Earth Syst. Dynam., 12, 671–687, 2021 https://doi.org/10.5194/esd-12-671-2021 © Author(s) 2021. This work is distributed under the Creative Commons Attribution 4.0 License. Earth system economics: a biophysical approach to the human component of the Earth system Eric D. Galbraith1,2,3 1 Department of Earth and Planetary Science, McGill University, Montréal, Canada 2 Institut de Ciència i Tecnologia Ambientals (ICTA-UAB), Universitat Autònoma de Barcelona, Barcelona, Spain 3 ICREA, Barcelona, Spain Correspondence: Eric D. Galbraith (eric.galbraith@mcgill.ca) Received: 27 August 2020 – Discussion started: 4 September 2020 Revised: 5 April 2021 – Accepted: 8 April 2021 – Published: 27 May 2021 Abstract. The study of humans has largely been carried out in isolation from the study of the non-human Earth system. This isolation has encouraged the development of incompatible philosophical, aspirational, and method- ological approaches that have proven very difficult to integrate with those used for the non-human remainder of the Earth system. Here, an approach is laid out for the scientific study of the global human system that is intended to facilitate seamless integration with non-human processes by striving for a consistent physical basis, for which the name Earth system economics is proposed. The approach is typified by a foundation on state variables, cen- tral among which is the allocation of time amongst activities by human populations, and an orientation towards considering human experience. A framework is elaborated which parses the Earth system into six classes of state variables, including a neural structure class that underpins many essential features of humanity. A working example of the framework is then illustrated with a simple numerical model, considering a global population that is engaged in one of two waking activities: provisioning food or doing something else. The two activities are differentiated by their motivational factors, outcomes on state variables, and associated subjective experience. While the illustrative model is a gross simplification of reality, the results suggest how neural characteristics and subjective experience can emerge from model dynamics. The approach is intended to provide a flexible and widely applicable strategy for understanding the human–Earth system, appropriate for physically based assess- ments of the past and present, as well as contributing to long-term model projections that are naturally oriented towards improving human well-being. 1 Introduction entirely new approaches – such as global carbon cycle sci- ence (Falkowski et al., 2000) – and, in return, has brought Over the past 4 decades, Earth system science has devel- fresh thinking to the previously isolated disciplines from oped a rich understanding of interactions between the myriad which it was born. physical, chemical, and biological components of our planet Yet, despite being motivated by the human impacts on (Steffen et al., 2020). By considering the Earth as a single the planet, Earth system science has done relatively lit- system, which is itself comprised of a hierarchy of mechanis- tle to directly incorporate humans themselves (Motesharrei tically interacting subsystems, Earth system science has fa- et al., 2014; Donges et al., 2017; Calvin and Bond-Lamberty, cilitated the challenge of thinking across vast scales of space 2018). For example, although the seminal textbook by Kump and time and contextualized global change within the long- et al. (2004) discusses human impacts on the planet at length, term evolution of life (Lenton et al., 2011). In its quest to there is no mention of human demographics, societal dynam- understand planetary functioning, this new science has suc- ics, or well-being. Instead, the impacts of the human system ceeded in crossing many disciplinary boundaries, developing Published by Copernicus Publications on behalf of the European Geosciences Union. 672 E. D. Galbraith: Earth system economics are viewed as external forcings on the non-human Earth. This ing knowledge systems, social behaviour, and experience – to exclusion is particularly clear when considering Earth sys- physical embodiments. This may reflect the historical devel- tem models (ESMs), the numerical flagships of Earth sys- opment of social sciences and humanities, originating as they tem science. ESMs encapsulate the current understanding of did when virtually all people believed in an eternal, disem- the planet by representing the component systems in a sim- bodied soul (McDonald, 1993). Thus, although many modern plified fashion, integrated within a seamless framework and social scientists probably do not subscribe to this belief, the discretized on a global grid. Because all component systems underlying conceptual frameworks and approaches remain co-exist within the same spatial framework and because they aligned with its tacit implications, and many core features of are based on common foundations of biology, chemistry, and social science, such as values, beliefs, and norms, continue to physics, the means of exchange between the component sys- depart from non-physical starting points (Bouchaud, 2008). tems are obvious, so that they can be integrated as a whole Differences in these non-physical starting points have led, in to provide a synoptic global view. But ESMs do not include turn, to a plethora of fields of human study, among which the global human system within the same common founda- there is little common ground, hampering interdisciplinary tions, and, as a result, the synoptic perspective of Earth sys- progress. tem science typically fails to include its most rapidly chang- Yet, like all living organisms, humans are physical beings. ing and disruptive component (Mote et al., 2020). This is The biological reality of human bodies embeds us within not to say there are no efforts in this direction; for exam- ecosystems and links us to biogeochemical cycles through ple macroeconomic models are frequently run in parallel our food production, material fluxes, and waste flows (Haberl with ESMs while exchanging information on greenhouse gas et al., 2019). The fact that each of us can be only in one emissions and increasingly sophisticated land use changes place at a time and engage in a limited number of activi- (Rounsevell et al., 2014), an approach known as integrated ties per day places fundamental physical constraints on our assessment modeling. But this approach is targeted primar- economies (Becker, 1965). In addition, advances in neuro- ily toward weighing 21st-century climate change impacts vs. science now provide rich and compelling evidence that ev- greenhouse gas mitigation strategies (Nordhaus and Yang, erything that once appeared to be attributable to a disem- 1996; Van Vuuren et al., 2007; Anderson, 2019) rather than bodied soul is actually formed “by the meat”, i.e. as emer- to provide fundamental insight into the global human sys- gent properties of our brains (Clark, 2015). The intricate net- tem. “Sectoral” models often resolve geographically explicit work of synapses in each of our heads determines what we interactions between humans and specific Earth system com- think, how we feel, and who we are (LeDoux, 2003). These ponents, such as the agricultural system (Rosenzweig et al., synapses are continually changing as we go through our daily 2014) or global marine fishery (Galbraith et al., 2017), but experiences, at rates that are biologically constrained (As- the fragmented sectoral approach does not naturally build to- coli, 2015). Thus, just as knowledge of the molecular pro- ward what Voinov and Shugart (2013) call an “integral” per- cesses occurring within leaves can help to predict aspects of spective on the human Earth. the global terrestrial ecosystem (Stocker et al., 2020), there Why has there not been more effort devoted to the human is good reason to hope that many aspects of humanity, histor- component of Earth system science? I suggest three reasons ically considered unquantifiable, can actually be better un- (though there are certainly more). First, biologically iden- derstood by considering how they emerge from the physical tical humans have interacted with the Earth system in very constructs of synaptic networks. Neuroscience still has much different ways when living under different social and tech- to learn about the functioning of the brain, but it holds great nological contexts. For example, a hunter–gatherer society promise as a common ground to help unify the fragmented has extremely different per capita impacts on the Earth sys- domains of social science (Boyer, 2018). tem than does a 21st-century urbanized society. Thus, assum- The lacklustre development of the human component of ing a fixed set of functional characteristics for our species – Earth system science is also evident in its failure to enrich a strategy that works quite well for other organisms within the scientific understanding of humans themselves. This is in the Earth system – fails to address the most remarkable fea- contrast to integrative Earth system approaches such as ocean tures of humanity. Second, we care a lot about what humans biogeochemistry, which has provided important insights into think and how they feel, which can make scientists hesitate to marine ecology (e.g. Follows et al., 2007). Early work un- simplify features of humanity in the way frequently done for der the name of human ecology made significant progress other components of the Earth system. Third, there is a vast toward modeling hunter–gatherers through their interactions cultural gulf between natural and social sciences that is very with the environment (e.g. Winterhalder, 1993), providing difficult to bridge due to profoundly incompatible literatures. valuable insights for anthropology and sociology, but these This gulf has left each culture largely ignorant of the other, works were not widely seized upon. In contrast, there have a problem that was identified decades ago (Snow, 1959) and been very few Earth-system-scale studies that ask fundamen- continues to persist. tal questions about the physical coupling of humans with the At the root of the natural–social science divide lies the dif- ecosystem (Motesharrei et al., 2014) and even fewer that ex- ficulty of linking the essential features of humanity – includ- plore the implications for the quality of human existence. Earth Syst. Dynam., 12, 671–687, 2021 https://doi.org/10.5194/esd-12-671-2021 E. D. Galbraith: Earth system economics 673 Figure 1. ESE provides a bridge between the Earth system science approach, typified by Earth system models, and the diverse fields of human study. Yet human well-being is of central importance to social sci- to physical constraints, with the ability to resolve high ence and policymakers and could either improve or deteri- spatial resolution; orate dramatically in future, depending on societal choices (Barrington-Leigh and Galbraith, 2019). 2. suggesting new non-monetary metrics that capture key Here a strategy is pursued to provide a seamless integra- aspects of the global human system; despite long- tion of human and non-human parts by representing humans standing calls for moving beyond gross domestic – like the rest of the Earth system – on a biophysical founda- product (GDP) as the dominant measure of progress tion. The strategy aims to facilitate new forms of communi- (Van den Bergh, 2009; Costanza et al., 2014; Das- cation at the juncture of natural and social sciences, with the gupta, 2021), proposed alternatives are often converted aspiration of providing new insights for both human and non- to monetary units or subject to criticisms of arbitrari- human systems (Fig. 1). The name Earth system economics ness. By focusing on physical quantities and their rates (ESE) is proposed for the endeavour, though as discussed be- of change, ESE can offer an alternative set of objective, low, it differs from mainstream economics in a number of im- quantitative metrics to inform the human–Earth status; portant ways. Section 2 gives a few examples of the types of 3. testing hypotheses in historical dynamics (Turchin, problems that could be addressed with ESE. Section 3 pro- 2018); in the same way that Earth system models can be vides an overview of the key guiding principles that motivate used to test hypotheses about past climate changes, ESE the ESE approach. Section 4 details a high-level conceptual models could be applied to test hypotheses about past framework for the global human system. Section 5 describes changes in the human system – for example, what emer- a numerical model of the global human system, inspired by gent features are required to accurately hindcast the spa- simple models of the global carbon cycle (e.g. Sarmiento and tial progressions of key societal transitions in history, Toggweiler, 1984), as an illustration of how the framework such as the Neolithic transitions or industrialization? can be operationalized. Section 6 provides analysis and dis- cussion of the model. Section 7 offers concluding comments. 4. the spatial and temporal dynamics of human interac- tions with ecosystems and consequences for biodiver- sity; tight coupling with physically based biodiversity 2 Topics that could be readily addressed with Earth models can provide new tools with which to test hy- system economics potheses regarding early mass extinctions or the con- trols on future threats to ecosystem stability in a spa- As ESE pursues a new approach, it is difficult to foresee all tially explicit context; applications that may arise from its development. Some of 5. mechanistic linkages between subjective well-being and the more obvious applications might include the following: the biophysical consequences of societal actions; how could human lived experience vary given different soci- 1. gaining a mechanistic, bird’s eye perspective on the etal pathways and within physical constraints, including global human system that allows for seamless analysis coupled Earth system impacts such as climate change across scales, essentially capturing the human system as and biodiversity loss? an integrated part of global material cycles and subject https://doi.org/10.5194/esd-12-671-2021 Earth Syst. Dynam., 12, 671–687, 2021 674 E. D. Galbraith: Earth system economics Most of these complex problems have been addressed by 3.2 Quantification of activities other means, especially at local scales, but all remain incom- pletely resolved. The ESE approach aims to provide a novel The diversity of human endeavours can be overwhelming and global, integrated view, while prompting new avenues for might appear to defy a recourse to conserved quantities in mechanistic insight. The approach may ultimately be more the way that the motion of fluids is linked to momentum and widely applicable than indicated by this short list. density through the Navier–Stokes equations. However, there is no question that the amount of time available to each hu- man is a strictly conserved quantity. All humans engage in 3 Guiding principles of Earth system economics some form of activity for exactly 24 h d−1 . The activities in which a population is engaged determine its impact on the In a nutshell, ESE aims to quantify physical aspects of the biophysical reality and also play a major role in determining human–Earth system (state variables), including how the the subjective experience of its individuals. Thus, activities physical state is dynamically altered by human actions (time are employed here as the central feature of ESE. allocation) and consequences for the nature of human expe- There does not exist a universal system for classifying ac- rience. In this section, a few general principles are discussed, tivities. Even the activity of a reader of a scientific article as motivation for the framework which follows. can be described in many ways, which may include reading, working, thinking, learning, sitting, or using a screen or com- 3.1 Striving for physical foundations puter. The activity may be subjectively enjoyable or unpleas- Foremost, ESE strives for a grounding in quantifiable, phys- ant, depending on the quality of the text and disposition of ical terms. Physical variables exhibit persistence over time, the reader. The optimal strategy to classify activities would and physical processes impose firm limits on possible rates of involve as little subjective interpretation as possible and be change, leading to dynamic predictability. Physical variables grounded as firmly as possible in physical features, a pos- also lend themselves to strict definitions, which can prevent sibility that could be further developed elsewhere. For the double counting while simultaneously helping to ensure in- moment, it is sufficient to consider this a difficult and incom- clusivity. Much of the predictive success of natural sciences pletely resolved problem. lie in their ultimate recourse to physical variables, which pro- In the absence of a universal lexicon of activities, applica- vide pathways to diverse insights whether starting from biol- ble to all humans at all times, a lexicon must be constructed ogy, physics, or chemistry. For example, the conservation of for a particular purpose. An activity lexicon must identify, mass, momentum, and energy play essential roles in many as unambiguously as possible, a set of mutually exclusive branches of Earth system science, from atmospheric circu- activities that together include all possible activities avail- lation to ice sheet motion and sea level rise. Ecosystem and able to the population. Thus, the fractional distribution of biogeochemical models benefit from the understandings of time between the activities must sum to exactly 1. For ex- living processes as molecular interactions writ large. ample, a simple two-activity lexicon would be sleeping and That said, the aim to ground the human system in phys- not-sleeping. To be useful, the lexicon should align activities ical quantities is not trivial. For some things – population with the outcomes that motivate them, by considering how demographics, cars, infrastructure, fossil fuel consumption they modify state variables. – it is straightforward. But for many of the most fascinat- ing and important aspects – such as behavioural motivations 3.3 Subjective experience and subjective experience – the biophysical bases remain vaguely described. Operationally, it will always be neces- Humans live rich inner lives, and individuals can be either sary to use coarse approximations for these (and other) un- filled with joy or tormented by suffering, depending on what resolved or poorly understood processes, a common strategy circumstances befall them. Improving the inner life experi- in Earth system models, such as the representation of cloud ence of humans has pre-occupied much of society for gener- physics with empirical parameterizations. These parameteri- ations and remains a central goal of global society, as exem- zations are always unsatisfying, but the fact that they are ex- plified by the UN Sustainable Development Goals: 11 of the plicitly recognized as unsatisfactory and can ultimately be 17 goals are oriented towards improving the life experience replaced by more physically grounded mechanistic under- of humans, while only six are oriented toward maintaining standings identifies a direction for progress. Resolutely ab- non-human aspects of the planet. Given that subjective expe- stract variables, on the other hand, resist connection to com- rience appears to be a top priority for most of humanity, it plementary scientific insights and reinforce disciplinary si- is explicitly included as an essential component of the ESE los. Thus, the important thing is that strengthening the phys- approach. ical foundation is ever present as a central goal of ESE: that Despite its importance, the biophysical understanding of long-term progress can be made by improving the physical subjective experience remains rudimentary (e.g. Alexander representation of all aspects of the human system through et al., 2020). It will take many years of additional research improved observations and theoretical development. before quantifications are available to assess human experi- Earth Syst. Dynam., 12, 671–687, 2021 https://doi.org/10.5194/esd-12-671-2021 E. D. Galbraith: Earth system economics 675 ence that rival, for example, our ability to quantify the con- Greek oikonomia, referred to managing the home in a ratio- centrations of trace gases in the atmosphere. Nonetheless, the nal way in order to benefit its occupants (Leshem, 2016). The field of subjective well-being has made great strides in pro- root oiko is also the basis of “ecology”, study of the home. viding large datasets on how people themselves evaluate their The aim of the current proposal is to provide an additional life experiences (Diener et al., 2018). These can be consid- means for holistic, science-based perspectives to assist in ra- ered along two axes: tional decision-making that can improve the management of the wealth of our common home, the Earth system, for the 1. Affect: this refers to the momentary emotions felt benefit of its inhabitants. Hence, the usage here is consistent throughout the day, sometimes assessed by asking a with the original Greek term. Nonetheless, it should be born subject whether they felt positive or negative emotions in mind that ESE is only distantly related to mainstream eco- (e.g. laughed, cried, felt angry) over some preceding nomics. time interval (Csikszentmihalyi and Larson, 2014) or by asking a candidate to rank the pleasantness (Gershuny and Sullivan, 2019) or unpleasantness (Kahneman and 3.5 Focus on population-level interactions Krueger, 2006) of different activities. ESE focuses on humans at the population level as the pri- mary interactive unit. Of course, human behaviours and ex- 2. Cognitive life evaluation and eudaemonia: these reflect periences all actually happen at the individual level. But just time-integrated rather than momentary aspects of well- as the dynamics of a fluid can be usefully described with- being. For cognitive life evaluations, the subject is asked out resolving the motions of individual molecules within it, to consider their life as a whole and evaluate their level population characteristics can be usefully described without of satisfaction with it, usually on a 10-point scale. The resolving individual interactions, and symmetry breaking can results are often correlated reasonably well with af- lead to fundamentally different behaviour across scales (An- fect and can be predicted to some degree from mate- derson, 1972). What is more, these population characteristics rial and non-material variables (Helliwell et al., 2012; can show greater predictability when the emergent result de- Barrington-Leigh and Galbraith, 2019). The term eu- pends on well-behaved statistical distributions of individual daemonia refers to a fulfillment of purpose and is often behaviour, as illustrated by the dynamics of human mobility oriented towards philosophical goals of what life ought (Simini et al., 2012; Alessandretti et al., 2020). to be rather than one that is desirable on purely hedonic Focusing on the population level does not mean that vari- terms (Ryan and Deci, 2001). Although a major con- ability within the population must be ignored. Variability can cern of society on historical timescales, often addressed be incorporated as additional information that describes the through religion, eudaemonia has been less studied in variability in a parameter, such as a probability distribution recent years, with less effort dedicated to developing function. For example, the distribution of wealth within a quantitative indices. population can often be approximated as a power law, for These axes of subjective well-being do not capture all that which only a single parameter (the exponent) needs to be de- is important to human experience, and the difficulty of com- fined (Wold and Whittle, 1957). paring assessments between cultures and languages cannot be taken lightly. But it appears likely that the quantitative ba- 3.6 Emphasis on the ultimate “what” and “why” of sis for constructing population-level assessments of life ex- activities rather than the “how” perience will continue to improve as time progresses. A great deal of human study is oriented towards understand- ing how social activities are coordinated and the means by 3.4 Drawing on all fields of human-related science which the cooperative activities of many individuals can Many disciplines study humans, including the core social sci- be optimized. The mechanisms by which this coordination ences of economics, anthropology, sociology, and psychol- occurs underpin many fascinating aspects of culture, eco- ogy, as well as history, medicine, law, business, and educa- nomics, management, and law, but are not the target of en- tion. All of these disciplines can provide useful insights on quiry here. the global human system. For this reason, ESE aspires to es- Instead, ESE is characterized by a focus on the what and tablish common ground that is compatible with aspects of all why of human activities. Here, “what” refers to the final net fields of human study, by explicitly considering the physical outcome of an activity, or complex of cooperative activities, foundations that underly them. in physical terms. The “why” refers to the ultimate motiva- So why use the term “economics”? In its modern use, tions for undertaking the activity, again in relation to the fi- this term has become narrowly associated with the distri- nal net outcome in the case of a complex of coordinated ac- bution of scarce resources, the production and consumption tivities. For example, the final outcome of creating farming of goods and services by firms and households, and mone- tools, tilling soil, sowing seeds, tending plants, and harvest- tary exchanges. However, the origin of the word, from the ing crops is to provide food (what). The motivation for this https://doi.org/10.5194/esd-12-671-2021 Earth Syst. Dynam., 12, 671–687, 2021 676 E. D. Galbraith: Earth system economics is a hunger-driven need for food (why). Thus, ESE aims to can be done through countless ways, none of which is per- circumvent the complexity of immediate, individual moti- fect but some of which can be useful. And, as written in a vations for component activities (such as whether the work discussion of Box’s aphorism by Truran (2013), “it may be is done for pay, which could itself be motivated by material necessary to create a model that takes a totally different per- consumption, which itself could be motivated by a desire to spective in order to improve upon currently accepted mod- raise social standing) by considering the net outcome of any els.” set of activities as the relevant motivating factor. Here, a new perspective on the human system is proposed that is consistent with the ESE principles outlined in Sect. 3 3.7 Applicability to any point in time and that forms an intuitive and inclusive structure that aligns well with observational data. To be tractable at the global It could be easier to design a conceptual system exclusively scale, the framework definition is hierarchical, so that it can for the present day, with which we are intimately familiar, be used at a high level of aggregation. The framework is in- than one which works equally well back to medieval times clusive, encapsulating the entirety of the global human sys- or the Late Pleistocene. Yet, if we aspire to consider the tem, while aiming to facilitate the representation of its mech- distant future, many decades or centuries hence, this abil- anistic properties. At the same time, the categories are con- ity must be a bare minimum requirement, since presumably ceptually straightforward to expand into disaggregated de- the future could hold many revolutionary changes that defy tail, with as little ambiguity as possible, and spatial disaggre- the imagination today. The ESE approach strives to be ap- gation should be easy to apply. This proposed framework is plicable across the full temporal scope of human existence, intended as a superstructure within which analyses or models enabling hindcasts to test dynamical hypotheses against his- could be developed through further work. torical observations as well as to explore hypothetical future projections. 4.1 State variables 3.8 Focus on emergent consequences of predictable The ESE framework is defined by state variables. Each state aspects variable represents some physical aspect of the human–Earth system, living or non-living. Each variable could be mea- Most aspects of complex systems, including the human– sured and quantified over some spatial and temporal domain Earth system, are unpredictable. But within this sea of un- (at least in theory, even if impractical or impossible with cur- predictability lie islands of predictability. For example, the rent technology) and is subject to physical constraints. chaotic processes that determine daily weather can be ap- The highest-level grouping of six variable classes, pro- proximated well enough to provide a very detailed forecast posed here, is illustrated in Fig. 2 and elaborated below. The over the next 12 h but are almost completely unpredictable five variable classes in the outer ring include everything on on a timescale of 1 month. Yet, on a coarser scale, seasonal the surface of the Earth and can therefore be thought of as and even decadal climate forecasts are now reasonably good a conceptual superstructure within which more detailed sub- (Smith et al., 2019). Similarly, societal dynamics include a divisions can be resolved. The final state variable class, time vast variety of interacting, nonlinear processes that are ex- allocation, is not physically embodied but is nonetheless sub- tremely challenging to predict but within which occur more ject to the limitation of 24 h d−1 and is unambiguously de- predictable aspects. Thus, ESE strives to identify the more fined for a population within a given spatial–temporal do- robust, least unpredictable aspects of the system, seeking main. insights on the emergent results of their interactions. Soci- Soma. This refers to the living ensemble of human bodies etal, cultural, and economic characteristics of populations are and their biophysical characteristics, including microbiota. described through the simplifying lens of how they impact The Soma determines the biogeochemical fluxes required to physical variables and time allocation. The roles of the more maintain the population, including food and water consump- unpredictable aspects can then be assessed through the quan- tion as well as the production of heat and waste. It also in- tification of structural and parameter uncertainty, the use of cludes properties reflecting the health status of the population probability distributions, and the inclusion of tipping points (including symbiotic and pathogenic microbes) and physical if they are identified through other means. fitness. Example state variables here could include the to- tal population biomass (kg), an age-structured population de- 4 Earth system economics conceptual framework scription (number and age), or detailed information on body compositions (e.g. C : N ratio, Fe content). Humans have an intellectual ability to foresee the future that Neural structure. It is because of the dynamical processes is unparalleled amongst other forms of life and an apparently in our brains that we are the dominant species on the planet. infinite scope to modify their biophysical surroundings. How Our neurons encode networks that are highly plastic, and this could these features possibly be captured in a numerical as- plasticity forms the foundation of our ability to learn (As- sessment? To paraphrase George Box, the answer is that it coli, 2015) as well as our responses to stimuli (Lindquist Earth Syst. Dynam., 12, 671–687, 2021 https://doi.org/10.5194/esd-12-671-2021 E. D. Galbraith: Earth system economics 677 Neural activation. Existing neural structures are activated by sensory stimulus and result in what we experience as thoughts, emotions, and feelings. The sensory stimulus in- cludes external factors such as the landscape, music, food, mobile phone screens, and conversation, as well as interocep- tive body status such as hunger and thermal comfort, and in fact both types of sources generally co-occur (Barrett, 2017). This class of state variables represents manifestations of this neural activation. Because the details of the activation it- self remain difficult to observe, emergent properties such as subjective well-being measures are most usable at present, though observation technologies are rapidly improving. Neu- ral activation is also conceptually useful as the pathway by which neural structure is modified. Things. Humans are clever, but it is not through individual cleverness alone that we have become the dominant species on the planet (Henrich, 2017). Rather, we leverage our abil- ity to think by creating entities with novel properties, con- structed through shared knowledge and social coordination, that then amplify our ability to modify the physical environ- ment. This includes the fabrication of tools, the construction Figure 2. The superstructure of the conceptual framework. Each of of buildings and infrastructure, the making of vehicles and the six shaded regions corresponds to a state variable class. Vari- airplanes, and the writing of books and computer code. The ables in the blue regions relate to the persistent structure and instan- Things class includes all of these and is defined as all non- taneous activation of human neural systems, the orange region to living entities which are brought into existence as a desired human bodies, the brown region to human-created things, the grey to the distribution of time amongst activities, and the green to the outcome of human activity. As such, the Things class does remainder of the Earth system. not include livestock or genetically modified organisms, nor does it include waste. Instead, these are regarded as modifi- cations of the remainder of the Earth system. et al., 2012). The biophysical characteristics of our brains Remainder of the Earth system. This includes all living lie at the foundation of core societal traits such as knowledge organisms other than humans (including agricultural plants and behaviours, as well as subjective experience (Lindquist and livestock), the atmosphere, regolith, soil, rock, the ocean, et al., 2012; Boyer, 2018). Thus, state variables describing and the cryosphere. These fall within the traditional domain the brains of humans within a population can be used to of Earth system science. Although the variables within this represent these essential features. One type of state variable class can all be affected by the human system, and many may could quantify aspects of associative links within the popu- be very strongly modified (e.g. cows, grapefruit), they do not lation connectome (the ensemble of all synaptic connections require human activity in order to persist and/or are living in a population, sensu Sporns et al., 2005), such as the num- organisms, thereby differentiating them from Things. ber of associations encoded by synapses. Other possibilities Time allocation. The allocation of time between activities would be topological descriptions of the neural structure, e.g. is a complex topic, which has been studied in many branches the degree of diversity of associations within the popula- of social sciences (see Gershuny and Sullivan, 2019, for a tion, the quality of the predictive capacity of associations, useful overview). In a simple form, the allocation of time can or links with hormonal and emotional responses. Although be regarded as the emergent outcome of competing motiva- the processes underlying the formation of new synapses and tions, expressed at the population level. As discussed above, their selective destructions remain incompletely understood, a motivation is strictly defined as the reason to undertake an they are certain to happen at finite, biologically constrained activity (the why) that relates to the set of physical outcomes rates, placing limits on possible rates of learning and mod- caused directly by the activity (the what). The consequent ulating the persistence of behaviour, values, and emotional population-level time allocation, which emerges from the features within a population. Conceptualizing neural struc- balance of competing motivations, causes changes in state tures as real, persistent aspects of the Earth system prompts variables including subjective experience according to the a novel perspective on the consequences of human time al- context (e.g. the presence of Things, neural structures, cli- location and points toward the underlying physical basis of mate). The variables in this class are simply the fraction of how systemic, population-level societal changes can occur. time (e.g. h d−1 ) devoted to each activity by the population. https://doi.org/10.5194/esd-12-671-2021 Earth Syst. Dynam., 12, 671–687, 2021 678 E. D. Galbraith: Earth system economics 5 Illustrative model Table 1. Model state variables. Next, a simple model is presented that illustrates how the Variable Symbol Value and ESE framework could be operationalized in a global model. units The model bears some similarity to simple ecological mod- Human biomass Smass kghuman els that have previously aimed for direct coupling of humans Neural structure, provisioning-activated Nprovision no. and ecosystems (e.g. Henderson and Loreau, 2018) but us- Neural structure, other-activated Nother # ing the ESE conceptual basis. The model description follows Provisioning tools Tprovision kgtools the ODD (Overview, Design concepts, Details) protocol for Food biomass Efood kg Time allocated to provisioning activity Aprovision d d−1 describing individual- and agent-based models (Grimm et al., Time allocated to other activity Aother d d−1 2006), as updated by Grimm et al. (2020). Because the model Experienced affect Xaffect unitless simulates changes in state variables according to differential equations, rather than being agent-based, a number of the standard design concepts are omitted from the description. Despite these omissions, the description is rather long, and the time-conscious reader may wish to skip most of this sec- tion. 5.1 Purpose and patterns The purpose of the model is to illustrate how linkages can be simulated among the six classes of state variables through dynamical interdependencies. The model focuses on the interaction between a human population and an ecosys- tem that provides food, with feedbacks mediated by state- dependent motivations that re-allocate time between the ac- tivities of collecting food (provisioning) and doing some- thing else (other). The model generates time-varying patterns in population-level state variables, including neural structure and affect, and generalizes many societal features through their influences on motivating time allocation. This simple illustrative model is not intended to realistically capture any particular period of human history but to give an example of how the ESE conceptual framework could guide model con- struction. Figure 3. Model architecture, shown within the conceptual frame- work of Fig. 2. Dependencies are shown as arrows: (i) food supply 5.2 Entities, state variables, and scales vs. metabolic demand influences the time allocated to provisioning, Aprovision ; (ii) time allocation to provisioning and the availability The model is not agent-based and therefore does not sim- of provisioning Things Tprovision influence the per capita extraction ulate interactions amongst entities. Instead, the model con- rate of edible food Efood ; (iii) the extraction rate of Efood deter- siders dynamical changes in state variables, starting from an mines the change in human biomass Smass ; (iv) time allocation be- initial state, according to ordinary differential equations. This tween activities Aprovision and Aother influences population affect distinction is analogous to the distinction between what are Xaffect ; (v) time allocation influences the relative formation rates of referred to as Eulerian and Lagrangian models in fluid dy- neural structure types, either Nprovision or Nother . namics (Vallis, 2017). The approach taken here is equivalent to the Eulerian method, modeling the human population as a field, characterized by its state, rather than trying to resolve The entire planet is represented here as a single entity the motions of individual particles. without any spatial resolution, for ease of illustration, sim- The model includes state variables that fall within the six ilar to the seminal model by Forrester (1971). This “zero- classes introduced in Fig. 2, as listed in Table 1 and shown dimensional” spatial form is common for a proof-of-concept in Fig. 3. Each state variable has a single scalar value at any in Earth system modeling (e.g. Wolfe et al., 2016), though it point in time. The two time allocation variables are Aprovision is envisioned that the most useful application for this frame- and Aother . Each of these represents a compound activity, work would be on a global grid. The model is discrete in time comprised of many unresolved component activities. and is solved numerically through finite differences from a prescribed initial state using forward time steps of 1 week. Earth Syst. Dynam., 12, 671–687, 2021 https://doi.org/10.5194/esd-12-671-2021 E. D. Galbraith: Earth system economics 679 5.3 Process overview and scheduling Finally, one metric of subjective well-being is used here: the affect balance associated with different activities. It is as- First, the population motivational factors are calculated, sumed that a population-average level of affect αx occurs un- based on the food surplus and time allocation from the prior der each activity x, as determined by many factors that are time step, and these motivational factors are used to calcu- not resolved here. Basically, one of the activities is bound to late the time allocation for the current time step. Next, the be more enjoyable than the other. Because provisioning gen- mass flux of provisioned food is calculated as a function of erally falls under the category of work, whether or not it is the biomass variables (Efood and Smass ) and the time alloca- done through the formal economy, it is assumed to incur a tion to provisioning (Aprovision ), and the biomass is updated lower level of affect. The other activity, although sure to in- to conserve mass while accounting for metabolism. Finally, clude many sub-activities that are unpleasant, is assumed to the neural characteristics are updated, and the subjective ex- incur a higher overall average affect. Note that this analysis perience is recorded, according to the time allocation. ignores any sense of eudaemonia that may result from either activity and is purely hedonic. 5.4 Principles The model considers a few physical features of a global hu- 5.5 Emergence man population, tied together by a novel use of time allo- cation as the central dynamical variable. Two activities are The model generates emergent outcomes through the integra- defined in terms of their physical outcomes (the “what” of tion of the state-dependent equations. These emergent fea- the activities). The allocation of available time between the tures include the temporal evolution of food biomass, human two activities is determined by the relative strengths of their population size, time allocation, neural structure, and experi- motivational factors within the population. ence. The details of these features are dependent on parame- The first activity, provision, includes all activities required ter choices, which were not exhaustively explored here. to extract edible organic matter from the environment and distribute it to the population in an edible form. Conceptually 5.6 Adaptation this could include hunting, fishing, or farming, as well as any necessary processing and transportation of food. The transfer The human population dynamically allocates the available of biomass between edible food and the human population is time between the two available activities. This allocation is strictly conserved, and the model uses realistic values for hu- determined by the motivation to provision food, which is a man metabolic and growth rates. The second activity, other, function of the food supply rate relative to the population includes all other non-essential activities in which the popu- metabolic requirements, in competition with the motivation lation is engaged. to do something other than provisioning food. Motivational factors are conceptualized as originating Human behaviour is exceedingly complex and cannot be from the individual level (at which state-dependent neural ac- predicted from first principles. Here, the motivational re- tivation would occur), scaled up to population-level changes sponses at the population scale are approximated by smooth in time allocation as modulated by social structures. Moti- response functions, reflecting competing tendencies to alter vations are formulated as competing influences on time al- time allocation between activities in response to changes in location, similar to the dynamics-of-action approach used relevant state variables. The simple model considers two mo- by Atkinson et al. (1977) but applied here at a population tivational factors – mprovision and mother – each of which has level rather than an individual level. I am not aware of prior a value between 0 and 1, indicating the relative drive to in- works that use the same formulation, but the precise means crease in the fraction of time devoted to the corresponding by which this is achieved are not the focus here. Rather, the activity. To represent saturating responses to an input vari- goal is simply to simulate internally consistent dynamical able, the Holling type 2 formulation is used, since it provides links. stability and each usage introduces only one additional pa- A simple model is used for changes in the neural struc- rameter (k). Each motivational factor is then weighted by a ture of the population, based on two simplifying assump- response coefficient r, which reflects the strength with which tions: first, that the rate of new synapse formation within the time is reallocated to the activity in response to the motiva- population is randomly distributed at a constant rate within tional factor. individual cortexes and, second, that synapses that are be- The r and k parameters are intended to reflect the com- ing fired through activation are more likely to strengthen and bined outcomes of individual psychology and societal pro- persist (Ascoli, 2015). Under these two assumptions, the de- cesses (where societal processes include all cultural, social, velopment of strong synapses, which then become important political, and economic interactions). Neither of these param- pathways for future thoughts, are dependent on engagement eters has a direct equivalence that can be measured precisely, in relevant activities. In this way, the time allocation to activ- a common occurrence in ecological modeling, and their val- ities contributes to the modification of the neural structure. ues are chosen in order to produce reasonable model be- haviour. Although they are held constant in the individual https://doi.org/10.5194/esd-12-671-2021 Earth Syst. Dynam., 12, 671–687, 2021 680 E. D. Galbraith: Earth system economics Table 2. Model parameters. Variable Symbol Default value Net primary production NPP 54 Pg C yr−1 Edible fraction φedible 0.005 Human maximum growth rate µmax 0.03 yr−1 Human metabolism ω 0.02 d−1 Initial mass per human β 6 kg C human−1 Food distribution parameter σ 0.014 kgfood kg−1 human d −1 Available time favail 0.6 d d−1 Sensitivity to food shortage kfood 0.4 Sensitivity to other time shortage kother 0.3 d d−1 Reactivity to food shortage rfood 1 Reactivity to other shortage rother 1 Maximum efficiency of provisioning max provision 3 × 10−14 kg−1 human d −1 Provisioning Things half-saturation constant kT provision 50 kgT provision human−1 Food decay rate λ 0.05 d−1 Normalized synapse formation rate µsynapse 1/75 human−1 yr−1 Synapse destruction rate λsynapse 1 yr−1 Affect during provisioning activity αprovision 0.2 Affect during other activity αother 0.5 simulations shown here, they would not in reality be static characterized by a standard deviation σ . The cumulative dis- properties of the population but could in theory be dynami- tribution function provides the fraction of the population cally linked to neural structure state variables. However this shortagefrac that would experience some level of food short- would go beyond the simple scope of the current illustration. age, as illustrated for two values of σ in Fig. 4. The preva- lence of experienced food shortage is assumed to drive the 5.7 Initialization response of time allocation, according to the motivational pa- rameters. Initial values for state variables were chosen to ensure sta- ble integration. These initial values included a small human population and large food biomass, in order to allow the pop- 5.9.2 Motivations ulation to grow continually for a couple of centuries under The motivation to allocate time to provisioning is a function typical parameter values. Parameter values are given in Ta- of the food shortage, given by ble 2. shortagefrac mprovision = . (2) 5.8 Input data kfood + shortagefrac The model does not use input data to represent time-varying The value of the half-saturation constant kfood determines processes. the relative strength with which the population is motivated to respond to a given shortage, with a smaller value respond- 5.9 Submodels ing more strongly to smaller undernourished fractions and saturating more quickly (Fig. 4). 5.9.1 Food shortage Obtaining food is a primary concern for all animals, but The mass-specific population average food shortage they also tend to spend some fraction of time doing other (shortageave ) is calculated as the difference between the total things. Depending on the species, they might invest time de- provisioned food for the time step and the food required to veloping burrows or nests, engaging in courtship and mat- meet metabolic needs plus additional growth: ing, or resting in a safe place. Humans, more than any other animal, are characterized by the wide range of activities in shortageave = ((ω + µ)Smass − provisioned)/Smass . (1) which they are motivated to engage, other than obtaining In order to illustrate how population state variables can cap- food. The associated motivational factor is termed here mother ture within-population variations, the available food is as- and assumed to increase in intensity as the time allocated to sumed to be unequally partitioned within the population fol- Aother is decreased. This motivation could be construed at an lowing a normal distribution about the average shortage, individual level, such as the individual desire to do some- Earth Syst. Dynam., 12, 671–687, 2021 https://doi.org/10.5194/esd-12-671-2021 E. D. Galbraith: Earth system economics 681 Figure 4. Food shortage and provisioning motivation. Panel (a) shows the fraction of population experiencing food shortage for two values of σ . A larger value of σ corresponds to greater inequality (red) compared to the value used in the simulations (blue). Panel (b) shows the population motivation factor for provisioning as a function of the shortage fraction. The relative motivation strength is shown for highly responsive (low k, purple) and weakly responsive (high k, yellow) societies. P thing more enjoyable, rewarding, or relaxing than food pro- i A, distributing the available time among the possible ac- vision. It could also occur through social mechanisms, such tivities according to their strength of motivation, while re- as cultural norms to engage in rituals or constructing reli- specting the conservation of total time. gious buildings, or through economic mechanisms such as a decrease in provision labour wages as food availability is 5.9.4 Effectiveness multiplier increased. No attempt is made here to represent these mecha- nisms (i.e. the how); instead all are simply bundled in a non- The provisioning activity is associated with a multiplier provision motivation, given by the following equation: provision that represents how effectively the activity is carried P out by the population. This effectiveness multiplier could re- i A − Aother flect the culturally transmitted knowledge and skill within mother = P , (3) kother + i A − Aother a population, as well as the availability of tools and ma- P chinery, the quality of those tools and machinery, and non- where i A is the total time available for the defined activi- human aspects of the activity. The multiplier has units of ties and kother represents the rate at which mother approaches (kg Chuman )−1 d−1 , representing the fraction of available edi- saturation with increasing time spent provisioning. ble material that would be provisioned per human time. The value of provision increases with the per capita mass of pro- 5.9.3 Time allocation visioning Things, Tprovision , approaching a maximum that re- Each of the timePallocation terms Ai is constrained to vary flects the saturation of the active population with available between 0 and i A. For the illustrative model here, un- tools. This saturation is analogous to the saturation of en- resolved essential activities (e.g. sleep, meals, personal hy- zymes with reagents, so the Michaelis–Menten or Holling giene) type 2 formulation is also used here, so that P are assumed to occupy 40 % of the total time, so that i A = 0.6. The model is constructed to place multiple max Tprovision /(Aprovision Smass /β) motivations in competition for the remaining available time provision = provision . (6) kT provision + Tprovision /(Aprovision Smass /β) within the population. As such, the competing motivations exist in tension with each other, and the outcome represents In these simulations, Tprovision is held at a constant per a dynamic balance between them. The changes in motivated capita value. time allocation are given by 5.9.5 Mass fluxes dA0provision = rprovision mprovision , (4) The dynamical change in edible biomass Efood is given by dt dA0other dEfood = rother mother . (5) = φedible NPP − λEfood − Aprovision Smass provision Efood . (7) dt dt Thus, the motivational factors act upon the previously al- The fraction of net primary productivity (NPP) that is al- located time. The individual motivated time allocation terms located to edible material, φedible , would vary with many A0i are then divided by the sum of all A0i and multiplied by factors including the ecosystem type, climate, and human https://doi.org/10.5194/esd-12-671-2021 Earth Syst. Dynam., 12, 671–687, 2021 682 E. D. Galbraith: Earth system economics agency. For example, human activity can modify φedible , in- pends on cultural and societal characteristics as well as expo- creasing it through deliberate modifications including agri- sure to pathogens. Because the cultural and societal aspects culture and aquaculture or decreasing it by destructively har- of both fertility and mortality are complex, the model sim- vesting and overhunting or overfishing. Human activity could ply considers how the net result decreases below the poten- also modify NPP, increasing it by changing vegetation to tial maximum when assuming zero growth among the frac- more productive varieties or by fertilizing and irrigating or tion of the population experiencing a food shortage, so that decreasing it by causing soil erosion, nutrient loss, or other µ = (1 − shortagefrac )µmax . Food waste is not treated explic- forms of ecological degradation. For simplicity, all of these itly but could be regarded as an implicit component of ω, factors are conceptually bundled within a constant value of along with egested food. φedible as modifications of edible NPP relative to a “pristine” The human biomass then varies as state (i.e. untouched by humans). For reference, the present- dSmass day global annual production of edible material is approxi- = min(µ, Aprovision Smass provision Efood − ω))Smass . (8) dt mately 0.9 Pg C (the mass of carbon within all edible primary crops, processed crops, and animal products) according to the 5.9.6 Neural structure analysis of Alexander et al. (2017), implying a global φedible of roughly 0.02 for a global NPP of 54 Pg C (Running, 2012). The basic dynamic by which time allocation modifies the In the second term, λ (d−1 ) represents the consumption neural structure is crudely approximated here by of potentially edible material by all non-human organisms dNx synapse such as other mammals, birds, insects, fungi, or bacteria. This = µx Smass − λsynapse (1 − Ax )Nx , (9) non-human consumption is assumed to be first order with dt respect to Efood , for simplicity. The decay constant would be where Nx is the number of synapses in the population as- expected to vary with food type and environment but would sociated with activity x (normalized to the lifetime synapse generally be on the order of weeks. synapse production of an average individual), µx is the biomass- The final term represents the collection and essential pro- specific growth rate of new synapses that can be activated cessing of edible material by humans, which is the outcome by activity x, and λ is the synapse-specific rate of synapse of the provisioning activity. The term depends on Efood , the destruction (arbitrarily chosen to provide a 1-year e-folding size of the human population, Smass (kg Chuman ), the fraction timescale, for illustration). of time allocated to provisioning, Aprovision (d d−1 ), and the In this formulation, synapses are defined by their associ- effectiveness with which the population provisions the ex- ated activity. This does not imply that they are exclusively isting edible material per unit time provision (kg C−1 −1 human d ). related to the functional core of the activity but simply that The use of linear dependences is sure to be inappropriate, they are activated and thereby strengthened during the activ- given that optimal resources will be harvested first, and di- ity. There could also be overlap between the neural structures minishing returns would be expected to lead to a sublin- of different activities due to commonalities, which are not re- ear dependence on Aprovision Smass (note this is equivalent solved here. to labour in the similar Cobb–Douglas production function, It is essential that this quantification says nothing about which typically has an exponent < 1; Cobb and Douglas, the functional utility of the structural changes. Many of the 1928). This approach could potentially be improved upon in accumulated synapses may contribute little or even be delete- future. rious. The processes by which the brain selects and amplifies Next, the human capacity for food ingestion is given by the functional utility of certain synaptic modifications, while the product of the human population Smass and the sum of dampening others, remains an important topic of research in the population average biomass-specific metabolic rates ω neuroscience (Richards and Frankland, 2017). Nonetheless, and the potential net growth rate µ. Any excess of food pro- the fact that synapses are strengthened in response to activa- visioned beyond this limit is assumed to be discarded. An tion is well-established (Ascoli, 2015), and it is expected that average value of ω is calculated assuming a per capita ener- future work can improve on this crude representation. getic requirement of 10 MJ d−1 and food energy content of 30 kJ g C−1 (Alexander et al., 2017). The value of the max- 5.9.7 Subjective state imum growth rate µmax – the population growth rate when the rate of food provisioning is non-limiting – influences The instantaneous average affect of the population at time t the transient behaviour of the model but not the steady-state is given by the time-weighted mean of the activity-specific outcome. Because µmax is the maximum net growth rate, affects: equal to the birth rate minus the death rate (for constant in- 1 X dividual body size), its value reflects both the fertility rates xaffect = αx Ax , (10) favail of the population and the mortality due to disease, violent deaths, and old age. The fertility rate is dependent on cul- which can be rewritten for this two-activity model as tural and societal characteristics, while the rate of death de- Earth Syst. Dynam., 12, 671–687, 2021 https://doi.org/10.5194/esd-12-671-2021 E. D. Galbraith: Earth system economics 683 Figure 5. Time series of a typical model experiment. In (f), the dashed line indicates the metabolic cost of maintaining the population (i.e. ωhuman Smass ) and the shaded green area represents the food surplus. rate in advance of food limitation but is perhaps realistic for Aprovision populations in which reproductive rates do not decline in xaffect = αother − (αother − αprovision ) , (11) favail response to declining food surpluses. Note that mass is not strictly equivalent to the number of humans, since the mass giving a linear decrease with Aprovision below a maximum per human could change. affect αother . 6.2 Dependence of population size on rprovision or rother 6 Model analysis and discussion Figure 6 shows the same experiment shown in Fig. 5, as well The transient dynamics of the model can assume many forms as a second experiment in which a single parameter value was depending on the input parameters as a result of the non- changed: rother was increased by a factor of 4. This increase linear interactions between the societal aspects (motivational reflects a greater motivation within the population to engage factors) and the human–ecosystem coupling (food provision- in Aother rather than Aprovision . Such a motivation could re- ing). The following sections describe typical features of the flect a desire for leisure, a societal focus on monumental ar- transient behaviour that are robust across a breadth of rea- chitecture, or a culture of learning – these distinctions are not sonable parameter choices. resolved here. The higher rother (relative to rprovision ) results in a smaller 6.1 Approach to steady-state population population size at steady state (Fig. 6a). This occurs even though the food shortage experienced by the population is When initialized from a population density well below the same at steady state: the population simply decides to al- the steady-state value, the human population grows near- locate less time to provisioning because their priority is to exponentially (Fig. 5a). The food biomass is drawn down engage in other activities. Because they provision less inten- (Fig. 5b), generating decreasing yields for the same effort sively, the food biomass remains more abundant (Fig. 6b), (Fig. 5c). Hunger increases in response (Fig. 5d), which resulting in a greater provisioning efficiency (Fig. 6g). The drives a greater Aprovision (Fig. 5e). The surplus (differ- greater allocation of time to other activities results in a large ence between the solid and dashed line in Fig. 5f) grad- contrast in the neural structure, with Nother much greater than ually shrinks, until after a couple of centuries the surplus Nprovision in the population with high rother (Fig. 6i). Ad- reaches the point at which it constrains the growth rate. At ditionally, the steady-state affect is greater with high rother this point the population growth rapidly declines to zero and (Fig. 6j), given the assumption that other activities provide a Smass reaches a plateau (Fig. 5a). The transition from growth higher affect than provisioning. Thus, the high rother experi- to plateau happens more sharply than under logistic growth ment produces a smaller population of happier people with a because the modeled growth rate remains large even as the more diverse neural structure. food surplus shrinks, and the constraint of food limitation on growth is imposed abruptly. This could be unrealistic for populations that have sufficient foresight to slow their growth https://doi.org/10.5194/esd-12-671-2021 Earth Syst. Dynam., 12, 671–687, 2021 684 E. D. Galbraith: Earth system economics Figure 6. Time series of two model experiments with different values of rother . 6.3 Golden ages tion of the aspect we care most about – the human experience of life. Figure 6 illustrates an interesting nonlinear dynamic, particu- The first part of the paper outlined a set of principles, larly pronounced in the food-focused population (low rother ). proposed to form the basis of Earth system economics. It During the initial population growth phase, Aother remains was suggested that, despite the inherent challenges, there is relatively high, since food is abundant and hunger is low. promise in striving for improved physical understanding of This allows the development of Nother , indicating a more di- essential human features. A physically based understanding verse neural structure within the population, and supports a is less prone to ambiguity and could circumvent disciplinary high level of affect. However, as food limitation approaches, barriers, opening new opportunities for dialogue across many food shortages increase, and the low rother causes the popu- fields of human study. Section 4 proposed a framework of lation activity to shift rapidly to Aprovision . The Nother is no state variables to capture the entire human system in a way longer maintained at the high level, and the affect drops. One that is both inclusive and functionally consistent. Time allo- could conceive that such a century-timescale transient would cation provides a central, quantitative anchor for the other- be recalled by a society as having passed through a golden wise bewildering possible range of human activities, while age, such as that mythologized in ancient Greece (Baldry, the Soma, neural structures and Things are recognized as 1952) and frequently echoed throughout history. persistent, defining features of the human system. The small This dynamic does not occur under all parameter combi- number of high-level variable classes is intended to facilitate nations, and it should be borne in mind that the model is a synoptic global view, while subdivisions of these classes very simple. However, it serves to illustrate a straightfor- within the overarching framework can allow resolution of ward interaction that would be expected to produce tempo- important details. Sections 5 and 6 provided an illustration rary golden ages with abundant food and time for other ac- of how a numerical model might be constructed within the tivities such as learning, producing art, and building public framework. works. The ESE approach could be greatly advanced through fur- ther progress in three key domains in the short term. The first is a better understanding of global human time alloca- 7 Conclusions tion, including improved theoretical foundations and harmo- nized multinational datasets. The second is a corresponding mapping of human-created Things that is structurally consis- The global human system can appear overwhelmingly com- tent with the resolved human activities and their biophysi- plex, which has contributed to the general hesitance to in- cal outcomes. And the third includes insights on the process- clude it within Earth system science on a common footing oriented relationships that link activity and context to multi- with the atmosphere, ocean, terrestrial ecosystem, marine ple dimensions of subjective experience. In the longer term, ecosystem, and cryosphere. This paper has suggested that the an improved set of metrics for neural structures that goes global human system can be more effectively integrated with beyond the rudimentary approach used here could open the the remainder of the Earth system through an unremitting fo- door to realistically quantifying rates of change for key soci- cus on physical foundations, including an explicit considera- Earth Syst. Dynam., 12, 671–687, 2021 https://doi.org/10.5194/esd-12-671-2021 E. D. Galbraith: Earth system economics 685 etal characteristics. In addition, there already exists a wealth son, R. G., Mueller, S. C., Nusslock, R., Papadelis, C., Polnaszek, of complementary approaches to assessing aspects of the bio- K. L., Richter, S. H., Silton, R. L., and Styliadis, C.: The Neuro- physical reality of humans. These range from global spatial science of Positive Emotions and Affect: Implications for Culti- hunter–gatherer models (Timmermann and Friedrich, 2016) vating Happiness and Wellbeing, Neurosci. Biobehav. Rev., 121, to human mobility studies (Meekan et al., 2017), models 220–249, 2020. Anderson, K.: Wrong tool for the job: Debating the bedrock of physical labour capacity (Dunne et al., 2013), and the of climate-change mitigation scenarios, Nature, 573, 348–348, study of societal metabolism (Giampietro et al., 2014; Haberl 2019. et al., 2019). Combining these and other physically based ap- Anderson, P. W.: More is different, Science, 177, 393–396, 1972. proaches could yield powerful new insights. Ascoli, G. A.: Trees of the brain, roots of the mind, MIT Press, The ESE approach is intended to help with thinking across Cambridge, Massachussetts, 2015. scales and disciplines. Although the latter half of this paper Atkinson, J. W., Bongort, K., and Price, L. H.: Explorations using has focused on the use of ESE for numerical modeling, it computer simulation to comprehend thematic apperceptive mea- may prove more useful as a conceptual basis for analyzing surement of motivation, Motiv. Emotion, 1, 1–27, 1977. the global human system in general. By providing an overar- Baldry, H. C.: Who invented the golden age?, Classical Quart., 2, ching organizing framework, it is hoped that ESE may help 83–92, 1952. to unite disparate learnings from the social sciences and to Barrett, L. F.: How emotions are made: The secret life of the brain, Houghton Mifflin Harcourt, New York, 2017. bring them together with natural sciences to answer urgent Barrington-Leigh, C. and Galbraith, E.: Feasible future global sce- questions about the functioning of the human–Earth system. narios for human life evaluations, Nat. Commun., 10, 1–8, 2019. Becker, G. S.: A Theory of the Allocation of Time, Econ. J., 75, 493–517, 1965. Code availability. All MATLAB code used to run the model and Bouchaud, J.-P.: Economics needs a scientific revolution, Nature, generate figures is available for download from the Zenodo archive 455, 1181–1181, 2008. at https://doi.org/10.5281/zenodo.4660554 (Galbraith, 2021). Boyer, P.: Minds make societies: How cognition explains the world humans create, Yale University Press, London, UK, 2018. Calvin, K. and Bond-Lamberty, B.: Integrated human-earth sys- Competing interests. The contact author has declared that there tem modeling – state of the science and future directions, are no competing interests. Environ. Res. Lett., 13, 063006, https://doi.org/10.1088/1748- 9326/aac642, 2018. Clark, A.: Surfing uncertainty: Prediction, action, and the embodied Acknowledgements. I am very grateful to Maria Pastor, mind, Oxford University Press, Oxford, 2015. Viki Reyes-Garcia, Dan Zhu, Priscilla LeMezo, William Fajzel, Cobb, C. W. and Douglas, P. H.: A theory of production, Am. Econ. Ian Hatton, Sara Miñarro, and all members of the iESD lab- Rev., 18, 139–165, 1928. oratory for insightful and inspiring discussions. Kim Scherrer, Costanza, R., Kubiszewski, I., Giovannini, E., Lovins, H., Chris Barrington-Leigh, and Jeroen Van Den Bergh provided valu- McGlade, J., Pickett, K. E., Ragnarsdóttir, K. V., Roberts, D., able feedback on the paper. This project has received funding from De Vogli, R., and Wilkinson, R.: Development: Time to leave the European Research Council (ERC) under the European Union’s GDP behind, Nat. News, 505, 283–285, 2014. Horizon 2020 research and innovation programme (grant agreement Csikszentmihalyi, M. and Larson, R.: Validity and reliability of the no. 682602). experience-sampling method, in: Flow and the foundations of positive psychology, Springer, Dordrecht, 35–54, 2014. Dasgupta, P.: The economics of biodiversity: The dasgupta review, Financial support. This research has been supported by the HM Treasury, London, 2021. H2020 European Research Council (grant no. BIGSEA 682602). Diener, E., Oishi, S., and Tay, L.: Advances in subjec- tive well-being research, Nat. Human Behav., 2, 253, https://doi.org/10.1038/s41562-018-0307-6, 2018. Donges, J. F., Winkelmann, R., Lucht, W., Cornell, S. E., Review statement. This paper was edited by Christian Franzke Dyke, J. G., Rockström, J., Heitzig, J., and Schellnhuber, H. J.: and reviewed by Marcin Czupryna and one anonymous referee. Closing the loop: Reconnecting human dynamics to Earth Sys- tem science, Anthrop. Rev., 4, 151–157, 2017. Dunne, J. P., Stouffer, R. J., and John, J. G.: Reductions in labour References capacity from heat stress under climate warming, Nat. Clim. Change, 3, 563–566, 2013. Alessandretti, L., Aslak, U., and Lehmann, S.: The scales of human Falkowski, P., Scholes, R., Boyle, E., Canadell, J., Canfield, D., mobility, Nature, 587, 402–407, 2020. Elser, J., Gruber, N., Hibbard, K., Högberg, P., Linder, S., Alexander, P., Brown, C., Arneth, A., Finnigan, J., Moran, D., and Mackenzie, F. T., Moore III, B., Pedersen, T. F., Rosenthal, Y., Rounsevell, M. D.: Losses, inefficiencies and waste in the global Seitzinger, S., Smetacek, V., and Steffen, W.: The global carbon food system, Agr. Syst., 153, 190–200, 2017. cycle: a test of our knowledge of earth as a system, Science, 290, Alexander, R., Aragón, O. R., Bookwala, J., Cherbuin, N., Gatt, J. 291–296, 2000. M., Kahrilas, I. J., Kästner, N., Lawrence, A., Lowe, L., Morri- https://doi.org/10.5194/esd-12-671-2021 Earth Syst. Dynam., 12, 671–687, 2021 686 E. D. Galbraith: Earth system economics Follows, M. J., Dutkiewicz, S., Grant, S., and Chisholm, S. W.: McDonald, L.: Early Origins of the Social Sciences, McGill- Emergent biogeography of microbial communities in a model Queen’s University Press, Montreal, QC, 1993. ocean, Science, 315, 1843–1846, 2007. Meekan, M. G., Duarte, C. M., Fernández-Gracia, J., Thums, M., Forrester, J. W.: World dynamics, Wright-Allen Press, Cambridge, Sequeira, A. M., Harcourt, R., and Eguíluz, V. M.: The ecology Massachusetts, 1971. of human mobility, Trends Ecol. Evol., 32, 198–210, 2017. Galbraith, E.: Earth System Economics proof of concept model, Mote, S., Rivas, J., and Kalnay, E.: A novel approach to carrying ca- Zenodo, https://doi.org/10.5281/zenodo.4660554, 2021. pacity: from a priori prescription to a posteriori derivation based Galbraith, E. D., Carozza, D., and Bianchi, D.: A coupled human- on underlying mechanisms and dynamics, Annu. Rev. Earth Pl. Earth model perspective on long-term trends in the global marine Sc., 48, 657–683, 2020. fishery, Nat. Commun., 8, 1–7, 2017. Motesharrei, S., Rivas, J., and Kalnay, E.: Human and nature dy- Gershuny, J. and Sullivan, O.: What we really do all day: Insights namics (HANDY): Modeling inequality and use of resources in from the centre for time use research, Penguin, London, UK, the collapse or sustainability of societies, Ecol. Econ., 101, 90– 2019. 102, 2014. Giampietro, M., Aspinall, R. J., Ramos-Martin, J., and Nordhaus, W. D. and Yang, Z.: A regional dynamic general- Bukkens, S. G.: Resource accounting for sustainability as- equilibrium model of alternative climate-change strategies, Am. sessment: the nexus between energy, food, water and land use, Econ. Rev., 86, 741–765, 1996. Routledge, Abingdon, UK, 2014. Richards, B. A. and Frankland, P. W.: The persistence and tran- Grimm, V., Berger, U., Bastiansen, F., Eliassen, S., Ginot, V., Giske, sience of memory, Neuron, 94, 1071–1084, 2017. J., Goss-Custard, J., Grand, T., Heinz, S. K., Huse, G., Huth, A., Rosenzweig, C., Elliott, J., Deryng, D., Ruane, A. C., Müller, C., Jepsen, J. U., Jorgensen, C., Mooij, W. M., Muller, B., Pe’er, G., Arneth, A., Boote, K. J., Folberth, C., Glotter, M., Khabarov, Piou, C., Railsback, S. F., Robbins, A. M., Robbins, M. M., Ross- N., Neumann, K., Piontek, F., Pugh, T. A. M., Schmid, E., Ste- manith, E., Ruger, N., Strand, E., Souissi, S., Stillman, R. A., hfest, E., Yang, H., and Jones, J. W.: Assessing agricultural risks Vabo, R., Visser, U., and DeAngelis, D. L.: A standard protocol of climate change in the 21st century in a global gridded crop for describing individual-based and agent-based models, Ecol. model intercomparison, P. Natl. Acad. Sci. USA, 111, 3268– Model., 198, 115–126, 2006. 3273, 2014. Grimm, V., Railsback, S. F., Vincenot, C. E., Berger, U., Gallagher, Rounsevell, M. D. A., Arneth, A., Alexander, P., Brown, D. G., de C., DeAngelis, D. L., Edmonds, B., Ge, J., Giske, J., Groeneveld, Noblet-Ducoudré, N., Ellis, E., Finnigan, J., Galvin, K., Grigg, J., Johnston, A. S. A., Milles, A., Nabe-Nielsen, J., Polhill, J. N., Harman, I., Lennox, J., Magliocca, N., Parker, D., O’Neill, B. G., Radchuk, V., Rohwader, M.-S., Stillman, R. A., Thiele, J. C., Verburg, P. H., and Young, O.: Towards decision-based global C., and Ayllon, D. The ODD protocol for describing agent-based land use models for improved understanding of the Earth system, and other simulation models: A second update to improve clarity, Earth Syst. Dynam., 5, 117–137, https://doi.org/10.5194/esd-5- replication, and structural realism, JASSS-J. Artif. Soc. S., 23, 117-2014, 2014. 2020. Running, S. W.: A measurable planetary boundary for the bio- Haberl, H., Wiedenhofer, D., Pauliuk, S., Krausmann, F., sphere, Science, 337, 1458–1459, 2012. Müller, D. B., and Fischer-Kowalski, M.: Contributions of so- Ryan, R. M. and Deci, E. L.: On happiness and human potentials: A ciometabolic research to sustainability science, Nat. Sustainabil., review of research on hedonic and eudaimonic well-being, Annu. 2, 173–184, 2019. Rev. Psychol., 52, 141–166, 2001. Helliwell, J., Layard, R., and Sachs, J.: World happiness report, The Sarmiento, J. L. and Toggweiler, J.: A new model for the role of the Earth Institute, Columbia University, New York, 2012. oceans in determining atmospheric pCO2 , Nature, 308, 621–624, Henderson, K. and Loreau, M.: How ecological feed- 1984. backs between human population and land cover influ- Simini, F., González, M. C., Maritan, A., and Barabási, A.-L.: A ence sustainability, PLoS Comput. Biol., 14, e1006389, universal model for mobility and migration patterns, Nature, 484, https://doi.org/10.1371/journal.pcbi.1006389, 2018. 96–100, 2012. Henrich, J.: The secret of our success: How culture is driving human Smith, D., Eade, R., Scaife, A. A., Caron, L.-P., Danabasoglu, evolution, domesticating our species, and making us smarter, G., DelSole, T., Delworth, T., Doblas-Reyes, F., Dunstone, Princeton University Press, Princeton, 2017. N., Hermanson, L., Kharin, V., Kimito, M., Merryfield, W. J., Kahneman, D. and Krueger, A. B.: Developments in the measure- Mochizuki, T., Muller, W. A., Pohlmann, H., Yeager, S., and ment of subjective well-being, J. Econ. Perspect., 20, 3–24, 2006. Yang, X.: Robust skill of decadal climate predictions, npj Clim. Kump, L. R., Kasting, J. F., and Crane, R. G.: The earth system, in: Atmos. Sci., 2, 1–10, 2019. Vol. 432, Pearson Prentice Hall, Upper Saddle River, NJ, 2004. Snow, C. P.: The two cultures and the scientific revolution, Cam- LeDoux, J. E.: Synaptic self: How our brains become who we are, bridge University Press, Cambridge, 1959. Penguin, New York, 2003. Sporns, O., Tononi, G., and Kötter, R.: The human connectome: a Lenton, T., Watson, A., and Watson, A. J.: Revolutions that made structural description of the human brain, PLoS Comput. Biol., the Earth, Oxford University Press, Oxford, 2011. 1, e42, https://doi.org/10.1371/journal.pcbi.0010042, 2005. Leshem, D.: Retrospectives: What did the ancient Greeks mean by Steffen, W., Richardson, K., Rockström, J., Schellnhuber, H. J., Oikonomia?, J. Econ. Perspect., 30, 225–238, 2016. Dube, O. P., Dutreuil, S., Lenton, T. M., and Lubchenco, J.: Lindquist, K. A., Wager, T. D., Kober, H., Bliss-Moreau, E., and The emergence and evolution of Earth System Science, Nat. Rev. Barrett, L. F.: The brain basis of emotion: a meta-analytic review, Earth Environ., 1, 54–63, 2020. Behav. Brain Sci., 35, 121–202, 2012. Earth Syst. Dynam., 12, 671–687, 2021 https://doi.org/10.5194/esd-12-671-2021 E. D. Galbraith: Earth system economics 687 Stocker, B. D., Wang, H., Smith, N. G., Harrison, S. P., Keenan, T. Van Vuuren, D. P., Den Elzen, M. G., Lucas, P. L., Eickhout, B., F., Sandoval, D., Davis, T., and Prentice, I. C.: P-model v1.0: Strengers, B. J., Van Ruijven, B., Wonink, S., and Van Houdt, R.: an optimality-based light use efficiency model for simulating Stabilizing greenhouse gas concentrations at low levels: an as- ecosystem gross primary production, Geosci. Model Dev., 13, sessment of reduction strategies and costs, Climatic Change, 81, 1545–1581, https://doi.org/10.5194/gmd-13-1545-2020, 2020. 119–159, 2007. Timmermann, A. and Friedrich, T.: Late Pleistocene climate drivers Voinov, A. and Shugart, H. H.: “Integronsters”, integral and inte- of early human migration, Nature, 538, 92–95, 2016. grated modeling, Environ. Modell. Softw., 39, 149–158, 2013. Truran, P.: Models: Useful but Not True, Springer International Pub- Winterhalder, B.: Work, resources and population in foraging soci- lishing, Heidelberg, 61–67, https://doi.org/10.1007/978-3-319- eties, Man, 28, 321–340, 1993. 00452-5_10, 2013. Wold, H. O. and Whittle, P.: A model explaining the Pareto distri- Turchin, P.: Historical dynamics: Why states rise and fall, Princeton bution of wealth, Econometrica, 25, 591–595, 1957. University Press, Princeton, 2018. Wolfe, G. M., Marvin, M. R., Roberts, S. J., Travis, K. R., and Liao, Vallis, G. K.: Atmospheric and oceanic fluid dynamics, Cambridge J.: The Framework for 0-D Atmospheric Modeling (F0AM) v3.1, University Press, Cambridge, 2017. Geosci. Model Dev., 9, 3309–3319, https://doi.org/10.5194/gmd- Van den Bergh, J. C.: The GDP paradox, J. Econ. Psychol., 30, 117– 9-3309-2016, 2016. 135, 2009. https://doi.org/10.5194/esd-12-671-2021 Earth Syst. Dynam., 12, 671–687, 2021