Introduction: China’s Climate Change Dilemma

‘Why’, asks Boulton in a recent edition of E:CO (Boulton & Mitleton-Kelly, 2010: 2), ‘is climate change policy of such interest for complexity thinkers’? The reasons she offers—the scale of the issue, its sheer complexity and its centrality to questions of economic growth and social justice—apply nowhere more than in China.

As the world’s biggest emitter of greenhouse gases, China is not only a major contributor to global warming; it is itself a country where the effects of climate change are keenly felt. According to a 2008 government white paper ‘China is one of the countries most vulnerable to the adverse effects of climate change’ (State Council, 2008: 7). The paper itemizes the ‘adverse impacts’ climate change has already had on China: ‘increased instability in agricultural production’; ‘severe damages to crops and livestock production caused by drought and hot extremes’; ‘rising frequency of animal and plant diseases’; more frequent floods and worse droughts; and ‘sea-level rise’ which has caused ‘soil salinization and coastal erosion’ and has degraded fishery zones (State Council, 2008: 7-10).

Plainly, environmental degradation exacerbated by climate change is one of the biggest issues confronting China’s leadership. Calculating that environmental damage has cost eight to fifteen percent of GDP each year (excluding health costs), Ministry of Environmental Protection official Pan Yue has claimed that ‘China has lost almost everything it has gained since the late 1970s due to pollution’ (Byrnes, 2006). The OECD (2010: 205) estimated in 2007 that the ‘cost of inaction’ on climate change to China’s economy is around $100 billion annually.

So it is not surprising that the Chinese government has set ambitious targets for environmental protection, reduced national energy intensity and the greater use of clean energy (Minas, 2010: 34-36).

But top-down policies and targets have yielded limited returns. As a recent Carnegie Endowment paper put it, ‘China’s strong, centralized policy on clean and renewable energy has not translated into tangible incentives at the provincial level’ (Chandler, 2008: 6). According to the Climate Group’s Wu, ‘[w]hen local growth is the dominating performance indicator [for officials], and if no mechanism holds local leaders accountable for environmental protection, nationals laws and regulations and policy targets would remain ineffective and unenforced’ (2010: 178).

In practice, much has relied on the presence or absence of ‘enlightened local officials’ (Economy, 2004: 260). A recent study described China’s energy strategy as ‘somewhat confused and uncoordinated’ (Steinfeld, 2008: 144). According to the OECD (2010: 203-204), ‘laws are formulated in a way to let local administrations interpret them to their own advantage’.

Responding to climate change, then, is as much a problem of governance as of environmental protection. In China, the misaligned incentives of provincial and local officials have been a significant obstacle to progress. Complexity theory suggests some mechanisms for shifting these incentives and addressing China’s wider climate policy dilemma.

Complex Adaptive Systems

Complexity theory is concerned with the processes by which systems change. There are two key propositions: 1. Systemic change will be determined by competing ‘attractors’—forces which incentivize systemic actors to adopt or intensify certain characteristic behaviors; 2. The responses to changed attractors within a system tend to be ‘nonlinear’—i.e. ‘a very small change in the independent variable can result in a very large—even infinitely large—response in the dependent variable’ (Priesmeyer, 1992: 15, 21-22). Changed attractors may result in ‘regime shift’, whereby a system can ‘internally switch between different self-reinforcing processes that dominate how the system functions’. ‘Resilience’ refers to the degree of pressure a system can sustain before such a shift takes place (Norberg & Cumming, 2008: 2-3).

The concept of ‘complex adaptive systems’ focuses on the ‘capacity of the system to change in response to prevailing (and anticipated, where possible) conditions by means of self-organization, learning, and reasoning’ (Norberg & Cumming, 2008: 2). Complex adaptive systems share a number of key attributes: Their components will simultaneously be members of multiple sub- and supra-systems, referred to by Baskin (2007b: 60) as ‘nested networks’; the behavior of systemic actors will be governed by rule sets, or system control parameters; the systems and their members are adaptive, with both the members and the rule sets changing over time; and systemic actors will self-organize ‘based on internal interaction rules and external constraints, result[ing] in more or less stable patterns, with capacity for massive change’ (Minas, 2005: 34-35). As Baskin (2007b: 60-61) puts it, ‘[i]t is the interaction of its agents, on one hand, and the conditions of its environment, on the other, that determines the behavior of any CAS’.

Self-organization is driven by feedback loops which are nonlinear and which interact with each other (Richardson, 2008: 14)—‘in a highly interconnected world, phenomena have many causes, some of which may not be evident’ (Baskin, 2007b: 61). Novel behavior is thereby not compelled, but rather ‘emerges as a result of the rich interaction over time of multiple component agents and of the system with its context’ (Minas, 2005: 35). Once system parameters change, systemic adaptation occurs ‘for free’ (Kauffman, in Glor, 2007: 36). Adaptation does not result in an eventual, optimal state, as in Game Theory, but is rather an ongoing ‘chain of responses without any obvious end’ (Allen et al., 2006: 2). This has clear implications for policy-making: ‘Recognizing that there is no optimal solution … pushes us to see policy as experiment, to recognize that there must be built into any policy implementation processes a way of reviewing outcomes collaboratively from a number of perspectives, of learning, of modification, of keeping policies live’ (Boulton, 2010: 35, emphases in the original).

Society And Organizations As Complex Adaptive Systems

The insights derived from complexity theory have been applied to many disciplines across the physical, biological and social sciences, and have been useful in gaining a clearer understanding of phenomena as diverse as ‘evolution, meteorology, ecology, traffic congestion on roads, traffic congestion in computer networks, urban demographics, the frequency and scale of avalanches, stock prices, and the onset of financial crisis’ (Higgs, 2001: 4). Complexity theory has been usefully applied also in a variety of organization science contexts, such as nascent entrepreneurship (Lichtenstein et al., 2007), public health system reform (Minas, 2005), the development of research communities (Allen et al., 2006) and the tensions within multilateral institutions (Higgs, 2001). Characterizations of organizations (and societies) as complex adaptive systems stress the diversity of systemic actors, which are ‘loosely and often nonlinearly linked and [which] produce emergent patterns of systemic behavior’ (Meek et al., 2007: 25). Human complex systems, like their counterparts in nature, are characterized by multiple levels of organization—‘the individual, team, divisional and group level and also in a much larger web of external complex adaptive systems—their economic, social and political environments’ (Carlisle & McMillan, 2006: 3). The inflexible application of natural science complexity models to human organization would strain those models to breaking point. Instead of testing these models to destruction, complexity has been applied to social systems as an analogical ‘way of thinking’—a conceptual lens rather than a ‘theory of everything’ (Baskin, 2007a: 112).

Complex systems theory has yielded a number of insights into the management of human organization—firms, governments and whole societies:

Minimum Specifications

Because ‘[m]anagers cannot control environments’, the tendency to micromanage will often be counterproductive (Carlisle & McMillan 2006: 6). The existence of ‘rigid procedures, bureaucratic regulations and hierarchical controls’ tends to ‘hamper’ innovation (Carlisle & McMillan 2006: 5). Hierarchical organizational culture tends to discourage self-organizing innovation (Glor 2007: 44). Because causation in complex systems is nonlinear, the introduction of detailed and inflexible regulations will often produce unintended and unwelcome results. Instead, complexity suggests that

[p]rogress towards goals that are desirable but difficult to achieve can occur through applying to the system a few simple, flexible rules, sometimes referred to as minimum specifications … Minimum specifications leave room for creativity and innovation … If minimum specifications focus on system-wide targets, they encourage generative relationships and the emergence of solutions that are relevant to local conditions (Minas, 2005: 37).

Rich Interactions

Minimum specifications allow systemic actors ‘as much scope and support as possible to self-organize into “cells” or groups and to network’ beyond conventional ‘silos’ (Carlisle & McMillan, 2006: 7). The development of ‘mission-driven network[s]’ is encouraged (Meek et al., 2007: 33). Such interactions hold out the prospect of ‘yield[ing] new learning relevant for innovation’ (Surie & Hazy, 2006: 13). The encouragement of localized interactions, as in the PPP case study below, also allows managers to experiment with new incentive settings while limiting ‘the consequences of mistakes or underdeveloped ideas’ (Surie & Hazy, 2006: 19).


Whereas, in the conventional narrative, ‘ “change” is the dragon slain by each heroic incoming CEO [or president], or yoked to his triumphal chariot’ (Hodge & Coronado, 2007: 3), complexity suggests that management is essentially an enabler: by altering systemic parameters and encouraging rich interactions, managers can expedite shifts to new and more desirable ‘structural attractors’ (Allen et al., 2006: 15). Complexity-aware management utilizes self-organization, rather than attempting to override it.

Emergent Leadership

Finally (and most dramatically), complexity challenges conventional management thinking by holding out the possibility that leadership is itself ‘an emergent event’, ‘in which knowledge, action preferences, and behaviors change, thereby provoking an organization to become more adaptive’ (Lichtenstein et al., 2006: 2, 4). This possibility is illustrated by Allen’s example of the emergence, in academic research, of ‘clusters’ of researchers and papers which endorse particular ideas and ignore or dismiss others. The publication of a ‘single “outlying” paper’ can begin this process but cannot shape or direct it (Allen et al., 2006: 15). Similarly, Kelly’s account of the development of open source software concludes that innovation is driven by ‘a spectrum of attitudes, techniques and tools that promote collaboration, sharing, aggregation, coordination, ad hocracy and a host of other newly enabled types of social cooperation’ (2009: 122). The implication of this perspective is that ‘leadership can occur anywhere within a social system’ (Lichtenstein et al., 2006: 4). Innovation and the setting of system parameters are not the exclusive preserve of government, as the case studies below illustrate.

Richardson suggests that the key features of complex adaptive systems ‘may be common sense to the experienced manager’ (2008: 25). Perhaps so. The role of incentives in determining behavior and the creative capacity of non-hierarchical organization are not observations unique to complexity theory. But the formalization of suppositions can itself be useful, as ‘a check on our intuition. Sometimes it shows that our intuition is wrong’ (Barrett, 1998: 318). In any case, the complex systems perspective is arguably a ‘profoundly different way’ of conceptualizing management and policy, one which breaks with the received wisdom of organization science (Richardson, 2008: 25).

Chinese society and complexity theory: There is an argument that Chinese society should be particularly receptive to these management insights of complexity theory. This argument rests on the observation that the ‘similarities between the principles of Chinese philosophy and complexity theory are deep and profound’ (Baskin, 2007b: 62). Baskin has identified a set of areas where complexity principles correspond to Chinese philosophy (Daoism & Neo-Daoism). These include a shared awareness of inter-connectedness and nonlinear causality (Baskin, 2007b: 62-64).

Other analysts have credited to Confucius the ‘essential ideas and predictions of the science of complexity’ (Jones & Culliney, 1998: 395), while Davis argues that the Daoist notion of wuwei (translated as ‘effortless action’) holds lessons for contemporary management (2004: 55-57).

Notwithstanding the controversy over the particular meaning of wuwei (Ivanhoe, 2007) and the concept’s obvious tension with China’s long history of centralized rule, these analyses suggest that complexity-aware policy could encounter a lower threshold of resistance in China than in the West, with its intellectual tradition of ‘linear causality’ (Baskin, 2007a: 112). The possible correspondence between complexity theory and ancient principles of Chinese management constitutes fertile ground for further research.

Rich interactions: Harnessing Complexity In Climate Change Policy

The following case studies hint at the enormous potential for effective, complexity-appropriate climate change initiatives in China. The initiatives described share several characteristics: Rather than prescribing detailed rules, they involve the setting of broad, systemic incentives; the actors involved self-organise, with minimal direction from national government; leadership, or movement towards desired ends, is indeed a collaborative rather than a centralised act; and the projects involved are (to varying degrees) localized, responding to local conditions but with the potential for expansion.

Nanjing Voluntary Energy Efficiency Agreements

Beginning with an initial phase to assess feasibility in 2005-2006, a voluntary energy efficiency programme has been developed in Nanjing. Firms entered into agreements with the Nanjing Environmental Protection Bureau (EPB) in 2008. Each committed to an energy intensity reduction target of 3-5% by 2009 on 2007 figures. This project is one of the first of its kind in mainland China and is based on the recognition that, with ‘few exceptions, implementation of top—down policies at the local level is weak’ (Eichhorst & Bongardt, 2009: 1859). Project-designers have therefore concentrated on setting incentives to encourage effective participation.

The agreements were flexible, accounting for the actual capacity of firms to address issues, as well as particular local and sectoral requirements. Firms were motivated to participate by several factors, among which improving relations with the Nanjing EPB ‘turned out to be a major driver for all of them’ (Eichhorst & Bongardt, 2009: 1863). Additionally, the agreements gave firms opportunities to improve their own technical know-how, by exchanging knowledge with the public authorities on specific problems (Eichhorst & Bongardt, 2009: 1857). During the pilot phase, close ‘technical assistance’ and ‘process management’ assistance were provided by Dutch SenterNovem experts (Eichhorst & Bongardt, 2009: 1857). ‘International expertise and experience’ in reducing GHG emissions was reported as a major incentive for government in the pilot programme (Eichhorst & Bongardt, 2009: 1863). Additionally, because Nanjing has prestigious ‘Environmental Model City’ status, its government has more than the usual incentive to achieve environmental targets.

Following the Dutch experience, each agreement focused on ‘strong process management’, with each company setting up an ‘Energy Action Team’ (with senior management participation) responsible for implementation, monitoring and liaising with the EPB (Eichhorst & Bongardt, 2009: 1858). The project has been supported, rather than frustrated, by the Ministry of Environmental Protection, which has avoided detailed implementation guidelines and has ‘encouraged local innovation’, even allowing for ‘informal contracts’ between the EPB and firms (Eichhorst & Bongardt, 2009: 1861).

Eichhorst and Bongardt report that the Nanjing model may be expanded to other localities if it is judged successful. In September 2008, the National Development and Reform Commission flagged the possibility of including voluntary agreements in future national energy policies. Eichhorst and Bongardt conclude that the

partnership approach of voluntary agreements showed to have the potential to overcome both the institutional and policy barriers to implementation of top—down policies by developing new spaces of communication and dialogue between EPBs (environmental protection) and state-owned industry (economic development). In this context, focusing on the management procedures for implementation was essential (Eichhorst & Bongardt, 2009: 1864).

The Nanjing case shows that effective voluntary agreements have the potential to align the incentives of environmental and economic agencies, attract firms to participate, encourage knowledge transfer from the West and design appropriate local solutions. The voluntary agreement model also responds to the limited reach of national command-and-control measures. As the OECD found, the Nanjing programme ‘shows that voluntary agreements can complement national measures and create incentives for public participation’ (2010: 205).

Local Incentive Shifts—Private-Public Partnerships And ‘Issue-Bundling’

In mainland China, the central government has seen its ability to control emissions practices reduced by the privatization of many polluting state-owned enterprises (Koehn, 2008: 56). As a consequence, sub-national government initiatives have assumed greater importance. However, sub-national government agencies will often be dependant on revenues from polluting enterprises, reducing their willingness to take such initiatives.

Two developments have had the effect of incentivizing sub-national governments to improve local energy practices. The first development resets economic incentives. Following China’s accession to the WTO, local governments have been empowered to solicit and authorise foreign and joint-venture investments in green technologies. Such investment creates employment and can further innovation. For example, provincial governments have subsidized Suntech silicon photovoltaic cell factories in their localities. Governments have also been active in securing the capital and knowledge transfer necessary for localized wind turbine projects. The injection of foreign capital and technology, together with the growing ability of clean energy installations to provide employment and efficient energy, weakens government dependence on polluting installations (Koehn, 2008: 69).

The second development concerns the political pressures on sub-national governments. Concerns over public health and environmental degradation are responsible for many of the thousands of protests that occur in China each year. Koehn suggests that this mounting unrest has ‘opened a new window for sub-national framing of GHG-emission mitigation’ (Koehn, 2008: 63-64). For example, officials from Liaoning province and Benxi city linked emissions mitigation with public health concerns. Citing inordinately high respiratory disease rates in industrial areas of Benxi, the officials launched a ‘green schools’ initiative and transformed the city, which previously ‘could not be seen on satellite images due to the heavy smog’, into a ‘model’ for ‘source-point pollution control and prevention’ (Koehn, 2008: 68). The motivation of avoiding local unrest is likely to grow if China’s environmental problems intensify. Importantly, the urge to act is complemented by other incentives, notably the desire to earn ‘model city’-style plaudits from provincial and national authorities. As the Benxi example shows, popular unrest can create new opportunities for policy entrepreneurs in sub-national governments to further their agenda.

In both developments, emerging attractors (WTO accession and the availability of foreign capital and expertise in the first case and intensifying health concerns in the second case) have prompted unpredictable responses: the creation of climate-friendly projects. The nature and strength of such emerging attractors will vary from place to place, as will the degree of resistance to them (determined, inter alia, by the reliance of sub-national governments on revenues from polluting firms).

City-To-City Networks

Local governments have also begun to engage in a form of ad hoc sub-national diplomacy, forming associations amongst themselves to reduce GHG emissions. Two such initiatives are the Cities for Climate Protection (CCP) programme and the C40 Cities Climate Leadership Group (C40).

The CCP programme is run by the International Council for Local Environmental Initiatives (ICLEI), which was established in 1990. The CCP programme was established in 1993 and involves more than 676 local authorities from around the world. The total area that they cover accounts for over eight percent of global GHG emissions (Betsill & Bulkeley, 2006: 143).

CCP participants commit to five ‘milestones’: (1) Conduct a baseline emissions inventory and forecast; (2) Adopt an emissions reduction target for the forecast year; (3) Develop a Local Action Plan; (4) Implement policies and measures; and (5) Monitor and verify results. In return, participants receive ICLEI support in the form of software for monitoring GHG emissions and information regarding best practices.

There are indications that the CCP programme has value. Its United States participants reduced their GHG emissions by an average of 100,000 metric tonnes per city in 1999. Australian participants reduced emissions by 78,182 metric tonnes during 2000-2001. In both countries, GHG emissions increased on average during these periods (Betsill & Bulkeley, 2006: 144). In 2007, ICLEI released a ‘local government operations protocol for the quantification and reporting of greenhouse gas inventories’ (Dodman 2009: 188).

Betsill and Bulkeley argue that the CCP programme exemplifies ‘a new form of environmental governance’, by creating ‘a new sphere of authority through which the governance of climate change is taking place and which is not bound to a particular scale’ (Betsill & Bulkeley, 2006: 144, 151). Certainly, in some senses, the CCP programme challenges the traditional state-centric model of organization. CCP bypasses states in an issue area of national concern. Further, by creating linkages across national borders and establishing regional secretariats, ICLEI could be depicted as usurping the national government prerogative of conducting external affairs. Rather than perceiving a threat, however, national governments should look on the activities of ICLEI and its peers as necessary complements to their own efforts, adding a ‘crucial layer to the complexity of global climate governance’ (Pattberg & Stripple, 2008: 379). Self-organizing public networks have stepped into the breach left by inadequate national leadership. By their very existence, they demonstrate that if the old ways of command-and-control management worked, they would have worked by now.

Currently, no Chinese municipality participates in the CCP programme. CCP participation has the potential to augment sub-national government initiatives of the kind described above. Additionally, there is scope for Chinese municipalities to create their own networks, if political considerations preclude involvement in the CCP.

The C40 was formed in 2006 as a partnership between the Large Cities Climate Leadership Group and the Clinton Climate Initiative (CCI). The CCI is pledged to provide the C40 cities with emission mitigation ‘solutions’, including: ‘Pooling the buying power of cities’ (to lower the price of energy saving products); ‘Mobilizing expert assistance to help cities develop and implement programmes’; and ‘Creat[ing] and deploy[ing] common measurement tools so that cities can establish a baseline on their greenhouse gas emissions, track reductions and share best practice’. The C40 has pooled the ‘climate change action plans’ of its participating cities, which include Beijing, Shanghai and Hong Kong (which hosted the latest C40 workshop, in November 2010).

The Future: Climate Change Policy For Complex Adaptive Systems

The application of complex systems theory suggests what some of the contours of China’s solution to its climate change dilemma will be. Complexity also suggests what the solution will not be.

The solution will not be solely to introduce new regulations, for China already has ‘some of the most advanced laws’ on renewables, clean production, environmental impact assessment and pollution control (as Ministry of Environmental Protection vice minister Pan Yue says: ‘In theory, we have solved the problems. Now, the challenge is to make this compulsory’) (Warburton & Horn, 2007: 51).

Nor is more rigorous enforcement, on its own, the appropriate solution. Such is the strength, and stability, of local economic growth incentives that the adversarial enforcement of national regulations is unlikely to succeed (and not without incurring prohibitive costs): ‘If resistance is seen as the reason [for thwarted change] then the solution is to battle against and to overcome resistance, wherever it is to be found. However, in [complex adaptive systems], behavior follows attractors in the system’ (Minas, 2005: 37).

Additionally, complicated (and perhaps competing) regulations can have baneful consequences: the ‘[n]umerous experiments, competing standards, and alternative microlevel approaches’ adopted in China in recent years have, according to Steinfeld, compromised the ‘internalization of the externalities associated with national energy choices’ (2008: 145). ‘Complexity thinking’, according to Boulton, ‘suggests this uncertainty [of consequences] is the norm rather than an aberration in difficult circumstances’ (Boulton, 2010: 35).

Policy solutions, as the above case studies suggest, will involve interventions to create, identify and strengthen incentives that attract local governments, firms and civil society toward climate-friendly behavior. Such interventions should encourage the creativity of systemic actors (in the form of, for example, non-hierarchical networks and Corporate Social Responsibility) and should allow for local variation. Policy-makers should be flexible in their initiatives and be prepared to respond to inherently unpredictable outcomes, both desirable and malign.

Major trends augur well for the development of complexity-appropriate climate change policy. Intensive technology transfer and internal innovation have resulted in the development of substantial clean energy industries in China. It is now possible to speculate that China may become a market leader in energy efficiency standards:

In setting and enforcing tough energy-efficiency standards for consumer appliances and vehicles, the government leaves foreign producers little choice but to comply and innovate. Similarly, to the extent that domestic producers are forced to meet these standards, they develop core competencies in the design, development, and production of energy-efficient products, competencies for which global markets will only grow as energy constraints become more binding on all nations in the future (Steinfeld, 2008: 145-46).

A complementary development has been the rapid and impressive upgrading of China’s scientific and diplomatic personnel involved in climate change. Policy-makers now commonly recognise that, in the long-term, mitigating climate change and continuing economic growth are not imperatives pointing in opposite directions, but are two sides of China’s developmental coin (Yu, 2008: 88).

These developments have made the economic incentives favouring further clean energy development stronger, with richer creativity within the system on which to draw. This suggests that complexity-aware policy is becoming not only more appropriate, but also easier to do.


The lesson of complex systems theory is that system behavior changes when incentives (attractors) shift, with innovation often emerging from surprising places. In climate change policy, the task of the policy-maker is to devise a set of minimum specifications or rule sets that foster system connectedness and allow for the promise of non-hierarchical creativity to be realized. Innovations will emerge from the interactions of national and sub-national governments, businesses, epistemic communities and civil society.

The examples of voluntary energy efficiency agreements, local issue-bundling, public-private partnerships and city-to-city networks should encourage policy-makers to apply a deft, but light, touch to the climate change dilemma. In each case, creativity beyond government was harnessed—the expertise of foreign advisers, the know-how of local innovators, the pooled experience of civil societies and non-government organizations and the outrage of citizens protesting health problems. A continuing challenge is how to propagate or scale up these innovations so that they have national and supra-national impact.

It is a challenge that complexity theory can help to meet. As Boulton recently put it, ‘[c]omplexity thinking emphasizes the impossibility of separating out economic from other considerations’ (2010: 36). Policy-makers charged with responding to climate change—in China and elsewhere—should bear this reality firmly in mind.