The promise and perils of transformative urban climate policy in German and American cities (2025)

Climate scientists have warned of the urgent need to address the climate crisis. National governments have made agreements to respond, however, they are not on track^1,2. Some governments, like the United States, have had and again have a federal administration that rejects the science and will not act³. This has led cities around the world to take up the call to action. Both scholars and local government officials posit that cities are the most appropriate sites for addressing climate change due to their population density, governance capacity and resources, infrastructure, innovation, and economic activity^4,5,6. People from many sectors of society believe solving the climate crisis requires transformative action, but what determines what that action will look like, and to what extent are cities open to and capable of transformative change?

An emerging idea is that reform happens through ‘mainstreaming,’ a process of normalizing ways of thinking that involves the deliberate and cumulative actions of various actors⁷. Discourses reflect key storylines and imaginaries of the desired future and can become mainstreamed as they gain traction in the public debate, and influential discourse coalitions (groups of actors who share the usage of a particular set of storylines) form around them. In the sphere of urban climate policy and governance, climate urbanism has emerged as a conceptual framework for understanding how climate policy is being integrated into city management at the local scale⁸. Castán Broto and Robin⁹ define climate urbanism as an approach to urban life, planning, and development that prioritizes climate action. The scholarship offers a climate urbanism typology, including reactive, entrepreneurial, and transformative approaches. Reactive strategies manage hazards without changing governance priorities, while entrepreneurial approaches adopt sustainability actions when they can be oriented toward economic development. Transformative climate urbanism breaks from the first two traditional governance practices, focusing on radical systems change in transportation, energy, and building while keeping justice and equity outcomes in mind. A transformative urban agenda for climate change requires models that are able to address the root causes of climate change and exchange narrow, market-driven, and technocratic approaches for broader systemic approaches^10,11. Thus, transformative urban climate action requires alternative imaginaries and storylines to status quo approaches to urban governance. We situate our study within the growing body of scholarship on climate urbanism^8,9 and transformative climate action¹².

We frame discourse as a key arena of struggle over the potential for transformative climate action. By combining the climate urbanism approach with discourse analysis, we are able to analyze which priorities are set and the coalitions of actors who steer transformative and non-transformative climate strategies. Maarten Hajer’s framework¹³ on discourse coalitions refers to the idea that groups of actors (such as policymakers, scientists, and activists) form alliances based on shared narratives and interpretations of social and environmental issues¹⁴. These coalitions influence policy and public opinion by promoting specific storylines that shape how problems and solutions are understood and addressed. Westman and Castán Broto draw attention to urban climate imaginaries, which they define as “as collective discourses surrounding the urban that reflect the aspirations of future visions, which also are embedded in institutions and socio-material practices.” (p. 85, ref. ¹⁵). These narratives are shown to be powerful as they can be “deployed to justify certain climate policies in urban environments” (p. 84, ref. ¹⁵). Urban climate imaginaries represent future-oriented thoughts, actions, and desires about how cities should be governed in relation to climate change and the types of interventions that may or may not be supported. Thus, public discourse is involved in the policy process and has the power to both reflect and challenge prevailing policy decisions. It can shape how policies are perceived, debated, and ultimately adopted.

In this research, we conducted a discourse analysis of climate action policy debates in four case study cities located in Germany and the U.S. The discourse indicates that these four cities aspire to transformative modes of climate urbanism. Yet, the debate also reveals storylines and imaginaries that serve as barriers to transformation. Our aim is to show how and why the four case study cities are moving towards transformative climate action and what they need to do to succeed.

This leads us to address two research questions:

1. 1.

   What actors and themes influence the key transformative discourse coalitions emerging from cities’ debates about climate action? and

2. 2.

   What does the discourse reveal about the barriers to successful transformation?

We adopt Hajer’s argumentative discourse analysis framework for analyzing discourse within the context of the practices that produce those discourses¹⁶. Discourse, “defined as an ensemble of ideas, concepts, and categories through which meaning is given to phenomena” (p. 63, ref. ¹⁶), is “constitutive of the realities of environmental politics” (p. 72, ref.¹³). Actors contribute to the public discourse with their statements and try to influence, convince, or learn from each other, creating relations between actors. Certain discourses emerge as dominant and in opposition to alternative discourses to reveal ideas about what policy actions are possible. They help to answer the questions of what should be done, who should be responsible, and what is politically feasible¹⁶. Discourse network analysis (DNA) combines content-based text analysis and relational network analysis to visualize and analyze policy positions made by individual speakers and organizations engaging in the debate and the dynamic interactions of discourse coalitions.

The process of political change is worked out through actors to the extent that they reproduce existing storylines or change to new ones. Herein lies the possibility for transformative change. We draw on Hajer’s conceptual tools of discourse coalition and storyline¹³ to identify the groups of actors and organizations in various sectors of society that provide rationales for similar policy positions and struggle to attain discursive hegemony. These discourse coalitions form when their member actors adopt similar storylines (i.e., particular ways of thinking and talking about climate action politics) while engaging in debate about what to do. Actors in the coalition weave their shared discourses through storylines that become the dominant rationales for the topics that rise to the top of the political agenda of decision-makers. Hajer (p.13, ref. ¹³) states that “once the storyline gets enough socio-political resonance, it starts to generate political effects.” In our four case study cities, we identify the discourse coalitions and storylines that reflect transformation in urban governance in response to climate change.

To realize the transformative potential of urban climate action, approaches must be intersectional and take seriously the inextricable linkages between environmental protection and human health and encompass diverse pollution sources that are often co-produced alongside GHGs¹⁷. Transformative approaches are attuned to justice concerns and recognize connections among climate change, structural disadvantages, local environmental inequities, uneven urban development, and the differential impact of climate change on disadvantaged communities^10,17,18. Urban climate transformation models also include the ability to engage with diverse cultural norms and identities, processes for the inclusion of multiple forms of knowledge as legitimate and equal, center empowerment and protection of the vulnerable and dispossessed, reshape the urban relationship with nature, consider methods for resource redistribution, integrate scientific knowledge, and appreciate the capacity of community-driven action and the capacity of non-state actors to spark change^10,18. These models inform the aspirations of cities as agents of climate action, encouraging them to lead by adopting and implementing the best practices for achieving resilient and sustainable futures.

As cities pursue climate urbanism and transformative strategies, scholars have evaluated their progress. For example, in their case study of pathbreaking climate change policy in Portland, Oregon, Rutland and Aylett¹⁹ show how local government can achieve transformation but also why it is so difficult to do. Between 1993 and 2005, Portland reduced its per capita emissions by 12.5% despite population growth of 27% by convincing city residents to internalize energy efficiency practices in their own homes. The city produced knowledge (discourses), that became a storyline in Hajer’s terms, about the co-benefits of cost and emissions reduction that oriented the actions of residents in alignment with the desires of the municipal government. While the campaign for voluntary collective action transformed how Portland residents related to their home energy use, it also constricted transformation to the more palatable approach of using energy more efficiently rather than reducing consumption outright, and the city did not fully meet its intended reduction target. Even in successful cases of transformative climate urbanism, barriers to change dampen outcomes.

Additional research elaborates on barriers to transformative climate action. Lamb et al.²⁰ propose a typology of climate delay discourses that undermine support for climate policies by emphasizing negative impacts, raising doubts about mitigation, and highlighting obstacles. These discourses include redirecting responsibility, advocating non-transformative solutions, emphasizing downsides, and surrendering to inaction²⁰. Building on Lamb et al.’s discourses of delay framework, Painter et al.²¹ introduce ‘response skepticism’ discourse, which serves to obstruct climate action by questioning, dismissing, or contesting climate policy proposals. Through an empirical analysis of media coverage of the 2021 IPCC report, Painter et al.²¹ reveal two prominent discourses of climate obstruction: taking action will harm the economy and jobs; and climate action would involve too much personal sacrifice.

In this international comparative study, we use DNA²² to examine local policy discourses focused on climate action in four medium-sized cities: two in Baden-Württemberg, Germany (Karlsruhe and Mannheim), and two in California, United States (Long Beach and Oakland). To answer our first research question, we identify the actors or actor groups associated with different dominant themes of climate urbanism using text data from local newspapers and council meeting minutes (an additional source for the CA cities as the number of newspaper articles was lower). We describe our findings using one-mode network diagrams of actors (see Figs. 1–8) in the four cities visualized using time periods T1 and T2. We define the key actors by their organization affiliation and general sector of society to which the organization may be classified (Supplementary 2: Organizational types).

Following Hajer²³, we identify the discourse coalitions of organizations that are most involved in the discourse and share the same policy positions. To pinpoint the dominant themes within each city and understand the potential for transformative climate action, we identify dominant themes by ranking the top 15 climate change policy-related concepts using an index measure of degree centrality and frequency (see Supplementary Table 4 and ref. ²⁴). Additionally, we examine the themes that sparked debate by analyzing the number of statements in agreement and disagreement for each theme. Unlike previous studies that use DNA to examine climate discourses²⁵, we do not find polarization among the actors for any climate-related policy action. There are no instances in our data set denying the existence of anthropogenic climate change. Overwhelmingly, the key actors in all four cities orient their policy positions towards achieving greater climate sustainability. As a result, we focus on the discourse coalitions and storylines that demonstrate how our four cities are creating visions for future transformative climate urbanism.

To answer our second research question, we compare our empirical findings to the characteristics of transformative climate urbanism and reflect on the presence/absence of climate delay discourses that are barriers to transformation.

By articulating the transformation discourses circulating within the public debate that have the potential to create transformative climate action policies in our four cities, this research contributes to the climate governance literature by extending Castán Broto et al.’s climate urbanism typology, specifically the characteristics of transformative climate urbanism, and Hajer’s discourse coalitions approaches by showing that an analysis of discourses and discourse coalitions provides insight into the origins of forces for change. Transformative climate urbanism, as an approach to urban governance, can shape the way cities operate in a decarbonizing world if ideas and framings of transformative policy are frequent and central in the public debate. Discourse matters to policy creation and implementation because it is the source of the ideas for policymakers. Support builds for particular ideas as discourses circulate and different actors start to agree with particular themes and framings. This process of gaining socio-political resonance (e.g., the mainstreaming of transformation discourses) influences the local government decision makers to act in accordance with the dominant ideas⁷.

Actors pushing for themes of transformative climate urbanism

In our discourse analysis, we observed in all four case study cities concepts characteristic of transformative climate action. Dominant discourse coalitions engaging in the policy debate include actors with the capacity to govern (such as city administrations) and storylines about climate action that aspire to transformative change. In all four cities, city administrations emerged as the most central and frequent actors in climate policy debates. The implementation of climate policy is carried out by the city administrations and is often restricted by limited financial resources. Political actors, including elected officials and political parties, along with civil society actors, such as grassroots citizen movements, also formed significant discourse coalitions. Civil society actors and grassroots movements are strong forces that influence the public debate. In Long Beach, Mannheim, and Karlsruhe, public-sector industries, such as port administrations and energy utilities, were involved, but industry participation, especially from the private sector, was minimal in general. Notably, in Karlsruhe, the scientific organization KIT played a predominant role, actively producing climate science and disseminating findings to the public. Locally based civil society organizations such as Oakland’s Climate Action Coalition, a collaboration of community-based, faith, labor, and environmental advocacy organizations, pushed city administrators and elected officials to center social justice and equity in the city’s approach to climate action. Overall, the high level of engagement by city administrations, who have the institutional capacity for legislative action and implementation, coupled with the active participation of political and civil society actors who prioritize transformative action, create favorable conditions for transformative climate urbanism.

In California cities, the dominant discourse coalitions consist of city government administrations, local politicians, and civil society organizations with smaller contributions from non-governmental organizations (NGOs), scientific institutions, and economic enterprises. Civil society organizations included youth organizations focused on climate justice and members of the category “Resident” (i.e., individuals identified in the data as living, working, owning a business, or going to school in the city with no organizational affiliation). Examples of regional, national, and international NGOs include Communities for a Better Environment, San Francisco Bay Keeper, Sierra Club, and Greenpeace. The NGOs in Oakland are particularly notable for pushing transformative climate action because they promote strategically integrative efforts to transform urban governance. For example, The Greenlining Institute advocates equally for building economic wealth and climate resilience in communities of color²⁶. In addition, the City of Oakland has tried to hold fossil fuel producers accountable for causing climate change by suing for damages, which played out in a lawsuit in the U.S. District Court for the Northern District of California, making the court a key actor in the discourse²⁷. Even though both California cities have seaports, the Port of Long Beach is more involved in the climate action discourses, likely related to their aspiration to be branded as “The Green Port” and “distinguish the Port as a leader in environmental stewardship and compliance”²⁸.

In Karlsruhe and Mannheim, the dominant discourse coalitions consist of the city government administration, the Green Party, climate protection agencies, and Fridays For Future (FFF) climate strike activists. The Green Party is represented in both cities because it plays a substantial role in Germany. Also, in both cities, protests organized by FFF, starting in March 2019, resulted in a noticeable increase in participation from civil society actors (FFF, Parents for Future, Climate Alliance). Karlsruhe stands out compared to the other three cities because the Karlsruhe Institute of Technology (KIT), a science organization, has a high-frequency presence and centrality within the discourse as it promotes diverse and transformative policy positions. In Mannheim, the city administration is especially active. In the second part of our time period examined, the Federal Horticultural Show (BUGA) and the MVV (energy supply company) gain importance as actors. BUGA is a German exhibition on horticulture, which includes thematic areas such as landscape architecture. It takes place in a two-year cycle in different German cities. The event has a strong focus on sustainability topics such as environmental and climate protection and efficient energy use. Also uncommon compared to the other cities is the participation of the Evangelical Church in Mannheim’s discourse.

Across the four cities, we find similarities in the dominant themes connecting the actors in the discourse coalitions. They are the following policy-related discourses: 1) climate change impacts are affecting cities; 2) action must occur at the city level; 3) cities must adapt to climate change impacts; 4) there is an urgency to act now; 5) energy transition away from fossil fuels is necessary; 6) a transition in mobility modes/patterns is necessary; and 7) climate action must be prioritized over the economy. While these policy themes are not necessarily transformative, we explain them first and the level of agreement/disagreement with the policy positions they represent before turning to the elements of the discourse that have the potential to support future transformation.

Figure 9 illustrates levels of agreement/disagreement with the key discursive themes. The dominant discourse coalitions in all four cities recognize the negative impacts of climate change. Actors frequently comment on the hazards their urban populations and infrastructures are facing from current and future instances of excessive heat and drought, as discussed prominently in the BW cases, and of sea-level rise and flooding, as discussed primarily in the CA port cities. They also share the notion that urban governance practices should be modified to effectively address climate change and that cities are indeed “viable and appropriate sites for climate mitigation and adaptation,”(p. 993, ref. ⁴) as or more important than action at other governance levels. This discourse aligns with the foundational premise of climate urbanism⁴. Furthermore, there is agreement in pursuing adaptation strategies to react to these threats. Examples of the discourse are found in direct statements made by actors in our data set, such as the LB mayor, who stated, “Climate change is real, […] Long Beach is a coastal city. It is affecting us, and we have to be serious about protecting not just the planet, but our community. We are in danger as a city 50 years from now if we’re not planning today” (March 2018). “I think that we’re already recognizing that [the effects of climate change] are a reality, and we need to be planning for them” said the managing director of planning and environmental affairs for the Port of LB (April 2019). Sustainability Manager Daniel Hamilton said about Oakland’s proposed climate action plan that it “goes significantly farther than California or other cities. It’s one of the most ambitious plans anywhere on the planet. There are realistic ways to go deeper faster” (Nov 2019). “We are the ones implementing climate protection on the ground” stated Klaus Hofmann, president of the Chamber of Crafts in MA (May 2021), an NGO that represents the interests of small and medium-sized craft businesses. Christoph Schnaudigel, District Administrator in KA, pronounced that “We are aware of our obligation and are working hard to implement the climate protection concept adopted by the district council” (Nov 2018). These quotes and additional assertions by the four city government administrations that they hope to achieve climate resilience through adaptation establish why they are motivated to pursue transformative action: They see the impacts, they believe they can act, and they know they must adapt.

Furthermore, actors from multiple sectors (e.g., government, political, civil society, NGO, and public-sector industry) are mostly in agreement about the urgency to act now. This discourse is typically invoked to mobilize advocacy and push decision-makers to act more quickly. Its regularity in the discourse suggests that there is pressure from constituents for cities to prioritize climate action, providing another reason why cities are motivated to transform.

The discourses surrounding how cities are pursuing transformative climate action reflect several key areas for mitigating GHG emissions: energy, mobility, and building and housing transitions. These discourses focus on reducing the use of fossil fuels such as ending oil production (Long Beach), closing coal-fired power plants (Mannheim), and switching to electricity and renewable energy such as solar (Karlsruhe and Oakland). The mobility transition discourse emphasizes electrification of various forms of transport (e.g., passenger vehicles, public buses, city-owned fleet vehicles, and freight trucks), building electric charging infrastructure, and to a much lesser extent, getting people out of their cars by improving public transit and making cities more bicycle- and pedestrian-friendly. Reducing traffic in city centers and improving public transportation were more commonly discussed changes in German cities.

In contrast to these commonalities, we found key differences in policy discourses between the BW and CA cities: more discussion of building and housing transition, creation and preservation of green space, and traffic-calmed city centers in BW, and in CA, emphasis on environmental justice and criticism towards the sectors of society accountable for climate action. Environmental justice refers to remedying the disproportionate pollution burdens and unequal environmental amenities faced by low-income, BIPOC, and historically marginalized communities.

Another observed difference reflects the important role played by specific events in shaping policy discourse. The FFF climate protests impacted the debates in the German cities, while the discourses in California cities reflected the Black Lives Matter (BLM). social justice protests. In Mannheim, debate centered around the shutdown of the GKM, the coal-fired power plant in Mannheim. The media extensively covered the highly controversial discussion of the planned shutdown.

The theme prompting the most disagreement is the position that climate action should be prioritized over the economy, with conflict frequently emerging within discussions of energy transition. A diversity of industry, government, politics, NGOs, and civil society organizations either do not support climate action that harms the economy or at least desire for economic concerns to be balanced with climate action. City administrators, some elected officials, and industry groups, in particular, favor solutions that “align those needs [for addressing climate change] with the city’s budget” (LB Jan 2021), avoid “unnecessary burdens on … local businesses” (LB Mar 2018) and do not “harm the domestic economy” (KA Jan 2020). Government actors reference oft-repeated idealized definitions: “[sustainability] has always been about balancing and improving the three E’s of economy, equity, and environment” (LB Oct 2020). Another disagreement expressed about energy transition is that eliminating fossil fuels as either direct energy sources or generators of electricity will increase energy costs and decrease the reliability of the supply of energy. These arguments often link to concerns with transitioning too quickly, reflecting disagreements over the urgency of phasing out fossil fuels.

In contrast, those stating that addressing climate change is more or as important as economics aim to “avoid one-sided influence by economic interests” (MA Nov 2020) by focusing on the health of the planet and of humans. Despite the advocacy by NGOs and civil society groups to prioritize humanitarian values, economic interests remain deeply entrenched in the discourse across all four cities. Our findings indicate that policy actors acknowledge the necessity of addressing climate change impacts and demonstrate strong support for mitigation efforts through energy and mobility transitions. However, there is a pervasive concern about potential adverse effects on the existing economy, which consequently moderates the urgency of their calls to action.

Following Hajer and the results of our discourse network analysis, we summarize the key storylines about transformative climate action in each of our case study cities (Table 1) We identified four aspects of transformative climate urbanism in the dominant climate action storylines in our cities: 1) prioritize energy transition for emissions reduction over economic concerns; 2) center justice concerns and empowerment of vulnerable and dispossessed communities; 3) pursue intersectional solutions that focus on the relationship between climate and human health; and 4) privilege scientific knowledge in shaping climate action.

Long Beach is pursuing transformative action to the extent that they are implementing an energy transition away from fossil fuels despite harming the local city economy. The discourse reflects a desire to transform the economy in conjunction with climate action. A portion of actors have the political will to advocate for this transformation: “Our planet is worth more than oil company profits” (LB resident Jan 2018). The city administration and politicians are receptive: “We have to …dig in more on how we really reduce our fossil fuel consumption and how we really transition from an oil economy that has been, for many years, very beneficial from a revenue perspective to the city,” (Mayor of LB, Jan 2021).

Oakland’s key approach to transformation emphasizes community-driven action that protects and empowers vulnerable communities and creates a sustainable and socially inclusive city. Environmental justice is a key theme in the climate action discourse, so much so that their climate action plan is titled “Equitable Climate Action Plan.” It calls for a just transition to a low-carbon economy: As Colin Miller of the civil society organization Oakland Climate Action Coalition stated, “It’s up to all of us to work together to chart a path of just transition from an extractive, destructive and racist economy towards equitable, regenerative and local living economies that uphold human rights and the life support systems of the earth (Oct 2018)” As Youth vs. Apocalypse activist Isha Tobis Clarke stated, “The fight against this climate crisis is really a fight against all systems of oppression that are fundamental to our world. We fight against environmental injustice because if working-class communities of color like West Oakland were not viewed as disposable, we would not have this climate crisis” (Jan 2020). In addition, Oakland tried to hold the industry accountable by participating in a lawsuit against oil companies for damages from fossil fuel production. While unsuccessful, the effort demonstrates transformative climate action in that the city tried to get the judicial system to agree that “oil and gas producers should help … Oakland pay for the costs of adapting to [global climate change]” (Richard Wiles, Washington, DC, based Center for Climate Integrity).

Aspirations for a justice approach are part of the discourse in the other cities too, but to a lesser extent. Health equity is one such concern: “We must reconcile environmental racism. The Black community is subjected to 66% more air pollution than other races” (LB elected official, Aug 2020). In addition, the youth activist movement FFF has brought attention to the idea of intergenerational justice.

Putting increased attention on human health is another aspect of transformative climate urbanism storylines. Oakland attempted to prioritize human health by banning the transportation of coal through the city’s low-income neighborhoods. Oakland City Council member Dan Kalb stated, “There is no doubt that the scientific evidence shows there are substantial safety risks and health impacts of handling and moving nine million tons of dirty coal each and every year into and out of Oakland” (May 2018). Even though the judge did not rule in Oakland’s favor, the city linked climate action to improving air quality and the health of the city’s vulnerable and socially marginalized residents.

In Mannheim, and to a lesser extent in Karlsruhe, the discourse reflects the transformative aim to reconfigure human-nature relations with urban designs that prioritize human health and healthy ecosystems. Actors support “more urban greenery overall” to respond to increasing extreme urban heat (KA head of a government regional association, May 2020), “green[ing] areas so that water can evaporate” (MA head of city environmental department, April 2018), and “measures ranging from green facades and roofs to rainwater retention basins to protect against flooded basements” (MA city administrator, June 2018). The environmental administrator of Mannheim underscores sustainable urban design: “Open and green spaces bring more quality of life. …what we plant is important” (MA April 2018). Acceptance of efforts to widely implement nature-based solutions demonstrates one way Mannheim is pursuing transformative climate urbanism.

In contrast to the other three cities, Karlsruhe is the only one that overtly privileges scientific knowledge in shaping transformative climate action. Scientific organizations are particularly active in the debate: the KIT, the South German Climate Office, which specializes in climate science communication, and the international environmental group Scientists for the Future (S4F), an interdisciplinary collective of scientists, researchers, and academics who support the student movement FFF. Listening to science is emphasized by public officials and decision-makers, especially for determining city-scale climate neutrality goals, quantifying GHG reduction targets, and testing different climate action scenarios. S4F is active in both German cities. In Karlsruhe, a S4F scientist activist argued for the inclusion of researchers in co-producing urban climate targets, “Measures to achieve the international climate targets must therefore be developed and implemented together with researchers” (KA Nov 2019). The President of KIT emphasizes the necessity for science integration in policy and decision-making stating, “We use the great technical and specific socio-scientific potential of KIT to test climate protection measures together with citizens and introduce them into society” (KA March 2021).

The influence of science organizations and science-derived discourses on urban governance was not prevalent in the CA cities, even though actors from major research universities make statements in the media explaining climate change impacts and scientific studies are referenced in climate action plans. The difference in BW cities stems from the participation of scientists explicitly in policy recommendations as well as city administrators and politicians specifically referencing their studies for designing policy. For example, the MA mayor recognized the importance of the Wuppertal Institute for Climate, Environment, and Energy to “define and further specify the fields of action for a climate-neutral Mannheim” (MA March 2021). In relation to Mannheim’s work to quantify GHG reduction targets, the city councilor from the Free Democratic Party lauded the Wuppertal study as “an important first step towards a climate-neutral city” (March 2021).

Discursive barriers to successful urban climate transformation

To answer our second research question, we analyze our empirical findings in relation to the characteristics of transformative climate urbanism discourse. While our results reveal storylines that aspire to transformative change in each of the four case study cities, a range of discursive barriers to urban climate transformation persists. Following Castán Broto et al.⁸, we find evidence of discourses aligned with reactive and entrepreneurial modalities of climate urbanism. In contrast to radical systems change and intersectional approaches that characterize transformative climate urbanism, across the four case studies, we find various actors reproduce reactive storylines: climate policy discourses that aim to enable climate adaptation while protecting the existing economic and political status quo. In the reactive mode, discursive barriers to successful transformation include the prioritization of short-term adaptations and focus on risk management to defend existing economic and social life rather than addressing the root causes of climate change. In a coastal city vulnerable to sea-level rise, such as Long Beach, this is illustrated by raising sea walls with the aim of defending and protecting the wealthiest neighborhoods’ property interests.

In Long Beach, elected officials are particularly concerned with replacing oil field revenue that funds city services: “We depend a lot on oil” (LB, Jan 2020). “As an oil producer … we are different … than just about every other city as we manage the city’s oil interests and are the primary operator of the Wilmington oil field… (LB, Aug 2020). A large portion of the Wilmington Oil Field, the third largest oil field in the contiguous US, is in Long Beach city limits. An LB resident stated, “I also like the idea of … continuing oil and gas operations as an income stream for …additional police resources” (Jan 2018). In Karlsruhe, some of the political parties share a similar reactive position, such as a city council member of the conservative CDU party who declared: “They will not protect the environment with their [climate action] proposal, but only harm the domestic economy” (KA, Jan 2020).

Entrepreneurial approaches to climate urbanism also pose discursive barriers to transformative climate action. The entrepreneurial mode “is associated with the deployment of neoliberal principles of urban management to address climate change”(p. 248, ref. ⁸). We found little evidence for purposeful entrepreneurial climate urbanism, but rather more defense and reproduction of the status quo of competitive urban protectionism strategies. While some actors mention the creation of good-paying green jobs as important for successfully addressing climate change at the city level, promoting economic competitiveness through the development of green economies was not prominent in the debate. Furthermore, a few actors called for their city to become a “leader in climate change” (LB May 2021), a “nationwide leader in sustainability” (LB Mar 2021), or “play a pioneering role in climate protection” (KA July 2019) to boost its reputation.

A persistent barrier in our findings is the reproduction of neoliberal logic related to entrepreneurial climate urbanism. For example, we find evidence of the entrenchment of resource-intensive forms of urban development. The CEO of the Karlsruhe Energy and Climate Protection Agency (KA Oct 2019) expresses: “Heat pumps and electric cars need electricity. And because electricity consumption will increase massively with the expansion of electromobility, at the least, we need to pack the roofs with solar modules today.” This quote also illustrates the premise that reducing consumption (as a root cause of climate change) is contested. As a chief executive of a carbon capture technology company stated in The Los Angeles Times, “This [carbon capture] will make a dramatic difference, but it does not affect the way you and I behave. You can keep using the same car, keep driving your kids to hockey practice”²⁹.

Another discursive barrier related to entrepreneurial climate urbanism is the promotion of green solutions aimed at improving the urban esthetics of formerly industrial landscapes that could attract new businesses and residents through amenity-based city boosterism³⁰. The emphasis on human health and green space access in urban design indicates an engagement with and the potential for transformative climate urbanism. However, greening responses focused primarily on solutions that attract new forms of capital investment can instead serve to promote more incremental rather than radical change, lack discussion of empowerment of vulnerable communities, and ignore the equitable distribution of new green urban infrastructure.

Beyond discursive barriers related to the reactive and entrepreneurial modalities of climate urbanism, we also observed forms of Lamb et al.’s²⁰ climate delay discourses in our qualitative data, especially in the two German case study cities. One obstacle related to climate delay is to reject the accountability of certain actors or redirect responsibility to other actors involved. For example, the mayor of MA pronounced that “… nation-states and their traditional institutions are still unable to solve the challenges of our time” (Nov 2019). In KA, the Upper Rhine Mineral Oil Refinery and the Petroleum Industry Association call for more support from the federal government and the EU (Jan 2020). In Oakland, the judge presiding over the lawsuits to hold fossil fuel producers accountable for causing climate change rejected the accountability storyline. The judge deemed it not fair or reasonable to “place the blame for global warming on those who supplied what we [society] demanded” (June 2018). The outcomes of the two lawsuits in Oakland also reflect a failure to achieve transformative climate action related to the limitations of enforcing government regulations at the city level.

Pushing for non-transformative solutions is another delay strategy that is observable in some of the cases, such as the CDU political party’s call for a cost cap for climate protection in KA. Such non-transformative solutions are justified by deploying a logic that emphasizes the downsides of climate action, namely that it is too expensive. “A sum of 500 million Euros is hardly feasible in a normal budget,” stressed CDU city councilor Sven Maier. “The costs must remain manageable, and targeted measures must be tackled first” (April 2020). Even when the desire for energy transition is expressed explicitly, we find evidence of delay tactics related to the storyline that disruptive change is not necessary. This type of discursive barrier reflects faith in implementing technological solutions that are no more disruptive than what people have come to expect from modern technologies. For example, the CEO of the energy supply company in KA expressed his support for the energy transition storyline if electrification did not cause a major change to the functioning of the system: “The energy transition is not only a response to the challenges of global climate change. It is also a powerful modernization push for our energy system and thus also for our country” (KA, March 2019). In the LB case, a city staff member reported, “Our goal is to convert every single bus to electric here in Long Beach, and we will be one of the first cities to do so” (Mar 2019).

Based on our empirical findings, we observed additional forms of delay discourses that extend delay typologies introduced by Lamb et al.²⁰ and Painter et al.²¹. First, we found evidence in our policy discourse analysis of a delay tactic characterized by climate compromise leading to maladaptation. Adaptation to climate change that has the opposite of its intended effect is referred to as maladaptation. Barnett and O’Neill³¹ identify five ways in which unsuccessful adaptation may increase vulnerability to climate change impacts for other systems, sectors, or social groups. In comparison to alternatives, maladaptation actions “increase emissions of greenhouse gases, disproportionately burden the most vulnerable, have high opportunity costs, reduce incentives to adapt, and set paths that limit the choices available to future generations”³¹. We identified actors criticizing city governments for making compromises to settle opposing positions leading to maladaptation. In the LB case study, to restore coastal wetlands located on land owned by an active oil company, the city government negotiated a land swap that stops oil extraction on certain wetland parcels but allows new oil drilling on other parts of the wetland complex as compensation. Actors in the debate, commenting at a city council meeting in March 2018, ardently criticized the maladaptive action of continuing to contribute to GHG emissions and setting a path that limits the choices of future generations. “If you do this, you are culpable in extending this oil extraction… what is killing us, what is killing our planet” (LB resident). They also identified the costs to air and water systems and the disproportionate impact on vulnerable communities, asking, “How can the city knowingly enter into a program designed to generate new and unsafe levels of toxic air and water pollution?” (LB business owner); and “When that new pipeline and the new oil drilling leaks, you are going to be responsible for their sickness and cancer, at least partially” (LB resident).

In the BW cities, actors reflected on how climate action in one area may create vulnerability to negative impacts on another ecosystem. In KA, a representative of the Federal Waterways Engineering and Research Institute, warned, “…to counteract extreme water levels such as in 2018, such measures would have to intervene in the river to an exceptionally large extent. … adverse effects cannot be ruled out” (May 2020). Similarly, a city councilor in MA opposed energy transition through wind power “because of the high level of interference in ecosystems that are already severely affected by climate change” (Jan 2021). In addition, responses to weather extremes—excessive heat or droughts—may turn to energy- or water-intensive maladaptation, especially to maintain existing economic activity. The operator of a state winery noted, “Due to the increasing drought, irrigation systems are needed for young vines” (KA Feb 2021). Policy debates resolved through compromise between opposing positions may lead to a form of climate urbanism that is not just ineffective but has adverse effects.

An additional discursive climate delay barrier that we observed in the data in the CA cases, though was not present in the German cities, is climate paralysis. Climate paralysis reflects an intention or desire to act, but indecision about how to act stalls policy creation, adoption, or implementation. Actors in the policy debate realize that everyone is responsible, and everything is interconnected, such that thinking through the ripple effects of climate action becomes overwhelming. The result is being stuck in terms of deciding where to start and what will be the best solution. Statements like the following illustrate the waffling, “This recommendation is not that there’s something wrong with the [plan to jointly purchase electricity] or that we shouldn’t do it, it’s just that the risk of things not working out as expected is too high at this point” (LB, director of financial management, Sept 2020). Echoing delay discourses that emphasize the downsides of climate action, other actors point out the enormity of the task. As the lawyer representing Chevron in the Oakland lawsuit argued, “the cities were asking the court to create “a new regime to regulate oil and gas production around the U.S. and around the world. That is, to say the least, a big ask’” (May 2018). Other actors point out negative side effects and disruptions to people’s widely held expectations: “Restricting consumer choice and affordability through this ban [on natural gas in new homes] will make us more dependent on expensive and unreliable energy sources,” said the chief executive of the California Independent Petroleum Association (Oak, Oct 2019). A LB elected official pointed out that as the owner of the Wilmington oil fields, the city is “‘the Oil Industry’,” and the city’s perception is “that we would then miss out if the city phased out oil too quickly” (Aug 2020). Confusion over unintended consequences and discomfort with sacrificing expectations can lead to paralysis such that well-intentioned climate policy debates end with postponing action.

In this section, we synthesize the findings of our two research questions: 1) What actors and themes influence the key transformative discourse coalitions emerging from cities’ debates about climate action? and 2) What does the discourse reveal about the barriers to successful transformation? We then situate these findings within the current literature. Guided by Hajer’s discourse coalition framework, informed by climate urbanism literature with a focus on transformative discourse and discursive barriers, and operationalized through discourse network analysis, this study offers a novel perspective on how transformative discourse can enable change.

Previous literature has evaluated the success of particular sustainability discourses in supporting transformation³². The most impactful climate actions in addressing global warming do more than merely transition (i.e., shift from one form to another), but rather transform (i.e., become something new), such as fundamentally transforming our systems to reduce GHGs globally and protect against the impacts of a warming planet. Frantzeskaki et al. suggest that for transformative change to occur, it is essential to both recognize the necessity for change and envision a desired future. They emphasize that having a positive vision is crucial for creating a sense of place and facilitating sustainability transitions³³). Additionally, they argue that transformative change demands both “disruptive and conforming innovations that challenge the status quo in ways of thinking, organising, doing, and knowing” (p. 19, ref. ³³). An emerging concept is that reform occurs through ‘mainstreaming,’ which is described as a process-oriented approach involving deliberate and cumulative actions by various actors⁷. We argue that the vision of a desired future arises from discourses and becomes mainstreamed as discourse coalitions coalesce around them. We identified discourse coalitions in all four cases, with city administrations being the most central and frequent actors, alongside civil society actors and grassroots movements as powerful forces influencing the public debate. The strong involvement of city administrations, equipped with the ability to enact and implement legislation, along with the proactive engagement of political and civil society actors committed to transformative initiatives, fosters an environment conducive to transformative climate urbanism.

Whether a city adopts transformative climate urbanism as its mode for achieving its urban imaginary depends on the discourses that gain traction and replace the taken-for-granted ideas that prevent transformative climate action. Jennifer Rayner³⁴ identifies three prevalent discourses shaping the perspectives of policymakers, politicians, business leaders, and the general public on climate action. These are: 1) the belief that economic growth must continue at its current pace while decarbonizing; 2) the idea that net-zero accounting can maintain global temperatures within safe limits; and 3) the notion that immediate climate action is too costly. Andersson and Gyberg³² identify six climate change discourses in the broader literature: weak ecological modernization (EM), strong EM, green governmentality, civic environmentalism, and climate justice/radical green discourses. Their analysis of climate action policy documents from three Swedish towns revealed that strong EM was the most prevalent. However, they also noted the presence of transformation discourses within civic environmentalism and climate justice.

Our study confirms these findings of previous research and further refines them by incorporating relational and dynamic empirical data. We observed across the four cities common policy-related themes among the actors in the discourse coalitions: 1) cities are affected by climate change; 2) action is needed at the city level; 3) cities must adapt; 4) urgent action is required; 5) transitioning from fossil fuels is essential; 6) changes in mobility are necessary; and 7) climate action should be prioritized over economic concerns.

Our findings highlight key discursive themes for transforming climate action in energy, mobility, and building transitions. The prevalent actors aim to reduce fossil fuel use by ending oil production (Long Beach), closing coal plants (Mannheim), and switching to renewable energy (Karlsruhe and Oakland). Discursive themes on mobility transitions focus on electrifying transport, building charging infrastructure, and promoting public transport and bicycle-friendly cities. Reducing traffic and improving public transportation are more emphasized in German cities. We found key differences in the policy discourses between BW and CA cities. BW cities focus more on building and housing transitions, green space creation, preservation, and traffic-calmed city centers. In contrast, CA cities emphasize environmental justice and criticize sectors of society responsible for climate inaction. Environmental justice addresses the disproportionate pollution burdens and unequal environmental amenities faced by low-income, BIPOC, and historically marginalized communities. Our study further reveals that specific events significantly influenced policy discourse. In German cities, FFF climate protests shaped debates, while in California cities, BLM protests were more impactful. In Mannheim, the controversial shutdown of the GKM coal-fired power plant received extensive media coverage. The most contentious theme is prioritizing climate action over the economy, especially in energy transition discussions, which is in line with previous research^32,34. Various stakeholders, including industry, government, politics, NGOs, and civil society, either oppose climate action that harms the economy or seek a balance between economic and climate concerns. City officials, a number of elected representatives, and industrial organizations, specifically, favor solutions that preserve the domestic economy.

Following Hajer’s¹³ framework, we summarize the key storylines for each city: In Long Beach, the transformative storyline focuses on reducing emissions by eliminating oil extraction and decreasing the city’s reliance on related revenue, while also improving health outcomes for vulnerable populations. In Oakland, the storyline centers on empowering vulnerable communities to achieve climate justice and hold the fossil fuel industry accountable for health impacts and the costs of adapting to climate change. Karlsruhe’s transformative storyline emphasizes creating a culture that listens to science and implements its recommended solutions. In Mannheim, the focus is on redesigning the city with nature-based solutions to reduce emissions, adapt to climate change impacts, and inspire individual behavioral change.

We found evidence of discursive barriers that hinder the transformation, according to Castán Broto et al.⁸, which are discourses aligned with reactive and entrepreneurial modalities of climate urbanism. For example, short-term adaptation and risk management to defend existing economic and social life instead of addressing the roots. Another persistent barrier is the reproduction of neoliberal logic related to entrepreneurial climate urbanism. Instead of reducing consumption as the root cause of climate change, technological solutions such as carbon capture or the expansion of electromobility are preferred. Beyond these discursive barriers, we also observed forms of Lamb et al.’s²⁰ climate delay discourses, especially in the German cases. One example is the refusal of accountability by certain actors or the shifting of responsibility to other involved actors. Advocating for non-transformative solutions is a delay strategy seen in some cases, such as the CDU party’s call for a cost cap on climate protection in Karlsruhe. These solutions are justified by emphasizing the high costs of climate action. Our policy discourse analysis revealed a delay tactic involving climate compromise leading to maladaptation, where adaptation efforts have the opposite of their intended effect. Additionally, we observed climate paralysis in the CA cases, but not in the German cities.

This study has limitations, including its focus on four case cities and reliance on potentially biased data sources such as newspaper articles and council meeting transcripts³⁵. We believe our approach, guided by Hajer’s discourse coalitions framework¹³, informed by climate urbanism literature with a focus on transformative discourse, and operationalized through discourse network analysis, offers a novel perspective on how transformative discourse can enable change. Nevertheless, discursive barriers remain and may hinder this transformative potential.

This analysis offers valuable lessons for advocates of transformative change at the city level:

1. 1.

   Discourse matters and public discourse reveals how stakeholders perceive a policy issue, how their interactions shape these perceptions, and which policy visions and imaginaries become mainstreamed.

2. 2.

   Discursively aligning and supporting the dominant transformative discourse coalition helps to increase discursive hegemony and the potential to influence policymaking.

3. 3.

   A city’s adoption of transformative climate urbanism depends on discourses that gain traction and replace entrenched ideas that hinder transformative action.

4. 4.

   Specific events significantly influence policy discourse: FFF protests shaped debates in German cities, while BLM protests were more impactful in California cities.

5. 5.

   Be mindful of the most contentious theme: prioritizing climate action over economic concerns, particularly in discussions about energy transition.

6. 6.

   Discursive barriers include storylines around short-term adaptation and risk management that defend existing economic and social life instead of addressing root causes. Another barrier is the reproduction of neoliberal logic related to entrepreneurial climate urbanism.

7. 7.

   Delay discourses hinder or delay transformation, such as the refusal of accountability or shifting responsibility to other actors, emphasizing the high costs of climate action, or leading to maladaptation through climate compromise.

Data and operationalization

This research utilizes a comparative case study design to examine two cities in California (CA), United States, and two cities in Baden-Württemberg (BW), Germany. “Case studies are superior to large-N studies in helping the researcher to understand the perceptions and motivations of important actors and to trace the processes by which these cognitive factors form and change”(p. 6, ref. ³⁶. Our study includes a within and between-country comparison. We draw on earlier studies applying a matching approach^37,38 and focus on two pairs of structurally similar, though not identical, cities situated in similar national contexts. The subnational (i.e., state) contexts are also similar: CA and BW have a high share of national GDP, and the residents are mostly politically progressive and environmentally conscious with some conservative ideology among middle-class suburban and rural voters (Table 2).

Selection of time periods

We examined the public discourse about acting on climate change in the four cities between 1 January 2018, and 1 June 2021. For cleaner visualization of the discourse networks, we established two time periods, T1 and T2, with 1 June 2019, as the cut point. We selected this date to reflect the before and after of a globally significant event related to inspiring action on climate change: the Global Student Climate Strikes held on 15 March 2019. We incorporated a 2.5-month lag time for the impact to emerge and to make the time periods closer to equivalent (17 and 24 months, respectively). Also, during T1, both Long Beach and Oakland held community workshops to gather input on their proposed climate action plans. Long Beach started community outreach earlier (June 2018) but finished later (final plan approved August 2022), while Oakland started outreach later (January 2019) but finished sooner (final plan approved in July 2020). For the California cities, the two time periods reflect the shift in focus at the local level from early planning stages to specific planning processes that commanded considerable city government attention and resources.

In contrast to the CA cities, Karlsruhe and Mannheim adopted climate action plans a decade earlier than our data collection time frame, so different events shaped the context of the policy discourse. However, the city councils in both cities debated passing climate emergency declarations. Karlsruhe’s city council passed one in 2019, while Mannheim’s did not. A more important context for debate in the BW cities emerged from the weekly student climate strikes, known as FFF, which drew attention to climate change across Europe. In both BW cities, the participation of FFF groups in the climate action discourse increased, especially in T2. Furthermore, in Mannheim, the closing of a coal-fired power plant, the largest energy production site in Baden-Württemberg, garnered considerable media attention regarding climate policy. This coal-fired power plant was originally scheduled to be shut down in 2020, but in response to the energy crisis related to the war in Ukraine, it will remain open until 2025 to ensure energy security in case of a gas shortage.

Discourse network analysis

In policy debates, actors contribute to the discourse with their statements and try to influence, convince, or learn from each other; this creates relations between the actors. DNA is a method that combines content-based text analysis and relational network analysis³⁹. DNA involves the qualitative coding of statements about policy positions made by individual speakers and organizations engaging in the discourse.

Within the discourse, the actors and the connections between these actors can be visualized in network graphs, with nodes and links⁴⁰. The size of the node represents the frequency of time the actor (organizations in our study) engages in the discourse; the links between the organizations indicate shared policy positions or themes. The links are stronger if organizations share multiple themes. DNA measures the dominance of policy position themes in the discourse using two measures. First, theme frequency refers to the number of times that theme is mentioned in the public discourse. The second measure, degree centrality, refers to the degree of connectedness between that theme and other policy themes. If degree centrality is high, the theme is central in the discourse. We define dominant actors as those with high centrality and high frequency.

Data selection and manual coding procedure

Within the 1 January 2018, to 1 June 2021, time frame explained above, we selected newspaper articles, primarily from municipal and local community newspapers, using the keywords “climate change” OR “climate protection” OR “sustainability.” (German expressions: “climate change” (Klimawandel) OR “climate protection” (Klimaschutz) OR “sustainability” (Nachhaltigkeit)). When screening the articles, we found duplicate or thematically irrelevant articles and excluded those. In addition, articles that dealt exclusively with national climate policy without reference to local issues were excluded. In the CA cities, the availability of newspaper articles was significantly lower, revealing differences between the media systems. Therefore, we also collected articles from regional (e.g. Los Angeles and San Francisco) and national US newspapers that discussed climate change in Long Beach and Oakland specifically. We also included transcripts of relevant city council meetings for the CA cities. Four types of information were manually coded⁴¹: actors, affiliated organizations, concepts, and agreement/disagreement. We coded agreement or disagreement based on the content of direct or indirect statements in the newspaper articles or statements made at the city council meetings. For example, if an actor appears in a newspaper article with a direct statement in quotation marks, then that statement was coded. If an actor makes a statement that is paraphrased in the text of the newspaper article, it is an indirect statement and has also been coded. In both cases of direct or indirect statements, we coded the content of the statement as a concept when it expressed a policy position on climate action. The list of concepts is included in the codebook (see Supplementary Table 1). This codebook was prepared in advance through deductive methods and further developed inductively during the coding process by the research groups in both BW and CA, as well as during joint weekly meetings of both country teams. The coding process presented challenges due to linguistic differences, but to ensure intercoder reliability, categories were constantly reconciled and discussed. All German statements in the text above were translated by the authors.

Policy network analysis and visualization

We first analyzed one-mode networks of actors by representing organizations in local climate policymaking as nodes that are connected through shared themes. If two actors share one or more themes within the discourse, this can be interpreted as an ideational connection with a strength depending on the number of shared themes. This means, the proximity of two actors’ opinions on climate change policy themes in the discourse is directly related to the number of shared concepts or themes. This is visualized through network graphs where the position of nodes denotes the actors. If actors share many themes, they appear closer in the network graphs due to the layout algorithm of the Visone network visualization software we used⁴⁰. For this network analysis, the “subtract function” from the DNA software was utilized⁴¹. This function subtracts the number of incongruent themes, where one actor disagrees on a theme and the other agrees, from the number of congruent themes, where actors agree on the same theme⁴². The subtract function is useful in identifying strongly connected groups of actors. To find out how dominant actors and groups of actors are in the discourse, we focused in our study on the top 15 organizations based on their degree centrality and frequency. Measures of high centrality and high frequency are indicators of strong connection, importance or dominance in the discourse²⁴.

Second, we analyzed two-mode networks consisting of themes that are shared by different actors in the discourses. Here we use the measures degree centrality and frequency as an indicator of importance of a theme in the discourse. Third, to trace the evolution over time, we compare different time slices as explained above.

Qualitative content analysis

In the last part of this study, we conducted a qualitative content analysis of the quotes from the different actors across the four cities in the data set. First, we sorted the top 15 themes based on degree centrality and frequency and reviewed the areas of highest agreement and disagreement in the debate. Then we qualitatively analyzed the individual quotes representing the top 15 themes and categorized them into the different modes of climate urbanism and climate delay discourses from the literature, while also identifying modes not depicted in the existing typologies.

The DNA data are publicly available and uploaded on GitHub https://github.com/MelNa18/CA-BW-climate-policy.git.

Authors and Affiliations

1. California State University, Long Beach, Department of Geography, Long Beach, CA, USA

   Christine L. Jocoy&Lily House-Peters

2. University of Tübingen, Geo and Environmental Research Center, Tübingen, Germany

   Melanie Nagel

3. University of Tübingen, Institute of Political Science, Tübingen, Germany

   Melanie Nagel

4. Heidelberg University, Heidelberg Center for the Environment (HCE), Heidelberg, Germany

   Melanie Nagel

Contributions

C.L.J.: conceptualization, data collection, and analysis, investigation, writing—original draft preparation, reviewing and editing, funding acquisition. M.N.: conceptualization, methodology, software, data collection and analysis, visualization, writing—original draft preparation, reviewing and editing, funding acquisition. L.H.-P.: conceptualization, data collection, and analysis, writing—original draft preparation, reviewing, and editing.

Corresponding author

Correspondence to Melanie Nagel.

Competing interests

The authors declare no competing interests.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions