Introduction

Developed through a collaborative international effort spanning climate science, economics and other social sciences, and technoeconomic modelling communities, the Shared Socioeconomic Pathways (SSPs) has for over a decade provided researchers with a comprehensive framework for integrated, cross-disciplinary assessment of climate change mitigation and adaptation efforts1. The SSPs were designed to make up the socioeconomic dimension of the climate science community’s scenario framework, by describing systemic factors such as demographics, technology, economic, policy direction and orientation, resources, and human behaviour, and allowing for understanding how these factors interact with climate change and climate policy. The SSP framework includes five qualitative storylines describing futures of different challenges for mitigation and adaptation2 and depicting how key socioeconomic factors under these storylines might evolve3,4. These factors have also been quantified in ‘marker’ scenarios to be used as inputs in scenario studies using integrated assessment models (IAMs), which have played a crucial role in the reports of the Intergovernmental Panel on Climate Change (IPCC)—including in the latest 6th Assessment Report (AR6). Initially, quantitative elements would include projections for high-level modelling drivers, such as population data, gross domestic product (GDP), and education levels5,6. Eventually, more quantitative elements became available, providing future projections of aspects such as governance quality4, rule of law7, gender inequality8, extreme poverty9 and human development10.

In spite of continuous efforts for quantification of SSP elements as well as the improvements reflected in three official updates to the SSP datasets, the SSP framework still falls short of quantitatively integrating political processes and institutions—particularly those related to democratic governance. A country’s governance structure can significantly impact its response to climate change: for instance, democracies with stronger public participation tend to create more effective and robust climate policies11, and these countries have been found to be better at meeting international commitments, reducing emissions while maintaining economic growth, and achieving greater overall CO₂ reductions12. Conversely, despite finding it easier to enforce climate policies, authoritarian regimes often lack mechanisms to correct ineffective policies and show suppression of dissent and elite-driven maladaptation13,14. Additionally, they not only lack mechanisms for public accountability and equitable adaptation planning but also amplify climate injustices and, by centralising power and dismantling checks and balances, they enable policy capture by vested interests15,16. The advantage of democratic governance in producing more effective and equitable climate action lies in its features that directly influence a society’s adaptive capacity by ensuring that policies reflect local needs, distribute costs fairly, and maintain public legitimacy over time14,17,18. Nonetheless, democratic systems also have weaknesses. The IPCC AR6 specifically identifies economic lock-ins and fossil-fuel lobbying as major barriers, where the very diversity of representation in democratic frameworks can allow established industries to delay or weaken climate measures11. The effectiveness of climate policy, therefore, depends more significantly on the quality of democracy, particularly in terms of participatory procedures and accountability systems, than on the type of regime alone11.

In its current form, the SSP framework fails to capture how dynamic democratic institutions fundamentally enable or constrain climate action, making the integration of democratic indicators essential for creating more realistic projections1. Recent efforts to incorporate political perspectives in the IPCC AR6 suggest growing institutional awareness of its limitations19. Moreover, the SSP framework’s neglect of justice principles undermines its legitimacy and policy relevance: it fails to consider who gets included in decisions, or how different groups are affected, making it less useful for real-world policy. Critically, while originally discussing institutional effectiveness2 or governance4, the framework does not accommodate hybrid regimes that maintain electoral processes while undermining institutional checks on power14 and/or create enabling environments for other actors and judicial interventions that shape real-world outcomes15,16, which is an emerging governance trend with significant implications for climate issues20.

This study analyses the evolution of democratic governance that is assumed in the SSP framework. In the first part of our analysis, we use the narrative policy framework (NPF) to examine which democratic characteristics are implicitly assumed or supported within the textual storylines of the SSPs, revealing the normative visions embedded in them. Towards this, we employ a comprehensive framework of democratic elements that are relevant to climate action, as defined by Brawley-Chesworth et al.21, including participation (enabling citizens to engage in governance) and representation (ensuring their influence shapes political outcomes). Furthermore, the framework assumes that effective decision-making depends on integrating diverse knowledge, expertise, and accountability to make leaders responsible for their actions, while deliberation, respectful dialogue, and assessment of the regime’s effectiveness in achieving its goals promote justice, equitable distribution of benefits and burdens, inclusive procedures, and recognition. Thus, the five SSP narratives are assessed in terms of references to democratic participation, representation, knowledge and expertise, accountability, deliberativeness, effectiveness, and justice. In the second part of our analysis, we identify quantitative SSP elements that could act as proxy variables for representative characteristics of democracy. We then examine projections of these proxy variables to uncover potential patterns across the marker SSP implementations and compare them with the democratic assumptions from the SSP storylines. Eventually, we conduct a meta-analysis of scenarios included in the IPCC AR6 Database to identify implied relationships between democratic characteristics and indicators of climate change mitigation performance. We finally integrate the results of our analysis to provide actionable recommendations for the scientific community, including climate-economy modellers and political scientists.

Results

Democratic characteristics in SSP narratives

To examine how the SSPs potentially construct assumptions about democratic governance, we use the narrative policy framework, a theory-driven approach to analysing policy narratives, viewing them as strategic tools embedded in policymaking processes rather than simple communication messages. At its core, NPF suggests that narratives influence public opinion, policy coalitions, and institutional design by shaping how actors assign meaning to political events and decisions, while simultaneously reproducing the broader political, cultural, and institutional contexts in which they are embedded22,23. In this sense, narratives both shape and are shaped by existing governance norms, recognising that while an objective policy world exists, its meaning is interpreted and constructed through political processes22. Narratives comprise recognisable elements such as settings, characters (heroes, victims, and villains), a plot, and a moral of the story, often implying or justifying a specific policy solution22. These narrative structures are consistently found across political discourse and provide a common grammar through which policy actors convey meaning and mobilise support. Our NPF-based analysis focuses on the full textual descriptions of the five SSP narratives, as presented in O’Neill et al.2, Riahi et al.24, and the IPCC AR6 Synthesis Report11. The “Methods” section offers an overview of our NPF application. Building on this framework, we examine below how each SSP storyline frames future societal developments.

SSP1 envisions a future where societies transition “gradually toward a more sustainable path”2, rooted in social inclusion, institutional cooperation, and human well-being. A broad coalition of actors, such as governments, international institutions, civil society and the private sector, acts as a central protagonist and collaborates to reshape global development in an increasingly effective way. While not explicitly mentioned, citizens are implied to participate actively through changing perceptions and consumption patterns. In this setting, strong and flexible institutions work in synergy to enable both mitigation and adaptation. Furthermore, “inequality is reduced both across and within countries”2, indicating not only a redistributive governance model, but also one that enhances representation and social solidarity. The plot follows a causally coherent and hopeful trajectory. “Social, cultural and economic costs of environmental degradation and inequality”2 serve as catalysts for structural change. Societal behaviours also evolve toward reduced consumerism, suggesting the public is willing to participate in and promote the transition. These outcomes are not framed as the result of top-down enforcement but rather as the consequence of democratic deliberation, shared values, and institutional responsiveness. The moral of the SSP1 narrative is that a better, fairer, and more sustainable world is achievable through deliberate, inclusive, and coordinated action across all levels of society. Thus, sustainability is not a trade-off with democracy but the result of it.

SSP2 portrays a world where existing trends continue without major disruptions or breakthroughs. The key characters are national governments, global institutions, and societies, which are positioned as moderately responsive yet constrained. Citizens appear indirectly, primarily as recipients of slow improvements in living conditions, but there is no mention of their participation in the transition. This marginalisation implies limited influence on shaping policy, reflecting a restricted form of representation within the democratic process. The setting reflects a form of procedural democracy, where formal institutions fail to structurally change. Societies are characterised by limited cohesion, with “income inequality that persists or improves only slowly”2. These features indicate structural constraints on distributive and recognitional justice, as democratic institutions fail to address systemic challenges. The SSP2 plot follows a gradual, almost stagnant progression with limited institutional ambition and uneven social investment. Although “technological development proceeds apace”2, the lack of breakthroughs underscores the absence of innovation. “Environmental systems experience degradation”2, with some slow improvements, a dynamic that reflects slow institutional learning and partial implementation of policy. The moral of SSP2 is that a business-as-usual democratic model results in insufficient progress to address the scale of climate and societal challenges. While there is no collapse of institutions, there is also no systemic step forward, with existing systems enduring but underdelivering.

SSP3 constructs a fragmented world order dominated by nationalism, regional competition, and authoritarian tendencies. The central protagonists are countries and governments that are portrayed as self-interested actors prioritising competitiveness and security over global cooperation. The storyline references a “limited number of comparatively weak global institutions, with uneven coordination and cooperation”2, suggesting that mechanisms for collective accountability and participation have largely eroded. The emergence of “more authoritarian forms of government with highly regulated economies”2 further constrains the scope for democratic participation and civil liberties. Citizens are framed as victims of systemic neglect, suffering from persistent inequality. The setting is one of de-democratisation, institutional fragmentation, and high social vulnerability. The narrative makes clear that democratic governance is in retreat as “countries move toward more authoritarian forms of government”2. Institutions capable of addressing transnational challenges are ineffective or absent. These conditions reflect a governance landscape of low justice and minimal accountability, where state priorities are driven by control, not inclusion. The plot of SSP3 unfolds through deterioration and fragmentation. As geopolitical tensions rise, states grow defensive, leading to a downward spiral of institutional isolation, inequality, and environmental degradation. The moral of the SSP3 story is a cautionary one: it warns that a fragmented and authoritarian global system not only fails to deliver sustainable development but also amplifies inequality, environmental crisis, and conflicts. The narrative does not present a path forward, but rather a breakdown of democratic legitimacy and capacity, where deliberative democracy, inclusive participation, and transnational accountability are not only neglected but actively dismantled.

SSP4 describes a stark division in representation and influence. The main actors are divided along sharp socioeconomic lines. On one side is an elite “internationally-connected, well-educated society”2, while on the other “a fragmented collection of lower-income, poorly educated societies”2. Unlike the collaborative protagonists of SSP1 or authoritarian villains of SSP3, this pathway defines characters through inequality itself, the systemic “villain” that limits democratic responsiveness and participation. The setting is defined by the erosion of democratic equity and the concentration of political and economic capital. It describes a world where “increasing disparities in economic opportunity and political power lead to growing stratification both across and within countries”2. This signals a severe decline in justice, representation, and participation. Large portions of the population are excluded from meaningful political engagement and, with “limited access to effective institutions for coping with economic or environmental stresses”2, their capacity to advocate for their needs or influence decisions is further diminished. The SSP4 plot is driven by systemic division and institutional inaction. As “power becomes more concentrated [and] vulnerable groups have little representation”2, governance mechanisms increasingly serve the interests of an elite class, sidelining deliberation and weakening the moral contract of inclusive democracy. The moral of the story in SSP4 is that progress for the few does not translate into justice for the many. While global elites can mobilise resources to mitigate emissions, large portions of the world’s population remain trapped in vulnerability. The fact that such disparities persist even in democratic societies suggests a collapse of the normative foundations of democracy, where procedural form is maintained but substantive equality, representation, and accountability are eroded.

SSP5 describes a world with strong formal participation and institutional effectiveness, centred on market-oriented solutions. The central characters of this narrative are industrialised and emerging economies, competitive markets, and participatory societies that are presented as agents of progress2. Governments mainly support market-based systems, and disadvantaged groups are acknowledged in the storyline but act mainly as beneficiaries of market inclusion. The setting is marked by institutional strength, economic optimism, and an overwhelming commitment to rapid growth and modernisation. Participation is encouraged, at least in procedural terms, while policy is geared toward liberalisation and ensuring institutional access. However, the global community places “relatively little effort to avoid potential global environmental impacts due to a perceived trade-off with progress on economic development”2. The tension here is between democratic inclusion and ecological neglect as, while participation is structurally enabled, intergenerational justice and environmental accountability are marginalised. The plot of SSP5 unfolds through a narrative of accelerating growth enabled by market mechanisms, innovation, and technocratic governance. “Faith in competitive markets”2 and technological progress drive social inclusion. Democratic deliberation and accountability exist, but focus primarily on economic performance rather than sustainability. The moral of this story is one of confidence in technocratic governance, market solutions, and engineered resilience. Human development goals and economic growth result in relatively low challenges to adaptation, but only “for all but a few”2. This shifts the moral orientation of the narrative away from justice and accountability and toward efficiency and innovation.

From our analysis, we may conclude that—despite different governance modes portrayed—narratives explicitly centring on democratic participation, institutional responsiveness, and justice are scarce within the SSP storylines and, subsequently, in the quantified scenario studies building on the data tied to them. SSP1 stands out as the only scenario to consistently embed strong democratic assumptions across all examined dimensions; in contrast, SSP2 presents only a vision of procedural democracy, one that is detached from transformation. While democratic institutions formally exist, they are portrayed as slow, uneven, and only partially responsive to societal needs. SSP3 and SSP4 exemplify scenarios of democratic decline, though in different forms: SSP3 narrates a collapse of democratic coordination, where participation and representation are actively suppressed, and citizens are framed as victims of state-led exclusion, while in SSP4, democratic institutions remain formally intact in some regions but are hollowed out by economic stratification, with vulnerable groups excluded from decision-making. In both narratives, democratic characteristics such as justice, deliberation, and accountability are either absent or structurally undermined. Finally, SSP5 depicts a hybrid model of democracy that is highly technocratic and market-driven, upholding formal participation and emphasising institutional effectiveness and innovation, while subordinating ecological and intergenerational justice to economic growth. Table 1 below summarises these findings across SSP narratives. Overall coding reflects the qualitative analysis described above and remains subjective, used primarily to synthesise the results of the analysis.

Table 1 Implied assumptions of the SSP narratives in terms of core democratic characteristics
Full size table

Implicit democratic assumptions in SSP quantifications

From the 25 currently available quantitative components of the SSPs [see Section S.1 in the Supplementary Information (SI)], we find proxy variables for five out of the seven characteristics of the analytical framework defined above (see Table 1): participation, representation, accountability, effectiveness, and justice (see “Methods” for details). Starting from the easiest proxy to define, government effectiveness has been directly estimated for almost every country using a comprehensive composite indicator across all SSPs4. Projections for indicators on the quality of the Rule-of-Law and Control of Corruption indices are available for all SSPs7 and can be directly associated with the principle of accountability25,26. While several SSP indicators relate to different dimensions of justice, we choose the Gini Income Inequality Coefficient and the share of population in extreme poverty9 as proxies to distributional justice. On political participation, education has been frequently considered as an enabler27 and is thus selected as a proxy; similarly, the human development index (HDI), a composite indicator that synthesises indices relating to education, health, and income, is also selected as it shows a positive correlation with participation28 and other democratic indicators29. In the case of representation, the selection of proxies is less straightforward. We select gender representation, proxied by the gender inequality index (GII), as the most fitting indicator provided in the SSP framework and its extensions8. We assume that regions with low gender inequality will have better gender representation in the political system and thus improve the quality of the representation overall30—acknowledging, however, that representation in democratic institutions is not only affected by gender representation.

Starting with participation, based on data from Cuaresma & Lutz10, HDI in 2020 is highest for most OECD countries in Europe, North America, and Oceania and lowest for many African nations. Figure 1 shows indicative results for selected countries in all world regions, and Section S.2 in the SI visualises all available results (Figs. S1–S8). In all SSPs, this indicator shows improvements from 2020 to 2050 for all countries (visualised with a green colour on the figure), with the highest improvements shown for African countries and India in SSP1 and SSP5, which converge with countries that already had high HDI. Assuming that democratic structures are an enabler of this increase in HDI in most countries28,29, this may indicate that democratic participation is also increasing in all SSPs around the world. This is somewhat at odds with the expected trajectories of political participation from the narrative analysis. For instance, SSP2 is expected to be a continuation of current trends but instead it seems that improvements in HDI (and, assumingly, in participation) are almost at the level of the high-development SSPs 1 and 5. Additionally, the conflicts implied in the narrative of SSP3 and the rising inequalities in SSP4 do not significantly hamper HDI in any country, with HDI levels in those SSPs remaining rather stable by 2050—and even increasing for some countries. Similar patterns can be extracted by looking only at education (Fig. S2 in the SI); for instance, the mean years of education for the female population between 25 and 29 years old seems to increase in all SSPs and across countries, in the same way as HDI.

Fig. 1: Projections of quantitative indicators from the SSP framework related to democratic principles.
figure 1

The human development index (HDI) is used as a proxy indicator for democratic participation, the gender inequality index (GII) for representation, the rule-of-law index for accountability, government effectiveness for effectiveness, and the Gini coefficient for justice. The green background shows that an indicator’s value improved between 2020 and 2050, yellow that it stayed the same, and red that it worsened. As the 2020 values slightly varied among SSPs, the minimum values among them are shown for illustrative reasons, apart from the GII and the Gini coefficient, where the maximum values are used. The countries are selected as representatives of different world regions, economies and emitter groups. Data from Cuaresma & Lutz10, Andrijevic et al.4,8 and Rao et al.31, available through https://ssp-extensions.apps.ece.iiasa.ac.at.

Full size image

When it comes to representation, based on Andrijevic et al.8, GII in 2020 was lowest for Western European countries, Australia and Canada (0–0.2) and highest among several African and Asian countries (0.5–0.8; see Fig. 1). GII is significantly improved in almost all countries in 2050 for SSPs 1 and 5 (reaching 0–0.2), agreeing with the narratives of these SSPs that inequalities are reduced between and within countries. Assuming that in countries with low gender inequality, all genders are adequately represented in democratic institutions30, we can assume that democratic representation is improving in these SSPs. However, SSPs 3 and 4 also show improvements in gender inequality between 2020 and 2050 (and, presumably, in representation). While gender inequality in these SSPs still varies significantly among countries throughout the time horizon, these improvements are at odds with the findings of the narrative analysis where inequalities seemingly deteriorate. Counterintuitively, SSP4 shows bigger improvements than SSP3, despite inequality being the definitive feature of the SSP4 storyline. SSP2 is better aligned with its narrative, showing moderate improvement and retaining patterns of inequality between countries; thus, it is assumed that representation also improves until 2050 across SSPs, with SSP1, SSP2 and SSP5 quantitative implementations better resonating their storylines.

From an accountability perspective, drawing from Andrijevic et al.4, the rule-of-law index for 2020 was highest in Canada, Australia and several Western European and Scandinavian countries and lowest for China, Saudi Arabia, Venezuela, Somalia and other countries in Africa. In contrast to the previous indicators, rule-of-law seems to decline by 2050 for most countries in SSPs 1–3 (Fig. 1), which are the only SSPs available for this indicator. Rule-of-law improves for China, Brazil, Saudi Arabia and several countries in the Global South in the context of SSP1, SSP2 and—in the case of Brazil, India and Bangladesh—even during SSP3. Overall, it seems that the modelled projection of this index—and, likely, accountability as a whole—among the three SSPs does not correspond to their respective storylines. While results indicate little difference in accountability between SSPs, the SSP1 narrative assumes high accountability in most countries, SSP2 low-to-medium and SSP3 very low. Differences between the SSPs are more pronounced for the Control of Corruption indicator, where SSP1 and SSP5 show improvements between 2020 and 2050 for all countries, with SSP3 and SSP4 showing similar levels of corruption in 2020 as in 2050, and SSP2 somewhere between these two groups.

Government effectiveness in 2020 follows similar patterns as with the HDI indicator, with the highest effectiveness scores (0.8–1.0) found in Europe, North America, Australia and New Zealand and the lowest (0.05–0.4) in some African countries (see Fig. 1, based on Andrijevic et al.4). In the quantified implementations of SSP1 and SSP5 for this indicator, effectiveness is significantly improved in all countries between 2020 and 2050, reaching a score higher than 0.6 in almost all countries and with many OECD countries reaching the maximum. The improvement in effectiveness is less pronounced in SSP2, although more than half of the countries reach a score of 0.6 in 2050. Interestingly, SSP4 is similar to SSP2 in terms of effectiveness patterns, while SSP3 patterns are slightly worse and reverting toward 2020 levels. While these patterns follow the ones implied by the narrative analysis (i.e., SSP1 and SSP5 have the highest effectiveness, SSP2 and SSP4 moderate and SSP3 is lowest), the quantified effectiveness projections do not vary significantly between SSPs. Additionally, there is no deterioration of effectiveness, even in the case of the extensive conflicts between countries that are implied in the SSP3 storyline.

As a proxy for the justice dimension of democracy, the Gini Income Inequality Coefficient for 2020 was lowest (about 0.25–0.35) for most of Western Europe, Scandinavia and Canada and highest (about 0.55–0.70) for South Africa, Brazil, Mexico and Colombia (Fig. 1, based on Rao et al.31). The Gini coefficient significantly improves in 2050 projections for SSP1 and, especially, SSP5 where nearly all countries achieve higher income equality apart from Canada and the USA that show opposite trends. Among the biggest winners is China, where the Gini coefficient falls from 0.47 in 2020 to 0.30 in 2050 for SSP5 and 0.34 for SSP1. These patterns of improvements from 2020 to 2050 are also evident in SSP2, albeit not to the same extent. SSP3 and SSP4 show a deterioration of the Gini coefficient, suggesting that incomes become much more unequal within most countries. The few exceptions to this include Mexico, Bangladesh, South Korea and Bolivia in SSP4, where inequality decreases. In contrast to the other proxies examined before, these patterns of the Gini coefficient better match expectations from the narrative analysis: inequalities are reduced in SSP1 and SSP5, stabilised in SSP2, and exacerbated in SSP3 and—especially—SSP4. This pattern does not hold, though, for the other indicator of justice, the share of population in extreme poverty (Fig. S8 in the SI), where poverty is reduced in most countries without much variation among SSPs.

Co-evolution of democracy and mitigation in IPCC AR6 scenarios

Here, we examine the relationship between the implicit democratic assumptions in the SSPs and the evolution of pertinent indicators for climate change mitigation, as projected in the IPCC AR6 scenario database. For instance, does high government effectiveness correlate with good performance in terms of climate action in most world regions? While answers to such questions have been partially provided by empirical studies32, it is helpful to understand whether such relationships exist within the scenario results that are assessed in the authoritative IPCC reports and, subsequently, inform global climate action. The purpose of such analysis is not to use the IPCC AR6 results to prove these relationships between democratic elements and climate action. Instead, it is an opportunity to understand gaps in the AR6 scenario space and identify entry points for future work that can enhance the representation of democratic elements in the mitigation and adaptation scenarios used in future assessments—including in the upcoming AR7.

Figure 2 shows relational plots between the democracy-related SSP proxies from the previous section and a selection of mitigation indicators. These proxies and indicators are quantified for the R10 world regions (geographical disaggregation used in AR6). For more information on the mitigation indicators and data sources, see “Methods”.

Fig. 2: Relationship between democracy-related indicators and mitigation indicators from the IPCC AR6 scenario ensemble (per R10 region and SSP).
figure 2

These relationships are represented through points on a scatterplot (one per R10 region of each scenario), while the lines represent linear regressions with 95% confidence intervals. More details are given in the “Methods”. Data from the AR6 Scenario Explorer and Database44 and the SSP Extensions Explorer (https://ssp-extensions.apps.ece.iiasa.ac.at).

Full size image

In most of the examined scenarios following SSP1, SSP2 and SSP5, global regions that have significant reductions in CO2 emissions between 2020 and 2050 also had a high HDI in 2020 (Fig. 2). This finding is less robust in SSPs 3 and 4, where 95% confidence intervals are too wide to indicate any meaningful correlation. Similar patterns can be seen in the reduction of energy intensity of GDP, which also seems to be positively correlated with HDI, even for scenarios following SSP3 and SSP4. In contrast, no clear patterns can be found for SSPs between HDI and the share of renewable energy sources (RES) in the electricity mix and between HDI and carbon intensity. Assuming a positive association between HDI and political participation, these findings indicate that high political participation may be positively correlated with CO2 emission cuts and improvements in energy efficiency in the examined scenarios.

Similarly, regions with low gender inequality in 2020 seem to perform better in terms of reducing their CO2 emissions and energy intensity by 2050 (Fig. 2). This is shown most prominently for scenarios following SSP1, SSP2 and SSP5. Patterns between GII and CO2 emissions reductions are less clear for SSP3 and SSP4, although some correlation is still indicated for these SSPs between GII and energy intensity. A positive correlation is found between GII and the carbon intensity of energy for SSP2, although it is weaker compared to the other mitigation indicators. The connection between GII and RES shares is unclear for all SSPs. Assuming again a positive association between gender equality and representation in democratic institutions, these findings suggest that countries with high representation today can potentially achieve significant reductions of their CO2 emissions and energy intensity.

Regions with high rule-of-law in the AR6 scenario ensemble show high reductions in CO2 emissions, energy intensity, and carbon intensity for SSP1- and SSP2-aligned scenarios (Fig. 2). There is an even more noticeable positive correlation between the rule-of-law index and RES share for these SSPs, compared to the more ambiguous findings for HDI and GII. Findings for SSP3-aligned scenarios are more uncertain, except for the positive relationship between rule-of-law and energy intensity reductions, as in the case of the other SSPs. Similar positive correlations are found between control of corruption and reductions in CO2 emissions and energy intensity, with less clear results shown for RES share and carbon intensity (see Fig. S9 in the SI). Overall, these results suggest that accountability may have a positive relationship with the examined mitigation indicators, especially CO2 emissions, energy intensity and carbon intensity.

Government effectiveness also shows a positive correlation with the assumed reductions in CO2 emissions and energy intensity in the AR6 scenarios that follow SSP1, SSP2 and SSP5 (Fig. 2), following similar trends as with all previous indicators. This is an intuitive finding as governments featuring high effectiveness in this decade may be in a better position to implement the policies required for the transition. The relationship is much more uncertain in SSP3 and SSP4, as well as in the associations between effectiveness and RES share and between effectiveness and the carbon intensity of energy.

In contrast to all other proxies, the relationship between the Gini coefficient and all four mitigation indicators within the AR6 scenario ensemble is less clear, as confidence intervals are wide for most SSPs. Indications of positive correlations are seen in SSP2 scenarios between the Gini coefficient and the change in CO2 emissions and energy intensity, albeit weaker than similar correlations for other proxies. Counterintuitively, the Gini coefficient also shows a positive association with the RES share, implying that regions with higher income inequality (higher Gini coefficient) have more renewable electricity but are expected to perform worse in decarbonisation (higher emissions and energy intensity). Similar indications are found for the other justice proxy based on extreme poverty (Fig. S10 in the SI), where regions with a higher share of extreme poverty in SSP2 have higher energy and carbon intensities as well as a higher RES share. Still, these signals are mostly weak and not evident for other SSPs, suggesting that the interaction between justice dimensions and climate change mitigation is much more complex than can be assessed here.

Discussion

Our study highlights opportunities to enhance the SSP framework toward more accurately reflecting democratic governance in modelling science. Given its central role in the IPCC’s 5th and 6th Assessment Reports, improvements to the SSP framework are essential to ensure that the modelled futures account for political realities and complexities, while also examining more participatory, just, and inclusive transition pathways. Our narrative analysis found that, apart from SSP1, most SSP narratives present futures that are technocratic or elite-driven and do not emphasise participation, institutional responsiveness, and justice qualities. These findings align with and reinforce recent calls to integrate questions of equity and inclusion more directly into scenario design. As Andrijevic et al.8 argue, current SSPs often underrepresent the political, institutional and distributive consequences of gendered access to resources, decision-making and public goods. Similarly, Leininger et al.33 highlight that political development, including institutional quality, regime type and accountability, is currently acknowledged but not meaningfully integrated in the SSPs, leading to an underestimation of the barriers to effective climate action. Both studies underscore the need to expand the scenario space to include political variables that capture not only climate feasibility, but also democratic legitimacy and institutional resilience. Our analysis complements these critiques by showing that many SSP narratives reproduce assumptions of elite-driven or technocratic governance while underemphasising participatory and justice-based pathways. These implications contribute to recent calls on the expansion of the scenario space34 and extend beyond scenario modelling, as narratives can interact with evolving political and governance norms, both influencing them and being influenced by them22. For instance, the technocratic assumptions embedded in most SSPs may be carried forward into the scenarios using them, which in turn can shape the scientific and policy discourse on climate action.

Turning to the quantitative analysis, we find that the SSP variables selected as quantitative proxies to democratic characteristics diverge significantly from the patterns implied by the SSP storylines. Across the world, most proxy variables would improve towards the middle of the century—even in SSP3 and SSP4, where democracy is expected to deteriorate according to their storylines. In contrast, even amid widespread regional conflicts assumed in SSP3, the Human Development Index, Government Effectiveness, and the share of population in extreme poverty do not decrease for almost any country between 2020 and 2050. Additionally, indicators proxying participation and accountability (GII, rule-of-law, and control of corruption) are expected to stabilise between 2020 and 2050 in SSP3 and SSP4 or, in the case of rule-of-law, slightly decline in some countries. Turning to SSP2, while it aims to reflect continuation of current trends, it tends to approach SSPs 1 and 5 for most proxy indicators—especially for HDI, rule-of-law and government effectiveness. Quantitative results for SSP1 and SSP5 show similarities for most selected indicators, implying similar progress in improving aspects of democratic government—that is, despite their storylines implying stark differences to one another in terms of quality of democracy. For instance, the SSP5 storyline describes a much more technocratic world, where efficiency and competitiveness are instead prioritised compared to justice aspects; such divergences would suggest different modelled trends of GII or share of population in extreme poverty, which is not found in our analysis.

Nonetheless, in most scenarios of the IPCC AR6 ensemble, regions scoring high in indicators related to democratic characteristics tend to perform well in mitigation: regions that have, on average, high government effectiveness, rule-of-law, and control of corruption in 2020, tend to achieve significant reductions of CO2 emissions and energy intensity by 2050. These findings intuitively relate to the impact of institutional quality in addressing the climate crisis (including aspects of effectiveness and accountability) and largely align with the available empirical evidence35. We note that these positive correlations are found among scenarios that follow SSPs 1, 2 and 5, while there is significant uncertainty in scenarios implementing SSPs 3 and 4. A possible explanation for this can be found in the uneven development or international conflicts among countries that are assumed in SSP4 and SSP3, respectively, which may accentuate barriers to climate action even for countries with high institutional capacity. In contrast, development is suggested to evolve in a relatively more uniform way in the other SSPs—especially in SSP1 and SSP5. Contrary to all other proxy variables, the relationship between the examined mitigation indicators and the two distributive justice proxies (the Gini Income Inequality Coefficient and the share of population in extreme poverty) shows high uncertainty across most SSPs. This may reflect the complex relationship between justice and climate action, at least in the way justice is represented in the models assessed in IPCC AR614. Nevertheless, these findings reflect implicit assumptions about democratic governance embedded in both model implementations and scenario results and thus do not necessarily serve as evidence of actual correlation between democracy and climate action in the real world.

Apart from strengthening the SSP narratives, the political science and climate-economy modelling communities should critically collaborate on considerably improving the alignment of SSP narratives with the quantified implementation of the framework in marker scenarios—at least in the context of governance and democracy. As shown in the quantitative analysis, the evolution of most democracy-related variables shows little diversity over the first half of the century between SSPs, with most displaying optimism that misrepresents the intended futures described in the SSP storylines (e.g., government efficiency never deteriorates even amid widespread conflicts embedded in the SSP3 storyline). In addition, most variables change linearly over the years and oftentimes in tandem with one another; for instance, while all SSP quantifications assume that human development improves alongside government effectiveness, there is no quantified narrative that explores a world with high government effectiveness but low human development. Building on empirical evidence from political science research, additional indicators can be incorporated to cover political elements beyond effective governance, such as democratic participation, deliberation and justice, but also democratic legitimacy and institutional resilience. Both new and existing indicators could be adjusted to exhibit greater variation across SSPs, thereby covering a larger part of the uncertainty space. Shocks and democratic backsliding, such as the ones recently witnessed in the USA and Europe, could also be explored in narratives and their quantitative representations. This could additionally help mainstream the study of discontinuities and extreme events in integrated assessment modelling36,37, thereby moving beyond the limitations of linear modelling approaches. Furthermore, additional narratives could go beyond the technocratic optimism of the existing ones, including narratives driven by participation or justice aspects, or even narratives that explore alternative economic models (e.g., degrowth38) along with their respective governance structures.

The cooperation between political and other social scientists on the one hand and the modelling community on the other can also extend beyond the enrichment of the SSP frameworks and onto shaping scenario design itself. While this analysis suggests that relationships between various democracy-related assumptions and mitigation impacts in IPCC AR6 scenarios align with empirical evidence (e.g., that effective governments can better enable or promote climate action), these relationships are mostly implicit in the models. Most modelling studies have hitherto used specific SSP datasets to inform basic drivers in the models, such as GDP and population, and not to explicitly study mitigation impacts of aspects such as governance in their scenarios. These recommendations for enhancing SSP quantifications can support modellers in exploring a broader range of scenario variations by more fully integrating governance and political dimensions, thereby reducing the currently observed overreliance on SSP2 in the scenario ensemble—in IPCC AR6 and more broadly in the published literature. Moreover, future modelling studies can explore the interactions between climate action and democratic governance in more detail by assessing scenarios with different democratic institutions and structures, levels of participation and expressions of justice in various countries.

Methods

Narrative policy framework

The analytical strength of the NPF lies in its ability to operate across different levels of analysis. At the micro-level, it highlights how individuals process narratives and form policy preferences. At the meso-level, it examines how policy subsystems and advocacy coalitions strategically construct narratives to mobilise allies, define conflict, and influence policy outcomes. At the macro-level, it connects narratives to broader cultural and institutional contexts that shape and are shaped by dominant policy narratives22. Democratic processes in governance are embedded in language, values and institutional norms that are often subtle and contested39, and the NPF provides a lens to make these dynamics visible and analytically tractable. This study focuses primarily on the meso-level, where narrative construction most directly intersects with institutional norms, legitimacy claims and models of participation. The SSPs are not political blueprints, yet their qualitative storylines play a central role in shaping climate modelling and policy assessments. As previous research has shown, these storylines contain implicit visions of how societies might organise and govern themselves in the face of climate disruption2,40.

Building on this, our analysis uses the NPF in order to reveal what implicit assumptions about democracy are embedded in the settings, characters and morals of each SSP narrative, and how these may shape perceptions of legitimacy and governance. Scenario storylines, including the SSPs, have been shown to embed expectations about governance, institutional development and the distribution of agency, even when presented as descriptive accounts of socioeconomic futures19,41. Similarly, transition narratives can carry moral and political judgements about responsibility and justice42. Our analysis followed a structured reading of each SSP storyline guided by the four core NPF elements. First, the setting was examined to identify how socio-political conditions, institutional capacities and development trajectories are portrayed. Second, the characters were analysed to determine which actors are granted agency, which appear marginal or vulnerable, and how responsibility is distributed. Third, we focused on the plot to examine how causal relationships between societal development, institutional dynamics and climate outcomes are framed. Finally, the moral of the story was interpreted to identify what forms of governance, cooperation or authority are implied as effective or legitimate.

While the NPF provides the structural elements through which the SSP narratives can be systematically examined, an interpretive dimension is necessary to uncover the political and normative assumptions embedded in their storylines. Recent work in narrative policy analysis demonstrates the value of combining attention to narrative form with an examination of narrative content to reveal how narratives organise meanings about responsibility, justice and institutional change42. In this context, we treat the SSPs as political imaginaries that encode assumptions about authority, participation, state capacity and collective agency.

Quantitative proxies for democracy in the SSPs

In this section, we examine the full range of available SSP variables from the SSP Scenario Explorer (https://data.ece.iiasa.ac.at/ssp) and the SSP Extensions Explorer (https://ssp-extensions.apps.ece.iiasa.ac.at) for potential links with the seven democratic characteristics that were proposed by Brawley-Chesworth et al.21. These links are based on relevant political science literature, for instance, empirical studies that identify correlations between the democratic characteristics and the socioeconomic variables in the SSPs. It is noted that these links are neither exhaustive nor aimed to cover the full range of dimensions that exist in each democratic characteristic. They instead aim to provide an indication of the direction of democracy in each SSP, based on the quantification of a relevant proxy. All available SSP variables, along with their suggested links to democratic elements, can be found in Table S1 in the SI.

As described in the respective sub-section in the Results, links have been identified for participation, representation, accountability, effectiveness and justice. Specifically: (1) effectiveness is directly represented using a composite indicator on government effectiveness, (2) accountability using the rule-of-law index and the control of corruption index, (3) justice using the Gini income inequality coefficient and the share of population in extreme poverty (as proxies to distributional justice), (4) participation through education and the human development index, and (5) representation through the closest related indicator which was gender inequality. Even though SSPs and most IPCC AR6 scenarios are rather technocratic and based on innovation learning curves, there was no explicit variable referring to the knowledge and expertise of the government. A potential link can be made with government effectiveness or even with the overall educational level of the population, but it is rather indirect and uncertain. Similarly, there is no variable related to deliberativeness in the SSPs nor a potential proxy variable for aspects of this principle.

To analyse the selected proxies, we first downloaded all related datasets from the SSP scenario explorer and the SSP extensions explorer and parsed them using the pyam Python package (https://pyam-iamc.readthedocs.io). Using the pandas (https://pandas.pydata.org) and geopandas (https://geopandas.org) libraries, we plotted these variables for selected countries in Fig. 1 and in maps showing all countries in Section S.2 in the SI. To focus on the future evolution of these proxy indicators, we assess their change between 2020 and 2050 across all SSP variants. This period captures the early and critical phase of the transition. The year 2050 is particularly significant, as many countries have set it as a target for achieving net-zero emissions.

Scenario analysis

For the scenario analysis, we jointly analyse the scenario data from the IPCC AR6 with the SSP indicators. A major challenge with this approach is the lack of variety of SSPs used in the AR6 scenarios, as almost 90% of these scenarios used the SSP2 as a pathway (Fig. S11 in the SI). A potential reason for this imbalance is that assumptions in such studies are usually entered as alternative scenarios rather than alternative runs of different SSPs. Another important point is that the set of AR6 scenarios is an ensemble of opportunity without structured ways of varying scenario ensembles34,43. This imbalance is even more pronounced across different policy categories explored in the scenarios (see Fig. S12 in the SI). Thus, it is not easy to compare insights directly between scenarios as most scenarios will have the same SSPs and, thus, assumptions related to democracy. A third challenge is the granularity of the results: AR6 scenarios provide results that are mostly global or refer to large global regions such as Europe, Latin America or Southeast Asia. Therefore, most scenarios do not provide details on impacts to individual countries, apart from major ones such as the USA or China.

To address these challenges, focus on the most granular level of results that is provided for most scenarios in the AR6 Scenario Explorer and Database44. This would be the R10 results, where the world is divided into ten regions. The R10 regions are (1) Africa, (2) Eastern Asia (primarily China), (3) Europe (including Turkey), (4) South Asia (primarily India), (5) Latin America and the Caribbean, (6) Middle East, (7) United States and Canada, (8) the Pacific OECD countries (Australia, Japan, New Zealand), (9) former Soviet Union countries not included in Europe (10) and the rest of Asian countries that are not included in the previous categories (e.g., Indonesia, Thailand and Viet Nam). The exact country grouping per region can be found on the R10 description in the common definitions (https://github.com/IAMconsortium/common-definitions). We then select four pertinent indicators for assessing the effectiveness of climate change mitigation. The main indicator is the change of CO2 emissions, as this is one of the main results provided by most models and a typical parameter used by the modellers as a target for the modelled transition. Since the majority of the emissions come from the energy system, most models submitted energy-related results to the IPCC AR6 report. The other three selected indicators are related to the energy system but examine it from different perspectives. First, we look at the share of renewable energy sources in the electricity mix, as this is a very frequent indicator used for policy analysis. We then look at the energy intensity of the GDP as an indicator of energy efficiency and the carbon intensity of energy as another measure of success in decarbonising the energy system. The last two indicators are also part of the widely used Kaya identity which describes the most important drivers of CO2 emissions.

As the last step in this process, we create a combined dataset where we join the dataset with the four mitigation indicators from the IPCC AR6 results and the proxies of democratic characteristics collected from the SSP Extensions. We then plot results for all R10 regions and calculate a linear regression using 95% confidence intervals to show the relationship between each proxy variable and each transition indicator. For instance, do regions with high rule-of-law perform well in terms of reducing CO2 emissions among the examined scenarios? As mentioned above, this relationship can help us identify implicit patterns between democratic governance and climate action in the AR6 scenarios, in the absence of potential explicit relationships due to real-world and modelling complexities. As such, our aim is to assess whether these patterns are consistent with expectations from empirical evidence. The analysis, therefore, offers insights into the current ability of the scenario modelling community to capture the nexus of democratic and climate governance (even implicitly), with the aim of informing future improvements through more explicit modelling of governance processes and institutions. It is important to emphasise that the IPCC AR6 scenario results are not a representative statistical sample as they are collected as an ensemble of opportunity. Thus, the linear regression mostly helps us visualise some underlying patterns rather than prove a correlation between variables. Nevertheless, we also use robust regression to address some of the assumptions of the linear regression and avoid issues with the sample, especially for SSPs 3 and 4 where available scenarios are limited.