Cambridge Centre for Climate Change Mitigation Research (4CMR)

4CMR Working Paper Series

Braving the Tempest: Methodological foundations of policy-making in

sustainability transitions
J.-F. Mercurea,b,, H. Pollittc , A. M. Bassid , J. E. Vinualesb , N. R. Edwardse
a Cambridge

Centre for Climate Change Mitigation Research (4CMR), University of Cambridge, 19 Silver Street, Cambridge, CB3 1EP, United
Kingdom
b Cambridge Centre for Environment, Energy and Natural Resource Governance (C-EENRG), University of Cambridge, 19 Silver Street,
Cambridge, CB3 1EP, United Kingdom
c Cambridge Econometrics Ltd, Covent Garden, Cambridge, CB1 2HT, UK
d KnowlEdge Srl, 2, via San Giovanni Battista, 21057 Olgiate Olona (VA), Italy
e Environment, Earth and Ecosystems, The Open University, Milton Keynes, UK

Abstract
Policy-makers currently face unprecedented challenges and uncertainty when taking decisions that simultaneously
affect economic development, technology and the environment. A lack of consensus on the impacts of climate policy
in this complex interacting system paralyses the policy-making processes, and policy indecisiveness generates stagnation in investment decision-making, at the core of economic development. It is not clear to policy-makers how to
reconcile economic policy supporting growth with climate change mitigation, and it is not clear how effective policies
are likely to be.
This paper argues that informing policy-making using conventional equilibrium or optimisation modelling of technology and economics is not fine-grained enough to capture the complexities of real-world human behaviour and its
diversity, leaving a wide uncertainty gap for policy-making. We suggest that the use of a dynamical methodology
involving complexity science coupled to behavioural science with sophisticated uncertainty analysis can provide appropriate tools to better understand policy issues in sectors that involve a high degree of interaction (collective effects)
or differentiation (heterogeneity/diversity) between agents. This is used in a feedback loop with researchers in policy
and law, in order to test potential policy packages while assessing their feasibility in the policy process.
We argue that a better representation in models of these complex feedbacks between three critical interrelated
areas could enable a paradigm shift for our understanding of how these aspects work together: technological change,
the macroeconomy, and the natural environment. We describe what impact these developments would have on our
predictive power and ability to constructively inform coupled technology-economy-environment policy-making aimed
at addressing climate change. We identify four areas of environmental policy where the high degree of agent interactions (positive re-inforcing feedbacks) and/or behavioural diversity (e.g. the income distribution) makes their analysis
impractical/inconsistent using conventional methods, and where the application of this methodology could be decisive
for gaining appropriate insight for policy making. These are (1) the effectiveness of policy for emissions reductions
in consumer based sectors (e.g. private transport) (2) the analysis of green growth, (3) cascading uncertainty across
models and (4) cross-sectoral impacts of sector specific policies (e.g. biofuels). By providing concrete examples, we
suggest how a wider adoption of this approach could lead to a step change in our ability to address the complex policy
4CMR Working Paper no 10

problems raised by sustainability transitions.

Keywords:
Environmental policy assessment, Climate change mitigation, Complexity, Behavioural science

1. Introduction
1.1. Investments in a storm of uncertainty
National and international public policy-making currently faces the unprecedented challenge of effectively managing the complex interaction of economic development, energy systems and environmental change (IPCC, 2014).
Expected outcomes of stringent climate policies are subject to uncertainty and disagreement, which acts as a disincentive to policy-makers, and the lack of policy clarity hinders industry decision-makers in their investment strategy, such
as for replacing ageing or environmentally damaging capital, which affects economic growth. In contexts of damaging
policy indecision, investors often find it worthwhile to wait before committing to new long-lived capital investment
decisions (IEA, 2007). Meanwhile, carbon budgets are increasingly consumed (IPCC, 2013a), and the likelihood of
avoiding dangerous climate change is rapidly decreasing. Consensus is urgently needed to identify effective climate
policy, and facilitate its rapid global adoption.
This situation originates from conflicts in our understanding of the complex whirlpool between the economy,
technology and the environment. It is perceived by policy-makers that strong trade-offs exist between improving the
sustainability of the economy, and adequately supporting fragile and fast changing economies that are in financial
crisis in some areas, and undergoing accelerated growth in others. Can economies around the world suffering from
the great recession afford the changes required to avoid climate change? Meanwhile, it is also perceived that wealth
could be generated by new green technology, a seeming contradiction. But the adoption of green technology is a
behavioural question, which is linked to the distribution of income, and the diversity of firms and consumers. It is not
clear to policy-makers how to reconcile economic growth with pressing environmental objectives, and whether the
two are mutually exclusive.
Successful policies of the scale required to stabilise emissions and avoid climate change have no precedent. We
highlight four primary areas where uncertainty contributes to the current level of indecisiveness in public policy
for economic development and environmental protection: (1) the effectiveness of policy instruments to instigate the
necessary changes to reduce environmental damage, (2) the economic implications of reducing environmental damage,
(3) the actual environmental damage, long-term impacts of human activity on the natural system, and the impacts of
future environmental change on society, and (4) the success of strategies for policy implementation and possible
unintended consequences of climate policies across sectors.
Solutions to environmental degradation lie in the diffusion of innovations, technologies and practices, in energy,
industry and households, for which low carbon alternatives mostly already exist. How their adoption can be incentivised in time to avoid dangerous environmental change, and whether this is economically or technically possible,
is an open question. Whether such dynamics could support economic development is not understood. Whether climate policies influence access to food, water and energy is not fully determined. As the world heads into a policy,
technical, economic and environmental storm, guidance on how to understand these complex interactions is becoming
increasingly important. This work aims to identify some research approaches that could significantly improve our
ability to predict the outcome of climate policies, by integrating behavioural and non-equilibrium complexity science
and environmental feedbacks into climate policy analysis, with a framework consistent across disciplines.
Corresponding

author: Jean-Francois Mercure

Email address: jm801@cam.ac.uk (J.-F. Mercure )
1 Tel: +44 (0) 1223337126, Fax: +44 (0) 1223337130

4CMR Working Paper no 10

July 3, 2015

Cambridge Centre for Climate Change Mitigation Research (4CMR)

1.2. Integrating complexity and behavioural sciences as a framework of analysis

Equilibrium and optimisation-based analyses, notably Cost-Benefit Analysis (CBA),2 have been used extensively
to inform climate policy-making and have contributed significant insight to systems-wide issues and possible solutions (e.g. GEA, 2012, Stern, 2007). These approaches are generally normative; they seek to identify optimal policies.3
Optimisation approaches (computable general equilibrium (CGE), partial equilibrium cost-optimisation) are sensitive
to the curvature of demand and supply or cost functions: when convexities are introduced, several solutions to the
optimisation problem emerge. Non-convexities in relevant cost functions are generally the result of self-reinforcing
processes and generally also imply the critical property of path-dependence (Anderson et al., 1989, Arthur, 1999).
Meanwhile, when optimising the aggregate utility (or cost) functions of a multitude of agents, finding a single optimal
point can only occur in a model where all agents have the same utility (or cost) function, i.e. an economy formed of
identical people.4 If, for the sake of obtaining a single solution, agents are approximated with an identical average
person, the effect is to neglect what we call here agent diversity. As we explain below, when either or both simplifications are used (neglecting self-reinforcing processes and/or diversity), we are describing an economy that behaves as
a single optimising person.5 We explain in this paper what types of policy problems these sets of simplifications do
not allow us to analyse self-consistently, and what can be done about it.
The conventional equilibrium perspective described above may be termed reductionist, in the sense of complexity
theory (Anderson, 1972): the macro system behaviour is aggregated from micro properties, but without considering
the interactions between agents. These interactions, however, determine the difference between a diversified economy,
and an economy behaving as a single person alone, or as a group of isolated people. In particular, interactions between
agents, or between agents and their environment, are responsible for all crucial self-reinforcing processes (i.e. crowd
effects). In the theory of complex systems, many properties emerge solely from the interactions between agents, not
from the behaviour of the agents themselves, and this includes important economic phenomena such as the profile of
diffusion of innovations, learning-by-doing, expectations in finance and economic fluctuations, trends and fashions,
technology lock-ins, and many more. These phenomena exist, but do not stem from microeconomic behaviour of
individual agents; therefore they cannot be studied with a theory that ignores interactions between agents. Thus while
there is a good rationale for desiring a macro theory built on micro-foundations, the latter must include inter-agent
interactions, and this is a great challenge.
The conventional equilibrium perspective is also homogeneous: people and firms are represented with the behaviour of a single average agent, which therefore gives no role to the distribution of income or other socio-economic
and industry parameters, and only one type of behavioural response exists. Where this could be important is in the
representation of consumer or investor choices, which is critical in the process of the diffusion of innovations, technologies and practices. As can be inferred from classical innovation diffusion theory by Rogers (2010), behavioural
response, its diversity, the diversity of social groups and unequal distribution of information is precisely what determines actual rates of adoption of innovations. This is basic knowledge in the field of marketing (e.g. Smith, 1956, on
product differentiation). Meanwhile, utility maximisation under budget constraint (expected utility theory) is known
to be an incorrect representation of attitudes towards risk, gains, losses and uncertainty (prospect theory Kahneman
and Tversky, 1979, Tversky and Kahneman, 1974).
Self-reinforcing processes and agent diversity both (independently) lead equilibrium/optimisation theory to produce multiple solutions. Thus, if we make the assumption that a simple unique equilibrium solution exists to the CBA
of climate change mitigation, we are neglecting the path-dependence produced by self-reinforcing phenomena, which
are at the source technology lock-ins, financial bubbles and crises, technology transitions, etc, and the diversity of
agents which determines the rates of adoption of innovations. We are then missing several key components for the
2 By this we mean all theories that involve either a social planner that organises society, or a subset of society, in such a way to minimise a total
cost or maximise an aggregate utility function, and/or theories that involve a unique stable economic equilibrium and a propensity for the economy
to return to this unique equilibrium after exogenous disturbances.
3 Optimal from a cost or utility viewpoint. Maximising utility is socially optimal in the neoclassical perspective, see below.
4 A subtlety should be noted here: the representative agent can be defined based on discrete choice theory, e.g. with a multinomial logit, in which
the elasticities represent the collective response of the underlying population of agents with different preferences. This requires all agents to have
identical full knowledge of all consumer goods in all markets, and no mutual interactions (isolated agents interacting through the economy only).
This is mathematically equivalent to a model of constant elasticity of substitution in the neoclassical perspective (see Anderson et al., 1992).
5 I.e. a single utility optimising person. Of course, individuals are much more complex than what is represented in the representative agent of
classical economics, as exemplified for instance with prospect theory (Kahneman and Tversky, 1979, Tversky and Kahneman, 1974).

4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

analysis of environmental policy. These simplifications in our analysis are reflected in the policies that we advocate:
we expect a simple internalisation of environmental costs, in the form of a price instrument (e.g. a carbon price or
tax), to optimally and effectively incentivise technology adoption and diffusion, and lack tools to understand why, in
the real world, such instruments do not behave and deliver in the expected way (Grubb et al., 2014). However, the
expected success of such a tax is actually a reiteration of our assumptions, not a result of our analysis.
Integrating complexity (e.g. Arthur, 2014) and behavioural sciences (e.g. Domencich and McFadden, 1975, Kahneman and Tversky, 1979) applied to economics, however, would bring path-dependence and heterogeneity to the
core of our analysis. Complexity science is the cross-disciplinary field that specifically studies properties that emerge
from interactions between system components, initially studied in physical, biological and computer sciences (e.g.
Anderson, 1972, Sigmund, 1993). By introducing descriptions of how agents behave and interact, theories and models turn from normative to positive/descriptive and their methodology evolves from optimisation to simulation, where
the analyst can only rely on what if approaches. It is understood that in simulations of complex systems, uncertainty
plays an important role in part because the theories do not necessarily predict a propensity to return to equilibrium.
This is an aspect well understood and controlled in the climate sciences, and no inherent reason exists that prevent us
working in the same way in economics. In practical terms, this means adopting a work structure where many scenarios
are created based on possible policy choices, and acceptable results are retained based on a multidimensional range
of outcomes that reach given objectives. For example, this can involve filtering scenarios, in multidimensional human
and biogeochemical space, to determine policy options that enable society to avoid exceeding planetary boundaries
(Rockstrom et al., 2009, Steffen et al., 2015) while ensuring continued human development globally (OXFAM, 2012).
1.3. Where complexity and heterogeneity become important
A sustainability transition inherently involves socio-technical change, which is a highly non-linear, self-reinforcing
process with lock-ins that drive expectations, propelled by choices of and adoption by diverse agents with differing
perspectives and incomes, for which complexity and behavioural sciences provide a good frame of understanding.
Such technology transitions have an inevitably significant influence on the evolution of the economy through productivity and structural change, which occurs with the development of new firms and industries, and the destruction of
others (e.g. Arthur, 1989, Freeman and Louca , 2001, Schumpeter, 1934). Meanwhile, an economic transformation
towards higher or lower sustainability takes place in direct interaction with the environment and its biogeochemical
cycles. Simultaneous representations of these three areas (technology, macroeconomics and environment) cannot reliably be optimised even on the largest super-computers: most particularly because diverse agents will have conflicting
definitions of optimality. However, there is no inherent reason why they could not be simulated like the climate, as
likely outcomes of chosen policies.
Four key areas in environmental policy analysis could gain significantly from methodological developments as
suggested here and described in more detail below:
1.
2.
3.
4.

Developing and deploying models of policy effectiveness for low-carbon technology diffusion,
Improving our understanding of the macroeconomic impacts of low-carbon policy,
Integrating environmental feedbacks into this techno-economic analysis,
Using (1-3) to study complex cross-sectoral issues (e.g. the energy-water-food nexus) in a close feedback
working structure with policy-advisors and policy-makers.

A complexity paradigm has been suggested as a practical approach to policy problems (Probst and Bassi, 2014).
In this paper, we provide a few examples of how complexity and behavioural sciences can be integrated and applied to
environmental policy assessment, with emphasis on model uncertainty analysis, in order to build a powerful approach
for improved policy analysis.
2. Material and methods: re-thinking four critically linked areas of analysis
2.1. Technology transitions and the diffusion of innovations
The choice of environmental technology lies with intrinsically diverse consumers and private investors. Indeed, the
choice of consumer technologies, such as vehicles, takes place within contexts of distributed income that span several
orders of magnitude (Mercure and Lam, 2015). In particular, the diversity of incomes, social groups and attitudes is
4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

known to determine the rates and profiles of diffusion of innovations (Rogers, 2010). Meanwhile, technology almost
invariably possesses the property of increasing returns to adoption, (see below, e.g. in Anderson et al., 1989, pp 9-10).
The response to policy is a key sustainability planning parameter, and modelling the impacts of policy incentives to
the consumer using averages over consumer group parameters that are wide-ranging, due to their diversity, would be
misleading. Clear interest exists in estimating the response to policies of diverse agents before they are implemented
(Grubb et al., 2014).
Modelling decision-maker choices is key to predicting the effectiveness of policy for emissions reductions. Extensive knowledge exists on the impacts of incentives to consumers with differing incomes and social groups, which
has not yet been widely applied to climate policy analysis. This lies in the fields of marketing research (empirically,
e.g. Bass, 1969, Fisher and Pry, 1971, McShane et al., 2012), anthropology (e.g. Douglas and Isherwood, 1979)
and behavioural economics (discrete choice modelling, e.g. Ben-Akiva and Lerman 1985, Domencich and McFadden
1975 and behavioural response, e.g. Kahneman and Tversky 1979, Tversky and Kahneman 1974). The relevance of
these fields is obvious. However, surprisingly, they have been overlooked in climate change mitigation modelling or
environmental assessment research. Their use thus remains a promising but emerging field (e.g. Axsen et al., 2009,
Giraudet et al., 2012, Rivers and Jaccard, 2006, Kohler et al., 2009). Grubb et al. (2014) emphasises the need for future research on behavioural aspects of emissions reductions; insufficient effort has been devoted to understanding the
aggregate behavioural response to environmental policy instruments, and this is related to a possibly overstated general expectation in the field and in policy spheres that an efficient equilibrium should emerge on its own in technology
markets.
Meanwhile, the diffusion of innovations is a highly non-linear process. In transitions theory, socio-technical
evolution is seen through a perspective involving niches, a socio-technical context and a broader environment (Geels,
2002). This view does not allow treating human-technology systems and practices as following a linear optimisation
or equilibrium; it is a continuous transformation with momentum. Transitions build onto themselves with increasing
returns to adoption (i.e. adoptions that increase the likelihood of further adoptions, and thus propagate like diseases,
e.g. Mansfield, 1961), featuring possible lock-ins into sub-optimal configurations (Arthur, 2014, 1989). They possess
complex non-linear behaviour due to crowd effects.
In the empirical marketing research literature, technology substitution has been extensively shown to follow simple
S-shaped diffusion profiles in a myriad of contexts (e.g. Fisher and Pry, 1971, Grubler et al., 1999, Mansfield, 1961,
Marchetti and Nakicenovic, 1978). More generally, the competition between several technologies for market space
was shown to follow coupled Lotka-Volterra systems (Bhargava, 1989, Karmeshu et al., 1985). Conversely, it has been
shown recently that Lotka-Volterra systems can be derived from simple statistics of industrial dynamics (Mercure,
2015). This can equivalently be expressed with the replicator dynamics system of evolutionary theory: natural
selection is carried out by the consumer, who filters successful innovations based on their fitness to the market,
while entrepreneurs strive to improve their products in order to increase their fitness by better matching consumer
tastes (e.g. Safarzynska and van den Bergh, 2010). This natural selection, however, involves decision-making by
consumers under bounded rationality, who are naturally highly diverse, and the diversity of consumers drives product
differentiation and increasing product diversity (for example with private vehicles, see below).
The effectiveness of a technological change policy rests on whether it is able to promote the onset of a wave of
technology adoption (e.g. incentivising the mass adoption of electric cars). The wider benefits of improving understanding and model representation of the effectiveness of policy could be significant. If one can determine, within
uncertainty ranges, the likelihood and magnitude of the effectiveness of specific policy instruments at incentivising
decisions of real consumers and investors, the uncertainty inherent in policy-making becomes better characterised and
greatly reduced. It moreover becomes possible to predict the complex and possibly unexpected cross-sectoral impacts,
some of which are currently hotly debated subjects. This includes:
1. The impacts on the electricity sector of support policies for electric vehicles
2. The impacts on land-use change of the development of new biofuel markets
3. Synergies and interference between different policy instruments within or across sectors
We simply do not know the possible far reaching consequences of some simple policies. On the energy supply
side, it has been shown recently (Mercure et al., 2014) that the use of a discrete choice theory-based decision-making
model applied in a technology diffusion framework in the electricity sector allows the exploration technology uptake
4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

and likely interactions between policy instruments (carbon pricing, feed-in tariffs, regulations and subsidies) for emissions reductions. We show that the collective impacts of combined policy can be much more (or much less) than the
sum of their individual impacts, emphasising possible synergies (or lock-ins), which cannot be observed using costoptimisation. This is an example where methodology provides a possible step forward for environmental assessment
since it helps move the field from generating feasible or desirable energy sector storylines to producing likely
outcomes of policy based on existing technology and knowledge of the market, but also technology diffusion dynamics. It also helps delineate normative from descriptive analysis methods as useful but distinct disciplines. However, as
is the case with any model involving non-linear dynamics, uncertainty over model outcomes due to uncertainty over
parameters increases exponentially with modelling time span, which needs to be dealt with (section 2.3 below).
2.2. Green growth: macroeconomics and the finance of innovations
Policy for the diffusion of low carbon innovations presents significant opportunities for industry and jobs (e.g.
Blyth et al., 2014, CBI, 2014). The feedback between the diffusion of new environmental technologies and economic
development is poorly understood and reported. There is an apparent contradiction between observations of highly
lucrative activity arising in successful low carbon ventures (e.g. Tesla electric and Toyota hybrid cars, wind turbines,
solar photovoltaic), and the perception that the use of lowest-cost fossil energy with conventional technology is indispensable for any sustained future growth. Indeed, within the conventional equilibrium economics set of possible
outcomes, maximising economic growth requires minimising the costs of energy and other services at the system
level, which yields the optimal equilibrium path (e.g. see various models reviewed in Edenhofer et al., 2010, 2006,
IPCC, 2014).6 In the broader complexity economics perspective, the incorporation of self-reinforcing mechanisms
(innovation, learning, adoption, expectations and herd effects) leads to many other possible outcomes (Arthur, 2014,
Keen, 2011, e.g. what are known as limit cycles). In particular, equilibrium theory cannot reproduce economic crises,
while the finance of technology and innovation is known to lead to productivity change, but also to speculation and
bubbles, which themselves lead to economic cycles (Freeman and Louca , 2001, Perez, 2001, for example the dotcom bubble). Understanding this requires studying the feedbacks between macroeconomic and technology diffusion
dynamics. The diffusion of low-carbon innovations can produce both economic externalities (e.g. higher costs) and
benefits (e.g. enhanced investment and employment), opposing forces with a net impact difficult to predict offhand.
Investment fluctuations have a strong influence on the level of economic development, dynamics closely linked to the
origin of the great recession (Keen, 2011).
Combining these notions, in the context of current economic theory, can become a messy affair with little consensus; however it can also be clarified. Consensus exists on the notion that productivity change in the economy
takes place through innovation, technological change and diffusion, economies of scale and the diffusion of new
management practices and organisation paradigms. This is a subject studied since Schumpeter (1934), whose centenary7 legacy is now carried forward by the field of evolutionary economics (e.g. Andersen, 1994, Nelson and Winter,
1982, Safarzynska and van den Bergh, 2010). New technologies feature important increasing returns to investment
and adoption, and involve cross-sectoral spillovers that often lead to economy-wide activity accelerators (Freeman
and Louca , 2001). For example, in the industrial revolution, new textile machinery required better iron and steel, in
which investment led to developing an industry and cost reductions that reshaped the design space for other products
(Freeman and Louca , 2001), enabling many related and clustered innovations to emerge and economically reach the
marketplace. A statistical analysis of the process of clustering of innovations is given by Arthur and Polak (2006).
Increasing returns here refers to productivity growth that results from the adoption of innovations; productivity
changes always come from the adoption of innovations (new technology, management practices, new lifestyles) but
adopting new technologies does not always ensure productivity growth. However, facing a large energy supply and
end-use sector to transform for environmental protection, policy-makers and society currently contemplate a radical
transformation of a scale not unlikely to result in substantial economic implications due to cross-sectoral spillovers
6 The collaborative project called the Innovation Modelling Comparison Project (Edenhofer et al., 2006), which concluded in 2006, aimed at
addressing specifically the introduction of induced technological change into economic models, predominantly of equilibrium or optimisation type.
This corresponds to including either technology learning curves (partial equilibrium models), or total factor productivity changes in production
functions (CGE models). These nevertheless involve an economic equilibrium, and thus as we argue, only partially illuminate the problem of
self-reinforcing processes, while they do not address the issue of agent diversity.
7 The German version of Schumpeter (1934) was published in 1911.

4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

(e.g. new materials, new design and engineering methods, etc). If the implications are beneficial, the transition could
be thought of as green growth, where new investment and employment are generated from technology policy albeit
possibly with important relative price changes. Significant uncertainty lies in determining whether this could be the
case.
Whether climate policy can lead to substantial economic impacts, and whether these would be beneficial or detrimental, is the subject of intense debate (of note is the Porter Hypothesis, suggesting that environmental regulation
could improve competitiveness of firms, see Porter 1991, Porter and van der Linde 1995. For the related debate see
e.g. Barker et al. 2006, Edenhofer et al. 2010, 2006, IPCC 2014, Nordhaus 2010, Stern 2007). Through this, as
we explain in the next few paragraphs, we find that it ultimately depends on whether or not economic resources are
assumed to be already used optimally in the economy (e.g. see Barker et al., 2006, Mercure et al., 2015). We first
note that the dichotomy and schism in the field is marked, and effectively dates back a hundred years: equilibrium
Walrasian economics takes income as given and divides it between investment and consumption (Walras, 1874), while
non-equilibrium Schumpeterian and Keynesian economics consider that income changes when investment changes,
such that investment is not necessarily made at the expense of consumption (Keynes, 1936, Schumpeter, 1934, 1939).
This may be argued to be a key element of Keynes General Theory.
Perhaps the key issue here is that the set of dynamics studied in equilibrium economic theory does not include those
that result from waves of clustered innovation, as often described in economic history, and the mechanisms unleashing
their finance (Freeman and Louca , 2001, Perez, 2001, Schumpeter, 1934). In Walras, when investment is forcefully
pushed towards a particular sector (e.g. by government expenditure), the price of finance increases (the interest rate)
and as a result, other sectors are outbid and suffer from under-investment (a crowding-out effect). Meanwhile, in
Keynes/Schumpeter, when an entrepreneur asks for a loan to a financial institution, based on expected returns, if the
latter decides to award it, it simultaneously creates a deposit and a liability, respecting the macroeconomic identity
while changing the amount of money in the economy and potentially having real effects. This aspect has recently been
a point of discussion in monetary theory (McLeay et al., 2014).
Investments in new technology are, effectively, often credit-financed,8 whether carried out by industry or consumers, which generates immediate financial flows to industry that are repaid back later during the lifetime of the
equipment purchased. The impact of replacing polluting technologies by low-carbon systems has two possible outcomes, depending on theory. In equilibrium theory, increased demand for finance leads to increases in the supply of
money for low-carbon investment purposes at higher cost, thus with a higher interest rate. Investment in low-carbon
systems is taken away from investment in other sectors through this price signal: investment resources are a fixed
proportion of income. Meanwhile, in non-equilibrium theory (i.e. post-Schumpeterian and post-Keynesian), income
is not fixed but rather depends on investment, which is also not fixed, fluctuating over time. But how can investment
fluctuate, if it is equal to savings, which are a fixed proportion of income? Based on the track-record of entrepreneurs
and their business plans, banks thus create money to finance technology ventures, which generates both investment
and savings in equal amounts (i.e. the theory of endogenous money, see e.g. McLeay et al., 2014). When particular
kinds of ventures are very successful, they attract lots of finance, and often lead to frenetic activity and bubbles (e.g.
the IT bubble in the 2000s, housing bubbles in 2007, see Perez, 2001). At moments when too many ventures fail, a
correction takes place, finance abruptly contracts, slowing down money flows to chains of connected businesses. This
sometimes leads to economic crises (e.g. as happened with housing investments in the financial crisis of 2007 and its
aftermath). These are unfulfilled expectations in the financial sector which, in equilibrium theory, is assumed away
by the theory of rational expectations (Muth, 1961), in which investor agents behave without interacting, and price
movements are uncorrelated.
2.3. Uncertainty analysis in human-environment interactions and climate change
Feedbacks between the economy and the natural world take place partly through environmentally damaging emissions (e.g. climate change), partly through direct human intervention (e.g. land-use change), interacting even with
one another in a highly complex system. The allocation of land for agriculture, including energy crop production,
interacts directly with the carbon cycle and climate through for instance deforestation, land-use change emissions and
desertification, but also indirectly through the economy. The extensive use of water further constrains and constricts
8A

minor part can be self- or equity-financed (e.g. in China, IMF, 2014)

4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

GUIDEPOST
a.

b.

c.

d.

Model input

Emulator 1

Emulator 2

e.g. emissions

e.g. carbon cycle

e.g. climate

Emulator 3
e.g. land productivity

Figure 1: Schematic representation of cascading uncertainty across statistical model emulators

the availability of natural resources for human use, in the now so-called energy-water-food nexus. However, natural resource use stems directly from economic processes, involving the demand for agricultural, energy and forestry
commodities that are traded in international markets.
Exploring complex interactions between the environment and the techno-economic system requires detailed representations of the role and behaviour of the natural environment, a highly ambitious endeavour. Indeed, modelling the
climate is carried out with the most powerful supercomputers available. The response of the environment to anthropological influence is increasingly well understood (IPCC, 2013a) and some of this knowledge has been reproduced
reliably enough that results of environmental simulations can be used directly without requiring detailed parallel simulations with supercomputers alongside an economic model. Statistical representations (emulators), which interpolate
climate responses based on large climate model output data have been used for some time (e.g. Meinshausen et al.,
2009), enabling social and economic modellers easy access to quantitative results from climate science, usually carried
out in different institutions.
In this context, the linear approach of pattern scaling (Tebaldi and Arblaster, 2014) is almost always used, which
assumes that the spatial pattern of change in the climate output of interest is invariant with respect to time and forcing.
This approximation is often inadequate, for instance in the case of land-use change, which not only has significant
impacts on the global climate through greenhouse gas emissions but also has a more localised (i.e. pattern changing)
impact on climate through changes in surface albedo, moisture transfer and river runoff (Myhre et al., 2013). A more
general approach that allows for changes in the pattern of climate impacts has been developed recently (Holden and
Edwards, 2010) and applied to emulate complex models of a range of climatic systems: the climate model PLASIMENTS (Holden et al., 2014), the carbon cycle model GENIE (Holden et al., 2013), and the land surface model LPJmL
(Oyebamiji et al., 2015). These emulators have already been applied in a range of integrated assessments (including
but not limited to IEA, 2013, Labriet et al., 2013, Mercure et al., 2014).
This methodology enables a useful approach for integrating large amounts of environmental data to socio-economic
and policy analyses: by concatenating such emulators in chains (see schema in Figure 1), one can obtain a representation of uncertainty cascading across sectors. This therefore avoids possible biases that may emerge from the use
of median trajectories when linking different models: this enables the exploration of representations of all likely
trajectories simultaneously, at low computational costs.
The cascading of uncertainty that increases with simulation time span stems from the general property of complex
non-linear models, which produce exponentially diverging scenarios for arbitrarily small changes in starting parameters (i.e. sometimes termed butterfly effect in chaos theory), typical of climate models. This, however, is also a
property of an economy based on complexity theory, and of diffusion-based technology models. Thus when we see
economic models in the same way as those of natural science, with complexity and path-dependence, this approach to
systematic uncertainty analysis becomes highly appropriate and attractive. Depending on the sensitivity of each subsystem to perturbations, i.e. the rate of divergence of scenarios, and conversely, the possible presence of attracting
states, the cloud of uncertainty may increase moderately or dramatically as it propagates through the chain.
4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

2.4. Policy effectiveness, behaviour and implementation

Tools from marketing research, behavioural economics and anthropology (see above) are extremely useful to
improve our understanding of consumer and investor behaviour. They provide a highly powerful addition to economyenvironment modelling: to predict quantitatively the likely behavioural response to policy instruments. Valuable
insights can be obtained this way: in the climate change policy area, time being limited, significant value lies with
any ability to predict the effectiveness of emissions reduction measures, especially when their costs and legal/political
feasibility are known in advance. However, predicting the feasibility of policies also requires knowledge of their legal
and political implications and constraints (e.g. political feasibility, legal consistency), which may be just as important.
Large numbers of policy analyses have been carried out where specific policies have been proposed to address
externalities such as pollution or environmental degradation, including taxes and other financial incentives imposed
on households, firms or consumers. The justification for taxes often lies in analyses of ethics and social justice (IPCC,
2013b, Ch. 3), however it rarely originates from analyses of their likely effectiveness (IPCC, 2013b, Ch. 15). Indeed,
policies are often chosen and adopted without much prior knowledge of their likely effectiveness, which ultimately
lies with investor and consumer decisions. Their impact is usually only assessed ex post.
Policy analysis for climate change mitigation is complex, and realistic propositions must take consideration of
existing contexts, treaties and law. For example, some forms of subsidies or feed-in tariffs, when preferential, are
inconsistent with international economic law and can attract significant legal difficulties (Vinuales, 2012). Furthermore, carbon pricing mechanisms can be politically very different from standard energy or emissions taxation, and
would thus have a different likelihood of success. Thus studying the effectiveness and feasibility of policy requires
integrating expertise on the mechanisms of policy-making and environmental law with expertise decision-making by
investors and consumers. This is, however, still uncommon in contemporary state-of-the-art climate change mitigation
modelling research.
We propose here, as part of the working structure of simulation-based environmental policy assessments, a twoway feedback with experts in environmental, investment and trade law and policy. This would effectively guide the
what if approach to scenario-creation for testing potential policies, as opposed to proposing policies that are already
optimal. Effectively, in a model that cannot be optimised, policies must be assessed based on criteria concerning
the likelihood of achieving particular objectives, when, as is typical, having to deal with several policy instruments
simultaneously that can interact with one another. To avoid any siloed approach, policy impacts across sectors must
be assessed, for which this approach is appropriate.9
3. Application to four key climate policy issues
3.1. The impact of diversity to the diffusion of new transport technology
The meaning and impact of characterising agent diversity for technology adoption can be demonstrated with data,
as shown in figure 2, using private passenger vehicle purchases as an example, taken from recent work (Mercure and
Lam, 2015). In the UK, the distribution of car purchases follows closely the income distribution (panel a). Cars of
different prices have different rated emissions, which are thus similarly distributed (panel b). Vehicles with alternative
engine technologies have yet a different distribution, and this stems from their gradual process of diffusion, which
takes place unevenly with respect to the distribution of conventional vehicles (panel c). Finally, rated emissions are
correlated with vehicle prices through a log-linear relationship.
Car purchase choices are most likely done with respect to each consumers respective social group, an assertion
demonstrated in empirical work (McShane et al., 2012), and this generally explains the relationship between consumption behaviour to the income distribution shown above. Consumers do not attempt to minimise their transportation
costs; instead they apparently purchase what is most common in their visible surroundings, and the diversity of social
groups is what forms the lognormal distribution shown above.
This therefore forces us to look at each band of income, and each type of consumer, separately. If we now explore
possible substitutions between models that could result from emissions reduction policies in the car market, we can
determine elasticities of substitution from these distributions by determining statistically which vehicle models are
9 This

approach is recommended by the European Commission in its Impact Assessment guidelines (EC, 2009).

4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

20

80

UK income distribution
UK car purchases
distribution

c) 500

2
1.5
1

0
0

20

40
60
Price ( kUSD/vehicle )

UK

200
100

d)

60
40
20

100

400

Petrol Econ
Petrol Mid
Petrol Lux
Diesel Econ
Diesel Mid
Diesel Lux
Hybrid Econ
Hybrid Mid
Hybrid Lux

350
Emissions ( gCO /km )

80

2000
4000
Engine Size (cc)

6000

All vehicles
Hybrid x 10

0
0

100

0
0

300

0
0

200

80
All vehicles
Hybrid x 10
Electric x 100

120
Sales (thousands)

300

400

100

0.5

b) 140

500

UK

400
Sales (thousands)

2.5
% population

Income ( kUSD/year )
40
60

Sales (thousands)

0
3

a)

300
250
200

200
300
Emissions (gCO2/km)

400

Sales numbers
10 000
1000
100

R2 = 0.71
R2 = 0.50
R2 = 0.63

150
100
50

20

40
60
Car Price (kUSD)

80

100

10

20
50
100
Vehicle Price ( k2012USD ), log scale

200

Figure 2: Data relating the impacts of consumer diversity and market structure on the effectiveness of emissions taxation in the private personal
mobility sector (reproduced from Mercure and Lam, 2015). a) Comparison of the UK income distribution to the price distribution of recent vehicle
purchases. b) Comparison of UK market coverage between conventional and unconventional vehicle engine technology. c) Associated distribution
of vehicle carbon intensities. d) Structure of the UK vehicle market for prices and emissions from which likely effectiveness of taxes can be
deduced. This data is available for 6 major economies (Mercure and Lam, 2015).

most likely to be chosen within the same new price band (after tax). Therefore, the rate of adoption of consumer
technology, including low carbon systems, stems precisely from this diversity, which varies across the world (as we
show in Mercure and Lam, 2015). This enables us to determine the effectiveness of market policies (taxes, subsidies)
at incentivising technology substitution, using market data. Indeed, this is exactly what is done in marketing research
in industry for the purpose of forecasting sales and placing new products in the market.
This information can also be integrated to technology diffusion models that aim to reproduce typical S-shaped
profiles (e.g. see Mercure, 2012) to provides expected rates of adoption. This picture is incomplete, as it does not
represent attitudes. However, just as with firms attempting to place new products in the market, the richness of this
type of information can give access to better characterised rates of technology adoption that could result from proposed
policies throughout the diffusion cycle, in comparison to incumbent models.
3.2. Green growth: employment and income impacts of low-carbon investments
Figure 3 shows an example of the possible causal chain in the process of technology finance impacting the economy in the electricity sector as portrayed by a post-Keynesian macroeconometric model (E3ME/E3MG Cambridge
Econometrics, 2014), the model and data further described in Mercure et al. (2015). In this model, no limit is imposed on finance for technology developments, which in contrast to an equilibrium model, has no direct relationship
to interest rates. In other words, financial resources are given to entrepreneurs by banks for low-carbon investments
through credit creation; for which our critical assumption is the solvency of both entrepreneurs and the viability of
low-carbon projects. Therefore we do not claim that the economic system has no limit at all on money flows; we are
4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

North America
Europe

a. Investment (2000bn$)

China
India
25
20

Difference ( bn$ )

100

Difference ( % )

1000

10

500

0
4

d. Electricity prices

15

40

0
1
2

10

0
h. Consumer price index

4 i. Real GDP
3
2

0
1

1
2020

2030

2040

2050

f. Real disposable income

10

20

5
2010

e. Employment

0
15 g. Consumption
Difference ( % )

b. Gov. Spending on subsidies 1500 c. Gov. income from ETS

15

50

60

Brazil
Rest of the World

2010

2020

2030

2040

2050

2
2010

2020

2030

2040

2050

Figure 3: Example of possible economic impacts of investment in low-carbon electricity generators, when studied using a non-equilibrium postKeynesian model, E3MG-FTT.

clear about our assumptions in which we ensure that all technology ventures that are financed in our scenarios are
profitable (as ensured by sustained policy and/or higher prices, e.g. higher electricity prices here), and that we do not
indefinitely increase government and/or private debt. This ensures that we do not simulate what could misleadingly
lead to a financial crisis.
In this example, higher costs of electricity generation technologies (e.g. wind turbines, solar panels, carbon capture
and storage) are passed on by utilities into their bills to customers, i.e. into a higher price of electricity, proportionally
to the evolving technology composition. The assumption is thus that entrepreneurs take loans to finance the new
capital, which is more expensive than fossil fuel generators. Larger loans are paid back over capital lifetime using a
higher price of electricity (and, for example possible feed-in tariffs). In the model, the higher level of money flows
from investment (panel a) creates jobs (panel e), increases disposable income (panel f) and consumption (panel g),
with possible effects on inflation (panel h). Meanwhile, higher prices of electricity increase operational costs of most
sectors, and thus decrease employment, household income and consumption (same panels), i.e. an opposing force.
These two effects were observed to roughly cancel each other in the model and scenario (not shown here). It appears
that a positive green growth effect, in this particular case, appears with the addition of a redistribution policy, where
revenue raised by fuel taxes to incentivise technological change (e.g. carbon pricing) is recycled to lower income
tax, which pushes the system balance towards higher income than in a baseline scenario (higher GDP, panel i). This
4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

20
10

2000

2020

2040

d.
1980

2000

2020

2040

15
1990

1980

2000

Nuclear
Oil

2020
Coal

Coal+CCS

2040
Gas

1980

Gas+CCS

2000
Biomass

Biomass+CCS

2020

Wind

Solar

2010

f.

2020

2030

2040

2050

2040

2060

2080

2100

4
3
2
1
0

2040
Hydro

400

20
5

450

30

Global Warming (oC)

Emissions ( Gt )

40

20

500

50

25

550

60

1990

0
c.
1980
30

10

e.

30

Projection

CO concentration (ppm CO )

Historical data

70

b.

Global emissions ( GtCO2 )

a.

Decarbonisation

2008

40

Business As Usual

2008

Electricity Generation ( GWh )

50

Geothermal

2020

Ocean

BAU

Mitigation

Figure 4: Example calculation of the environmental impacts of electricity policy instruments using E3MG/E-FTT (left 4 panels) with OU emulators
(right panels), from policy to global warming, with cascading uncertainty bounds from combined carbon cycle and climate system emulators.

effect subsides in later years when investment and redistribution declines, and may even reverse when the technology
transformation is completed and society faces servicing debt, which takes place through a higher price of electricity,
but at this point in time without benefiting from earlier money flows from banks issuing loans.
Note, therefore, that while economic growth is enhanced in this scenario, private debt is also increased until
beyond the end of the simulation. In non-equilibrium theory, borrowed investment flows indirectly contribute to
increased aggregate demand, in the short run when loans are issued, and decreased demand in the long run, when
loans are gradually paid back. Increasing the level of borrowing generates debt-based growth, the likely origin of the
recent banking crisis (see Keen, 2011), and possibly most economic cycles (Perez, 2001). This points to an intimate
relationship between the economic crisis, the financial sector and climate change mitigation policy that needs to be
better discussed and understood.10 We therefore conclude here that more research is required to be carried out on
understanding the levels of debt and risk, private and public, that economies are able to realistically undertake for
reducing global emissions. In a complexity perspective, this appears to be the key issue to explain, rather than short
run changes in income that are typically analysed with current models.
3.3. Model integration across processes with cascading uncertainty
Large models of the natural world have many imperfectly known parameters and, though multiple different possible combinations of those parameters, a collossal number of possible output values given variations in those parameters (even 10 settings of only 10 parameters would generate 1010 possibilities). Statistical modelling techniques are
10 As Keen (2011) argues, it is investment in assets, rather than investment in productive capacity, that leads to Ponzi-type schemes that can
become prone to sharp changes in value and lead to financial crises ; when expected increases in the value of existing assets is used to leverage
borrowing for investing further in assets, fragile Ponzi pyramids are built which inevitably later collapse.

4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

required to interpret such a large space of uncertain outcomes. When soft-linking several models in a causal chain,
the uncertainty of models further up the chain must inevitably generate higher uncertainty further down the chain
as we give ever wider ranges of parameters to models. Figure 4 gives an example of this with 3 models chained to
one another, starting with individual emissions scenarios from E3MG-FTT. In (a.), we have a baseline scenario of
the composition of the global electricity sector, calculated under 21 regions independently. Using a composite scenario of emissions reduction policies that include carbon pricing, technology support policies (subsidies and feed-in
tariffs) and regulations (available in Mercure et al., 2014), the electricity sector is transformed towards low-carbon
technologies (b.). Baseline emissions are projected to increase by 318% based on their 1990 level (c.), while in the
decarbonisation scenario, they are reduced by 90% (d.).
These emissions scenarios are fed to the carbon cycle emulator GENIEem, which generates GHG concentration
outputs with 95% probability ranges (e.). These are then fed to the emulator of the climate model PLASIM-ENTS,
producing a set of scenarios for global warming (f.), as well as other locally resolved changes of climate (on a 0.5
grid, not shown). This thus produces double uncertainty ranges from the concatenation of uncertainty. We find
that decarbonising the electricity sector by 90% is not sufficient to avoid exceeding 50% chance of exceeding the
international target of 2 C (Mercure et al., 2014). All sectors of the energy end-use system must be involved, notably
transport.
3.4. Cross-sectoral impacts of biofuels policy
In an optimisation model of land allocation, best allocations of farming activities are found across the surface of
the land studied, which includes the whole world for global models. This implies that farmers know instantaneously
what is too little and too much, and never generate excess product. Without excess product or shortages, price
fluctuations cannot occur. This is, of course, not what is observed (Piesse and Thirtle, 2009). However, building a
model that can reproduce observed price fluctuations is a significant challenge. These fluctuations are likely important
for understanding environmental degradation and biodiversity loss; indeed some deforestation occurs not just due to an
increasing consumption of agricultural commodities globally, but also due to commodity prices fluctuating globally,
due to and generating expectations of returns (e.g. Arima et al., 2011, Morton et al., 2006).
In a non-equilibrium model that includes both representations of decision-making in agriculture under expectations
of return and time delays, with a global model of the economy and bilateral trade of agricultural commodities, different
dynamics emerge (Arneth et al., 2014, Rounsevell et al., 2014). While price substitutions and changes in trade patterns
can occur with the consumption of agricultural commodities, massive land clearances and conversions can also take
place to accommodate expectations around changing prices. While this generates highly complex dynamics, it is
also useful to determine what types of transformations could occur in the future without further land management
policy to control land use change occurring as a result of mounting global pressures. Such dynamics however will
only be observed in a model that incorporates land use decision-making under expectations by heterogeneous agents
and non-equilibrium economic dynamics with time delays. If land-use decisions are assumed to be made based on
expectations and be influenced by neighbourhood effects, then similar contagion dynamics as to technology diffusion
would take place. This needs further investigation.
Biofuels policy brings quite complex uncertainties and dynamics into this picture. While it can be summarised
by stating that the willingness to pay for ethanol in some parts of the world could outbid the ability to pay for food
commodities in other parts of the world, it may also be the case that shortages of food could be felt through commodity
prices after land-use decisions are already taken and applied. Effectively, some biofuels policies have the potential
to open a door to substantial fluctuations in food prices, but we do not know which, and at which scale. This issue
requires models of a type that is currently not available.
4. Conclusion: a world of new possibilities for environmental assessment
Optimisation and equilibrium analysis and models are appropriate to use for normative exploration and identification of feasible and desirable future states of the technology-economy-environment system. In particular, since they
are comparatively highly detailed and tested, they can achieve this very well. This is because such activity does not
require knowing anything beyond physical and technical limits, i.e. no knowledge on human behaviour is required.
However, they provide ambiguous information on how (policy-wise) to achieve these ideal configurations, or on the
4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

likelihood of such scenarios happening, due to the lack of causal relationships with human behaviour. Producing
scenarios that predict the future course of events as a result of policy choices, with any likelihood (however low),
requires representations of human behaviour, its diversity and interpersonal interactions. These representations are
necessarily imperfect; however, they do bring models to cross over from normative to predictive.
It is often argued that this is not possible, and thus serious sustainability science can only express itself in the form
of exploratory scenarios (e.g. van Vuuren et al., 2011), not forecasts. We disagree, by taking the climate sciences
as an example where outcomes are expressed in terms of likelihood levels based on model statistics. Forecasts are
not always realised; they are nevertheless useful due to their careful quantification of probability given model theory,
parameters and their uncertainty. There does not exist any inherent reason why this is not possible in the social
sciences.11 The problem resides in the quantification of likelihood; however it is not possible with current mainstream
models. We do not claim, however, to be able to predict the onset of wars, election results, natural disasters, strategic
political choices or other unique events. Thus caveats appear in our claim, where some types of events are assumed
not to take place, for clarity of analysis (just like weather forecasts, which do not take into account possibilities of
volcanic eruptions, even in active areas).
The modelling paradigm shift proposed here opens many possibilities for environmental assessment. This involves
in particular the quantification of the expected effectiveness of specific policies at inciting particular agent groups to
take particular decisions (e.g. consumer purchases, investment choices, land-use decisions). High gains lie with any
ability to do this well: it reduces significantly the uncertainty involved in creating policy portfolios, and it makes
possible the exploration of their impacts, and in particular, across sectors when several sectoral models are coupled.
For example, one could study the impact of technology support policy for electric vehicles on electricity prices, which
will depend on the pace of their adoption. One could study the impact on food prices of large scale land regulations
to protect biodiversity. One could study the impact on deforestation of biofuels policy in transport. These subjects of
debate present major challenges to analysis, but require to be addressed rapidly, and this will require a new generation
of non-equilibrium models in order to be correctly addressed.
Acknowledgements
The authors wish to thank the full GUIDEPOST team for highly fruitful discussions over the length of funding
proposal developments. J.-F. M. thanks D. Crawford-Brown, T. Barker and 4CMR staff and students for support. J.-F.
M. and H. P. thank Cambridge Econometrics Ltd staff for support. J.-F. M. acknowledge the UK Engineering and
Physical Sciences Research Council (EPSRC), fellowship no EP/ K007254/1 and networking grant of the Newton
Fund EP/N002504/1.
References
Andersen, E. S., 1994. Evolutionary Economics, Post-Schumpeterian Contributions. Pinter Publishers.
Anderson, P. W., 1972. More is different. Science 177 (4047), 393396.
Anderson, P. W., Pines, D., Arrow, K. J., 1989. The economy as an evolving complex system. Westview press.
Anderson, S. P., De Palma, A., Thisse, J. F., 1992. Discrete choice theory of product differentiation. MIT press.
Arima, E., Richards, P., Walker, R., Caldas, M., 2011. Statistical confirmation of indirect land use change in the brazilian amazon. Environmental
Research Letters 6 (2).
Arneth, A., Brown, C., Rounsevell, M., 2014. Global models of human decision-making for land-based mitigation and adaptation assessment.
Nature Climate Change 4 (7), 550557.
Arthur, B., 2014. Complexity and the economy. Oxford.
Arthur, W. B., 1989. Competing technologies, increasing returns, and lock-in by historical events. The economic journal 99 (394), 116131.
Arthur, W. B., 1999. Complexity and the economy. Science 284 (5411), 107109.
Arthur, W. B., Polak, W., 2006. The evolution of technology within a simple computer model. Complexity 11 (5), 2331.
Axsen, J., Mountain, D. C., Jaccard, M., 2009. Combining stated and revealed choice research to simulate the neighbor effect: The case of
hybrid-electric vehicles. Resource and Energy Economics 31 (3), 221 238.

11 Of

the sciences, the only two theories that are inherently stochastic (truly random) are quantum physics and evolution in biology, surprisingly,
not social sciences. Furthermore, stochasticity in evolutionary theory stems from quantum physics. All existing theories of human behaviour
remain deterministic, and the difficulty in predicting human behaviour always resides in our lack of complete contextual information. While the
behaviour of individuals cannot be reliably predicted, the behaviour of large groups often follows clear trends, or ordering principles.

4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

Barker, T., Pan, H., Kohler, J., Warren, R., Winne, S., 2006. Decarbonizing the global economy with induced technological change: Scenarios to
2100 using E3MG. Energy Journal (Sp. Iss. 1), 241258.
Bass, F. M., 1969. New Product Growth for Model Consumer Durables. Management Science Series A-theory 15 (5), 215227.
Ben-Akiva, M. E., Lerman, S. R., 1985. Discrete choice analysis: theory and application to travel demand. MIT press.
Bhargava, S. C., 1989. Generalized Lotka-Volterra Equations and the Mechanism of Technological Substitution. Technological Forecasting and
Social Change 35 (4), 319326.
Blyth, W., Gross, R., Speirs, J., Sorrell, S., Nicholls, J., Dorgan, A., Hughes, H., 2014. Low carbon jobs: The evidence for net job creation from
policy support for energy efficiency and renewable energy. Ukerc reports, UKERC.
URL http://www.ukerc.ac.uk/publications/low-carbon-jobs-the-evidence-for-net-job-creation-from-policy-support-for-energy-e
html
Cambridge Econometrics, 2014. Energy-economy-environment model at the european level.
URL www.e3me.com
CBI, 2014. The colour of growth: Maximising the potential of green business. Cbi reports, CBI UK.
URL http://www.cbi.org.uk/media/1552876/energy_climatechangerpt_web.pdf
Domencich, T. A., McFadden, D., 1975. Urban travel demand - A behavioural analysis. North-Holland Publishing.
Douglas, M., Isherwood, B., 1979. The world of goods: Towards an anthropology of consumption. Routledge.
EC, 2009. Impact assessment guidelines. Tech. rep., European Commission.
URL http://ec.europa.eu/smart-regulation/impact/commission_guidelines/docs/iag_2009_en.pdf
Edenhofer, O., Carraro, C., Kohler, J., Grubb, M., 2006. Endogenous Technological Change and the Economics of Atmospheric Stabilisation,
Special Edition, Energy Journal. International Association for Energy Economics.
Edenhofer, O., Knopf, B., Barker, T., Baumstark, L., Bellevrat, E., Chateau, B., Criqui, P., Isaac, M., Kitous, A., Kypreos, C., Leimbach, M.,
Lessmann, K., Magne, B., Scrieciu, S., Turton, H., van Vuuren, D. P., 2010. The Economics of Low Stabilization: Model Comparison of
Mitigation Strategies and Costs. Energy Journal 31 (1), 1148.
Edenhofer, O., Lessmann, K., Kemfert, C., Grubb, M., Kohler, J., 2006. Induced technological change: Exploring its implications for the economics
of atmospheric stabilization: Synthesis report from the Innovation Modeling Comparison Project. Energy Journal (Sp. Iss. 1), 57107.
Fisher, J. C., Pry, R. H., 1971. A simple substitution model of technological change. Technological Forecasting and Social Change 3 (1), 7588.
Freeman, C., Louca , F., 2001. As Time Goes By. Oxford University Press.
GEA (Ed.), 2012. Global Energy Assessment. Cambridge University Press, Cambridge, UK and New York, NY, USA and the International Institute
for Applied Systems Analysis, Laxenburg, Austria.
Geels, F. W., 2002. Technological transitions as evolutionary reconfiguration processes: a multi-level perspective and a case-study. Research Policy
31 (8-9), 1257 1274.
Giraudet, L.-G., Guivarch, C., Quirion, P., 2012. Exploring the potential for energy conservation in french households through hybrid modeling.
Energy Economics 34 (2), 426 445.
Grubb, M., Hourcade, J.-C., Neuhoff, K., 2014. Planetary economics. Taylor Francis / Routledge, London, UK.
Grubler, A., Nakicenovic, N., Victor, D., 1999. Dynamics of energy technologies and global change. Energy Policy 27 (5), 247280.
Holden, P. B., Edwards, N. R., 2010. Dimensionally reduced emulation of an aogcm for application to integrated assessment modelling. Geophysical
Research Letters 37 (21).
Holden, P. B., Edwards, N. R., Gerten, D., Schaphoff, S., 2013. A model-based constraint on CO2 fertilisation. Biogeosciences 10 (1), 339355.
IEA, 2007. Climate policy uncertainty and investment risk. Tech. rep., IEA.
IEA, 2013. Global ev outlook - understanding the electric vehicle landscape to 2020. Tech. rep., IEA.
IMF, 2014. Peoples republic of china: 2014 article iv consultation, imf country report no. 14/235. Tech. rep., IMF.
URL http://www.imf.org/external/pubs/ft/scr/2014/cr14235.pdf
IPCC, 2013a. Climate Change 2013: The Physical Science Basis. Cambridge University Press.
IPCC, 2013b. Climate Change 2014: Mitigation of Climate Change. Cambridge University Press.
IPCC, 2014. Technical summary. In: Climate Change 2014: Mitigation of climate change. IPCC Working Group III Contribution to AR5. Cambridge University Press.
Kahneman, D., Tversky, A., 1979. Prospect theory: an analysis of decision under risk. Econometrica 47 (2), 263291.
Karmeshu, Bhargava, S., Jain, V., 1985. A rationale for law of technological substitution. Regional Science and Urban Economics 15 (1), 137141.
Keen, S., 2011. Debunking Economics - Revised and expanded edition. Zed Books.
Keynes, J. M., 1936. The General Theory of Employment, Interest and Money. Palgrave-Macmillan.
Labriet, M., Joshi, S., Vielle, M., Holden, P., Edwards, N., Kanudia, A., Loulou, R., Babonneau, F., 2013. Worldwide impacts of climate change on
energy for heating and cooling. Mitigation and Adaptation Strategies for Global Change, 126.
Mansfield, E., 1961. Technical change and the rate of imitation. Econometrica 29 (4), pp. 741766.
Marchetti, C., Nakicenovic, N., 1978. The dynamics of energy systems and the logistic substitution model. Tech. rep., IIASA.
URL http://www.iiasa.ac.at/Research/TNT/WEB/PUB/RR/rr-79-13.pdf
McLeay, M., Radia, A., Thomas, R., 2014. Money creation in the modern economy. Tech. rep., Bank of England.
URL http://www.bankofengland.co.uk/publications/Pages/quarterlybulletin/2014/qb14q1.aspx
McShane, B. B., Bradlow, E. T., Berger, J., 2012. Visual influence and social groups. Journal of Marketing Research 49 (6), 854871.
Meinshausen, M., Meinshausen, N., Hare, W., Raper, S., Frieler, K., Knutti, R., Frame, D., Allen, M., 2009. Greenhouse-gas emission targets for
limiting global warming to 2c. Nature 458 (7242), 11581162.
Mercure, J.-F., 2012. Ftt:power : A global model of the power sector with induced technological change and natural resource depletion. Energy
Policy 48 (0), 799 811.
URL http://dx.doi.org/10.1016/j.enpol.2012.06.025
Mercure, J.-F., 2015. An age structured demographic theory of technological change. In press, J. Evol. Econ.
URL http://arxiv.org/abs/1304.3602

4CMR Working Paper no 10

Cambridge Centre for Climate Change Mitigation Research (4CMR)

Mercure, J.-F., Lam, A., 2015. The effectiveness of policy on consumer choices for private road passenger transport emissions reductions in six
major economies. Environ. Res. Lett. 10 (064008).
Mercure, J.-F., Pollitt, H., Chewpreecha, U., Salas, P., Foley, A., Holden, P., Edwards, N., 2014. The dynamics of technology diffusion and the
impacts of climate policy instruments in the decarbonisation of the global electricity sector. Energy Policy 73 (0), 686 700.
Mercure, J.-F., Salas, P., Foley, A., Chewpreecha, U., Pollitt, H., Holden, P. B., Edwards, N. R., 2015. Complexity, economic science and possible
economic benefits of climate change mitigation policy. Under review, preprint available on ArXiv.
URL http://arxiv.org/abs/1310.4403
Morton, D. C., DeFries, R. S., Shimabukuro, Y. E., Anderson, L. O., Arai, E., del Bon Espirito-Santo, F., Freitas, R., Morisette, J., 2006. Cropland
expansion changes deforestation dynamics in the southern brazilian amazon. Proceedings of the National Academy of Sciences 103 (39),
1463714641.
Muth, J. F., 1961. Rational expectations and the theory of price movements. Econometrica: Journal of the Econometric Society, 315335.
Myhre, G., Shindell, D., Breon, F.-M., Collins, W., Fuglestvedt, J., Huang, J., Koch, D., Lamarque, J.-F., Lee, D., Mendoza, B., Nakajima, T.,
Robock, A., Stephens, G., Takemura, T., Zhang, H., 2013. Anthropogenic and natural radiative forcing. In: Climate Change 2013: The Physical
Science Basis. Cambridge University Press.
Nelson, R. R., Winter, S. G., 1982. An Evolutionary Theory of Economic Change. Harvard University Press, Cambridge, USA.
Nordhaus, W. D., 2010. Economic aspects of global warming in a post-Copenhagen environment. Proceedings of the National Academy of Sciences.
OXFAM, 2012. A safe and just space for humanity. Tech. rep., OXFAM.
URL https://www.oxfam.org/sites/www.oxfam.org/files/dp-a-safe-and-just-space-for-humanity-130212-en.pdf
Oyebamiji, O. K., Edwards, N. R., Holden, P. B., Garthwaite, P. H., Schaphoff, S., Gerten, D., 2015. Emulating global climate change impacts on
crop yields. Statistical Modelling.
Perez, C., 2001. Technological Revolutions and Financial Capital. Edward Elgar.
Piesse, J., Thirtle, C., 2009. Three bubbles and a panic: An explanatory review of recent food commodity price events. Food Policy 34 (2), 119
129.
Porter, M. E., 1991. Americas green strategy. Scientific American 264 (4), 168.
Porter, M. E., van der Linde, C., 1995. Toward a new conception of the environment-competitiveness relationship. Journal of Economic Perspectives
9 (4), 97118.
Probst, G., Bassi, A., 2014. Tackling Complexity. Greenleaf Publishing.
Rivers, N., Jaccard, M., 2006. Useful models for simulating policies to induce technological change. Energy Policy 34 (15), 2038 2047.
Rockstrom, J., Steffen, W., Noone, K., Persson, ., Chapin, F. S., Lambin, E. F., Lenton, T. M., Scheffer, M., Folke, C., Schellnhuber, H. J., et al.,
2009. A safe operating space for humanity. Nature 461 (7263), 472475.
Rogers, E. M., 2010. Diffusion of innovations. Simon and Schuster.
Rounsevell, M. D. A., Arneth, A., Alexander, P., Brown, D. G., de Noblet-Ducoudre, N., Ellis, E., Finnigan, J., Galvin, K., Grigg, N., Harman,
I., Lennox, J., Magliocca, N., Parker, D., ONeill, B. C., Verburg, P. H., Young, O., 2014. Towards decision-based global land use models for
improved understanding of the earth system. Earth System Dynamics 5 (1), 117137.
Safarzynska, K., van den Bergh, J. C. J. M., 2010. Evolutionary models in economics: a survey of methods and building blocks. Journal of
Evolutionary Economics 20 (3), 329373.
Schumpeter, J. A., 1934. The Theory of Economic Development - An inquiry into Profits, Capital, Credit, Interest and the Business Cycle (translated
from original 1911 version). Harvard University Press, Cambridge, USA.
Schumpeter, J. A., 1939. Business Cycles. McGraw-Hill.
Sigmund, K., 1993. Games of Life. Oxford University Press, Oxford, UK.
Smith, W. R., 1956. Product differentiation and market segmentation as alternative marketing strategies. The Journal of Marketing, 38.
Steffen, W., Richardson, K., Rockstrm, J., Cornell, S. E., Fetzer, I., Bennett, E. M., Biggs, R., Carpenter, S. R., de Vries, W., de Wit, C. A., Folke,
C., Gerten, D., Heinke, J., Mace, G. M., Persson, L. M., Ramanathan, V., Reyers, B., Srlin, S., 2015. Planetary boundaries: Guiding human
development on a changing planet. Science 347 (6223).
Stern, N., 2007. The Economics of Climate Change. Cambridge University Press, Cambridge, UK.
Tebaldi, C., Arblaster, J., 2014. Pattern scaling: Its strengths and limitations, and an update on the latest model simulations. Climatic Change
122 (3), 459471.
Kohler, J., Whitmarsh, L., Nykvist, B., Schilperoord, M., Bergman, N., Haxeltine, A., 2009. A transitions model for sustainable mobility. Ecological
Economics 68 (12), 29852995.
Tversky, A., Kahneman, D., 1974. Judgment under uncertainty: Heuristics and biases. science 185 (4157), 11241131.
van Vuuren, D. P., Edmonds, J. A., Kainuma, M., Riahi, K., Weyant, J., 2011. A special issue on the RCPs. Climatic Change 109 (1-2), 14.
Vinuales, J., 2012. Foreign Investment and the Environment in International Law. Cambridge University Press.
`
`
`
Walras, L., 1874. . . . lEments
dEconomie
politique pure; ou, ThEorie
de la richesse sociale. L. Corbaz & cie.

4CMR Working Paper no 10