. The estimated array capacity of each individual array ranged from 19 kW_p to 97 MW_p. A temporal synthesis of the input solar PV dataset, separated by array scale, is shown in Fig. 2b,c. The smaller commercial-scale arrays are roughly twice as common, yet account for one-tenth of the installed capacity and converted land area of utility-scale arrays. Note that commercial-scale arrays are predominantly fixed axis, whereas utility-scale arrays are more frequently single-axis tracking systems. There are also notable peaks in the number of installations for both array scales in 2016, potentially in response to the NEM 2.0 legislation timeline²¹. While there is some spatial clustering of converted crop types (Fig. 2a), converted crops were widely distributed across the CCV.

**Fig. 2: Study area and characterization of ground-mounted agrisolar PV installations.**

Offset food and nutritional production

The 925 agriculturally co-located arrays displaced 3,930 ha of cropland, which is ~0.10% of the CCV active agricultural land²². In the baseline scenario (Methods provide scenario details), nutritional loss was 0.16 trillion kcal (Tkcal) and 1.41 Tkcal foregone by commercial- and utility-scale arrays, respectively (Fig. 3). The total, 1.57 Tkcal, is equivalent to the caloric intake of ~86,000 people for 25 years (solar lifespan), assuming a 2,000 kcal d^–1 diet. The nutritional footprint of commercial-scale arrays (−21.2 million kcal (Mkcal) ha^–1 yr^–1) was greater than utility-scale arrays (−15.6 Mkcal ha^–1 yr^–1) and the total impact was primarily composed of grain (58%), orchard crops (21%) and vegetables (10%). Utility-scale arrays displaced the nutritional value of grain (60%) hay/pasture (16%) and vegetables (10%). Note that for displaced kcal production of hay/pasture, contribution was negligible despite dominating the converted area due to inefficient caloric conversion to human nutrition for feed and silage crops. Resource footprint, total lifespan impact and crop contribution is shown in Fig. 3. Cumulative resource impacts across the region through time are available in Supplementary Fig. 1.

**Fig. 3: Lifespan land use, food loss, electricity production and potential irrigation electricity offset and potential water conservation with agrisolar co-location in California’s Central Valley.**

Electricity production and consumption

We modelled the annual electricity generation for each array and offset irrigation electricity demand. Total cumulative electricity generation for these identified arrays by 2042 was projected to be 10 TWh for commercial-scale arrays and 113 TWh for utility-scale arrays. The potential electricity saved by not irrigating converted land was 11 GWh and 146 GWh for commercial- and utility-scale arrays, respectively. Note that this was three orders of magnitude less than the total electricity generation. For reference, the total lifespan impact of electricity production and potential irrigation electricity offset ( ~ 124 TWh) could power ~466,000 US households for 25 years (assuming 10.6 MWh yr^–1 per household).

Changes in water use

Most (74%) agriculturally co-located arrays in the CCV replaced irrigated croplands. On the basis of the business-as-usual change in total water-use budget (considering irrigation water-use offset and operation and maintenance—O&M water use), we estimate that agrisolar co-location in the region would reduce water use by 5.46 thousand m³ ha^–1 yr^–1 (total: 42.1 million m³) and 6.02 thousand m³ ha^–1 yr^–1 (total: 544 million m³) over the 25-year period for commercial- and utility-scale arrays, respectively. This could supply ~27 million people with drinking water (assuming 2.4 liters per person per day) or irrigate 3,000 hectares of orchards for 25 years. O&M water use on previously irrigated land was ~eight times less than irrigated crops—if offset irrigation water were conserved rather than redistributed. Irrigated crops that contributed the most to the offset irrigation water use were orchards (29%), hay/pasture (28%) and grain (27%) for commercial-scale installations and grain (37%), hay/pasture 31%), cotton (15%) for utility-scale installations.

Agricultural landowner cash flow

Adjacent agrisolar co-location is more profitable than the baseline agriculture-only scenario, regardless of how landowners are compensated (Fig. 4). For commercial-scale arrays, agrisolar landowners experience early losses from installation expenditure (−US$53,000 ha^–1 yr^–1). However, the lifespan cash flow was dominated by NEM, offset electricity costs and surplus generation sold back to the grid, resulting in a net positive economic footprint of US$124,000 ha^–1 yr^–1, 25 times greater returns than lost food revenue (−US$4,920 ha^–1 yr^–1). The resulting economic payback period was 5.2 years (best- and worst-case payback in 2.9 and 8.9 years respectively; Supplementary Fig. 2).

**Fig. 4: Lifespan economic footprint of commercial- and utility-scale agrisolar co-location.**

The net economic footprint for utility-scale agrisolar landowners (US$2,690 ha^–1 yr^–1) was 46 times less than the commercial-scale footprint (Fig. 4b). In contrast to commercial-scale arrays, utility-scale agrisolar landowners were not responsible for installation or O&M costs but still lost food revenue (−US$3,330 ha^–1 yr^–1) and were only compensated by land lease (US$1,940 ha^–1 yr^–1) and offset operational (US$3,830 ha^–1 yr^–1) and irrigation water-use costs (US$220 ha^–1 yr^–1). In the worst-case scenario, the total budget was negative (−US$432 ha^–1 yr^–1), suggesting that some landowners could lose revenue. There was no payback period for utility-scale agrisolar landowners because the net economic budget was always positive (baseline and best-case scenario) or always negative (worst-case scenario). Cumulative economic impacts across the region in Supplementary Fig. 3.

On average, estimated foregone farm operation costs exceeded forgone food revenue (Fig. 4). While this may be affected by reporting differences in agricultural revenue and farm operation cost sources, agricultural margins are known to be small, or negative, for certain croplands (for example, pastureland), with margins likely to decrease further under future climate change and water availability scenarios²³. For commercial-scale installations, cutting farm operation costs in half (highly conservative) resulted in a longer economic payback period of just a month. Cutting offset farm operation costs in half for utility-scale installations did not affect economic payback or the always-positive baseline and best-case budget.

Discussion

The effect of agrisolar co-location on food production

We found that displacing agricultural land with solar PV locally reduced crop production ( ~ 1.57 Tkcal), which may affect county- and state-level food flows. Fortunately, on national and global scales, food production occurs within a market where reduced production in one location creates price signals that can stimulate production elsewhere. For example, high demand and increased irrigation pumping costs in the CCV have resulted in higher prices received for specialty orchard crops. Thus, farmers have elected to switch from cereal and grain crops to specialty crops²⁴. Solar PV is also far more energy dense per unit of land than growing crops to produce biofuels¹⁸—a practice common across large swaths of agricultural farmland in the US and elsewhere. We show that conversion of feed, silage and biofuel croplands provides high irrigation water-use offsets while minimizing nutritional impacts due to the low or non-existent caloric conversion efficiencies of these crops (Fig. 3). Though, considering food waste and a lack of crop-specific nutritional-quality knowledge, we cannot evaluate end-point impacts of reported foregone kcal (calories) on human diets and health²⁵.

California produces 99% of many of the nation’s specialty fruit and nut orchard crops (for example, almonds, walnuts, peaches, olives)²⁶. Fields producing these crops were commonly converted to solar PV (270 ha of orchard crops), and it may be difficult to shift production of these crops to other locations due to their intensive water footprint, climate sensitivity and time to production^27,28. Altering global supply of these crops could lead to food price increases similar to biofuel land-use changes²⁹ with agricultural markets taking time to compensate³⁰. We found that these nutritionally dense, valuable and operationally costly crops are more commonly replaced by commercial-scale rather than utility-scale installations, resulting in a higher nutritional footprint at the site scale (Fig. 3). However, due to their smaller arrays size (Fig. 2), these arrays have a lower regional lifespan nutritional impact. The total solar PV area we consider (the area covered by panels and space between them) does not account for total cropland transformation by all solar energy infrastructure. Thus, total cropland area converted and associated caloric losses may be underestimated by up to 25%. We conducted a sensitivity analysis on this potential area bias for all area-based metrics and discuss the details of this underestimate in Supplementary Discussion.

Global food needs are projected to double by 2050^31,32. To meet these needs, yield per unit area must increase, agricultural land area under production must increase and/or food waste and inefficiency must be reduced. Reducing waste is feasible but requires a considerable change in dietary preferences³³ and supply chain pathways³⁴. Yield increases alone are unlikely to meet these needs³¹ and half of global habitable land is already agricultural³⁵. Cultivated lands are facing additional pressures due to soil quality deterioration, aridification, water availability, urban growth and threats to global biodiversity that will be exacerbated under a changing climate^36,37,38,39. Given these pressures on arable land, cropland selection for future agrisolar co-location, both commercial- and utility-scale, should be assessed at local, regional, national and international scales to maintain food availability and security.

Water security potential with agrisolar co-location

Here we show that solar PV installations preferentially displace irrigated land in the CCV (3,310 ha and 74% of co-located installations). Displacing this irrigated cropland enhances farmer cash flow while probably reducing overall water use by 5.46 and 6.02 thousand m³ ha^–1 yr^–1 for commercial- and utility-scale arrays, respectively. The total displaced irrigation water use was eight times the O&M use for those arrays. Thus, installing solar PV in water-scarce regions has substantial potential to reduce water use, which bolsters findings from previous studies^17,18,40,41. This analysis does not incorporate the additional hydrologic effects of modifying surface energy and water budgets, including reducing evapotranspiration and the potential for increased groundwater recharge^42,43.

Given that the cash flow benefits from utility-scale agrisolar co-location are relatively small, we evaluated how water-use limitations may be a factor in farmland conversion decisions. We hypothesize that fallowing land is largely a consequence of water shortages in the CCV^24,40, thus fallowing land proximal to an array (within 100 metres) may indicate an emergent agrisolar practice: intentional fallowing and irrigation water-use offset adjacent to arrays supported by revenue from the array. Each array was coded by the adjacent crop type before and post installation of the array. While we cannot know what landowners would have done with the array acreage absent the installation, this analysis provides evidence of broader land-use trends that might have been driving decisions. The transition of array acreage from before proximal post-installation land use for utility-scale arrays is displayed in Fig. 5.

**Fig. 5: Land-use change adjacent to utility-scale solar PV installations on previously irrigated cropland in the CCV.**

Understanding how economic incentives affect the replacement of valuable cropland with solar PV is essential to inform future energy landscape models and policies. Here we examined the transition to post-solar installation fallowing in adjacent irrigated cropland (Fig. 5). We observed fallowing of adjacent irrigated cropland at 58 utility-scale installations totalling 658 MW_p and 968 ha (27% of utility-scale area) composed of 410 ha of grain, 250 ha of hay and pasture, 225 of orchards, grapes and vegetables and 82 ha of cotton and other crops. The direct area of these arrays (968 ha) can be linked to a potential irrigation water-use offset of 195 million m³ over 25 years. If these arrays were on-farm plots of average size, 14,000 ha of fallowed land adjacent to these 58 arrays could displace an additional 120 million m³ of irrigation water use, each year, or 3,000 million m³ over 25 years (Supplementary Methods). Thus, if landowners choose to fallow farmland adjacent to leased land for utility-scale arrays, the water-use reductions are greatly amplified. We discuss several important limitations⁴⁴ of the Cropland Data Layer (CDL) regarding this analysis in Supplementary Discussion.

Intensely irrigated cropland in the CCV is vulnerable to drought, especially in southern basins that rely heavily on surface-water deliveries due to limited groundwater availability⁴⁵. The California Budget Act of 2021 provides financial support for fallowing to motivate farmers to reduce water use⁴⁶. Whereas fallowing land can help mitigate some hydrological problems, removing production can also result in large agricultural revenue losses⁴⁷. Converting land with solar electricity production, rather than simply fallowing could reduce risks to farmers while enhancing financial security¹⁷, especially during periods of extreme drought⁴⁰. Whereas this has implications for future installations, we show that farmers already appear to be practicing solar fallowing, probably resulting in long-term irrigation water-use reductions.

We acknowledge the potential issues in assuming that foregone irrigation water use due to solar PV installations was conserved rather than redistributed. However, a portion of this potential offset is probably real given three observations: (1) utility-scale installations correlate with newly fallowed land, which was not observed for commercial-scale arrays; (2) the 2014 Sustainable Groundwater Management Act (SGMA)⁴⁸ requires water-use reductions by the 2040s and (3) agriculturally co-located solar PV maintains Williamston Act Status under the Solar-Use Easement⁴⁹ (which has recently been revived⁵⁰), a California tax incentive common in irrigated lands highly suitable for solar⁵¹. In our dataset, 46% of utility-scale installations and 58% of commercial-scale installations were completed after SGMA was enacted (Fig. 2b,c). We also performed a sensitivity analysis where only 50% of irrigation water-use offset was conserved rather than redistributed, which still resulted in an estimated US$9 million and 246 million m³ conserved due to the regional change in water use from just direct area converted (Supplementary Discussion).

Given this potential for water-use offset, solar fallowing for water-use reduction presents an opportunity for incentivized solutions that are already of interest to landowning farmers in the region¹⁷. With suitable solar area in the CCV exceeding projected fallowing acreage to comply with SGMA⁵¹, implementing agrisolar co-location policies and incentives such as these could promote complementary land uses and enhance public support¹⁵.

Achieving economic security across return structures

Regardless of scale and related financial benefits, farmers are switching away from cultivating crops to cultivating electricity. This study empirically demonstrates that both NEM and land-lease incentive structures have been viable frameworks for PV deployment in some of the most valuable cropland in the US⁶. Critically, we incorporate farm-specific agricultural dynamics across a region (offset farm operation costs, irrigation costs and food revenue) into economic considerations for replacing cropland with solar.

By including these revenues and costs, this study clearly demonstrates the strong economic incentives to replace cropland with commercial-scale arrays (Fig. 4a). Under the grandfathered NEM 1.0 and 2.0 agreements, commercial-scale agrisolar landowners enhanced financial security by 25 times lost food revenue over the lifetime of the array, while simultaneously reducing water use. The resulting total net revenue, US$124,000 ha^–1 yr^–1, is potentially underestimated because post-lifespan module replacement, resale or continued use is likely, and property values could increase (terminal value) compared to the reference scenario. We also have not considered several programmes, credits and incentives (for example, Rural Energy for America Program) that could enhance net revenue (Supplementary Discussion). However, these returns are not unlimited due to NEM capacity limitations (<1 MW_p) and requirements to size the installation below annual on-farm load²¹.

Renewable energy policy evolves quickly, shifting incentives for new customer generators. Whereas climate change and decreasing water availability in the coming decades²³ will probably increase financial motivation to install solar in agriculture, future adoption and the co-benefits reported here will also depend on new business models for grid pricing⁵². Pricing structures have already and will inevitably continue to change as utilities, regulators and grid customers adapt to distributed renewable generation, avoid curtailment and avoid the utility death spiral⁵². Although future installations and policy are not the focus of this study, the newest policy, NEM 3.0, substantially reduces compensation for surplus generation and limits options for multiple metered connections⁵³, probably requiring future installations to add battery storage and other measures to maintain similar profitability⁵⁴. However, this study considers solar arrays that are grandfathered into their respective NEM 1.0 and 2.0 agreements. Additionally, our estimated load contributions suggest that revenue reported here mostly originates from offset demand rather than credit for surplus generation (Supplementary Notes and Supplementary Discussion). The bottom line is that owning solar PV, offsetting annual on-farm electric load and selling surplus electricity back to the utility under NEM 1.0 and 2.0 has increased economic and energy security for farmers with existing arrays and has probably promoted water-use reductions in the region. Importantly, we also assumed that all decisions were made by and returns received by landowning or partial-owning farmers. We do not have access to land-ownership data for the CCV, but nearly 40% of agricultural land in the region is rented or leased⁵⁵.

Utility-scale land-lease rates alone do not offset lost agricultural revenue. However, including offset farm operation costs results in a substantially lower but still profitable agrisolar economic footprint with no major up-front capital investment (Fig. 4b). In water-scarce regions, particularly where water-use reduction is required, the smaller returns from utility-scale agrisolar practices and potentially related fallowing of land may be more attractive than continued cultivation under water-supply uncertainty¹⁷. Thus, without profitable compensation, agrivoltaic practices may not be feasible if offset operational costs and water-use reductions are driving utility-scale agrisolar decision making. We also omit some agricultural dynamics (such as the environmental benefits of carbon reduction), which could reinforce resource and economic security for both commercial- and utility-scale installation (Supplementary Discussion).

Opportunities for agrisolar research

Whereas funding and incentives for co-location research have expanded rapidly in recent years, we advocate extending these to agrisolar co-location. Adjacent agrisolar replacement with barren or unused ground cover still falls short of the full potential of ecovoltaic and agrivoltaic multifunctionality^7,9,10,11. However, the regional resource and economic co-benefits of replacing irrigated land in water-stressed regions with solar PV here cannot be ignored. These findings are also immediately relevant to the Protecting Future Farmland Act of 2023⁵⁶, which set out a goal to better understand the multifaceted impacts of installed solar on US agricultural land. We discuss additional placement and management decisions that fall under the umbrella of agrisolar co-location in Supplementary Discussion.

We have shown that the goal of co-location, to enhance synergies between the co-production of agriculture and/or other ecosystem services and net-zero electricity production, is at least partially achievable with agrisolar co-location. Broader agrisolar research may also expose the consequences of not widely adopting agrivoltaics to retain agricultural production and protect food security. Given the ecosystem service benefits reported here, there may be an opportunity to broaden the scope of co-location research and incentives to include agrisolar co-location practices defined here.

Methods

Identifying agrisolar PV arrays across the CCV

We used remotely sensed imagery of existing solar PV arrays and geographic information system (GIS) datasets to develop a comprehensive and publicly available dataset of ground-mounted arrays co-located with agriculture in the CCV through 2018. We extracted all existing non-residential arrays from two geodatabases (Kruitwagen et al.^4,57 and Stid et al.^5,58) within the bounds of the CCV alluvial boundary⁵⁹. We removed duplicate arrays and applied temporal segmentation methods described in Stid et al.⁵ to assign an installation year for Kruitwagen et al.⁴ arrays. We acquired Kruitwagen et al.⁴ panel area within array bounds by National Agriculture Imagery Program imagery pixel area with solar PV spectral index ranges suggested in Stid et al.⁵ and removed commissions (reported array shapes with no panels). We then removed arrays with >70% overlap with building footprints⁶⁰ to retain only ground-mounted installations. Finally, overlaying historical CDL crop maps with new array shapes, we removed arrays in areas with majority non-agricultural land cover the year before installation (Supplementary Fig. 4 and Supplementary Discussion).

The resulting dataset (925 agrisolar co-located arrays) included 686 ground-mounted arrays from Stid et al.⁵ plus 239 from Kruitwagen et al.⁴. For these sites, we calculated array peak capacity (kW_p) by⁶¹:

(1)

where is the total direct area of PV panels in m², is the average efficiency of installed PV modules during the array installation year⁶² (Supplementary Fig. 5) and is the irradiance at standard test conditions (kW m^–2). Arrays were split into ‘Commercial-’ (<1 MW_p) and ‘Utility-’ (≥1 MW_p) scale arrays following the California Public Utility Commission NEM capacity guidelines⁶³.

Scenario summary and assumptions

We computed annual FEW resource and economic values for each ground-mounted agrisolar PV array identified across the CCV for four scenarios: (1) reference, business as usual with no solar PV installation and continued agricultural production on the same plot of land, (2) baseline, agrisolar PV installation with moderate assumptions related to each component of the analysis, (3) worst case, PV installation with high negative and low positive effects for each component, (4) best case, similar but opposite of the worst-case scenario. We compare baseline to the reference scenario to estimate the most likely FEW and economic effects and use the differences between best- and worst-case scenarios to estimate uncertainty. Supplementary Tables 2 and 3 provide an overview of scenarios for each resource and Supplementary Tables 4 and 5 for baseline agrisolar lifespan FEW resource and economic value outcomes, respectively.

Identified arrays were installed between 2008 and 2018 and were assumed to have a 25-year lifespan for arrays due to performance, warranties, module degradation and standards for electrical equipment^64,65. We assumed that land-use change effects ceased following 25 years of operation to simplify assumptions about module replacement, resale or continued use. We then summarized the FEW and economic effects of all arrays across the CCV and divided our temporal analysis into three phases: (1) addition (2008–2018) where arrays were arrays were being installed across the CCV, (2) constant (2019–2032) with no array additions but all arrays installed by 2018 are operating and maintained and (3) removal (2032–2042), where arrays are removed after 25 years of operation.

We performed several sensitivity analyses to address limitations in the available data and methods and to show how changes in future policy (NEM) could affect incentives displayed here. Sensitivity analysis included the capacity cut-off between commercial- and utility-scale (5 MW), solar PV lifespan (15 and 50 years), nominal discount rate (3%, 7% and 10%), solar PV direct area bias (proportional direct to total infrastructure area and a uniform perimeter buffer) and irrigation redistribution (assuming 50% of irrigation water-use offset is redistributed rather than conserved), all else equal (Supplementary Discussion and Supplementary Tables 6–20). We discuss additional assumptions and limitations in Supplementary Discussion.

Displaced crop and food production

Replacing fields (or portions thereof) with solar PV arrays affects crop production by (1) lost production of food, fibre and fuels and (2) reduced revenue from crop sales. We simplify the complex effects of lost production and include solely the foregone calories through both direct and indirect human consumption, which is justified because CCV crop production is largely oriented towards food crops. Future analyses could evaluate the lost fibre (primarily via cotton) or fuel (via biofuel refining) production.

We evaluated the economic and food production effects of displaced crops through a crop-specific opportunity cost assessment of land-use change, incorporating actual reported; yields, revenue, caloric density and regionally constrained caloric conversion efficiencies for feed/silage and seed oil crops. All crop type information was derived from the USDA National Agricultural Statistics Service (NASS) CDL²² for the array area in both prior- and post-installation years (Supplementary Fig. 4 and Supplementary Methods provide the adjacent fallowed land analysis). Each array was assigned a majority previous crop from the spatially weighted means of crop types within the array area for the five years before the installation.

We converted all eligible crop types to kcal (also called calorie) for human consumption after Heller et al.²⁵. Foregone food production ( in kcal) following PV installation was then defined for each array as:

(2)

where is in kcal kg^–1, is in kg m^–2 and of each array in m². Crop-specific caloric density data (kcal kg^–1) were derived from the USDA FoodData Central April 2022 release⁶⁶. FoodData food descriptions and nutrient data were joined and CDL specific crop groupings were made through a workflow described in Supplementary Fig. 6. Crop-specific yield data (kg m^–2) were derived from the USDA NASS Agricultural Yield Surveys⁶⁷. State-level (California) yield data were processed similarly, with missing crop data filled based on national average yields. We used caloric conversion efficiencies for feed, silage or oil crop to account for crop production that humans do not directly consume.

For each array, we calculated annual revenue of forgone crop production in real (inflation adjusted) dollars, calculated by:

(3)

where is in US$ kg^–1, is in kg m^–2 and of each array in m². We used the annual ‘price received’ for all crops in the USDA NASS Monthly Agricultural Prices Report for 2008 through 2018⁶⁸. For the baseline case, we assumed that food prices will scale directly with electricity prices through 2042 given that they respond to similar inflationary forces⁶⁹. Supplementary Fig. 6 and Supplementary Methods provide a more complete workflow including best- and worst-case scenario assumptions.

Change in irrigation water use and cost savings

Irrigation water use can only be offset by agrisolar co-location if the prior land use was irrigated. The presence of irrigation was inferred from the Landsat-based Irrigation Dataset (LanID) map for the year before installation^70,71 (Supplementary Fig. 4). If the array area contained irrigated pixels, then we assumed the cropland area and all respective crops within the rotation were irrigated.

We calculated the total forgone irrigation water use ( in m³) by:

(4)

where in m is the crop-specific irrigation depth, in m³ is the annual county-level irrigation water-use estimate and in m³ is the county-level irrigation water-use estimate for the respective survey year irrigation depths.

We estimated annual crop-specific county-level irrigated depths from survey and climate datasets for each array. Crop-specific irrigation depths () were taken from the 2013 USDA Farm and Ranch Survey⁷² and 2018 Irrigation and Water Management Survey⁷³, and logical crop groupings were applied (for example, almonds, pistachios, pecans, oranges and peaches were considered orchard crops). Because irrigation depths depend on the total precipitation in each survey year, we used multilinear regression to build annual county-level irrigation water-use estimates () from five-year US Geological Survey (USGS) water use⁷⁴, gridMET growing season average precipitation⁷⁵, with year as a dummy variable to incorporate temporal changes in irrigation technologies and practices. For the installation phase (2008 to 2018), these depths varied based on historical climate and survey data, whereas the projection phases (constant and removal) used a scenario-dependent moderate, wet (worst-case, least water savings) or dry (best case, most water savings) year estimate from the historical record (discussed in Supplementary Methods).

Assigning an economic value to water use is difficult and varies based on the temporally changing supply and demand⁷⁶. We calculated the economic value of the change in water use (Water in real US$) to the farmer by:

(5)

where (m³) is the offset irrigation water use for the co-located crop minus O&M projected water use, (MWh m^–3) is the irrigation electricity required to irrigate the co-located crop given local depth to water and drawdown estimates from McCarthy et al.⁷⁷, (US$ MWh^–1) is the utility-specific (commercial-scale) or regional average (utility-scale) annual price of electricity based on the electricity returns and modelled electricity generation described in Supplementary Methods and is a CCV-wide average water right contract rate of ~ US$0.03 m^–3 (ref. ⁷⁸). Here we assume that water (and thus energy) otherwise used for irrigation was truly foregone and not redistributed elsewhere within or outside the farm. Change in O&M water use was based on Klise et al.⁷⁹ reported values, described in Supplementary Methods.

Electricity production, offset and revenue

Installing solar PV in fields has three benefits: (1) production of electricity by the newly installed solar PV array, (2) reduction in energy demand due to reduced water use and field activities and (3) revenue generation via net energy metering (NEM) or land lease. Here we assume that on-farm electricity demand is dominated by electricity used for irrigation and ignore offset energy (embodied) used for fuel.

We modelled electricity generation for each array using the pvlib python module developed by SANDIA National Laboratory⁸⁰. Weather file inputs for pvlib were downloaded from the National Renewable Energy Laboratory (NREL) National Solar Radiation Database⁸¹. We also estimated annual on-farm load to differentiate offset electricity use and surplus generation. Not only is electricity generated by the arrays, but electricity consumption is foregone for each array due to not irrigating the array area. The annual change in electricity consumption due to water use ( in GWh) is calculated by:

(6)

where is the county-level rates for irrigation electricity demand in GWh m^–3 and is the change in water use in m³ from equation (5). County-level electricity requirements to irrigate were calculated using irrigation electricity demand methods described in McCarthy et al.⁷⁷ modified with a CCV-specific depth to water (piezometric surface) product for the spring (pre-growing season) of 2018⁸².

Revenue from electricity generation was calculated separately for each array depending on array size and the installation year. Commercial-scale arrays (<1 MW) were assumed to operate under an NEM 1.0 if installed before 2017 and NEM 2.0 if installed later, which allows for interconnection to offset on-farm load and compensation for surplus electricity generation (Supplementary Methods and Supplementary Table 21). Thus, for commercial-scale arrays, annual cash flow from solar PV (NEM in US$) is calculated as:

(7)

where is real US$ saved by offsetting annual on-farm electric load and is real US$ earned by surplus PV electricity generation sold to the utility under NEM guidelines. Both and are estimated based on pvlib modelled electricity generation and valued at the historical utility-specific energy charge retail rates. Historical energy charges are available either through utility reports^83,84,85 or the US Utility Rate Database via OpenEI⁸⁶. We made several assumptions that resulted in omission of fixed charges including transmission and interconnection costs from the analysis. Details about electricity rates and omitted charges are summarized in Supplementary Methods.

For utility-scale arrays (≥1 MW), annual revenue from agrisolar co-location (Lease in US$) was assumed to be given by:

(8)

where Lease is the economic value estimated to be paid to the farmer by the utility for leasing their land in US$ m^–2 and Area of each array in m².

We assumed commercial-scale arrays installed before 2017 were grandfathered into NEM 1.0 guidelines for the duration of their lifespan. However, arrays installed in 2017 and 2018 fall under NEM 2.0 guidelines which include a US$0.03 kWh^–1 non-bypassable charge removed from ^21,87,88. Annual on-farm operational load was estimated and distributed across the year based on reported California agricultural contingency profiles⁸⁹ and Census of Agriculture county-level average farm sizes^90,91,92 (Supplementary Figs. 7 and 8 and Supplementary Methods). With distributed hourly load estimations and modelled solar PV electricity generation, we delineated electricity and revenue contributing to annual load () from surplus electricity and revenue that would have been sold back to the grid and credited via NEM ().

Future electricity revenue was projected using 2018 conditions (contribution to annual load, to surplus) and energy charge rates, modelled electricity production described above (includes degradation, pre-inverter, inverter efficiency and soiling losses) and projected changes in the price of electricity. The Annual Energy Outlook report by the US Energy Information Administration (EIA) provides real electricity price projections annually between 2018 and 2050 for ‘Commercial End-Use Price’⁹³. This annual rate of change was used to estimate projected deviations from 2018 energy charges (2018 US$ kWh^–1) during the constant and removal phases (2019–2042), with projected solar PV generation including discussed losses.

We used solar land consultant and industry reports for solar land-lease () rates that ranged from US$750 ha^–1 yr^–1 to US$4,950 ha^–1 yr^–1, with high-value land averaging IS$2,450 ha^–1 yr^–1 in the CCV^94,95. Comparable lease rates (~US$2,500 to US$5,000 ha^–1 yr^–1) were reported by developers in the CCV region¹⁷ and used in a solar PV and biomass trade-off study in Germany¹⁸ (~US$1,000 to US$2,950 ha^–1 yr^–1).

Array installation and O&M costs

Historical installation costs (Installation) were taken from the commercial-scale PV installation prices reported in the Annual Tracking the Sun report where reported prices are those paid by the PV system owner before incentives⁶². The baseline scenario is the median installation price, whereas the best- and worst-case scenarios are the 20th and 80th percentile installation costs, respectively. These reported values are calculated using NREL’s bottom-up cost model and are national averages using average values across all states. Installation cost was not discounted, as it represents the initial investment for commercial-scale installations at day zero. All future cash flows, profits and costs are compared to this initial investment. We also included the 30% Solar Investment Tax Credit in the Installation for commercial-scale arrays⁹⁶. The system bounds of this impact analysis were installation through the operational or product-use phase. We, therefore, did not assume removal expenses or altered property value (terminal value) to remove uncertainty in decision making at the end of the 25-year array lifespan.

Historically reported and modelled O&M values (pre-2020) range from US$0 kW_p^–1 yr^–1 (best case) to US$40 kW_p^–1 yr^–1 (worst case) with an average (baseline) of US$18 kW_p^–1 yr^–1 (refs. ^97,98). Projected O&M costs were based on modelled commercial-scale PV lifetime O&M cost to capital expenditure cost ratios from historical and industry data that provided scenarios varying on research and development differences (conservative, moderate, advanced). The annual reported values are provided from 2020 to 2050 for fixed O&M costs including: asset management, insurance products, site security, cleaning, vegetation removal and component failure and are detailed in the Annual Technology Baseline report by NREL⁹⁷, which are largely derived from the annual NREL Solar PV Cost Benchmark reports.

Farm operation costs

Business-as-usual farm operation costs (Operation) were derived from the ‘Total Operating Costs Per Acre to Produce’ reported in UC Davis Agricultural and Resource Economics Cost and Return Studies⁹⁹. We removed operational costs to ‘Irrigate’ from the total because we estimate that as a function of electricity requirements and water rights (described in ‘Change in irrigation water use and cost savings’) while retaining ‘Irrigation Labour’ as this was not included in our irrigation cost assessment. Best- and worst-case scenarios for farm operation costs coincided with yield scenarios described in ‘Displaced crop and food production’.

Discounted cash flow for agrisolar co-location

For each commercial-scale array in the CCV, we computed the annual real cash flow as:

(9)

and for each utility-scale array as:

(10)

where Commercial is the real return in 2018 US$ for commercial-arrays (<1 MW_p) and Utility is the real return in 2018 US$ for utility-scale arrays (≥1 MW_p). Each of the terms on the right-hand side of these equations are defined in the sections above.

We then computed real annual discounted cash flow () for each array to estimate the total lifetime value of each array. The at any given year n is calculated for each array by:

(11)

where is the real annual cash flow at year n (either Commercial or Utility as relevant for each array) and is the real discount rate without an expected rate of inflation (i) from the nominal discount rate () calculated using the Fisher equation¹⁰⁰:

(12)

Vartiainen et al.¹⁰¹ clearly communicates this method in solar PV economic studies and discusses the importance of discount rate (in their case, weighted average cost of capital) selection. For i, we use 3%, which is roughly the average producer price index (PPI) and consumer price index (CPI) (3.4% and 2.4%, respectively) between 2000 and 2022 and comparable to other solar PV economic studies^101,102. We use a 5% ¹⁰³ and perform a sensitivity analysis using 3%, 7% and 10% and discuss discount rates used in literature in Supplementary Discussion. Separately from the sensitivity analysis for , we also calculated our best-case and worst-case scenarios for each array.

All prices were adjusted to 2018 US dollars for calculation of real cash flow terms in equations (11) and (9). We adjusted consumer electricity prices and installation costs for inflation to real 2018 US$ using the US Bureau of Labor Statistics Consumer Price Index for All Urban Customers¹⁰⁴. We adjusted all production-based profits and costs (all other resources) using US Bureau of Labor Statistics Producer Price Index for All Commodities¹⁰⁵.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The datasets and outputs generated in the current study are publicly available via Zenodo at https://doi.org/10.5281/zenodo.10023293 (ref. ¹⁰⁶) with all source data referenced in the published article and in its Supplementary Information files.

Code availability

The code used to generate and analyse the datasets reported here are hosted via GitHub at https://github.com/stidjaco/FEWLS_tool and are available via Zenodo at https://doi.org/10.5281/zenodo.10023281 (ref. ¹⁰⁷).

References

Ardani, K. et al. Solar Futures Study (US DOE, 2021); https://www.energy.gov/sites/default/files/2021-09/Solar%20Futures%20Study.pdf
Adeh, E. H., Good, S. P., Calaf, M. & Higgins, C. W. Solar PV power potential is greatest over croplands. Sci Rep. 9, 11442 (2019).Article Google Scholar
Hernandez, R. R., Hoffacker, M. K., Murphy-Mariscal, M. L., Wu, G. C. & Allen, M. F. Solar energy development impacts on land cover change and protected areas. Proc. Natl Acad. Sci. USA 112, 13579–13584 (2015).Article CAS Google Scholar
Kruitwagen, L. et al. A global inventory of photovoltaic solar energy generating units. Nature 598, 604–611 (2021).Article CAS Google Scholar
Stid, J. T. et al. Solar array placement, electricity generation, and cropland displacement across California’s Central Valley. Sci. Total Environ. 835, 155240 (2022).Article CAS Google Scholar
USDA Land Values 2022 Summary (NASS, 2022).
Sturchio, M. A. & Knapp, A. K. Ecovoltaic principles for a more sustainable, ecologically informed solar energy future. Nat. Ecol. Evol. 7, 1746–1749 (2023).Article Google Scholar
Hernandez, R. R. et al. Techno–ecological synergies of solar energy for global sustainability. Nat. Sustainability 2, 560–568 (2019).Article Google Scholar
Agrisolar Best Practice Guidelines Version 2 (SolarPower Europe, 2023).
Agri–Photovoltaic Systems–Requirements for Primary Agricultural Use (Deutsches Institut für Normung, 2021).
Macknick, J. et al. The 5 Cs of Agrivoltaic Success Factors in the United States: Lessons From the InSPIRE Research Study (NREL, 2022).
Barron-Gafford, G. A. et al. Agrivoltaics provide mutual benefits across the food–energy–water nexus in drylands. Nat. Sustainability 2, 848–855 (2019).Article Google Scholar
Choi, C. S. et al. Environmental co‐benefits of maintaining native vegetation with solar photovoltaic infrastructure. Earth’s Future 11, e2023EF003542 (2023).Article Google Scholar
Gomez-Casanovas, N. et al. Knowns, uncertainties, and challenges in agrivoltaics to sustainably intensify energy and food production. Cell Rep. Phys. Sci. 4, 101518 (2023).Article Google Scholar
Pascaris, A. S., Schelly, C., Rouleau, M. & Pearce, J. M. Do agrivoltaics improve public support for solar? A survey on perceptions, preferences, and priorities. Green Technol. Resilience Sustainability 2, 8 (2022).
McCall, J., Macdonald, J., Burton, R. & Macknick, J. Vegetation management cost and maintenance implications of different ground covers at utility-scale solar sites. Sustainability 15, 5895 (2023).Article Google Scholar
Biggs, N. B. et al. Landowner decisions regarding utility-scale solar energy on working lands: a qualitative case study in California. Environ. Res. Commun. 4, 055010 (2022).Article Google Scholar
Bao, K., Thrän, D. & Schröter, B. Land resource allocation between biomass and ground-mounted PV under consideration of the food–water–energy nexus framework at regional scale. Renewable Energy 203, 323–333 (2023).Article Google Scholar
Fujita, K. S. et al. Georectified polygon database of ground-mounted large-scale solar photovoltaic sites in the United States. Sci. Data 10, 760 (2023).Article Google Scholar
Knapp, A. K. & Sturchio, M. A. Ecovoltaics in an increasingly water-limited world: an ecological perspective. One Earth 7, 1705–1712 (2024).Article Google Scholar
Picker, M., Florio, M. P., Sandoval, C. J. K., Peterman, C. J. & Randolph, L. M. Decision Adopting Successor to Net Energy Metering Tariff (California Public Utilities Commission, 2016).
USDA National Agricultural Statistics Service Cropland Data Layer (USDA, 2023); https://nassgeodata.gmu.edu/CropScape/
Medellín-Azuara, J., Howitt, R. E., MacEwan, D. J. & Lund, J. R. Economic impacts of climate-related changes to California agriculture. Climatic Change 109, 387–405 (2011).Article Google Scholar
Gebremichael, M., Krishnamurthy, P. K., Ghebremichael, L. T. & Alam, S. What drives crop land use change during multi-year droughts in California’s Central Valley? Prices or concern for water? Remote Sens. 13, 650 (2021).Article Google Scholar
Heller, M. C., Keoleian, G. A. & Willett, W. C. Toward a life cycle-based, diet-level framework for food environmental impact and nutritional quality assessment: a critical review. Environ. Sci. Technol. 47, 12632–12647 (2013).Article CAS Google Scholar
Ross, K. & Honig, M. California State Fact Sheet (USDA Farm Service Agency, 2011).
Lobell, D. B., Field, C. B., Cahill, K. N. & Bonfils, C. Impacts of future climate change on California perennial crop yields: model projections with climate and crop uncertainties. Agric. For. Meteorol. 141, 208–218 (2006).Article Google Scholar
Alam, S., Gebremichael, M. & Li, R. Remote sensing-based assessment of the crop, energy and water nexus in the Central Valley, California. Remote Sens. 11, 1701 (2019).Article Google Scholar
Wise, M., Dooley, J., Luckow, P., Calvin, K. & Kyle, P. Agriculture, land use, energy and carbon emission impacts of global biofuel mandates to mid-century. Appl. Energy 114, 763–773 (2014).Article CAS Google Scholar
Gilbert, C. L. How to understand high food prices. J. Agric. Econ. 61, 398–425 (2010).Article Google Scholar
Ray, D. K., Mueller, N. D., West, P. C. & Foley, J. A. Yield trends are insufficient to double global crop production by 2050. PLoS ONE 8, e66428 (2013).Article CAS Google Scholar
Tilman, D., Balzer, C., Hill, J. & Befort, B. L. Global food demand and the sustainable intensification of agriculture. Proc. Natl Acad. Sci. USA 108, 20260–20264 (2011).Article CAS Google Scholar
Godfray, H. C. J., Poore, J. & Ritchie, H. Opportunities to produce food from substantially less land. BMC Biol. 22, 138 (2024).Article Google Scholar
Kummu, M. et al. Lost food, wasted resources: global food supply chain losses and their impacts on freshwater, cropland, and fertiliser use. Sci. Total Environ. 438, 477–489 (2012).Article CAS Google Scholar
Ritchie, H. & Roser, M. Land Use (Our World in Data, 2013); http://ourworldindata.org/land-use
Molotoks, A. et al. Global projections of future cropland expansion to 2050 and direct impacts on biodiversity and carbon storage. Glob. Change Biol. 24, 5895–5908 (2018).Article Google Scholar
Prăvălie, R. et al. Arable lands under the pressure of multiple land degradation processes. A global perspective. Environ. Res. 194, 110697 (2021).Article Google Scholar
Elliott, J. et al. Constraints and potentials of future irrigation water availability on agricultural production under climate change. Proc. Natl Acad. Sci. USA 111, 3239–3244 (2014).Article CAS Google Scholar
Flörke, M., Schneider, C. & McDonald, R. I. Water competition between cities and agriculture driven by climate change and urban growth. Nat. Sustainability 1, 51–58 (2018).Article Google Scholar
He, X. et al. Solar and wind energy enhances drought resilience and groundwater sustainability. Nat. Commun. 10, 4893 (2019).Article Google Scholar
Shirkey, G. et al. An environmental and societal analysis of the US electrical energy industry based on the water–energy Nexus. Energies 14, 2633 (2021).Article CAS Google Scholar
Sturchio, M. A., Kannenberg, S. A., Pinkowitz, T. A. & Knapp, A. K. Solar arrays create novel environments that uniquely alter plant responses. Plants People Planet 6, 1522–1533 (2024).Article Google Scholar
Yavari Bajehbaj, R., Cibin, R., Duncan, J. M. & McPhillips, L. E. Quantifying soil moisture and evapotranspiration heterogeneity within a solar farm: implications for stormwater management. J. Hydrol. 638, 131474 (2024).Article Google Scholar
Lark, T. J., Mueller, R. M., Johnson, D. M. & Gibbs, H. K. Measuring land-use and land-cover change using the US. Department of Agriculture’s cropland data layer: cautions and recommendations. Int. J. Appl. Earth Obs. Geoinf. 62, 224–235 (2017).Google Scholar
Medellín-Azuara, J. et al. Hydro-economic analysis of groundwater pumping for irrigated agriculture in California’s Central Valley, USA. Hydrol. J. 23, 1205–1216 (2015).Google Scholar
Skinner, N. Budget Act of 2021 SB 170 (California Assembly, 2021).
Medellín-Azuara, J. et al. Economic Impacts of the 2021 Drought on California Agriculture Preliminary Report Prepared for The California Department of Food and Agriculture (UC Merced, 2022); http://drought.ucmerced.edu
California Water Code § 10729 (State of California, 2015).
Wolk, L. Local Government: Solar-Use Easement SB-618 (State of California, 2011).
Committee on Governance and Finance. Local Government Omnibus Act of 2022 SB-1489 (State of California, 2022).
Ayres, A. et al. Solar Energy and Groundwater in the San Joaquin Valley (Public Policy Institute of California, 2022); http://www.ppic.org/?show-pdf=true&docraptor=true&url=https%3A%2F%2Fwww.ppic.org%2Fpublication%2Fsolar-energy-and-groundwater-in-the-san-joaquin-valley%2F
Laws, N. D., Epps, B. P., Peterson, S. O., Laser, M. S. & Wanjiru, G. K. On the utility death spiral and the impact of utility rate structures on the adoption of residential solar photovoltaics and energy storage. Appl. Energy 185, 627–641 (2017).Article Google Scholar
Cooke, M. Decision Addressing Remaining Proceeding Issues (California Public Utilities Commission, 2023).
Barbose, G. One Year In: Tracking the Impacts of NEM 3.0 on California’s Residential Solar Market (Lawrence Berkeley National Laboratory, 2024); https://escholarship.org/uc/item/4st8v7j0
Bigelow, D. US Farmland Ownership, Tenure, and Transfer (USDA Economic Research Service, 2016).
Baldwin, T. & Grassley, C. Protecting Future Farmland Act of 2023 S.2931 (US Senate, 2023).
Kruitwagen, L. et al. A global inventory of solar photovoltaic generating units—dataset. Zenodo https://doi.org/10.5281/zenodo.5005868 (2021).
Stid, J. T. et al. Spatiotemporally characterized ground-mounted solar PV arrays within California’s Central Valley. Figshare https://doi.org/10.6084/m9.figshare.23629326.v1 (2023).
Faunt, C. C. Alluvial boundary of California’s Central Valley. US Geological Survey https://doi.org/10.5066/P9CQNCA9 (2012).
Heris, M. P., Foks, N., Bagstad, K. & Troy, A. A National Dataset of Rasterized Building Footprints for the U.S. (USGS, 2020); https://doi.org/10.5066/P9J2Y1WG
Martín-Chivelet, N. Photovoltaic potential and land-use estimation methodology. Energy https://doi.org/10.1016/j.energy.2015.10.108 (2016).Article Google Scholar
Barbose, G., Darghouth, N., O’shaughnessy, E. & Forrester, S. Tracking the Sun Pricing and Design Trends for Distributed Photovoltaic Systems in the United States (Lawrence Berkeley National Laboratory, 2022); http://emp.lbl.gov/publications/tracking-sun-pricing-and-design-1
Perea, H. Electricity: Natural Gas: Rates: Net Energy Metering: California Renewables Portfolio Standard Program AB-327 (State of California, 2013).
Federal Energy Management Program 10 CFR (US DOE, 2017).
Best Practices for Operation and Maintenance of Photovoltaic and Energy Storage Systems 3rd edn (NREL, 2018); http://www.nrel.gov/docs/fy19osti/73822.pdf
USDA. FoodData Central (Agriculture Research Service, 2019).
Crop Production (USDA, 2022).
Agricultural Prices (USDA, 2019); http://usda.library.cornell.edu/concern/publications/c821gj76b?locale=en
Ringler, C., Bhaduri, A. & Lawford, R. The nexus across water, energy, land and food (WELF): potential for improved resource use efficiency? Curr. Opin. Environ. Sustainability 5, 617–624 (2013).Article Google Scholar
Xie, Y., Gibbs, H. K. & Lark, T. J. Landsat-based irrigation fataset (LANID): 30 m resolution maps of irrigation distribution, frequency, and change for the US, 1997-2017. Earth Syst. Sci. Data 13, 5689–5710 (2021).Article Google Scholar
Xie, Y. & Lark, T. J. LANID-US: Landsat-based irrigation dataset for the United States. Zenodo https://doi.org/10.5281/zenodo.5548555 (2021).
Farm and Ranch Irrigation Survey (2013). (USDA, 2013); http://agcensus.library.cornell.edu/wp-content/uploads/2012-Farm-and-Ranch-Irrigation-Survey-fris13.pdf
Irrigation and Water Management Survey. USDA NASS 2018 Irrigation and Water Management Survey (2017 Census of Agriculture) (USDA, 2018); http://www.nass.usda.gov/Publications/AgCensus/2017/Online_Resources/Farm_and_Ranch_Irrigation_Survey/fris.pdf
USGS Water Use Data for California (USGS, 2015).
Abatzoglou, J. T. Development of gridded surface meteorological data for ecological applications and modelling. Int. J. Climatol. 33, 121–131 (2013).Article Google Scholar
Medellín-Azuara, J., Harou, J. J. & Howitt, R. E. Estimating economic value of agricultural water under changing conditions and the effects of spatial aggregation. Sci. Total Environ. 408, 5639–5648 (2010).Article Google Scholar
McCarthy, B. M. Energy Trends in Irrigation: A Method for Estimating Local and Large-Scale Energy Use in Agriculture (Michigan State Univ., 2021).
Baldocchi, D. D. The cost of irrigation water and urban farming. Berkeley News (2018).
Klise, G. T. et al. Water Use and Supply Concerns for Utility-Scale Solar Projects in the Southwestern United States (Sandia National Laboratories, 2013); http://www.osti.gov/servlets/purl/1090206
Holmgren, W. F., Hansen, C. W. & Mikofski, M. A. pvlib python: a python package for modeling solar energy systems. J. Open Source Software 3, 884 (2018).Article Google Scholar
Sengupta, M. et al. The National Solar Radiation Data Base (NSRDB). Renewable Sustainable Energy Rev. 89, 51–60 (2018).Article Google Scholar
California Department of Water Resources. i08 GroundwaterDepthSeasonal contours. California Natural Resources Agency Open Data Platform (2022); https://data.ca.gov/dataset/i08-groundwaterdepthseasonal-contours
Pacific Gas and Electric. Electric rates: current and historic electric rates. (PG&E, accessed 6 July 2023); https://www.pge.com/tariffs/en/rate-information/electric-rates.html
Sacramento Municipal Utility District. CEO & GM report on rates and services. (SMUD, accessed 6 July 2023); https://www.smud.org/Corporate/About-us/Company-Information/Reports-and-Statements/GM-Reports-on-Rates-and-Services
Southern California Edison. Historical Prices and Rate Schedules. (SCE, accessed 6 July 2023); https://www.sce.com/regulatory/tariff-books/historical-rates
Zimny-Schmitt, D. & Huggins, J. Utility Rate Database (URDB). OpenEI https://data.openei.org/submissions/5 (2020).
Ratemaking, Solar Value and Solar Net Energy Metering—A Primer (SEPA, 2015); https://www.energy.gov/sites/prod/files/2015/03/f20/sepa-nem-report-0713-print.pdf
Gong, A., Brown, C. & Adeyemo, S. The Financial Impact of California’s Net Energy Metering 2.0 Policy (Aurora Solar, 2017); https://www.ourenergypolicy.org/wp-content/uploads/2017/07/Aurora_NEM2_Whitepaper_v1.01__1_.pdf
Olsen, D., Sohn, M., Piette, M. A. & Kiliccote, S. Demand Response Availability Profiles for California in the Year 2020 (Lawrence Berkeley National Laboratory, 2014); http://www.osti.gov/servlets/purl/1341727/
2007 Census of Agriculture (USDA NASS, 2009); https://agcensus.library.cornell.edu/wp-content/uploads/2007-United_States-State-usv1.pdf
2012 Census of Agriculture (USDA NASS, 2014); https://agcensus.library.cornell.edu/wp-content/uploads/usv1.pdf
2017 Census of Agriculture (USDA NASS, 2019); http://www.nass.usda.gov/Publications/AgCensus/2017/Full_Report/Volume_1,_Chapter_1_US/usv1.pdf
Annual Energy Outlook 2020 with Projections to 2050 (EIA. 2020).
Lease Rates for Solar Farms: How Valuable Is My Land? SolarLandLease https://www.solarlandlease.com/lease-rates-for-solar-farms-how-valuable-is-my-land (2020).
Van Trump, K. What You Need to Know… Big Money Leasing Farmland to Solar Operators. The Van Trump Report https://www.vantrumpreport.com/what-you-need-to-know-big-money-leasing-farmland-to-solar-operators/ (2020).
Energy Policy Act of 2005 (US Congress, 2005).
2022 Annual Technology Baseline (NREL, 2022).
Ramasamy, V., Feldman, D., Desai, J. & Margolis, R. U.S. Solar Photovoltaic System and Energy Storage Cost Benchmarks: Q1 2021 (NREL, 2021); http://www.nrel.gov/docs/fy22osti/80694.pdf
Current Cost and Return Studies: Commodities (UC Davis, 2022).
Fisher, I. Appreciation and Interest (AEA Publication, 1896).
Vartiainen, E., Masson, G., Breyer, C., Moser, D. & Román Medina, E. Impact of weighted average cost of capital, capital expenditure, and other parameters on future utility‐scale PV levelized cost of electricity. Prog. Photovoltaics 28, 439–453 (2020).Article Google Scholar
Liu, X., O’Rear, E. G., Tyner, W. E. & Pekny, J. F. Purchasing vs. leasing: a benefit–cost analysis of residential solar PV panel use in California. Renewable Energy 66, 770–774 (2014).Article Google Scholar
Kelley, L. C., Gilbertson, E., Sheikh, A., Eppinger, S. D. & Dubowsky, S. On the feasibility of solar-powered irrigation. Renewable Sustainable Energy Rev. 14, 2669–2682 (2010).Article Google Scholar
Consumer Price Index (CPI) Databases (US BLS, 2023).
Producer Price Index (PPI) Databases (US BLS, 2023).
Stid, J. T. Agrisolar food, energy, and water and economic lifecycle scenario (FEWLS) tool data. Zenodo https://doi.org/10.5281/zenodo.10023293 (2025).
Stid, J. T. FEWLS tool: initial release of FEWLS tool. Zenodo https://doi.org/10.5281/zenodo.10023281 (2023).
Uber Technologies Inc. H3: hexagonal hierarchical spatial indexing. GitHub https://github.com/uber/h3 (2019).
Cartographic Boundary File (US Census Bureau, 2019); http://Census.gov
Ong, S., Campbell, C., Denholm, P., Margolis, R. & Heath, G. Land-Use Requirements for Solar Power Plants in the United States NREL/TP-6A20-56290, 1086349 (OSTI, 2013); https://doi.org/10.2172/1086349

Download references

Acknowledgements

This work was supported by the USDA National Institute of Food and Agriculture (NIFA) INFEWS grant number 2018-67003-27406. We credit additional support from the USDA NIFA Agriculture and Food Research Initiative Competitive grant number 2021-68012-35923 and the Department of Earth and Environmental Sciences at Michigan State University. Any opinions, findings and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the USDA or Michigan State University. We are grateful to B. McGill for bringing the vision of agrisolar co-location to life through her artistic conceptual depiction.

Author information

Authors and Affiliations

Department of Earth and Environmental Sciences, Michigan State University, East Lansing, MI, USAJacob T. Stid, Anthony D. Kendall & Jeremy Rapp
Department of Civil and Environmental Engineering, Michigan State University, East Lansing, MI, USASiddharth Shukla & Annick Anctil
Department of Sustainable Earth System Sciences, School of Natural Sciences and Mathematics, The University of Texas at Dallas, Richardson, TX, USADavid W. Hyndman
Biological Systems Engineering, University of Wisconsin-Madison, Madison, WI, USARobert P. Anex

Uncategorized

Humanity’s Endgame

A new history of societal collapse by an expert in existential risk argues that our globalized society is edging toward the precipice.

Nicolas Ortega for Noema Magazine

By Henry Wismayer

Henry Wismayer is a writer based in London.

LONDON — There are 8 million artifacts in the British Museum. But to commence his tale of existential jeopardy, risk expert Luke Kemp made a beeline for just two items housed in a single room. On a visit in early fall, beyond a series of first-floor galleries displaying sarcophagi from pharaonic Egypt, we stopped beside a scatter of human bones.

The exhibit comprised two of the 64 skeletons unearthed from the sands of Jebel Sahaba, in northern Sudan, in 1964. Believed to be over 13,000 years old, the bodies in this prehistoric cemetery were significant for what they revealed about how their owners died. Of those 64 skeletons, at least 38 showed signs of violent deaths: caved-in skulls, forearm bones with parry fractures from victims staving off blows, or other injuries. Whether a result of organized warfare, intercommunal conflict or even outright massacre, Jebel Sahaba is widely considered to be some of the earliest evidence of mass violence in the archaeological record.

According to Kemp, these shattered bones were a foreshadowing of another object in this room. Ten feet away, displayed at knee-height, was the Palette of Narmer. Hewn from a tapering tablet of grey-green siltstone, the item on display was an exact cast of the 5,000-year-old original — discovered by British archaeologists in 1898 — that now sits in Cairo’s Egyptian Museum.

At the center of the stone stands the giant figure of Narmer, the first king of Egypt. His left hand clasps the head of an enemy, presumed to be a rival ruler of the Western Delta. In his raised right hand he holds a mace. The image is thought to depict Narmer bludgeoning his greatest opponent to death, an act that solidified his sovereignty over all Egypt. Beneath his feet lie the contorted bodies of two other victims, while overhead a falcon presents Narmer with a ribbon, believed to represent the god Horus bestowing a gift of the Western Nile. “Here we have perfect historical evidence of what the social contract is. It’s written in blood,” Kemp told me. “This is the first depiction of how states are made.”

In the British Museum’s repository of ancient treasures and colonial loot, the palette is by no means a star attraction. For the half hour we spent in the room, few visitors gave it more than a passing glance. But to Kemp, its imagery “is the most important artwork in the world” — a blueprint for every city-state, nation and empire that has ever been carved out by force of arms, reified in stone and subsequently turned to dust.

Systematizing Collapse

When Kemp set out seven years ago to write his book about how societies rise and fall — and why he fears that our own is headed for disaster — one biblical event provided him with the perfect allegory: the story of the Battle of the Valley of Elah, recounted in 1 Samuel 17. Fought between the Israelites and the Philistines in the 11th century BCE, it’s a tale more commonly known by the names of its protagonists, David and Goliath.

Goliath, we are told, was a Philistine warrior standing “six cubits and a span,” or around 9 feet, 9 inches, clad in the alloy of copper and tin armor that would give his epoch its name: the Bronze Age. As the rival armies faced off across the valley, the giant stepped onto the battlefield and laid down a challenge that the conflict should be resolved in single combat.

For 40 days, Goliath goaded his enemy to nominate a champion, until a shepherd named David came forward from the Israelite ranks, strung a stone into his slingshot and catapulted it into Goliath’s brow, killing him at a stroke, and taking his head with the giant’s own sword. For centuries thereafter, the story of David and Goliath has served as a parable challenging the superiority of physical might. Even the most impressive entity has hidden frailties. A colossus can be felled by a single blow.

According to Kemp’s new book, “Goliath’s Curse,” it’s a lesson we would do well to heed. Early on, he dispenses with the word “civilization,” because in his telling, there is little that might be considered civil about how states are born and sustained. Instead, he argues that “Goliath” is a more apposite metaphor for the kind of exploitative, hierarchical systems that have grown to organize human society.

“‘Goliath’ is a more apposite metaphor for the kind of exploitative, hierarchical systems that have grown to organize human society.”

Like the Philistine warrior, the Goliath state is defined by its size; in time, centralized polities would evolve to dwarf the hunter-gatherer societies that prevailed for the first 300,000 years of Homo sapiens. Ostensibly, it is well-armored and intimidating, exerting power through the threat and exercise of violence. And, in kind with the biblical colossus, it is vulnerable: Those characteristics that most project strength, like autocracy and social complexity, conceal hidden weaknesses. (A more modern allegory, Kemp writes, can be found in the early Star Wars movies, in which a moon-sized space station with the capacity to blow up a planet can be destroyed by a well-placed photon torpedo.)

Kemp is, of course, by no means the first scholar to try to chart this violence and vulnerability through the ages. The question of what causes societies to fail is arguably the ultimate mission of big-picture history, and a perennial cultural fixation. In the modern era, the historian Jared Diamond has found fame with his theories that collapse is usually a product of geographical determinism. The “Fall of Civilizations” podcast, hosted by the historian Paul Cooper, has over 220 million listens. Perusing a bookshop recently, I spotted a recent release, entitled “A Brief History of the End of the F*cking World,” among the bestsellers.

What distinguishes Kemp’s book from much of the canon is the consistencies he identifies in how different political entities evolved, and the circumstances that precipitated their fall. A panoramic synthesis of archaeology, psychology and evolutionary biology, “Goliath’s Curse” is, above all, an attempt to systematize collapse. Reviewers have hailed the book as a skeleton key to understanding societal precarity. Cooper has described it as “a masterpiece of data-driven collapsology.”

Moreover, it is a sobering insight into why our own globalized society feels like it is edging toward the precipice. That’s because, despite all the features that distinguish modern society from empires of the past, some rules hold true throughout the millennia.

Becoming ‘Dr. Doom’

In September, Kemp traveled down from Cambridge to meet me in London for the day. Given his subject, I half-expected a superannuated and eccentric individual, someone like Diamond with his trademark pilgrim-father beard and penchant for European chamber music. But Kemp, 35, would prove to be the antithesis of the anguished catastrophist. The man waiting for me on the concourse at King’s Cross was athletic, swarthily handsome and lantern-jawed. He’d signed off emails regarding our plans to meet with a puckish “Cheerio.”

Kemp’s background is also hardly stereotypical of the bookish scholar. He spent his early years in the dairy-farming town of Bega in New South Wales, Australia, where cattle outnumbered people three-to-one. It was “something of a broken home,” he told me. His father was an active member of the Hell’s Angels, involved in organized crime, a formative presence that would later germinate Kemp’s interest in power dynamics, the way violence is at once a lever for domination and for ruin.

Escaping to Canberra, after high school, Kemp read “interdisciplinary studies” at the Australian National University (ANU), where he found a mentor in the statistical climatologist Jeanette Lindsay. In 2009, it was Lindsay who persuaded him to join a student delegation heading to COP15 in Copenhagen, where Kemp found himself with a front row seat to what he calls “the paralysis of geopolitics.”

At one stage, during a symposium over measures to curb deforestation, he watched his own Australian delegation engage in endless circumlocutions to derail the debate. Representatives from wealthier countries, most notably America, had large teams that they could swap in and out of the floor, enabling them to filibuster vital, potentially existential questions to a deadlock. “If you’re from Tuvalu, you don’t have that privilege,” Kemp explained.

Afterward, Kemp became preoccupied by “a startling red thread” evident in so many spheres of international negotiation: the role of America as arbiter of, and all too often barrier to, multilateral cooperation. Kemp wrote his doctoral thesis on how pivotal issues — such as biodiversity loss, nuclear weapons and climate change — had grown captive to the whims of the world’s great superpower. Later, when he published a couple of academic articles on the same subject, “the ideas weren’t very popular,” he said. “Then Trump got elected, and suddenly the views skyrocketed.”

In 2018, Kemp relocated to the United Kingdom, landing a job as a research affiliate at Cambridge University’s “Centre for the Study of Existential Risk” (CSER, often articulated, in an inadvertent nod to a historical avatar of unalloyed power, to “Caesar”). His brother’s congratulatory present, a 3-D printed, hand-engraved mask of the Marvel character “Dr. Doom,” would prove prophetic. Years later, as Kemp began to publish his theories of societal collapse, colleagues at CSER began referring to him by the very same moniker.

“Goliath hierarchies select for assholes — or, to use Kemp’s preferred epithet, ‘dark triad’ personalities: people with high levels of psychopathy, narcissism and Machiavellianism.”

It was around this time that Kemp read “Against the Grain,” a revisionist history of nascent conurbations by James C. Scott. Kemp had always been an avid reader of history, but Scott’s thesis, which argued that the growth of centralized states “hadn’t been particularly emancipatory or even necessarily good for human wellbeing,” turned some of Kemp’s earlier assumptions about human nature on their head.

Such iconoclastic ideas — subsequently popularized in blockbuster works of non-fiction like Rutger Bregman’s “Humankind” (2019), and “The Dawn of Everything” (2021) by Graeber and Wengrove — would prompt years of research and rumination about the preconditions that enable states and empires to rise, and why they never last forever.

‘Hobbes’ Delusion’

“Goliath’s Curse” opens with a refutation of a 17th-century figure whose theories still cast a long shadow across all considerations of societal fragility. In “Leviathan” (1651), the English philosopher Thomas Hobbes proposed that the social contract was contingent on the stewardship of a central authority — a “Leviathan” designed to keep a lid on humanity’s basest instincts. Political scientists refer to this doctrine as “veneer theory.”

“Once civilization is peeled away, chaos spreads like brushfire,” Kemp surmises. “Whether it be in post-apocalyptic fiction, disaster movies or popular history books, collapse is often portrayed as a Hobbesian nightmare.”

For decades now, the predominant version of history has been beholden to this misanthropic worldview. Many of the most influential recent theories of collapse have echoed Hobbes’ grand theory with specific exemplars. Diamond has famously argued that the society on Rapa Nui, or Easter Island, unraveled due to self-inflicted ecocide before devolving into civil war. That interpretation, in which the islanders deforested the land in the service of ancestor worship, has since been held up as a species-wide admonition — evidence, as researchers John Flenley and Paul Bahn have written, that “humankind’s covetousness is boundless. Its selfishness appears to be genetically inborn.” In “The Better Angels of Our Nature” (2011), Steven Pinker estimated that 15% of Paleolithic people died of violent causes.

But Kemp was struck by a persistent “lack of empirics” undermining these hypotheses, an academic tendency to focus on a handful of “cherry-picked” and emotive case studies — often on islands, in isolated communities or atypical environments that failed to provide useful analogs for the modern world. Diamond’s theories about the demise of Rapa Nui — so often presented as a salutary cautionary tale —have since been debunked.

To further rebut such ideas, Kemp highlights a 2013 study by the anthropologists Jonathan Haas and Matthew Piscitelli of Chicago’s Field Museum. In what amounted to the most comprehensive survey of violence in prehistory, the authors analyzed almost 3,000 skeletons interred during the Paleolithic Era. Of the more than 400 sites in the survey, they identified just one instance of mass conflict: the bones of Jebel Sahaba. “The presumed universality of warfare in human history and ancestry may be satisfying to popular sentiment; however, such universality lacks empirical support,” Haas and Piscitelli wrote.

If there was any truth to the Hobbesian standpoint, the Paleolithic, with its absence of stratified social structures, should have been marked by mass panic and all-out war. Yet the hunter-gatherer period appears to have been a time of relative, if fragile, peace. Instead, conflict and mass violence seemed to be by-products of the very hierarchical organization that Hobbes and his antecedents essentialized. Cave art of armies wielding bows and swords dates only to around 10,000 years ago. “As soon as you start tugging on the threat of collapse, the entire tapestry of history unravels,” Kemp told me.

But if Hobbes was wrong about the human condition — if most people are averse to violence, if mass panic and mutual animosity are not the principal vectors of societal disintegration — what then explains the successive state failures in the historical record? Where or what, to mix metaphors, is Goliath’s Achilles’ heel?

What Fuels Goliath?

In seeking to disentangle a template of collapse from this historiography, Kemp turned to historical data, searching for traits of state emergence and disintegration shared by different polities. “When I see a pattern which needs to be explained, it becomes a fascination bordering upon obsession,” he told me.

A central pillar of his research was the Seshat Global History Databank, an open-source database incorporating more than 862 polities dating back to the early Neolithic. Named after the Egyptian goddess of wisdom, Seshat includes a range of metrics like the degree of centralization and the presence of different types of weaponry; it aggregates these to create nine “complexity characteristics” (CCs), including polity size, hierarchy, governmental framework and infrastructure.

“Wherever Goliath took hold, ‘arms races’ followed, as other status-seeking aspirants jostled for hegemony. And Goliaths were contagious.”

Using this and other sources, Kemp set out to collate his own novel dataset, this time focusing on the common features not of complexity, but of collapse. In keeping with Seshat’s old-god nomenclature, he dubbed it the “Mortality of States” index, shortened to “Moros”, after the Greek god of doom. Covering 300 states spanning the last five millennia, the resulting catalogue is, Kemp claims, “the most exhaustive list of state lifespans available today.”

To some extent, Kemp’s data told a story that has become received wisdom: As Earth thawed out from the last ice age, we entered the Holocene, a period of warmer temperatures and climatic stability. This shift laid the terrain for the first big inflection point: the advent of agriculture, which encouraged our previously itinerant species to settle in place, leading to greater population density and eventually proto-city-states. These early states rose and fell, often condemned by internal conflict, climatic shocks, disease or natural disasters. But gradually the organization of human societies trended toward higher levels of complexity, from the diffuse proto-city-states, through the birth of nations, then empires, to the globalized system of today. The violent paroxysms of the past were merely hiccups on a continuum toward increased sophistication and civility, and perhaps someday immortality. Such is the tale that is commonly framed as the arc of human progress.

But trawling through the data in more detail also revealed unexpected and recurrent patterns, leading Kemp to an early realization: states observably age. “For the first 200 years, they seem to become more vulnerable to terminating. And after 200 years, they stay at a high risk thereafter,” Kemp told me.

The other glaring commonality concerned the structure of these societies. “The common thread across all of them is not necessarily that they had writing or long-distance trade,” Kemp said. “Instead, it’s that they were organized into dominance hierarchies in which one person or one group gains hegemony through its ability to inflict violence on others.”

Kemp argues that dominance hierarchies arise due to the presence of three “Goliath fuels.” The first of these is “lootable resources,” assets that can be easily seen, stolen and stored. In this respect, the advent of agriculture was indisputably foundational. Cereal grains like wheat and rice could be taxed and stockpiled, giving rise to centralized authorities and, later, bureaucracies of the state.

The second Goliath fuel is “monopolizable weapons.” As weaponry evolved from flint to bronze, the expertise and relative scarcity of the source material required for early metallurgy meant that later weapons could be hoarded by powerful individuals or groups, giving those who controlled the supply chain a martial advantage over potential rivals.

The third criterion for Goliath evolution is “caged land,” territories with few exit options. Centralized power is predicated on barriers that hinder people from fleeing oppressive hierarchies.

In Kemp’s telling, every single political entity has grown from one of these seeds, or more commonly, a combination of all three. Bronze Age fiefdoms expanded at the tip of their metal weaponry. “Rome,” Kemp writes, “was an autocratic machine for turning grain into swords,” its vast armies sustained by crop imports from the Nile Valley, its endless military campaigns funded by the silver mines it controlled in Spain. In China, the Han dynasty circumscribed its territory with its Great Wall to the north, intended both to keep Xiongnu horseback raiders out and the citizenry in. Europe’s colonial empires were built, in Diamond’s famous summation, by “Guns, Germs and Steel.”

For millennia, the nature of forager societies kept these acquisitive impulses to some extent contained, Kemp argues. The evolutionary logic of hunting and gathering demanded cooperation and reciprocity, giving rise to “counter-dominance strategies”: teasing, shaming or exile. With the advent of Goliath polities, however, the “darker angels of our nature” were given free rein, yielding social arrangements “more like the dominance hierarchies of gorillas and chimpanzees.”

“Rather than a stepladder of progress,” Kemp writes, “this movement from civilization to Goliath is better described as evolutionary backsliding.” Moreover, Goliaths “contain the seeds of their own demise: they are cursed. This is why they have collapsed repeatedly throughout history.”

In Kemp’s narrative, our retrograde rush toward these vicious social structures has been less about consensus than the relentless ascent of the wrong sort of people. Goliath hierarchies select for assholes — or, to use Kemp’s preferred epithet, “dark triad” personalities: people with high levels of psychopathy, narcissism and Machiavellianism. Consequently, history has been shaped by pathological figures in the Narmer mold, dominance-seekers predisposed to aggression. Reinforced by exceptionalist and paranoid ideologies, these strongmen have used violence and patronage to secure their dominion, whether driven by a lust for power or to avenge a humiliation. Several of the rebellions that plagued dynastic China, Kemp points out, were spearheaded by aggrieved people who failed their civil service examinations.

“Whether societies collapsed through gradual depopulation, like Çatalhöyük, or abruptly, as with Teotihuacan’s conflagration, Kemp argues that the triggers were the same.”

Wherever Goliath took hold, “arms races” followed, as other status-seeking aspirants jostled for hegemony. And Goliaths were contagious. The growth of “one bellicose city-state” would often produce a domino effect, in which the threat of an ascendant Goliath would provoke other regional polities to turn to their own in-house authoritarian as a counterweight to the authoritarian next door.

In this way, humankind gravitated “from hunting and gathering to being hunted and gathered,” Kemp writes. Early states had little to distinguish them from “criminal gangs running protection rackets.” Many of the great men of history, who are often said to have bent society to their will, Kemp told me, are better thought of as “a rollcall of serial killers.”

The 1% View Of History

Back downstairs, on the British Museum’s ground floor, we walked into a long gallery off the central atrium containing dozens of megalithic totems from the great ages of antiquity. The giant granite bust of Rameses II sat beatific on a pediment, and visitors peered into a glass cabinet containing the Rosetta Stone. Kemp, slaloming through the crowds, murmured: “The 1% view of history made manifest.”

Along both walls of an adjacent corridor, we came upon a series of bas-reliefs from the neo-Assyrian city of Nimrud, in modern-day Iraq. Depicting scenes from the life of the Ashurnasirpal II, who ruled Nimrud in the 9th century BCE, the gypsum slabs were like an artistic expression of Kemp’s historical themes: Ashurnasirpal sitting on a throne before vassals bearing tribute; Ashurnasirpal surrounded by protective spirits; Ashurnasirpal’s army ramming the walls of an enemy city, rivals dragging themselves along the ground, backs perforated with arrows. The entire carving was overlaid with cuneiform script, transcribed onto signage below, with sporadic sentences translated into English: “great king, strong king, king of the universe. … Whose command disintegrates mountains and seas.”

Across the atrium, in a low-lit room containing a bequest from the Rothschild family’s antique collection, Kemp lingered over an assortment of small wooden altarpieces, with biblical scenes and iconography carved in minuscule, intricate detail. Elite status could be projected in the imposing size of a granite statue, he said. But it could just as well be archived in the countless hours spent chiseling the Last Supper into a fragment of boxwood.

It is, of course, inevitable that our sense of history is skewed by this elite bias, Kemp explained. While quotidian objects and utensils were typically made of perishable materials, the palaces and monuments of the governing class were designed to be beautiful, awe-inspiring and durable. In the hours that we spent on the upper floors, we spied just one relic of ordinary life: a 3,000-year-old wooden yoke from Cambridgeshire.

Likewise, early writing often evolved to reinforce the “1% view of history” and formalize modes of control. The predominance of this elite narrative has produced a cultural blind spot, obscuring the brutality and oppression that has forever been the lot of those living at the base of a pyramid, both figurative and actual.

From all this aristocratic residue, Kemp sought to extract a “people’s history of collapse” — some means of inferring what it was like to live through collapse for the average person, rather than the elites immortalized in scripture and stone.

The Curse Of Inequality

If Kemp’s research revealed that historical state formation appears to follow a pattern, so, too, did the forces that inexorably led toward their demise. To illustrate how the process works, Kemp provides the example of Çatalhöyük, a proto-city that arose on the Konya Plain in south-central Turkey around 9,000 years ago, one of thousands of “tells,” mounded remnants of aborted settlements found throughout the Near East.

Excavations of the site’s oldest layers suggest that early Çatalhöyük was notable for its lack of social differentiation. Crammed together in a dense fractal of similarly sized mud-brick dwellings, the settlement in this period exhibits no remnants of fortification and no signs of warfare. Analysis of male and female skeletons has shown that both sexes ate the same diet and performed the same work, indicating a remarkable degree of gender equity.

This social arrangement, which the Stanford archaeologist Ian Hodder has described as “aggressively egalitarian,” lasted for around 1,000 years. Then, in the middle of the 7th millennium BCE, the archaeological record starts to shift. House sizes begin to diverge; evidence of communal activity declines. Later skeletal remains show more evidence of osteoarthritis, possibly betraying higher levels of workload and bodily stress. Economists have estimated that the Gini coefficient, which measures disparities in household income, doubled in the space of three centuries — “a larger jump than moving from being as equal as the Netherlands to as lopsided as Brazil,” Kemp writes. Within a few centuries, the settlement was abandoned.

“In almost every case, [societal] decline or collapse was foreshadowed by increases in the appearance of proxies of inequality.”

The fate of Çatalhöyük established a template that almost every subsequent town, city-state and empire would mirror. Its trajectory resounds throughout the historical record and across continents. Similar patterns can be discerned from the remnants of the Jenne-Jeno in Mali, the Olmecs of Mesoamerica, the Tiwanaku in Titicaca, and the Cahokia in pre-Columbian North America.

Occasionally, the archaeological record suggests a fluctuation between equality and disparity and back again. In Teotihuacan, near today’s Mexico City, the erection of the Feathered Serpent Pyramid by an emergent priestly class in around 200 CE ushered in a period of ritual bloodletting. A more egalitarian chapter followed, during which the temple was razed, and the city’s wealth was rechanneled into urban renewal. Then the old oligarchy reasserted itself, and the entire settlement, beset by elite conflict or popular rebellion, was engulfed in flames.

Whether societies collapsed through gradual depopulation, like Çatalhöyük, or abruptly, as with Teotihuacan’s conflagration, Kemp argues that the triggers were the same. As Acemoğlu and Robinson explored in “Why Nations Fail” (2012), the correlation between inequality and state failure often rests on whether its institutions are inclusive, involving democratic decision-making and redistribution, or extractive: “designed to extract incomes and wealth from one subset of society to benefit a different subset.” Time and again, the historical record shows the same pattern repeating — of status competition and resource extraction spiraling until a tipping-point, often in the shape of a rebellion, or an external shock, like a major climate shift or natural disaster, which the elites, their decision-making fatally undermined by the imperative to maintain their grip on power, fail to navigate.

In almost every case, decline or collapse was foreshadowed by increases in the appearance of proxies of inequality. A rise in the presence of large communal pots indicates an upsurge in feasting. Deviation in the size of dwellings, preserved in the excavated footprints of early conurbations, is a measure of social stratification, as wealth accumulates among the elite. Graves of that same nobility become stuffed with burial goods. Great monuments, honoring political and religious leaders or the gods who were supposed to have anointed them, proliferate. Many of the most lucrative lootable resources throughout history have been materials that connote elevated social standing, an obsession with conspicuous consumption or “wastefully using resources,” that marked a break from the hunter-gatherer principle of taking only what was needed. (Kemp wears a reminder of the human compulsion to covet beauty as much as utility, an obsidian arrowhead, on his wrist.)

All the while, these signs of burgeoning inequality have tended to be twinborn with an increasing concentration of power, and its corollary: violence. War, often instigated for no more reason than the pursuit of glory and prestige, was just “the continuation of status competition by other means,” Kemp writes. On occasion, this violence would be manifested in the ultimate waste of all: human sacrifice, a practice custom-made to demonstrate the leadership’s exceptionality — above ordinary morality.

Better Off Stateless

As Kemp dug into the data in more detail, his research substantiated another startling paradox. Societal collapse, though invariably catastrophic for elites, has often proved to be a boon for the population at large.

Here again, Kemp found that the historiography is subject to pervasive and fallacious simplifications. In his book, he repudiates the 14th-century Tuscan scholar Petrarch, who promulgated the notion that the fall of classical Rome and Greece ushered in a “dark age” of cultural atrophy and barbarism. His was a reiteration of sentiments found in many earlier examples of “lamentation literature,” left behind on engraved tablets and sheaves of papyrus, which have depicted collapse as a Gomorran hellscape. One of Kemp’s favorites is the “Admonitions of Ipuwer,” which portrays the decline of Egypt’s Old Kingdom as a time of social breakdown, civil war and cannibalism. “But it actually spends a lot more time fretting about poor people becoming richer,” he said.

In reality, Kemp contends, Petrarch’s “rise-and-fall vision of history is spectacularly wrong.” For if collapse often engulfed ancient polities “like a brushfire,” the scorched earth left behind was often surprisingly fertile. Again, osteoarcheology, the study of ancient bones, gives the lie to the idea that moments of societal disintegration always spelled misery for the population at large.

Take human height, which archaeologists often turn to as a biophysical indicator of general health. “We can look at things like did they have cavities in their teeth, did they have bone lesions,” Kemp explained. “Skeletal remains are a good indicator of how much exercise people were getting, how good their diet was, whether there was lots of disease.”

“Societal collapse, though invariably catastrophic for elites, has often proved to be a boon for the population at large.”

Prior to the rise of Rome, for example, average heights in regions that would subsequently fall under its yoke were increasing. As the empire expanded, those gains stalled. By the end of the Western Empire, people were eight centimeters shorter than they would have been if the preceding trends had continued. “The old trope of the muscle-bound Germanic barbarian is somewhat true. To an Italian soldier, they would have seemed very large,” Kemp said. People in the Mediterranean only started to get taller again following Rome’s decline. (In a striking parenthesis, Kemp points out that the average male height today remains two centimeters shorter than that of our Paleolithic forebears.)

Elsewhere, too, collapse was not necessarily synonymous with popular immiseration. The demise of the extravagant Mycenaean civilization in Greece was pursued by a cultural efflorescence, paving the way for the proto-democracy of Athens. Collapse could be emancipatory, freeing the populace from instruments of state control such as taxes and forced labor. Even the Black Death, which killed as much as half of Europe’s population in the mid-14th century, became in time an economic leveler, slashing inequality and accelerating the decline of feudalism.

It’s a pattern that can still be discerned in modern contexts. In Somalia, the decade following the fall of the Barre regime in 1991 would see almost every single indicator of quality of life improve. “Maternal mortality drops by 30%, mortality by 24%, extreme poverty by 20%,” Kemp recounted from memory. Of course, there are endless caveats. But often, “people are better off stateless.”

Invariably, however, Goliaths re-emerged, stronger and more bureaucratically sophisticated than before. Colonial empires refined systems of extraction and dominance until their tentacles covered diffuse expanses of the globe. Kemp, never shy of metaphor, calls this the “rimless wheel,” a centripetal arrangement in which the core reaps benefits at the margins’ expense.

At times, such regimes were simply continuations of existing models of extraction. In 1521, when the Spanish conquistador Hernán Cortés unseated the Aztec ruler Moctezuma II, it was merely a case of “translatio imperi” — the handing over of empire. The European imperial projects in the Americas were an unforgivable stain, Kemp said. But, more often than not, they assumed the mantle from pre-existing hierarchies.

Endgame

In the afternoon, we walked north from the British Museum over to Coal Drops Yard, formerly a Victorian entrepôt for the import and distribution of coal, now a shiny vignette of urban regeneration. The morning rain had cleared, and Granary Square was full of tourists and office workers enjoying the late summer sun. Kids stripped to their underwear and played among low fountains; people chatted at public tables beneath a matrix of linden trees. Kemp and I found an empty table and sat down to talk about how it could all fall apart.

As “Goliath’s Curse” approaches its conclusion, the book betrays a sense of impending doom about our current moment. The final section, in which Kemp applies his schema to the present day, is entitled “Endgame,” after the stage in chess where only a few moves remain.

Today, we live in what Kemp calls the “Global Goliath,” a single interconnected polity. Its lootable resources are data, fossil fuels and the synthetic fertilizers derived from petrochemicals. Centuries of arms races have yielded an arsenal of monopolizable weapons like autonomous drones and thermonuclear warheads that are “50 trillion times more powerful than a bow and arrow.” The land — sectored into national borders, monitored by a “stalker complex” of mass surveillance systems and “digital trawl-nets” — is more caged than ever.

We have reached the apotheosis of the colonial age, a time when extractive institutions and administrative reach have been so perfected that they now span the globe. However, the resulting interdependencies and fetishes for unending growth have created an ever-growing catalog of “latent risks,” or accumulated hazards yet to be realized, and “tail risks,” or outcomes with a low probability but disastrous consequences. Kemp characterizes this predicament, in which the zenith of human achievement is also our moment of peak vulnerability, as a “rungless ladder.” The higher we go, the greater the fall.

“We have reached the apotheosis of the colonial age, a time when extractive institutions and administrative reach have been so perfected that they now span the globe.”

Under a series of apocalyptic subtitles — “Mors ex Machina,” “Evolutionary Suicide,” “A Hellish Earth” — Kemp enumerates the existential threats that have come to shape the widespread intuition, now playing out in our geopolitics, that globalized society is sprinting toward disaster. After the post-Cold War decades of non-proliferation, nuclear weapons stockpiles are now growing. The architects of artificial intelligence muse about its potential to wipe out humanity while simultaneously lobbying governments to obstruct regulation. Our densifying cities have become prospective breeding grounds for doomsday diseases. Anthropogenic climate change now threatens to shatter the stability of the Holocene, warming the planet at “an order of magnitude (tenfold) faster than the heating that triggered the world’s greatest mass extinction event, the Great Permian Dying, which wiped away 80–90% of life on earth 252 million years ago,” Kemp warns.

The culprits in this unfolding tragedy are not to be found among the ranks of common people. The free market has always been predicated on the concept of Homo economicus, a notional figure governed by dispassionate self-interest. But while most people don’t embody this paradigm, we are in thrall to political structures and corporations created in that image, with Dark Triad personalities at the wheel. “The best place to find a psychopath is in prison,” Kemp told me. “The second is in the boardroom.”

Now, deep into the Global Goliath’s senescence, several of the indicators that Kemp identifies as having historically presaged collapse — egalitarian backsliding, diminishing returns on extraction, the rise of oligarchy — are flashing red. Donning his risk analyst hat, Kemp arrives at the darkest possible prognosis: The most likely destination for our globalized society is “self-termination,” self-inflicted collapse on a hitherto unprecedented scale. Goliath is more powerful than ever, but it is on a collision course with David’s stone.

Lootable Silicon

All of this seemed hard to reconcile with the atmosphere of contented civility in Granary Square on this sunny September afternoon. I proposed that an advocate for global capitalism would doubtless view our current circumstances as evidence of the Global Goliath’s collective, trickle-down bounty.

“We should be thankful for a whole bunch of things that started, by and large, in the Industrial Revolution,” Kemp said. “Vaccines, the eradication of smallpox, low infant mortality and the fact that over 80% of the population is literate. These are genuine achievements to be celebrated.”

Kemp argued that most redistribution has been a product of “stands against domination”; for example, the formation of unions, public health movements and other campaigns for social justice. Meanwhile, underlying prosperity still depends on the rimless wheel: the hub exploiting the periphery. “If we were here 150 years ago, we’d be seeing child laborers working in these courtyards,” he said, gesturing at the former coal warehouses that are now an upmarket shopping mall and that once served as a nerve center of the fossil fuel industry that built the modern age.

The same dynamics hold sway today, albeit at a further remove. Just south of us, across the Regent’s Canal, sat the London headquarters of Google, a billion-dollar glass edifice. At first glance, Kemp gave the building an enthusiastic middle finger.

Later, he explained: “The people sitting in that building are probably having a pretty good time. They have lots of ping pong tables and Huel. But the cobalt that they’re using in their microchips is still often dug up by artisanal miners in the Democratic Republic of Congo, getting paid less than a couple of dollars a day.”

Like much of the oligarchic class, the boy-gods of Silicon Valley still cleave to Hobbesian myths to justify their grip on wealth and power. Their techno-Utopian convictions, encapsulated in Bill Gates’ mantra that “innovation is the real driver of progress,” are merely a secular iteration of the divine mandates that Goliaths once used to legitimize their rule. Promises of rewards in the afterlife have been supplanted by dreams of a technological singularity and interplanetary civilization.

Another plausible eventuality, which Kemp dubs the “Silicon Goliath,” is a future in which democracy and freedom are crushed beneath the heel of advanced algorithmic systems. He is already at work on his next book about the evolution of mass surveillance, an inquiry that he told me “is in many ways even more depressing.”

Slaying Goliath

Toward the end of “Goliath’s Curse,” Kemp imagines a scenario in which the decision of whether to detonate the Trinity atomic bomb test in New Mexico in 1945 was made not by a Department of War but by a “Trinity jury,” an assembly of randomly selected members of the public.

“Now several of the indicators that Kemp identifies as having historically presaged collapse — egalitarian backsliding, diminishing returns on extraction, the rise of oligarchy — are flashing red.”

In such a counterfactual, with the Nazis defeated, Japan already inches from surrender and Manhattan Project physicists warning of a non-zero possibility that the test could ignite the whole atmosphere and exterminate all life on Earth, Kemp contends that a more inclusive decision-making process would have changed the course of history. “If you had a random selection by lottery of 100 U.S. citizens and asked them, ‘Should we detonate the bomb?’ What decision do they come to? Almost certainly ‘No,’ he told me.

As Kemp sees it, the widespread adoption of such open democracy is the only viable route to escape the endgame. These citizen juries wouldn’t be free-for-alls, where the loudest or most outrageous voice wins, but deliberative procedures that necessitate juror exposure to expert, nonpartisan context.

Such assemblies wouldn’t be enough to “slay Goliath” on their own, Kemp told me. “Corporations and states … [must] pay for the environmental and social damages they cause … to make the economy honest again.” Per capita wealth, Kemp added, should be limited to a maximum of $10 million.

I challenged Kemp that this wish-list was beginning to sound like a Rousseauvian fever-dream. But seven years immersed in the worst excesses of human folly had left him in no mood for half-measures. “I’m not an anarcho-primitivist,” he said. There was no point trying to revivify our hunter-gatherer past. “We’d need multiple planet Earths!” Kemp conceded. And yet the urgency of our current circumstances demanded a radical departure from the existing status quo, and no less a shift in mindset.

His final demotic prescription, “Don’t be a dick,” was an injunction to everyone that our collective future depends as much on moral ambition as political revolution. Otherwise, Goliath won’t be just a Bible story. It could also be our epitaph.

Uncategorized

Human agency in the Anthropocene

Ilona M. Otto ^a, Marc Wiedermann ^a, Roger Cremades ^b, Jonathan F. Donges ^ad, Cornelia Auer ^a, Wolfgang Lucht ^ac

https://doi.org/10.1016/j.ecolecon.2019.106463 Get rights and content

Abstract

The human species has been recognized as a new force that has pushed the Earth’s system into a new geological epoch referred to as the Anthropocene. This human influence was not conscious, however, but an unintended effect of the consumption of fossil-fuels over the last 150 years. Do we, humans, have the agency to deliberately influence the fate of our species and the planet we inhabit? The rational choice paradigm that dominated social sciences in the 20th Century, and has heavily influenced the conceptualization of human societies in global human-environmental system modelling in the early 21st Century, suggests a very limited view of human agency. Humans seen as rational agents, coordinated through market forces, have only a very weak influence on the system rules. In this article we explore alternative concepts of human agency that emphasize its collective and strategic dimensions as well as we ask how human agency is distributed within the society. We also explore the concept of social structure as a manifestation of, and a constraint on, human agency. We discuss the implications for conceptualization of human agency in integrated assessment modelling efforts.

1. Introduction

The Sustainable Development Goals and the Paris Agreement set very ambitious goals that, if taken seriously, would result in a rapid transformation of human-environmental interactions and decarbonization of the global socio-economic system (United Nations, 2015a, United Nations, 2015b). What the agreements do not specify, however, is how the transformation should be achieved and who the transformation agents would be. In most modern scientific assessment of global human-environmental interactions, including Integrated Assessment Models (IAMs), alternative futures do not evolve from the behavior of the population in the simulated region or market, but are externally chosen by the research teams (e.g. Moss et al., 2010). The human agency that can be broadly understood as the capacity of individual and collective actors to change the course of events or the outcome of processes (Pattberg and Stripple, 2008) is only weakly represented in the commonly used global system models. For example, Integrated Assessment Models are not capable of modelling abrupt changes and tipping points in both natural and human systems (e.g. van Vuuren et al., 2012) that may imply severe and non-linear consequences for the Earth system as a whole (Lenton et al., 2008). There is, however, a relatively rich body of literature in social sciences, primarily in political science and institutional theory, that conceptualizes human agency in the governance of social-ecological systems (e.g. Ostrom, 2005; Kashwan et al., 2018) and in Earth system governance (e.g. Biermann et al., 2012, Biermann et al., 2016). The aim of this paper is to assess the representation of human agency in Earth system science and integrated assessment modelling efforts and to examine how the rich body of literature on human agency in social sciences could be used to improve the modelling efforts.

The cornerstones of social sciences are built on the tension between agency and structure in social reproduction – the force of self-determination versus the embeddedness of social institutions (Dobres and Robb, 2000). Just as bio-physical laws determine the coupling between chemical and mechanical processes, social structures, including norms and institutions, impose constraints on the shaping of human interactions (North, 1990); they specify what people may, must, or must not do under particular circumstances and impose costs for non-compliance (Ostrom, 2005). Social institutions also have a function in expressing common or social interest and in channeling human behavior into what is socially desired (Coleman, 1990). Unlike bio-physical laws, however, social institutions are man-made structures and they are constantly being transformed by human action. In general, the smaller the social entity the less durable it is. The size, scale, and time-frame of the social entity push it towards a durable structure and stability (Fuchs, 2001). Numerous authors have contributed to this long and fruitful debate on micro- and macro-level social structures and interactions within social sciences. However, very little of that knowledge has so far been applied by the global environmental change modelling community. To give an example, the IPCC Report on Mitigation of Climate Change underlines the role of institutional, legal, and cultural barriers that constrain the low-carbon technology uptake and behavioral change. However, the diffusion of alternative values, institutions, and even technologies are not incorporated in the modelling results (Edenhofer et al., 2014). Little is known about the potential for scaling-up of social innovations, let alone the possible carbon emission reductions they could drive if applied on a larger scale. How quickly would such innovations diffuse into virtual and face-to-face social networks, and what would the agency of different actors, and groups of actors, be in such a diffusion process? The purpose of this work is to analyze how social theory could be better integrated into the global environmental change assessment community, and how relevant social theory could be incorporated in modelling efforts.

The paper is structured as follows. We start by reviewing how human agency has been incorporated within Earth system science and integrated modelling efforts so far. We then move to the exploration of the concept of human agency and social structure and review the relevant social stratification theories. We propose how the concept of human agency could be incorporated in global human-environmental system models, and finally we conclude.

2. Human agency in Earth system science and integrated assessment modelling

The recognition of the human species as the driving force of modern global environmental challenges, occurring at the end of the 20th Century, brought a new perspective to environmental and Earth system sciences. Lubchenco (1998) called directly for the integration of the human dimensions of global environmental changes with the physical-chemical-biological dimensions. In this context, Crutzen (2006) proposed the distinction of the Anthropocene as a new geological epoch, where the human species becomes a force outcompeting natural processes. As one possible framework to assess human agency in the Anthropocene, Schellnhuber (1999) developed the notion of “Earth System” analysis for global environmental management in which the human force has been conceptualized as a “global subject”. The global subject is a real but abstract force that represents the collective action of humanity as a self-conscious force that has conquered the planet. The global subject manifests itself, for instance, by adopting international protocols for climate protection.

The conceptualization of the human species as the global subject has been applied in Integrated Assessment Models (IAMs). IAMs refer to tools assessing strategies to address climate change and they aim to describe the complex relations between environmental, social and economic factors that determine future climate change and the effects of climate policy (van Vuuren et al., 2011). IAMs have been valuable means to set out potential pathways to mitigate climate change and, importantly, have been used in the IPCC’s assessments of climate change mitigation (Clarke et al., 2014). However, the development of Integrated Assessment Models (IAMs) coincides in time with the supremacy of the rational choice paradigm. Rational choice theory emphasizes the voluntary nature of human action and the influence of such actions on decisions, assuming human beings act on the basis of rational calculations of benefits and costs (Burns, 1994). According to this paradigm, rationality is a feature of individual actors and the world can be explained in terms of interactions of atomic entities. Humans are rational beings motivated by self-interest and consciously evaluate alternative courses of action. Markets are seen as the mechanisms linking the micro and macro levels and allow the combination of the concrete actions of individuals, e.g. buyers and sellers (Jaeger et al., 2001). The rational choice paradigm is reflected in welfare maximization assumptions underpinning the development of computable general equilibrium (CGE) models that are widespread in IAMs. CGE models are computer-based simulations which use a system of equations that describe the whole world economy and their sectoral interactions. The analysis of scenarios in CGE models compares a business-as-usual equilibrium with the changes introduced by one or several policies and environmental shocks — e.g. a carbon tax or emissions trading scheme under several climate scenarios — which generate a new equilibrium (Babatunde et al., 2017). It is important to understand that the policy shock in such models is introduced externally; it does not evolve from the model and does not consider the dynamics behind the agency of different actors and groups of actors. In fact, human societies in CGE models are only reflected in aggregated population numbers by world region. The institutional settings within the human societies operate are given and cannot be endogenously changed. CGE models place a strong emphasis on the market as a solution to all kinds of problems including environmental and social issues (Scrieciu, 2007). Furthermore, state-of-the-art IAMs model aggregate datasets of sub-continental size. For instance, the IAM known as REMIND considers just 11 world regions, while the energy component of IMAGE considers only 26. The order of magnitude of the population of each of these regions is between 287 M and 680 M inhabitants (ADVANCE, 2017). Similarly, in the global land use allocation model MAgPIE, the food energy demand for ten types of food energy categories (cereals, rice, vegetable oils, pulses, roots and tubers, sugar, ruminant meat, non-ruminant meat, and milk) in ten world regions differentiated in the model is determined exogenously by population size and income growth, assuming that, for example, higher income is related to a higher demand for meat and milk (Popp et al., 2010). The impacts of changing lifestyles and the implications of demand-side solutions can be explored only manually by varying the underlying assumptions.

In context of the definition of human agency used above, IAMs reflect an agency of a rational consumer who decides on the choice of an optimal action having access to perfect information about the alternatives. By analyzing energy, land use, and their implications on global emissions (e.g. van Vuuren et al., 2012; Hibbard et al., 2010) IAMs can compute an economic setup to maximize welfare functions. Nevertheless, the welfare functions do not cover the diversity of human preferences. Complex distinctions of qualitative aspects, such as networks or influencers that can drive these processes, do not exist.

This drawback has been noted by the IAM community and attempts have been made to integrate human agency related behavior towards the political economy, social behavioral and interaction patterns (Riahi et al., 2017), or regimes of effort sharing (van den Berg et al., 2019) have been made. Some models also consider inequality and a diversity of consumption patterns (Hasegawa et al., 2015; McCollum et al. 2018). However, these approaches are still driven by exogenous quantifications and are unable to sufficiently inspect dynamics of human agency. Although IAMs are able to design pathways combining multiple strategies to achieve the 1.5 °C target of the Paris Agreement, which include human agency related actions such as lifestyle changes (van Vuuren et al., 2018), many questions remain. For example, how can human agency be triggered to achieve the lifestyle changes, at an individual level, necessary to achieve the 1.5 °C target? Also, how can the necessary institutional dynamics be brought into play? So far, these aspects are rarely considered in IAMs.

Novel and promising modelling approaches to incorporate human agency are being developed in complex network science (Borgatti et al., 2009) and social-ecological system modelling (Pérez et al., 2016). Complex networks usually consist of a set of nodes representing individual agents or representative aggregations thereof (such as business parties, geographical regions or countries) which are connected by different types of linkages, such as business relations, diplomatic ties, or even acquaintance and friendship (Newman, 2018). This type of framework has been developed in the past, and applied successfully to describe heterogeneous datasets from the social sciences, and to establish conceptual models for socio-economic and socio-ecological dynamics (Filatova et al., 2013). Nevertheless, most of such models are still based on theoretical assumptions with weak links to empirical data. A closer link with empirical data has so far only been achieved at case study level, focusing on particular local socio-environmental phenomena such as fishery or water management with agents representing local resource users or managers (e.g. Suwarno et al., 2018; Troost and Berger, 2015). The questions driving this work are: (i) how can similar models be conceptualized in order to represent the whole World-Earth system of human societies and their bio-physical environment (Donges et al., 2018) and (ii) how can they be linked with empirical data?

3. The concept of human agency in social sciences

Dellas et al. (2011) refer to agency in the governance of the Earth system as the capacity to act in the face of earth system transformation or to produce effects that ultimately shape natural processes. Agency in Earth system governance may be considered as contributing to problem solving, or alternatively it could include the negative consequences of the authority to act. Lister (2003) and Coulthard (2012), in their research on agency related to environmental and citizenship problems, distinguish two dimensions: (i) ‘everyday agency’ being the daily decision-making around how to make ends meet, and ‘strategic agency’ involving long-term planning and strategies; and (ii) ‘personal agency’ which reflects individual choices and ‘political and citizenship agency’ which is related to the capacity of people to affect the wider change (Lister 2003). Personal agency varies significantly across human individuals. However, there are powerful examples of social protests and movements demonstrating that even individually disempowered people can have a strong voice if they act collectively (Kashwan, 2016). In the context of natural resources and environmental management, there are empirical examples of self-organized local and regional communities and grassroots movements crafting new institutions that limit the control of national authorities (García-López, 2018; Dang, 2018). To give an example, civil society groups in Mexico managed to shape the REDD+ policies to protect the rights of agrarian communities (Kashwan, 2017a). In this context, Bandura (2006) proposes the differentiation of individual, proxy and collective agency (2006: 165). Individual agency refers to situations in which people bring their influence to bear through their own actions. This varies substantially from person to person with respect to individual freedom to act and the consequences of action. Individual agency is influenced by a whole set of socio-economic characteristics including gender, age, education, religion, social, economic and political capital. In many cultures, the individual agency of women is limited, for example, by inheritance law or by informal norms restricting their mobility or educational opportunities (Otto et al., 2017). However, individual agency also varies with an individual’s ability to change the system rules. For example, very wealthy or influential people might find it easier to set new market trends or influence public decision-making processes than those with fewer resources (Otto et al., 2019). Proxy, or socially mediated agency refers to situations in which individuals have no direct control over conditions that affect their lives, but they influence others who have the resources, knowledge, and means to act on their behalf to secure the outcome they desire. Collective agency refers to situations in which individuals pool their knowledge, skills, and resources, and act in concert to shape their future (Bandura 2006: 165). These dimensions of agency are visualized in Fig. 1.

The dominant view of human agency in Earth system science and integrated modelling approaches has so far focused on the left upper corner of Fig. 1, i.e. on the everyday agency of individual human agents. This would correspond to, for example, modelling the effects of food consumption on land use patterns (e.g. Popp et al., 2010). Interestingly, although opinion formation and election models are well advanced in game theory (e.g. Penn, 2009; Ding et al., 2010), they have not yet been applied to the formation of international environmental policy in IAMs. At the same time the recent so-called protest voting shows that a small fraction of voters can push public policy down a radically different pathway. Some studies link the protest voting and rising populism with increasing inequalities and the political and social exclusion of the poor and underprivileged (Becker et al., 2017). In some cases, radical policy changes might also be achieved by individual acts of civil disobedience and, in a destructive manner, by terrorist attacks. Civil disobedience represents the peaceful breaking of unjust or unethical laws and is a technique of resistance and protest whose purpose is to achieve social or political change by drawing attention to problems and influencing public opinion. Terrorism is defined as an act of violence for the purpose of intimidating or coercing a government or civilian population.

Furthermore, radical policy changes and social tipping points can emerge due to changes in the collective behavior and preferences. The term ‘tipping point’ “refers to a critical threshold at which a tiny perturbation can qualitatively alter the state or development of a system” (Lenton et al., 2008), hence the mere existence of tipping points implies that small perturbations created by parts of such a system can push the whole system into a different development trajectory. Examples of tipping-like phenomena in socio-economic systems include financial crises, but could also include the spread of new social values, pro-environmental behavior, social movements, and technological innovations (Steffen et al., 2018). To give an example, social movements and grassroots organizations played an important role in the German energy transition that was initiated in 2011 as a reaction to the nuclear disaster in Fukushima in Japan. It was, however, preceded by about 30 years of environmental activism (Hake et al., 2015). Finally, tipping-like phenomena can also be achieved by consumer boycotts and carrotmob movements. Consumer boycotts coupled with environmental NGO campaigns led, in Europe, to changes in the animal welfare regulations and the implementation of fair trade schemes (Belk et al., 2005). Carrotmobs refer to consumers collectively swarming a specific store to purchase its goods in order to reward corporate socially responsible behavior (Hoffmann and Hutter, 2012).

At the same time, cultural values and the ethical interpretation of behavior might vary in some respects across different countries and world regions and will lead to different manifestations of agency. Cultural values provide a strong filter of the actions perceived as good or responsible, as well as the consequences of violating norms (Belk et al., 2005). In the climate change context, some authors link the public acceptance of climate policy instruments to the belief and value systems in place, and the perceptions of the environment (Otto-Banaszak et al., 2011).

4. The manifestation of human agency: the layers of social structure

Biermann and Siebenhüner (2009) propose a distinction between actors and agents in Earth system governance. Actors are the individuals, organizations, and networks that participate in the decision-making processes. Agents are those actors who have the ability to prescribe behavior. The collective prescriptions and constraints on human behavior are usually referred to as the social structure (Granovetter, 1985; Dobres and Robb, 2000). The social structure is composed of the rule system that constitutes the “grammar” for social action that is used by the actors to structure and regulate their transactions with one another in defined situations or spheres of activity. The complex and multidimensional normative network is not given, but is a product of human action; “human agents continually form and reform social rule systems” (Burns and Flam, 1986: 26). The social rule system can also be framed as social institutions that are involved in political, economic, and social interactions (North, 1991). Similarly, Elinor Ostrom defines institutions as “the prescriptions that humans use to organize all forms of repetitive and structured interactions. Individuals interacting within rule-structured situations face choices regarding the actions and strategies they take, leading to consequences for themselves and for others” (Ostrom, 2005: 3). Social norms are shared understandings of actions and define which actions are obligatory, permitted, and forbidden (Crawford and Ostrom, 1995). Social order is only possible insofar as participants have common values and they share an understanding of their common interests and goals (King, 2009). Williamson (1998) proposes differentiating different informal institutions such as norms, beliefs and traditions, and formal institutions that comprise formal and written codes of conduct.

The process of shaping of the social rule system formation is not always fully conscious and intended. Lloyd (1988: 10) points out that a social structure is emerging from intended and unintended consequences of individual action and patterned mass behavior over time “Once such structures emerge, they feedback on the actions” (Sztompka, 1991: 49). For Giddens (1984) human action occurs as a continuous flow of conduct and he proposed turning the static notion of structure into the dynamic category of structuration to describe the human collective conduct. Human history is created by intentional activities but it is not an intended project; it persistently eludes efforts to bring it under conscious direction (Giddens, 1984: 27). As pointed out by Sztompka (1994), Giddens, embodies human agency in the everyday conduct of common people who are often distant from reformist intentions but are still involved in shaping and reshaping human societies. This process of the formation of social structure takes place over time; the system which individuals follow today have been produced and developed over a long period. “Through their transactions social groups and communities maintain and extend rule systems into the future” (Burns and Flam 1987: 29).

Another element of the social structure that is identified by several authors corresponds to the network of human relationships that, just like the shapes in geometry, can take different forms and configurations (Simmel, 1971). The network of relationships among the social agents is also referred to as governance structures, or sometimes as organizations. North (1990: 73) defines organizations as “purposive entities designed by their creators to maximize wealth, income, or other objectives defined by the opportunities afforded by the institutional structure of the society.” Williamson (1998), focusing on the types of contracts, distinguishes three basic types of governance structures: markets, firms, and hybrids. In markets, transaction partners are autonomous; in firms, partners are inter-dependent and integrated into an internal organization. Hybrids are intermediate forms in which contract partners are bilaterally dependent but to a large degree maintain autonomy (Williamson 1996: 95–98). Studying communication networks and social group structures allows us to distinguish more social network relationship patterns (Sztompka, 2002: 138).

Finally, the social structure is also shaped and influenced by large material objects such as infrastructure and other technological and industrial structures, that some authors call the technosphere (Spaargaren, 1997: 78). Herrmann-Pillath (2018) defines the technosphere as the encompassing aggregate of all artificial objects in opposition to the natural world, and more specifically, establishes the systemic separateness of the technosphere relative to the biosphere. Just as social norms impose on one hand certain constrains on human behavior, however, on the other hand, structure the human interactions and also provide certain opportunities, the technosphere can be viewed as a humanly designed constructs that provide certain opportunities as well as they limit certain choices of individuals operating at different geographical and time scales (Donges et al., 2017a).

The system is fully interconnected, and the social structure layers are interrelated. The slow changing layers of social structure impose constraints on the layers that change more quickly. The faster changing layers of social structure, however, are also able to change the slow slayers through feedback mechanisms (c.f. Williamson, 2000). Human agency is manifested through the maintenance, reproduction and modifications in the social structure layers (Burns, 1994). Interestingly, infrastructure objects in the technosphere layer show a similar order of change as the informal and formal institutions, and thus might constrain the social change in the faster changing levels. Thus artefacts become co-carriers of agency (Herrmann-Pillath, 2018). Nevertheless, sharp brakes from the established procedures rarely happen. Such defining moments are an exception to the rule and usually emerge from massive discontents such as civil wars, revolutions, or financial crises (Williamson, 1998). Institutions can also lock the society into a path-dependence (Beddoe et al., 2009). The capacity to undergo a radical restructuring, however, is a unique feature distinguishing social systems from organic or mechanical ones. Restructuring the social structure is a product of human agency and is grounded in the interaction between structures and human actions that produces change in a system’s given form, structure or state (Archer, 1988: xxii). However, the transition of institutions is frequently driven by crises (Beddoe et al., 2009).

Burns (1994: 215-216) introduces the notion of ‘windows of opportunity’ that are very relevant for analyzing social transformations. Interactive situations lacking social equilibria, which typically occur after catastrophes and other shocks, usually give rise to uncertainty, unpredictability, and confusion, and motivate actors to try, individually or collectively, to restructure the situation. In such restructuring activities, actors typically engage in reflective processes and make “choices about choice” and participate in meta-games (Burns 1994: 208). The actors may structure and restructure their preferences, outcomes, and outcome structures, and occasionally also the entire decision and game systems in which they participate. Through such structuring activity, human agents also create, maintain and change institutions and collective or organized agents such as movements, the state, market and bureaucratic organizations (Burns and Dietz, 1992; Burns, 1994: 215–216).

Transformations are the moments in history when the meta choices – “choices about choices” are made. The outcomes of such choices and the new type of system depend largely on the agents that get involved in the collective process of designing the new system. This process could be exclusive and incorporate only a narrow group of decision-makers as frequently happens in “quiet” transitions to authoritarian regimes. Alternatively, they can be more open and include representatives of various social groups, as happened in the political and economic transformation in Eastern Europe. Taking this example, Burns (1994) proposes that transformations are a co-evolutionary process sometimes driven by contradicting actors’ interests. Transformations might entail shifts in core societal organizing principles and systems of rules. As a result, agents with vested interests may struggle to maintain established systems or to limit the changes within them. Other agents act openly or covertly to modify or transform the system. Table 1 summarizes the above discussion and tries to link the social structure layers to the dominant type of human agency that can to be used to transform them.

Table 1. The layers of social structure, the dominant type of agency and the order of change.

Structure layer	Sub-components	The dominant type of human agency	The order of change
Institutional	Informal rules: norms, religion, tradition, customs	Collective and strategic	30 to over 100 years
Institutional	Formal rules: constitutions, written codes of conduct, judiciary, property rights	Collective and citizenship	10 to 50 years
Organizational	Governance structures	Proxy and strategic	5 to 10 years
	Organizations	Proxy, strategic	5 to 10 years
	Networks	Proxy, individual, everyday	Continuous
Technosphere	Infrastructure	Proxy, strategic	10–50 years
Technosphere	Technology	Proxy, individual and everyday	Continuous

(Following Williamson, 1998).

Even in periods of radical change, however, the actors never start from scratch. They cannot choose a completely new system and they always depart from the ongoing social order in which they are embedded. The future evolves from practical activities, experiments, learning, conflict and struggle (Burns, 1994: 216). A similar point of view is presented by evolutionary institutional economists, in which transformations are seen not as a simple replacement of old institutions by new ones, but as a recombination and reworking of old and new elements and groups of actors (e.g. Stark, 1996; Bromley, 2000).

5. Distribution of human agency: differentiating socio-metabolic agent classes

Following the rational choice paradigm could lead us to a conclusion that the society is a sum of individuals (Burns, 1994) and that any forms of agency should be equally distributed among the individuals in the society. Such an approach is typical for integrated assessment models in which human systems are usually separated into population and economic sectors. The parameters that describe population are usually mainly population number, and economic production determines the use of resources and pollution emissions in the model (e.g. van Vuuren et al., 2012).

It is, however, enough to observe the world to know that such assumptions are very simplistic. People’s resource use and pollution emissions differ according to income, place of abode, type of occupation, and possessions. Moreover, their goals and interests, and the likelihood of them being fulfilled also differ. There are powerful individuals and groups in society who successfully strive for their interests, and there are individuals and groups who, despite struggling, never achieve their objectives. There are also masses of individuals who just strive to make ends meet. The questions are what types of agents or organizations can be incorporated in the models and what sort of agency do they have? Is there a need for a new social class theory taking access to energy and related carbon emissions as the base of social stratification?

Most social differentiation theories follow either the Marxist distinction between physical and capital endowments or the Weberian approach which differentiates classes through inequalities in ownership and income (Kozyr-Kowalski, 1992: 53). Some class theorists also highlight the development stages and inequalities across different countries and world-regions (Offe, 1992: 122). One more dimension that has not been discussed so far by social differentiation theories is the socio-metabolic profile of social classes, which constitutes the common ground for social and natural sciences. Social metabolism refers to the material flows in human societies and the way societies organize their exchanges of energy and materials with the environment (Fischer-Kowalski, 1997; Martinez-Alier, 2009). Social classes can be differentiated based on their metabolic profiles (Martinez-Alier, 2009). The use of energy by human beings can be divided into two main categories. The first one refers to the endosomatic use of energy as food, and the second one refers to the exosomatic use of energy as fuel for cooking and heating, and as power for the artefacts and machines produced by human society. Thus one person a day must eat the equivalent of 1500 to 2500 kcal to sustain their life functions, which is equivalent to about 10 MJ (megajoules) of energy per day or 3.65 GJ per year (Martinez-Alier, 2009). This amount varies only slightly among human beings. A rich person physically cannot eat much more, and even poorer individuals need the equivalent energy in the form of food to survive. Dietary composition and the amount of waste produced, however, will differ across the social strata. Nevertheless, there are still people suffering from hunger, unable to meet their basic needs.

The exosomatic energy use varies to a greater degree. The poorest social groups, who have no permanent access to electricity in their homes, who obtain energy for cooking and heating from the combustion of biomass products, who use overcrowded buses and trains to travel, use in total about 10 GJ of energy per person per year (Martinez-Alier, 2009) and constitute the lowest, socio-metabolic underclass. A more detailed picture can be derived by comparing the carbon footprint of different socio-economic groups. Personal CO₂ emissions are released directly in fuel combustion processes in vehicles, airplanes, heating and cooking appliances, and indirectly through electricity use and consumption of products that generated emissions in the upstream production processes. The authors include CO₂ emissions from energy used directly in homes (for space heating, lighting, etc.), for personal transportation (including personal vehicles and passenger aviation), and from the energy embedded in the production of goods consumed. Kümmel (2011) proposes the term “energy slaves” to describe the exosomatic energy use from fossil fuels by modern human society. On average, the daily energy consumption of a human being is equivalent to the men power of 15 people. Inhabitants of the most energy intensive Western Societies (i.e. the U.S.) consume, per person, the equivalent of the work of 92 people every day.

The results from UK households show that CO₂ emissions are strongly income, but also location, dependent. The highest emissions can be generated by people living in suburbs, mostly in detached houses, and having two or more cars. Emissions of such households equated to about 26 CO₂ tonnes in 2004. This amount was 64% higher than the emissions of the group with lowest emissions of 16 CO₂, which comprised mostly of older and single person urban households as well as the unemployed living mostly in urban areas (Druckman and Jackson, 2009). UK household emissions can be compared with emissions from households located in less developed countries. For example, household emissions in Malaysia, as in the UK, are strongly dependent on income and location. However, Malaysian households with the lowest emissions were found in villages as well as in low-income urban squatter settlements. The urban squatter settlement households emitted on average 10.18 CO₂ tonnes. The village households emitted on average 9.58 CO₂ tonnes per year. Households with the highest CO₂ emissions were located in high cost housing areas and they were responsible on average for 20.14 CO₂ tonnes per year (Majid et al., 2014).

On the other end of the social ladder, there are super-rich hyper-mobile individuals with multiple spacious residences, and whose live-styles are characterized by conspicuous consumption patterns. They are less than 1% of global population and their consumption related greenhouse gas emissions could be over 170 times higher than the world’s poorest 10% (Oxfam, 2015). They can be characterized by extremely high levels of all types of agency. The influence and roles of many super-rich in the world of politics, media, culture, business and industry are often inter-related. In contrast to the super-rich in pre-industrial societies they have almost unlimited mobility, owning properties in different counties, with their homes being guarded and fortified. They have the ability to switch countries of residence, taking the advantage of ‘nondomiciled’ tax status, i.e. being the national of a certain country while not actually living there (Paris, 2013). Table 2 presents a first attempt to stratify the global population according to their socio-metabolic profiles that is based on disaggregated data on consumption related carbon emissions (Oxfam, 2015; Otto et al., 2019).

Table 2. Socio-metabolic class differentiation.

Empty Cell	Percent of global population	Percent of life-style CO₂ emissions	The level of human agency
Socio-metabolic underclass	20%	2.5%	Extremely low
Socio-metabolic energy poor class	30%	7.5%	Low
Socio-metabolic lower class	30%	22%	Moderate level of collective agency
Socio-metabolic middle class	10%	19%	Moderate to high
Socio-metabolic upper class	9.5%	35.4%	Very high
Super-rich	0.54%	13.6%	Extremely high

(Based on: Oxfam, 2015; Otto et al., 2019).

The proportions in Table 2 are striking. The top 10% of the global population is responsible for almost 50% of global consumption related greenhouse gas emissions. The wealthiest 0.54% of the human population is responsible for more lifestyle carbon emissions than the poorest 50% (Otto et al., 2019).

Energy use, as well as carbon dioxide emission, can also be used to analyze the socio-metabolic profile of economic sectors, companies and other organizations. From 1854 to 2010 12.5% of all industrial carbon pollution was produced by just five companies – Chevron, ExxonMobil, British Petroleum, Shell and Conoco Philipps (Union of Concerned Scientists, 2018). To give an example from a different sector – in 2015 Saint-Gobain, a French multinational building materials manufacturer emitted 9.5 million metric tonnes CO₂e (Carbon Disclosure Project, 2016: 22). For a comparison, emissions from industrial processes in France in 2013 equated to 17.6 million tonnes CO₂e (General Directorate for Sustainable Development, 2016: 25) (GTM, 2018).

The socio-metabolic profile of social classes, nations, and organizations can be directly linked with their agency in the Earth system. The global socio-metabolic underclass is obviously characterized by a very low degree of agency. There are rare exceptions of mass protests initiated by the poorest social groups that can collectively influence formal institutions and change their governance (Kashwan, 2017b). However, these people are mostly occupied with making ends meet and have low organizational capabilities. In contrast, the global socio-metabolic upper classes are those who are characterized by a high level of individual agency as well as having the organizational capabilities to actively exercise their agency. Due to their resource incentive life-style they also have the moral obligation to be the agents of a transformation in global sustainability.

6. Improving the representation of human agency in integrated assessment modelling

In this section we ask how the above conceptual discussion could be summarized into guidelines improving the operationalization of human agency in Earth system science and integrated assessment modelling. In order to incorporate the different aspects of human agency as discussed in the previous sections, there is a need to introduce agents with heterogeneous goals, opinions and preferences into the models. The agents should be able to form networks that represent their mutual interrelationships and interactions between them. These system interaction rules should ideally refer to the social structure layers differentiated in Table 1, forming a nested hierarchical embeddedness of each agent.

Conceptual models, that incorporate the above requirements have been successfully developed and studied in the recent past. Their core properties might thus form a proper basis for extending IAMs to include heterogeneous agency on the level of (representative) individuals. Such models have been utilized to study opinion, and the associated consensus-formation specifically under the assumption of heterogeneous agents. Most of these works are based on the voter model in which agents exchange discrete (sets of) opinions in order to reach some consensus on a given (possibly abstract) topic or problem (Clifford and Sudbury, 1973; Holley and Liggett, 1975). Acknowledging that in its standard version the voter model considers all agents to have identical agency, extensions have been based on social impact theory (Latane, 1981) that specifically include heterogeneous relationships between single actors or groups (Nowak et al., 1990). Such extended models generally account for proximities between agents in some abstract space of personal relationships which is commonly modeled by assigning agents unique values of persuasiveness and supportiveness, describing their agency with respect to influencing as well as supporting others. While being of generic nature such classes of models can be easily modified to account for various kinds of processes related to social behavior, such as social learning (Kohring, 1996) or leadership (Holyst et al., 2001), which are again directly related to the notions of (heterogeneous distributions of) human agency. Certain models include additional layers of complexity by also accounting for the heterogeneous distribution of different group sizes (Sznajd-Weron, 2005) and certain majorities within those groups (Galam, 2002) when determining criteria for consensus in opinion dynamics.

One particular model of general cultural dynamics that has attracted great interest in the social science community, and that should be highlighted here, is the so-called Axelrod model (Axelrod, 1997). In its core, it accounts for two commonly observed tendencies in large groups of individuals or aggregations thereof: social influence (i.e. agency) and homophily (a process that dynamically influences each individual’s agency over time). The Axelrod-model not only specifically accounts for heterogeneity in the different agents but also (and to some degree unintuitively) allows emerging cultural diversity to be modeled in its convergent state. In general, such flexible approaches allow incorporating individual human agency in terms of the different ties an agent might have with others (Emirbayer and Goodwin, 1994; Granovetter, 1977). Additionally, each tie can be associated with different strengths, thus also incorporating heterogeneity in the human agency (Castellano et al., 2009). Network modelling approaches further allow us to explicitly resolve the associated social structure (as well as the temporal evolution thereof) through an evaluation of the overall topology of the network on the meso- or macroscale (Costa et al., 2007).

A necessary step in operationalizing human agency in IAMs includes differentiating global socio-metabolic agent classes with heterogeneous metabolic profiles linking them with the material and energy flows in the bio-physical environment as well as heterogeneous social profiles that specify their preferences, opinions, and positions in social networks. Such efforts could be linked to the emerging research on downscaling planetary boundaries (Häyhä et al., 2016) as well as the established research on differentiating social milieus (e.g. Bauer and Gaskell, 1999). Some authors also propose model co-development, together with citizens and citizen groups (Figueres et al., 2017). Some authors also recommend abandoning the search for one gold-standard model, and instead explore future pathways based on a multitude of different concepts and representations of people and human agency (Donges et al., 2017b). For example, Donges et al. (2018) propose a modelling framework allowing incorporation of large sets of different models and concepts, in a standardized form, in order to assess and compare different future trajectories.

7. Conclusions

The Anthropocene has emerged unintentionally as a side effect of the industrialization of human societies (Crutzen, 2006). There are only a few examples of the human ability to internally interact with planetary geological forces, with the Montreal Protocol being the most often referred to example (Velders et al., 2007). At the same time historical examples show that there are instances of rapid transitions in societies (Bunker and Alban, 1997). Achieving policy challenges as outlined in the Sustainable Development Goals require a certain degree of societal transformation. The concept of agency is central to implementing transformations needed to limit global warming and achieve the SDGs. Most of the IAMs that dominate the scientific assessments of global environmental changes do not include a representation of human societies that would have a capacity to undertake system transformations. At the same time, there is a relatively rich social science theory that can be used to improve the operationalization of human agency in integrated assessment modelling efforts.

In this paper we show that human agency can actively shape the World-Earth system (c.f. Donges et al., 2018) through interventions at different layers of social structure. Human agency, however, is not evenly distributed across all human individuals and social groups. We postulate a differentiation of socio-metabolic agent classes that could be integrated into integrated assessment modelling efforts. More socio-economic sub-national and sub-population group data is needed for this purpose (c.f. Otto et al., 2015). Social institutions for sustainable management of global, regional, and local ecosystems, however, do not generally evolve spontaneously, but have to be consciously designed and implemented by the resource users (Gatzweiler and Hagedorn, 2002; Kluvankova-Oravska et al., 2009). Each social transformation contains a disruptive component that implies a destruction of existing patterns of social interaction and institutional structures, and creation and emergence of new patterns and structures. Introducing more dimensions of human agency into IAMs, and co-creating scenarios and pathways for modelling exercises together with citizens and institutions, would help break the barriers that disconnect peoples’ actuality and agency with models, a discourse which has been gaining weight among policy makers (Figures, 2016). This disconnection can be broken by co-developing with citizens and various resource users the elements of global human-environmental system models, and by considering the people behind the numbers and the possible ways of funneling their agency. We encourage the integrated modelling community to work more closely with social scientists as well as we encourage social scientists to explore the methods and concepts applied in natural sciences.

Acknowledgments

The authors are grateful to two anonymous reviewers for their feedback that helped them to improve the paper. I.M.O., J.F.D., and R.C. are grateful for financial support by the Earth League’s EarthDoc programme. I.M.O. is supported by funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 821010. J.F.D. is grateful for financial support by the Stordalen Foundation via the Planetary Boundary Research Network (PB.net). M.W. is financially supported by the Leibniz Association (project DominoES). This research has been carried out within the COPAN – Co-evolutionary Pathways Research Group at PIK.

References

ADVANCE, 2017ADVANCEThe common integrated assessment model (IAM) documentation(2017)http://iamcdocumentation.eu/index.php/IAMC_wiki Google Scholar
Archer, 1988M.S. ArcherCulture and AgencyCambridge University Press, Cambridge (1988)Google Scholar
Axelrod, 1997Robert AxelrodThe Complexity of Cooperation. Agent-based Models of Competition and CollaborationPrinceton University Press, Princenton (1997)Google Scholar
Babatunde et al., 2017Kazeem Alasinrin Babatunde, Rawshan Ara Begum, Fathin Faizah SaidApplication of computable general equilibrium (CGE) to climate change mitigation policy: a systematic reviewRenew. Sust. Energ. Rev., 78 (October) (2017), pp. 61-71, 10.1016/j.rser.2017.04.064 View PDF View article View in Scopus Google Scholar
Bandura, 2006Albert BanduraToward a Psychology of Human Agency‎Perspect. Psychol. Sci., 1 (2) (2006), pp. 164-180Crossref View in Scopus Google Scholar
Bauer and Gaskell, 1999Martin W. Bauer, George GaskellTowards a paradigm for research on social representationsJ. Theory Soc. Behav., 29 (2) (1999), pp. 163-186, 10.1111/1468-5914.00096 View in Scopus Google Scholar
Becker et al., 2017Sascha O. Becker, Thiemo Fetzer, Dennis NovyWho voted for brexit? A comprehensive district-level analysisEcon. Policy, 32 (92) (2017), pp. 601-650, 10.1093/epolic/eix012 View in Scopus Google Scholar
Beddoe et al., 2009Rachael Beddoe, Robert Constanza, Joshua Farley, Eric Garza, Jennifer Kent, et al.Overcoming systemic roadblocks to sustainability: the evolutionary redesign of worldviews, institutions, and technologiesPNAS, 106 (8) (2009), pp. 2483-2489Crossref View in Scopus Google Scholar
Belk et al., 2005Russell Belk, Timothy Devinney, Giana EckhardtConsumer ethics across culturesConsum. Mark. Cult., 8 (3) (2005), pp. 275-289, 10.1080/10253860500160411 Google Scholar
Biermann and Siebenhüner, 2009F. Biermann, B. SiebenhünerManagers of Global Change: The Influence of International Environmental BureaucraciesMIT Press (2009)Google Scholar
Biermann et al., 2012F. Biermann, K. Abbott, S. Andresen, K. Bäckstrand, S. Bernstein, M.M. Betsill, H. Bulkeley, et al.Navigating the Anthropocene: improving earth system governanceScience, 335 (6074) (2012), pp. 1306-1307, 10.1126/science.1217255 View in Scopus Google Scholar
Biermann et al., 2016Frank Biermann, Xuemei Bai, Ninad Bondre, Wendy Broadgate, Chen-Tung Arthur Chen, Opha Pauline Dube, Jan Willem Erisman, et al.Down to earth: contextualizing the AnthropoceneGlob. Environ. Chang., 39 (July) (2016), pp. 341-350, 10.1016/j.gloenvcha.2015.11.004 View PDF View article View in Scopus Google Scholar
Borgatti et al., 2009Stephen P. Borgatti, Ajay Mehra, Daniel J. Brass, Giuseppe LabiancaNetwork analysis in the social sciencesScience, 323 (5916) (2009), pp. 892-895, 10.1126/science.1165821 View in Scopus Google Scholar
Bromley, 2000Daniel BromleyMost Difficult Passage: The Economic Transition in Central and Eastern Europe and the Former Soviet Union(2000)(In . Berlin)Google Scholar
Bunker and Alban, 1997B.B. Bunker, B.T. AlbanLarge Group Interventions: Engaging the Whole System for Rapid ChangeJossey-Bass Publications, San Francisco (1997)Google Scholar
Burns and Flam, 1987T.R. Burns, H. FlamThe Shaping of Social OrganizationSage Publications, London (1987)Google Scholar
Burns, 1994Tom R. BurnsTwo conceptions of human agency: rational choice theory and the social theory of actionPiotr Sztompka (Ed.), Agency and Structure. Reorienting Social Theory, Gordon and Breach Science Publishers, Yverdon, Camberwell, Paris (1994)Google Scholar
Burns and Dietz, 1992Tom R. Burns, Thomas DietzCultural evolution: social rule systems, selection and human agency, cultural evolution: social rule systems, selection and human agencyInt. Sociol., 7 (3) (1992), pp. 259-283, 10.1177/026858092007003001 View in Scopus Google Scholar
Burns and Flam, 1986Tom R. Burns, Helena FlamThe Shaping of Social Organization: Social Rule System Theory and its ApplicationsSage, London (1986)Google Scholar
Carbon Disclosure Project, 2016Carbon Disclosure ProjectEmbedding a carbon price into business strategy(2016)https://b8f65cb373b1b7b15feb-c70d8ead6ced550b4d987d7c03fcdd1d.ssl.cf3.rackcdn.com/cms/reports/documents/000/001/132/original/CDP_Carbon_Price_2016_Report.pdf?1474269757 Google Scholar
Castellano et al., 2009Claudio Castellano, Santo Fortunato, Vittorio LoretoStatistical physics of social dynamicsRev. Mod. Phys., 81 (2) (2009), pp. 591-646, 10.1103/RevModPhys.81.591 View in Scopus Google Scholar
Clarke et al., 2014Leon Clarke, K. Akimoto, M. Babiker, et al.Assessing transformation pathwaysClimate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, USA (2014)https://www.ipcc.ch/pdf/assessment-report/ar5/wg3/ipcc_wg3_ar5_chapter6.pdf Google Scholar
Clifford and Sudbury, 1973Peter Clifford, Aidan SudburyA model for spatial conflictBiometrika, 60 (3) (1973), pp. 581-588, 10.1093/biomet/60.3.581 Google Scholar
Coleman, 1990James S. ColemanCommentary: social institutions and social theoryAm. Sociol. Rev., 55 (3) (1990), pp. 333-339, 10.2307/2095759 Google Scholar
Costa et al., 2007L. da F. Costa, F.A. Rodrigues, P.R. Villas BoasCharacterization of complex networks: a survey of measurementsAdv. Phys., 59 (1) (2007), pp. 167-242, 10.1080/00018730601170527 View in Scopus Google Scholar
Crawford and Ostrom, 1995Sue E.S. Crawford, Elinor OstromA grammar of institutionsAm. Polit. Sci. Rev., 89 (3) (1995), pp. 582-600, 10.2307/2082975 View in Scopus Google Scholar
Crutzen, 2006Paul J. CrutzenThe ‘Anthropocene’Earth System Science in the Anthropocene, Springer, Berlin, Heidelberg (2006), pp. 13-18, 10.1007/3-540-26590-2_3 View in Scopus Google Scholar
Coulthard, 2012S. CoulthardCan we be both resilient and well, and what choices do people have? Incorporating agency into the resilience debate from a fisheries perspectiveEcol. Soc., 17 (1) (2012), p. 4Google Scholar
Dang, 2018Wenqi DangHow culture shapes environmental public participation: case studies of China, the Netherlands, and ItalyJ. Chin. Gov., 0 (0) (2018), pp. 1-23, 10.1080/23812346.2018.1443758 Google Scholar
Dellas et al., 2011Eleni Dellas, Philipp Pattberg, Michele BetsillAgency in earth system governance: refining a research agendaInt. Environ. Agreements, 11 (1) (2011), pp. 85-98, 10.1007/s10784-011-9147-9 View in Scopus Google Scholar
Ding et al., 2010Fei Ding, Yun Liu, Bo Shen, Xia-Meng SiAn evolutionary game theory model of binary opinion formationPhysica A, 389 (8) (2010), pp. 1745-1752, 10.1016/j.physa.2009.12.028 View PDF View article View in Scopus Google Scholar
Dobres and Robb, 2000Marcia-Anne Dobres, John E. RobbAgeny in ArchaeologyRoutledge, London and New York (2000)Google Scholar
Donges et al., 2018F. Donges Jonathan, Jobst Heitzig, Wolfram Barfuss, et al.Earth system modelling with complex dynamic human societies: the Copan:CORE World-Earth modeling frameworkEarth Syst. Dyn. Disc. (2018), 10.5194/esd-2017-126 Google Scholar
Donges et al., 2017aJ.F. Donges, W. Lucht, F. Müller-Hansen, W. SteffenThe Technosphere in Earth system analysis: a coevolutionary perspectiveAnthropocene Rev., 4 (1) (2017), pp. 23-33, 10.1177/2053019616676608 View in Scopus Google Scholar
Donges et al., 2017bJ.F. Donges, R. Winkelmann, W. Lucht, S.E. Cornell, J.G. Dyke, J. Rockström, J. Heitzig, H.J. SchellnhuberClosing the loop: reconnecting human dynamics to Earth system scienceAnthropocene Rev., 4 (2) (2017), pp. 151-157, 10.1177/2053019617725537 View in Scopus Google Scholar
Druckman and Jackson, 2009Angela Druckman, Tim JacksonThe carbon footprint of UK households 1990-2004: a socio-economically disaggregated, quasi-mulit-regional input-output modelEcol. Econ., 68 (2009), pp. 2066-2077View PDF View article View in Scopus Google Scholar
Edenhofer et al., 2014O. Edenhofer, R. Pichs-Madruga, Y. SokonaMitigation of Climate Change. Contribution of Work-ing Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate ChangeCambridge University Press, Cambridge, UK and New York, USA (2014)Google Scholar
Emirbayer and Goodwin, 1994Mustafa Emirbayer, Jeff GoodwinNetwork analysis, culture, and the problem of agencyAm. J. Sociol., 99 (6) (1994), pp. 1411-1454, 10.1086/230450 Google Scholar
Figueres et al., 2017Christiana Figueres, Hans Joachim Schellnhuber, Gail Whiteman, Johan Rockström, Anthony Hobley, Stefan RahmstorfThree years to safeguard our climateNature, 546 (7660) (2017), pp. 593-595Crossref View in Scopus Google Scholar
Figures, 2016Christina FiguresPlenary talk by Ch. Figueres, executive secretary of the United Nations Fremework convention on climate changePlenary Talk at the Adaptation Futures Conference, Rotterdam, 10–13 May 2016, Amsterdam (2016)https://www.youtube.com/watch?v=LOvDeIpVM8w Google Scholar
Filatova et al., 2013Tatiana Filatova, Peter H. Verburg, Dawn Cassandra Parker, Carol Ann StannardSpatial agent-based models for socio-ecological systems: challenges and prospectsEnvironmental Modelling & Software, Thematic Issue on Spatial Agent-Based Models for Socio-Ecological Systems, vol. 45 (2013), pp. 1-7, 10.1016/j.envsoft.2013.03.017(July)View PDF View article View in Scopus Google Scholar
Fischer-Kowalski, 1997M. Fischer-KowalskiSociety’s metabolism: on the childhood and adolescence of a rising conceptual starThe International Handbook of Environmental Sociology, Edward Elgar Publishing, Cheltenham, UK (1997)Google Scholar
Fuchs, 2001Stephan FuchsBeyond agencySociol. Theory, 19 (1) (2001), pp. 24-40, 10.1111/0735-2751.00126 View in Scopus Google Scholar
Galam, 2002S. GalamMinority opinion spreading in random geometryEur. Phys. J. B, 25 (4) (2002), pp. 403-406, 10.1140/epjb/e20020045 View in Scopus Google Scholar
García-López, 2018Gustavo A. García-LópezRethinking elite persistence in neoliberalism: foresters and techno-bureaucratic logics in Mexico’s community forestryWorld Dev. (2018), 10.1016/j.worlddev.2018.03.018MarchGoogle Scholar
Gatzweiler and Hagedorn, 2002F.w. Gatzweiler, K. HagedornThe evolution of institutions in transitionInt. J. Agric. Resour. Gov. Ecol., 2 (1) (2002), pp. 37-58, 10.1504/IJARGE.2002.000021 View in Scopus Google Scholar
General Directorate for Sustainable Development – SOeS, 2016General Directorate for Sustainable Development – SOeSKey Figures on Climate France and Worldwide(2016 edition) (2016)http://www.statistiques.developpement-durable.gouv.fr/fileadmin/documents/Produits_editoriaux/Publications/Reperes/2015/highlights-key-figures-climate-2016-edition.pdf Google Scholar
Giddens, 1984Anthony GiddensThe Constitution of Society. Outline of the Theory of StructurationUniversity of California Press, Berkley and Los Angeles (1984)Google Scholar
Granovetter, 1985Mark GranovetterEconomic action and social structure: the problem of embeddednessAm. J. Sociol., 91 (3) (1985), pp. 481-510, 10.1086/228311 Google Scholar
Granovetter, 1977Mark S. GranovetterThe strength of weak ties11this paper originated in discussions with Harrison White, to whom I am indebted for many suggestions and ideas. Earlier drafts were read by Ivan Chase, James Davis, William Michelson, Nancy Lee, Peter Rossi, Charles Tilly, and an anonymous referee; their criticisms resulted in significant improvementsSamuel Leinhardt (Ed.), Social Networks, Academic Press (1977), pp. 347-367, 10.1016/B978-0-12-442450-0.50025-0 View PDF View article View in Scopus Google Scholar
GTM, 2018GTMAnalysis: CO2 emissions at the World’s largest companies are risingMay 2, 2018(2018)https://www.greentechmedia.com/articles/read/analysis-firms-with-leading-climate-strategies-have-better-stock-performanc#gs.Y6dytuE Google Scholar
Hake et al., 2015Jürgen-Friedrich Hake, Wolfgang Fischer, Sandra Venghaus, Christoph WeckenbrockThe German Energiewende – history and status quoEnergy, Sustainable Development of Energy, Water and Environment Systems, vol. 92 (2015), pp. 532-546, 10.1016/j.energy.2015.04.027(December)View PDF View article View in Scopus Google Scholar
Hasegawa et al., 2015Tomoko Hasegawa, Shinichiro Fujimori, Kiyoshi Takahashi, Toshihiko MasuiScenarios for the risk of hunger in the twenty first century using shared socioeconomic pathwaysEnviron. Res. Lett., 10 (1) (2015)Google Scholar
Häyhä et al., 2016Tiina Häyhä, Paul L. Lucas, Detlef P. van Vuuren, Sarah E. Cornell, Holger HoffFrom planetary boundaries to national fair shares of the global safe operating space — how can the scales be bridged?Glob. Environ. Chang., 40 (September) (2016), pp. 60-72, 10.1016/j.gloenvcha.2016.06.008 View PDF View article View in Scopus Google Scholar
Herrmann-Pillath, 2018Carsten Herrmann-PillathThe case for a new discipline: technosphere scienceEcol. Econ., 149 (July) (2018), pp. 212-225, 10.1016/j.ecolecon.2018.03.024 View PDF View article View in Scopus Google Scholar
Hibbard et al., 2010Kathy Hibbard, Anthony Janetos, Detlef P. van Vuuren, Julia Pongratz, Steven K. Rose, Richard Betts, Martin Herold, Johannes J. FeddemaResearch priorities in land use and land-cover change for the earth system and integrated assessment modellingInt. J. Climatol., 30 (13) (2010), pp. 2118-2128, 10.1002/joc.2150 View in Scopus Google Scholar
Hoffmann and Hutter, 2012Stefan Hoffmann, Katharina HutterCarrotmob as a new form of ethical consumption. The nature of the concept and avenues for future researchJ. Consum. Policy, 35 (2) (2012), pp. 215-236, 10.1007/s10603-011-9185-2 View in Scopus Google Scholar
Holley and Liggett, 1975Richard A. Holley, Thomas M. LiggettErgodic theorems for weakly interacting infinite systems and the voter modelAnn. Probab., 3 (4) (1975), pp. 643-663Google Scholar
Holyst et al., 2001Janusz A. Holyst, Krzysztof Kacperski, Frank SchweitzerSocial impact models of opinion dynamicsAnnual Reviews of Computational Physics IX, vol. 9 (2001), pp. 253-273Annual Reviews of Computational Physics, Volume 9. WORLD SCIENTIFIChttps://doi.org/10.1142/9789812811578_0005 Crossref Google Scholar
Jaeger et al., 2001Carlo C. Jaeger, Ortwin Renn, Eugene A. Rosa, Thomas WeblerRisk, Uncertainty, and Rational ActionEarthscan Publications Ltd, London and Sterling (2001)Google Scholar
Kashwan, 2016Prakash KashwanInequality, democracy, and the environment: a cross-national analysisEcol. Econ. (131) (2016), pp. 139-151Google Scholar
Kashwan, 2017aPrakash KashwanDemocracy in the woodsEnvrionmental Conservation and Social Justice in India, Tanzania, and Mexico, Oxford University Press, New York (2017)Google Scholar
Kashwan, 2017bPrakash KashwanInequality, democracy, and the environment: a cross-national analysisEcol. Econ., 131 (2017), pp. 139-151View PDF View article View in Scopus Google Scholar
Kashwan et al., 2018Prakash Kashwan, Lauren M. MacLean, Gustavo A. García-LópezRethinking power and institutions in the shadows of neoliberalism: (an introduction to a special issue of world development)World Dev. (2018), 10.1016/j.worlddev.2018.05.026MayGoogle Scholar
King, 2009Anthony KingOvercoming structure and agency. Talcott Parsons, Ludwig Wittgenstein and the theory of social actionJ. Class. Sociol., 9 (2) (2009), pp. 260-288View in Scopus Google Scholar
Kluvankova-Oravska et al., 2009Tatiana Kluvankova-Oravska, Veronika Chobotova, Lenka Slavikova, Sonja TrifunovovaFrom government to governance for biodiversity: the perspective of central and eastern European transition countries – 0f3175391c2efbd10d000000.PdfEnviron. Policy Gov., 19 (2009), pp. 186-196Crossref View in Scopus Google Scholar
Kohring, 1996G.A. KohringIsing models of social impact: the role of cumulative advantageJ. Phys. I, 6 (2) (1996), pp. 301-308, 10.1051/jp1:1996150 View in Scopus Google Scholar
Kozyr-Kowalski, 1992Stanislaw Kozyr-KowalskiEconomic ownership and partial and combined classes of society. An attempt at a postive critique of post-Marxian MarxismOn Social Differentation. A Contribution to the Critique of Marxis Ideology, Adam Mickiewicz University Press, Poznan (1992)Google Scholar
Kümmel, 2011Reiner KümmelThe Second Law of Economics: Energy, Entropy, and the Origins of WealthSpringer Science & Business Media (2011)Google Scholar
McCollum et al., 2018David L. McCollum, Charlie Wilson, Michela Bevione, Samuel Carrara, Oreane Y. Edelenbosch, Johannes Emmerling, Céline Guivarch, et al.Interaction of consumer preferences and climate policies in the global transition to low-carbon vehiclesNat. Energy, 1 (2018), 10.1038/s41560-018-0195-zJulyGoogle Scholar
Nowak et al., 1990A. Nowak, J. Szamrej, B. LatanéFrom private attitude to public opinion: a dynamic theory of social impactPsychol. Rev., 97 (3) (1990), p. 362View in Scopus Google Scholar
Latane, 1981Bibb LataneThe psychology of social impactAm. Psychol., 36 (4) (1981), pp. 343-356Crossref View in Scopus Google Scholar
Lenton et al., 2008T.M. Lenton, H. Held, E. Kriegler, J.W. Hall, W. Lucht, S. Rahmstorf, H.J. SchellnhuberTipping elements in the earth’s climate systemProc. Natl. Acad. Sci., 105 (6) (2008), pp. 1786-1793, 10.1073/pnas.0705414105 View in Scopus Google Scholar
Lister, 2003Ruth ListerWhat is citizenship?Ruth Lister, Jo Campling (Eds.), Citizenship: Feminist Perspectives, Macmillan Education UK, London (2003), pp. 13-42, 10.1007/978-0-230-80253-7_2 Google Scholar
Lloyd, 1988C. LloydExplanation in Social HistoryBlackwell, Oxford (1988)Google Scholar
Lubchenco, 1998Jane LubchencoEntering the century of the environment: a new social contract for scienceScience, 279 (5350) (1998), pp. 491-497, 10.1126/science.279.5350.491 View in Scopus Google Scholar
Majid et al., 2014M.R. Majid, S.N. Moeinzadeh, H.Y. TifwaIncome-carbon footprint relationship for urban and rural households of Iskandar MalaysiaIOP Conf. Ser., 18 (012164) (2014), pp. 1-5Google Scholar
Martinez-Alier, 2009Joan Martinez-AlierSocial metabolism, ecological distribution conflicts, and languages of valuationCapital. Nat. Social., 20 (1) (2009), pp. 58-87Crossref View in Scopus Google Scholar
Moss et al., 2010Richard H. Moss, Jae A. Edmonds, Kathy A. Hibbard, Martin R. Manning, Steven K. Rose, et al.The next generation of scenarios for climate change research and assessmentNature, 463 (2010), pp. 747-756, 10.1038/nature08823 View in Scopus Google Scholar
Newman, 2018Mark NewmanNetworksOxford University Press (2018)Google Scholar
North, 1990Douglass C. NorthInstitutions, Institutional Change and Economic PerformanceCambridge University Press, Cambridge (1990)Google Scholar
North, 1991Douglass C. NorthInstitutionsJ. Econ. Perspect., 5 (1) (1991), pp. 97-112, 10.1257/jep.5.1.97 Google Scholar
Offe, 1992Clauss OffeThe welfare state, new class differentiation and diminution of social conflictOn Social Differentation. A Contribution to the Critique of Marxis Ideology, Adam Mickiewicz University Press, Poznan (1992), pp. 97-130Google Scholar
Ostrom, 2005Elinor OstromUnderstanding Institutional DiversityPrinceton University Press, Princenton (2005)Google Scholar
Otto et al., 2015Ilona M. Otto, Anne Biewald, Dim Coumou, Georg Feulner, Claudia Köhler, Thomas Nocke, Anders Blok, et al.Socio-economic data for global environmental change researchNat. Clim. Chang., 5 (2015), pp. 503-506Crossref View in Scopus Google Scholar
Otto et al., 2017Ilona M. Otto, Diana Reckien, Christopher P.O. Reyer, Rachel Marcus, Virginie Le Masson, Lindsey Jones, Andrew Norton, Olivia SerdecznySocial vulnerability to climate change: a review of concepts and evidenceReg. Environ. Chang. (2017), 10.1007/s10113-017-1105-9FebruaryGoogle Scholar
Otto et al., 2019Ilona M. Otto, Kyoung Mi Kim, Nika Dubrovsky, Wolfgang LuchtShift the focus from the super-poor to the super-richNat. Clim. Chang., 9 (2) (2019), pp. 82-84, 10.1038/s41558-019-0402-3 View in Scopus Google Scholar
Otto-Banaszak et al., 2011Ilona Otto-Banaszak, Piotr Matczak, Justus Wesseler, Frank WechsungDifferent perceptions of adaptation to climate change: a mental model approach applied to the evidence from expert interviewsReg. Environ. Chang., 11 (2) (2011), pp. 217-228, 10.1007/s10113-010-0144-2 View in Scopus Google Scholar
Oxfam, 2015OxfamExtreme carbon inequalityOxfam Media Briefing (2015)https://www.oxfam.org/en/research/extreme-carbon-inequality Google Scholar
Paris, 2013Chris ParisThe homes of the super-rich: multiple residences, hyper-mobility and decoupling of prime residential housing in global citiesGeographes of the Super-rich, Edward Elgar, Cheltenham, UK; Northhampton, MA, USA (2013)Google Scholar
Pattberg and Stripple, 2008Philipp Pattberg, Johannes StrippleBeyond the public and private divide: remapping transnational climate governance in the 21st centuryInt. Environ. Agreements, 8 (4) (2008), pp. 367-388, 10.1007/s10784-008-9085-3 View in Scopus Google Scholar
Penn, 2009Elizabeth Maggie PennA model of farsighted votingAm. J. Polit. Sci., 53 (1) (2009), pp. 36-54, 10.1111/j.1540-5907.2008.00356.x View in Scopus Google Scholar
Pérez et al., 2016Irene Pérez, Marco A. Janssen, John M. AnderiesFood security in the face of climate change: adaptive capacity of small-scale social-ecological systems to environmental variabilityGlob. Environ. Chang., 40 (September) (2016), pp. 82-91, 10.1016/j.gloenvcha.2016.07.005 View PDF View article View in Scopus Google Scholar
Popp et al., 2010Alexander Popp, Hermann Lotze-Campen, Benjamin BodirskyFood consumption, diet shifts and associated non-CO2 greenhouse gases from agricultural productionGlob. Environ. Chang., 20 (3) (2010), pp. 451-462, 10.1016/j.gloenvcha.2010.02.001 View PDF View article View in Scopus Google Scholar
Riahi et al., 2017Keywan Riahi, Detlef P. van Vuuren, Elmar Kriegler, Brian C.O. Jae Edmonds, Neill Shinichiro Fujimori, Nico Bauer, et al.The shared socioeconomic pathways and their energy, land use, and greenhouse gas emissions implications: an overviewGlob. Environ. Chang., 42 (January) (2017), pp. 153-168, 10.1016/j.gloenvcha.2016.05.009 View PDF View article View in Scopus Google Scholar
Schellnhuber, 1999H.J. SchellnhuberEarth system’ analysis and the second Copernican revolutionNature, 402 (6761supp) (1999), pp. C19-C23, 10.1038/35011515 View in Scopus Google Scholar
Scrieciu, 2007S.Serban. ScrieciuThe inherent dangers of using computable general equilibrium models as a single integrated modelling framework for sustainability impact assessment. A critical note on Böhringer and Löschel (2006)Ecol. Econ., 60 (4) (2007), pp. 678-684, 10.1016/j.ecolecon.2006.09.012 View PDF View article View in Scopus Google Scholar
Simmel, 1971Georg SimmelOn Individuality and Social Forms. Selected WritingsThe University of Chicago Press, Chicago and London (1971)Google Scholar
Spaargaren, 1997Gert SpaargarenThe Ecological Modernization of Production and Consumption. Essays in Environmental SociologyUniverstiy of Wageningen, Wageningen (1997)Google Scholar
Stark, 1996David StarkRecombinant property in east European capitalismAm. J. Sociol., 101 (4) (1996), pp. 993-1027, 10.1086/230786 View in Scopus Google Scholar
Steffen et al., 2018Will Steffen, Johan Rockström, Katherine Richardson, Timothy M. Lenton, Carl Folke, Diana Liverman, Colin P. Summerhayes, et al.Trajectories of the earth system in the AnthropoceneProc. Natl. Acad. Sci. (2018), 10.1073/pnas.1810141115August, 201810141Google Scholar
Suwarno et al., 2018Aritta Suwarno, Meine van Noordwijk, Hans-Peter Weikard, Desi SuyamtoIndonesia’s forest conversion moratorium assessed with an agent-based model of land-use change and ecosystem services (LUCES)Mitig. Adapt. Strateg. Glob. Chang., 23 (2) (2018), pp. 211-229, 10.1007/s11027-016-9721-0 View in Scopus Google Scholar
Sznajd-Weron, 2005Katarzyna Sznajd-WeronSznajd Model and Its ApplicationsArXiv:Physics/0503239, March(2005)http://arxiv.org/abs/physics/0503239 Google Scholar
Sztompka, 1991Piotr SztompkaSociety in Action. The Theory of Social BecomingUniversity of Chicago Press, Chicago (1991)Google Scholar
Sztompka, 1994Piotr SztompkaEvolving focus on human agency in contemporary social theoryAgency and Structure. Reorienting Social Theory, Gordon and Breach Science Publishers, Yverdon, Camberwell, Paris (1994)Google Scholar
Sztompka, 2002Piotr SztompkaSocjologia: Analiza SpoleczenstwaZnak, Krakow (2002)Google Scholar
Troost and Berger, 2015Christian Troost, Thomas BergerDealing with uncertainty in agent-based simulation: farm-level modeling of adaptation to climate change in southwest GermanyAm. J. Agric. Econ., 97 (3) (2015), pp. 833-854, 10.1093/ajae/aau076 View in Scopus Google Scholar
Union of Concerned Scientists, 2018Union of Concerned ScientistsLargest producers of industrial carbon emissionsMay 2, 2018(2018)https://www.ucsusa.org/global-warming/science-and-impacts/science/largest-producers-industrial-carbon-emissions.html#.Wumo_n–mUl Google Scholar
United Nations, 2015aUnited NationsResolution Adopted by the General Assemly on 25 September 2015. Transforming Our World: The 2030 Agenda for Sustainable Development(2015)http://www.un.org/ga/search/view_doc.asp?symbol=A/RES/70/1&Lang=E Google Scholar
United Nations, 2015bUnited NationsUN Sustainable Development GoalsUnited Nations (2015)http://www.un.org/sustainabledevelopment/sustainable-development-goals/#Google Scholar
van den Berg et al., 2019Nicole J. van den Berg, Heleen L. van Soest, Andries F. Hof, Michel G.J. den Elzen, Detlef P. van Vuuren, Wenying Chen, Laurent Drouet, et al.Implications of Various Effort-sharing Approaches for National Carbon Budgets and Emission PathwaysClimatic Change, February (2019), 10.1007/s10584-019-02368-y Google Scholar
van Vuuren et al., 2011Detlef P. van Vuuren, Jason Lowe, Elke Stehfest, Laila Gohar, Andries F. Hof, Chris Hope, Rachel Warren, Malte Meinshausen, Gian-Kasper PlattnerHow well do integrated assessment models simulate climate change?Clim. Chang., 104 (2) (2011), pp. 255-285, 10.1007/s10584-009-9764-2 View in Scopus Google Scholar
van Vuuren et al., 2012Detlef P. van Vuuren, Laura Batlle Bayer, Clifford Chuwah, Laurens Ganzeveld, Wilco Hazeleger, Bart van der Hurk, Twan van Noije, Brian O’Neil, Bart J. StrengersA comprehensive view on climate change: coupling of earth system and integrated assessment modelsEnviron. Res. Lett., 7 (2012), pp. 1-10Crossref View in Scopus Google Scholar
van Vuuren et al., 2018Detlef P. van Vuuren, Elke Stehfest, David E.H.J. Gernaat, Maarten van den Berg, David L. Bijl, Harmen Sytze de Boer, Vassilis Daioglou, et al.Alternative pathways to the 1.5 °C target reduce the need for negative emission technologiesNat. Clim. Chang., 8 (5) (2018), pp. 391-397, 10.1038/s41558-018-0119-8 View at publisher View in Scopus Google Scholar
Velders et al., 2007Guus J.M. Velders, Stephen O. Andersen, John S. Daniel, David W. Fahey, Mack McFarlandThe importance of the Montreal protocol in protecting climateProc. Natl. Acad. Sci., 104 (12) (2007), pp. 4814-4819, 10.1073/pnas.0610328104 View at publisher View in Scopus Google Scholar
Williamson, 1996Oliver E. WilliamsonThe Mechanisms of Governance, Oxford University Press (1996)Google Scholar
Williamson, 1998Oliver E. WilliamsonTransaction costs economics: how it works; where it is headedDe Economist, 146 (1) (1998), pp. 23-58View in Scopus Google Scholar
Williamson, 2000Oliver E. WilliamsonThe new institutional economics: taking stock, looking ahead