Determinants of emissions pathways in the coupled climate–social system

The positive and negative feedback processes operating within the coupled climate–social system are critical to understanding system behaviour and dynamics. In a two-step process, the feedback processes that are represented by the model were identified. First, the workshop was a four day interdisciplinary workshop that identified potentially relevant system feedback processes. Second, eight key feedback processes were identified by targeted searches across relevant literatures in psychology and economics. These results were compiled from a variety of disciplines including law, sociology, law and political science. This section briefly explains each feedback process and Table 1Fig. 1Describe how these feedback processes are linked in the model and the structure.

The components are shown in black, and the model feedback processes are in green. Feedback processes are identified as positive (+) (that is, reinforcing) or negative (−) (that is, dampening). The black arrow represents a link between components (policy–adoption effect), that is not directly connected to a particular feedback mechanism. Table contains descriptions of the components and key parameters that govern both component behaviour and feedback strength. 1.

Social-conformity feedback

Social networks that are created at work, home, school, or for leisure can have a significant impact on opinions and behavior.^30,31. Social norms, that is, the dominant or accepted practices or opinions within an social group, are expensive to violate and can have a lasting impact on individuals’ identities, habits, world-views and lifestyles.^32,33. Study in the USA has shown that belief in climate change and support of climate policies can partly be explained by perceived social consensus.³⁴. A large amount of literature has also demonstrated that social norms are an important determinant of whether or not someone is pro-environmental. This includes adopting solar panels or conserving energy.^35,36,37. A tendency toward social conformity can lead towards tipping-point-type dynamics where a system transitions suddenly out of a previously stable state if there is enough support for the alternative norm^24,38. The social conformity effect can be described in two ways. It involves the formation of public opinion on climate policy and individual actions to adopt pro-climate behavior (Fig. 1).

Climate change perception feedback

The anthropogenic influence on the Earth’s climate system is increasingly apparent^39,40,41. It is becoming more common to assess the contribution of anthropogenic global warming to the likelihood of extreme events.⁴². It has been hypothesized that this emerging signal of climate change in people’s everyday experience of weather might lead to widespread acknowledgement of the existence of global warming and possibly, by extension, support for mitigation policy⁴³. Numerous studies have linked global warming belief with local temperature anomalies. People seem to be able identify local warming^44,45If the weather is unusually warm, or is perceived as being so, people are more likely to believe in climate change.^46,47,48,49. People seem to be using their personal experience with weather to support their belief in climate changes.⁴⁹.

However, this so-called ‘local warming effect’ is complicated⁵⁰. Numerous papers have shown that interpretations of weather events can be filtered through pre-existing ideologies or partisan identities.^45,51,52. This suggests the presence of motivated reasoning (that is, the rejection of new information that contradicts pre-existing beliefs) in the processing of climate-change-related information^53,54. Moreover, the perception of weather anomalies might well be complicated by a ‘shifting-baselines’ effect in which people’s perception of normal conditions is quickly updated on the basis of recent experience of weather⁵⁵.

Feedback on political interest

Individual actions cannot make the large-scale reductions in emissions necessary to stabilize the climate. It is therefore crucial to consider how support for or opposition to climate policy can translate into collective action through the political process. This process is not straightforward—it is subject to political–economic constraints operating through complex political and government institutions and cannot be modelled as a simple linear function of public opinion^56,57,58. According to the political economy literature, there is a positive feedback effect whereby policy changes are initially implemented by powerful interests that can lobby against policy reversal or for further change. Texas’s establishment of the wind industry is one example.^26,27. While most examples are of reinforcing feedback processes in the literature, Stokes is a notable exception.²⁷ also documents instances of balancing feedback processes—where small policy changes activate powerful incumbents to lobby against further changes that threaten their interests.

Credibility-enhancing display feedback

Individuals may not be able to alter the trajectory or emissions of greenhouse gases, but individual pro-environmental behavior can have a ripple effect on the larger social network. Changing behaviour to better align one’s consumption or practices with how one believes society ought to function can strengthen this moral identity and send a normative signal to other community members about desirable collective outcomes^59,60. Engaging in costly personal actions aligned with collective goals can act as ‘credibility enhancing displays’, increasing the persuasiveness of the actor. Kraft-Todd et al.⁶¹This framework can be used to explain why community ambassadors for solar panel installation are more effective if the have installed solar themselves. Attari Krantz, Weber and Weber discuss climate change in general.^62,63It was found that the credibility and impact of climate policy advocates’ messages is affected by their personal carbon footprints.

Feedback on the expressive force of law

Legal or judicial institutions that are perceived as legitimate can give information about the desired or common attitudes of the population. These changes can feed back to reinforce or change the behaviour or attitudes of the society that produced them. Tankard and Paluck⁶⁴One of the three sources of information about community norms is signals from governing bodies. Legal scholars have developed the theory of the ‘expressive function’ of law—the idea that law and regulation work on society not only by punishing undesirable behaviour but also by signalling what kind of behaviour is praiseworthy and what is reprehensible^65,66,67. This signal is especially important if individuals do not have complete information about the distributions in attitudes or behaviours within a reference group^67,68. Numerous papers have shown evidence of feedback from changes to laws and regulations on the perception of social norms. This includes the legalization of gay marriage.^69,70, smoking bans⁷¹COVID-19 lockdowns⁷².

Endogenous cost-reduction feedback

New energy technologies are not only expensive but also tends to show a decline in price with increased installed capacity. This ‘learning-by-doing’ effect has been widely documented in the energy systems literature and is incorporated into some energy system models⁷³. Falling costs can be attributed to economies of scale, lower input prices, and efficiencies in production and design.⁷⁴. This is a feedback loop that encourages further deployment. Small initial deployments, which may be driven by subsidies or regulatory requirements, can lower costs. Rubin et al.⁷⁵ reviewed estimated learning rates (that is, the fractional reduction in cost for a doubling of installed capacity) for 11 generation technologies and found ranges between −11% and 47% with many estimates falling in the 2% to 20% range.

Temperature–emissions feedback

Climate change will have a wide impact on economic sectors and geographical regions. These effects could have an impact on the economy’s ability to produce emissions. Some studies have shown that warming has significant effects on economic growth.^76,77This could lead to a significant reduction in economic production and a consequent decrease in greenhouse gas emissions. There are other effects from warming on energy demand.⁷⁸Or on the carbon intensity of energy generation^79,80This effect might be partially or completely offset. Woodard et al.⁸ provide a central estimate of these combined effects of a 3.1% decline in emissions per degree of warming, with upper and lower bounds ranging from −10.2% to 0.1%.

The model developed here is designed to investigate the complex, emergent behaviour of the coupled climate–social system, including the feedback processes described above. Figure 1The six major model components are displayed at four interconnected scales: Individual (cognition component), Social (opinion and Adoption components), National (Policy component), and Global (Emissions and Climate Components). Table gives descriptions of the key parameters and processes in each component. 1, and equations are fully documented in Methods and the ‘Model documentation’ section of the Additional Information.

Tipping points, interactions, thresholds

The model components that are subject to coupled feedback can create complex, highly nonlinear behaviour. This behaviour can depend on the interactions between social, technical, and political systems. We will demonstrate this behaviour by performing three systematic explorations within the model parameter space. This is to highlight interactions between scales as well as model components. These values were chosen to highlight tipping point and threshold behaviour in our model. They are not necessarily the most representative or likely values. In the next section, we will discuss the constraints that affect the distribution of parameter values. Each panel of Fig. 2 shows model output, systematically varying 2–3 parameters while keeping all of the other model parameters fixed at the values given in Extended Data Table 1.

A, Illustration of a tipping point associated with individual adoption of behavioural change by climate policy supporters through the credibility-enhancing display feedback. b, The interactions between endogenous energy sector cost reductions and the opinion (fraction climate policy supporters) or policy (status-quo bias) components. c, The effects of the climate perception feedback on public opinion and specific cognitive biases. The Extended Data Table shows that the model parameters not listed in each figure panel will be maintained constant for all runs of the model at the same values. 1.

Individual behavioural change

Figure 2a demonstrates the potential for tipping points associated with individuals’ adoption of behavioural change. Because individuals have limited control over how societies produce and consume energy, the primary effect of behaviour changes on emissions is minimal. The temporary reduction of global CO2 emissions was temporarily caused by the COVID-19 lockdowns. This lockdown, which represents a major and unprecedented global change in mobility patterns and consumption patterns, has been extremely successful.₂Emissions by anywhere between 9% to 17% (refs. ^81,82This gives an upper limit on the potential impact of behavioral change on reducing carbon footprints. Our RCP7 baseline shows that global emissions will nearly double by 2100. This clearly indicates that it is not sufficient to provide the deep decarbonization required to stabilize global temperatures.

However, Fig. 2aThis study shows that climate policy supporters’ willingness to make costly pro-climate behavioural changes can be crucial in triggering positive feedback processes that tip it into a sustainable state. This interaction operates through the credibility-enhancing display feedback from adoption to opinion; if this feedback is small or absent, then no amount of individual action can drive major emissions reduction. However, if this feedback is strong, then behavioural change by climate policy supporters persuades more people to support climate policy, an effect that triggers a cascade of positive feedback processes in the opinion (social-conformity feedback) and mitigation (learning by doing) components that drive emissions to zero by 2100.

Learning by doing

Figure 2bIt illustrates the interaction between technological changes in the energy system and public opinion dynamics. On average, greater emissions reductions are associated with higher endogenous cost reductions. However, as this technological feedback must be initiated by climate policy, there is a threshold effect—a large nonlinear change in model behaviour at a particular parameter value—associated with the fraction of the population supporting climate policy. Below a threshold level, there is no policy that will drive the initial deployment necessary to kickstart the cost reduction feedback. Even beyond this threshold, higher levels support result in faster deployment and a greater effect of endogenous costs reductions (as shown by the steepening curve lines at the top). The two panels in Figure. 2bThis illustration shows how the characteristics and relationships of political institutions can affect this relationship: those that respond less to public opinion (or high status quo bias), (Fig. 2b bottom) have a higher threshold for policy support and ramp up climate policy more slowly, leading to higher cumulative emissions over the twenty-first century, even in the presence of a strong cost-reduction feedback in the energy sector.

Perception of climate changes

Figure 2cThe following illustration illustrates how information about the climate system could influence public opinion dynamics, if observation of weather affects support climate policy (that’s, the climate perception feedback). This feedback can have a significant impact on opinion dynamics as shown by the threshold behaviour at zero. Even for small effects of perceived weather on climate policy opinions, model behaviour is significantly different from model behaviour without any perception effect. However, this effect is mitigated in the presence cognitive biases which can fully offset the cognition-feedback. Model runs that used a fixed baseline to perceive temperature anomalies (Fig. 2cLeft), the majority of people support climate policy regardless of bias assimilation. This is because they believe that the perceived changes in weather are so large.

The presence of shifting baselines (Fig. 2cThis effect is complicated by the fact that it is not always obvious. A stronger perception feedback can lead to more climate policy opponents in 2050, especially if it is biased assimilation. This is because if baselines change and people compare current weather to the past 8 year, they will occasionally perceive unusually cold anomalies caused by natural weather variability. Even though temperatures are relatively warm relative to a preindustrial fixed baseline.⁵⁵. These perceived cold anomalies are reinforced by climate policy opponents’ belief in their position. This leads to the persistence of this opinion.

Limiting the parameter range

The illustrations in the previous section highlight how coupled socio-politico-technical feedback processes across components and scales in the climate–social system can produce nonlinear behaviour leading to a wide range of twenty-first century emission trajectories. This complexity is a result of climate policy being modelled as an exogenous product of fundamental social and political forces. To identify the most likely outcomes within this range, you need to set some parameters.

The model is a highly aggregated and abstracted representation of the coupled climate–social system, meaning that parameterization is not straightforward. Two exercises were performed using hindcasting performance to constrain the parameter space. The first exercise used the population weighted time series of public opinions on climate change in nine OECD nations (the USA, Canada France, Germany Italy, Spain, Spain, Australia, Japan, and the UK) between 2013-2020 from Pew Research Center.⁸³The average carbon price per metric ton for the same countries, over the same time period.⁸⁴To jointly constrain nine parameters in the cognition and opinion components.

The second exercise used the most recent Swedish carbon prices estimates to constrain two parameters of the emissions component.⁸⁵. Although only a tiny fraction of global emissions, the Swedish case is important because Sweden has had the world’s highest carbon price for several decades⁸⁴This allows for estimates of the impact of high and sustained carbon prices upon emissions. The model uses a single abatement price function. This implicitly assumes that Swedish abatement costs can be generalized more widely, which could be a weakness.

Monte Carlo mode is used to run relevant model components, each one sampling independently from the available parameter values. Model output for each run is then compared to the observed time-series and parameter combinations are weighted on the basis of the distance between model output and observed data (Methods). The differences between the weighted and unweighted parameter distributions give an indication of how much observations can constrain the parameter value.

Extended Data Figures 1 2These exercises result in the following: Extended Data Figure 1aThe dynamics of public opinion have some impact on cognition feedback as well as social conformity. Public opinion over the past decade on climate policy suggests that the population is socially sorted within opinion group (that is, slightly lower network homophily parameters) with relatively slow movement among groups (that’s, low persuasive power) and a relatively minor role for individual perceptions about climate change in opinion formulation (low evidence parameter). The exercise provides less information regarding the policy component’s parameters, but there is evidence for status quo bias in politics.

The exercise also constrains covariance among parameters (Extended Data Figure. 1b). For example, there is covariance between the network homophily, persuasive force and shifting baseline parameters—consistency with observed changes in OECD climate opinion over time requires that opinion groups are socially separated, movement between opinion groups is slow or cognitive biases like shifting baselines limit the role of observed climate change in driving public opinion. Extended Data Figure 2This shows the second hindcasting exercise for emissions parameters. It suggests that there is a low value to the contemporaneous policy effect on emissions (maximum emission rate), but not much information about the persistence and duration of those emissions reductions.

Future emission pathways

To probabilistically analyze emissions trajectories for the twenty-first century, we used the partially constrained parameter area. We performed 100,000 runs of the model, drawing from the joint distribution of the set of hindcast parameters and sampling uniformly over an additional 11 parameters, mostly within the adoption component (with the exception of a triangular distribution for the temperature–emissions feedback based on Woodard et al.⁸). The 2020 public opinion is used to initialize the model⁸³and emissions data, and run until 2100 with fixed parameter values for each model run. We used kClustering is a way to group together models with similar trajectories in climate policy and emissions over 2021, thereby identifying five distinct pathways (Methods). Focusing on clusters of similar policies and emissions pathways allows for exploration and explanation of a variety of model behaviours, while also avoiding a focus on the extremes or the central tendency.

Figure 3This graph shows the average policy and emissions trajectories of the five clusters. The model parameter values characteristic to each cluster indicate the socio-politico-technical states determining each policy–emissions trajectory. Extended Data Fig. shows these parameter values visually. 3. Table 2The following describes the different pathways and the end-of-century warming in each cluster based on the mean emissions scenario.

Policy (left) & global CO₂ emissions (right) trajectories from 100,000 Monte Carlo runs of the coupled climate–social model, clustered into 5 clusters using k-means clustering. The size of the cluster is determined by the line thickness.

Source data

The modal policy–emissions trajectory emerging from the model, 48% of model runs, has global emissions peaking in the 2030s and dropping steeply over the 2040–2060 period, resulting in 2100 warming of 2.3 °C above 1880–1910 levels. The 2030–2050 emissions pathway displays a perhaps remarkable similarity to recent estimates of the effect of current climate policies or stated nationally determined contributions. Sognnaes et al.¹¹These results can be interpreted as fossil-fuel CO₂ emissions between 30–36 Gt CO₂ in 2030 and between 23–40 Gt CO₂In 2050. Assuming that fossil fuels comprise 90% of total CO₂ emissions, equivalent values for the modal path trajectory are 38 Gt CO₂ in 2030 and 30 Gt CO₂In 2050. This congruency is despite the fact current and declared climate policies are not input to the model and do no constrain model behaviour.

The feedback mechanisms discussed above are highlighted by the third and second most common clusters. The ‘aggressive action’ trajectory is characterized by a strong social-conformity feedback in the opinion component through a high persuasive force parameter, leading to rapid diffusion of support for climate policy that—combined with effective and globally deployed mitigation technologies—drives emissions down faster than in the modal path, limiting warming to below the 2 °C temperature target. By contrast, the ‘technical challenges’ trajectory is characterized by a weak or absent learning-by-doing cost reduction feedback within the energy sector, as well as expensive and ineffective mitigation technologies. This pathway has the same climate policy trajectory as the modal path, but the absence of the technical-change feedback driving costs down over time leads to much greater emissions and warming of 3 °C by 2100.

Two other trajectories (‘delayed recognition’ and ‘little and late’) exhibit multi-decade delays in climate policy, producing higher emissions over the century. These trajectories make up just over 5% percent of model runs. They are characterized by low social conformity feedback in public opinions (high network homophily, low persuasive force), cognitive biasedes that limit any effect on perceived climate change in increasing support of climate policy and an unresponsive policy system (high statusquo bias), which slows climate policies even as public support grows.

Examining the list of parameters that separate the clusters of policy trajectories and the emissions trajectories (Table). 1Extended Data Figure 3) reveals an important role for parameters associated with the opinion, mitigation, cognition and policy components, particularly the strength of social conformity (for example, network homophily and persuasive force), the strength of mitigation technology feedback and effectiveness (for example, learning by doing, mitigation rate and lag time), the responsiveness of political institutions (for example, status quo bias) and the role of cognitive biases (for example, shifting baselines and biased assimilation). Not surprisingly, parameters from the adoption component don’t tend to be distinguishing characteristics for policy and emissions pathways. The model can show tipping-point behavior in which individual behavioural change can be crucial in driving the system towards zero emission (Fig. 2a), the particular conditions that are necessary for this model behaviour do not appear to be common after constraining the model parameters using the hindcasting exercise.

Drivers of variance in model behaviour were further explored by fitting random-forest models to two outputs of the 100,000 Monte Carlo runs of the calibrated model: policy in 2030 and cumulative emissions over 2020–2100. Explanatory variables are normalized values for the 22 model parameters. Extended Data Figure 4This table shows the minimum depth distributions for 10 variables that are most important to each model. Similar to the clustering analysis, variables relating to opinion dynamics (persuasive power and network homophily), responsiveness in the political system (status-quo bias and political interest feedback), individual perceptions of climate change (shifting baselines or evidence effect) and mitigation technologies are important in explaining variation of policy and emissions trajectories in the twenty-first century.