key: cord-0843809-ezn4ai8d authors: Bruun, Ole; Rubin, Olivier title: Authoritarian Environmentalism—Captured Collaboration in Vietnamese Water Management date: 2022-04-26 journal: Environ Manage DOI: 10.1007/s00267-022-01650-7 sha: 1ee32bb5319beb0e98f454fef264d865ae583f0a doc_id: 843809 cord_uid: ezn4ai8d This article examines collaborative environmental governance under authoritarian political structures. Building on the theoretical frame of authoritarian environmentalism, it peruses fieldwork material collected during 2009–2019 to determine the most prominent features of recent collaborative governance efforts in the field of water management in Vietnam, a historically seasonal flooding-dependent country. A key feature is technocratisation, where top-down management structures and practices prioritise technocratic solutions to environmental challenges over deliberation, awareness raising, and integration of local knowledge. Another equally important feature is authoritarian intensification, by which increasingly complex environmental management functions, coupled with the state’s determination to retain political control, reinforce authoritarian governance. We jointly refer to these features as captured collaboration, signifying a strong authoritarian regime dominance in both vertical and horizontal relations of environmental governance. However, while captured collaboration still appears to be a defining collaborative characteristic, the article acknowledges rising calls for deliberative government in Vietnamese society. This is particularly outspoken in relation to the highly contested issues of hydropower construction and enhanced floods, debates that simultaneously have paved the way for a burgeoning, though much delayed, paradigm shift. Collaborative governance has increasingly become mainstream in academia and institutionalised in political practices over recent decades (e.g., Bartz et al. 2021; Emerson et al. 2012; Ansell et al. 2020; Sorensen and Torfing 2009, 2021) . The recent compilation of the Collaborative Governance Case Database, in particular, has allowed for more comparative research in the field (see Douglas et al. 2020; Parker et al. 2020; Ulibarri et al. 2020, among others) . However, most collaborative governance research continues to be conducted in pluralistic settings and high-income countries (Chen et al. 2021) . 1 This pattern also holds true with respect to collaborative governance in the field of environment management, where only a few studies have been conducted in authoritarian (mainly Chinese) settings (Huang et al. 2017; Dai et al. 2020; Zhou and Dai 2021; Chen et al. 2021) . To begin closing this gap, using water management in Vietnam as a case in point, this article explores those special features of collaborative governance that may arise under authoritarian political structures. The case of Vietnam represents a typical one-party system, since the Communist Party of Vietnam (CPV) controls the highest governance organ, the National Assembly. Being the only legal party in Vietnam, the CPV has assigned itself a leading role in society and appoints the Central Committee as the top decision-making body. However, as a response to international calls for policy adherence and greater measures of inclusion and to domestic calls for democratisation and accountability, in the early 2000s a process of adopting elements of participatory and collaborative governance was initiated. We argue that the resulting instances of collaborative governance still bear the imprints of one-party rule. While water governance relies on horizontal networks of actors that engage in repeated interactions at multiple levels, the Vietnamese state restricts the autonomy and selfdetermination of the participating actors by exercising direct control from within. As compared to a mainstream understanding of collaborative governance, therefore, several authoritarian features stand out. Rather than using collaborative governance as a means of seeking trust-based knowledge sharing, compromises and agreements, it primarily serves the purpose of building legitimacy and public support through increased technical and administrative efficiency in the implementation of centrally defined policies. Similarly, rather than encouraging deliberative discussions of problems and responses in the pursuit of bottom-up accountability, collaboration is directed towards issues, contexts, and processes that do not challenge the basic allocation of power or resources in society. We jointly refer to these manifestations of collaboration under authoritarianism as captured collaborative governance. Drawing on substantial survey data and fieldwork experience from 2009 to 2019 in Vietnamese water management, we further suggest that two core features of Authoritarian environmentalism (AE) tend to restrain policy intentions of increasing collaborative engagement. One longterm feature is technocratisation, where top-down management structures and practices prioritise technocratic solutions to environmental challenges over deliberation, awareness raising, and integration of local knowledge. Another more recent feature is authoritarian intensification, by which increasingly complex environment management functions, coupled with the state's determination to retain political control, generate self-strengthening authoritarian governance. Within the methodological confines of a single country case (Vietnam) and a specific type of environmental management (water management), we suggest that the dynamics of AE set distinct criteria for collaborative governance. Thus, water management in Vietnam might formally check some of the boxes of collaborative governance in that it involves a broad range of public agencies, levels of government, mass associations, and local communities. But without robust accountability and actor autonomy it remains separate from the deliberative and consensus-seeking process often associated with collaborative governance (see Ansell and Gash 2008; Doberstein 2016; Sorensen and Torfing 2021) . The most recent research on collaborative governance in authoritarian settings seeks to reconcile authoritarian state structures with some form of collaborative governance, which obviously gives rise to varying practices (e.g., Zhou and Dai 2021; Chen et al. 2021) . Thus, in line with these other studies on collaborative governance in authoritarian settings, we understand collaborative governance as a broader analytical construct, void of deliberative and trust-based features. Collaborative governance may thus be characterised as a system where actors "across the boundaries of public agencies, levels of government, and/or the public private and civic spheres" carry out a public purpose that could not otherwise have been accomplished Nabatchi 2015:18, Emerson et al. 2012) . AE is used as an analytical entry-point to engage with the scope and characteristics of collaborative governance in Vietnam, specifically with reference to water management. The concept of authoritarian, or coercive environmentalism grew out of debate on the extent to which global climate and ecological crises justified tougher policies, and thus focused on the pursuit of an alternative generic model (Shearman and Smith 2007; Beeson 2010) . It was further spurred by green development policies in authoritarian states, not least China's ecological civilisation programme (Li and Shapiro 2020) . AE has, therefore, been understood as a public policy model that can concentrate authority in a few executive agencies operated by capable and honest elites seeking to improve environmental outcomes (Gilley 2012: 288; Li 2018: 6) . As further discussed below, the AE model does not preclude collaborative structures of governance involving a broad range of stakeholders but tends to have clearly designated roles for each according to a predetermined scheme as given by the nature of the regime. According to the authoritarian model, environmental policy formulation takes place within a political elite-led technocratic and societal management discourse. In this setup, inter-agency collaboration and technical efficiency is highlighted while public participation is relegated primarily to embracing stateproduced knowledge and complying with state policies (Gilley 2012: 291) . Participation therefore more commonly is oriented towards service and awareness-raising roles, including assisting in implementing government policies and programmes (e.g., Ahlers and Shen 2018: 316) . Consequently, a combination of enlightened state policies and technological change are the primary instruments of environmental management. Consulting "communities", "social organisations", and other stakeholders is increasingly institutionalised but remains inconsequential. According to Gilley, with reference to China, authorities commonly express distrust in the public, and the public's low level of concern, in turn, is cited by state institutions to justify their dominant role. This potentially restrains the AE model's capacity of rapid policy responses and certainly the potential for top-down mobilisation: in its raw form, the exclusion of social actors may create a malign lock-in effect "in which low social concern makes authoritarian approaches both more necessary and more difficult" (Gilley 2012: 300) . Weak public backing to environmental protection is no less a challenge in Vietnam, although policy makers are increasingly aware of a participatory deficit (Bruun 2020a ). An emphasis on policy expediency and a belief in ecoelites being best placed to make environmental decisions unavoidably favour large-scale state environmental interventions. Studies conducted over several decades show that, while these interventions almost categorically favour direct state and elite economic benefits, they at the same time disproportionately affect the livelihoods of smaller remote communities and ignore local sustainability impacts (Lo 2021) , giving rise to calls for policy change in society. They may further have devastating socio-economic impacts and exacerbate an unequal distribution of economic wealth and environmental harm between rural and urban communities (Bruun 2012; Lo 2020) . As associated with political resilience priorities and narrowing the scope for collaboration, several studies suggest that nature and forest management are contingent on social management rather than attributed independent merit (McElwee 2016: 5; Bruun 2020a). The authoritarian approach to green development has not least been one of simultaneously expanding the regulatory scope of the state to encompass a growing range of environmental issues and co-opting non-state actors into the state agenda (Li and Shapiro 2020: 11) . A growing body of studies has shown, for instance, the doubtful consequences of policies that builds exclusively on technocratic and regulatory discourse with little reference to society (Gilley 2012) , a considerable discrepancy between strong targetbased policies and weak outcomes (Li 2018) , and technocratic policies fostering a "command without control" situation (Kostka 2016) . Other studies have indicated that inherent authoritarian fragmentation and complex policy processes at local levels give rise to weaknesses in horizontal collaboration (Eaton and Kostka 2018) and allow considerable space for economic interests (Carlitz and Povitkina 2021) . Just like environmental and climate issues can topple governments in democracies, they have an equal power to set new agendas in authoritarian regimes and threaten their legitimacy, although they remain subjected to the forces of regime resilience. However, authoritarian regimes have continued to develop complex and strategic new forms to grow more resilient (e.g., Diamond 2021; Svolik 2012); in particular, one-party regimes have shown remarkable resilience and ability to maintain elite coherence (Svolik 2012: 163) . The challenge of collaborative governance in authoritarian settings may be compared to the "dictator's dilemma", articulated by US secretary of state George Schultz in 1985, in which openness is necessary to sustain economic growth but at the same time may erode the legitimacy of a closed regime. Similarly, collaborative governance benefits from the inclusion of a range of independent actors as well as free-flowing information between them. This collaborative effort contributes to the state's ability to effectively manage environmental challenges by adding expertise, capacity, and knowledge. However, at the same time, the participation of independent actors with free access to information may create multiple auditors with the capacity to expose policy failures and untapped policy alternatives that could embarrass leaders and lead to the public inference that the authoritarian rulers are inept and should be replaced (Egorov et al. 2009 ). Rather than strengthening top-down coercive measures (which is expensive) or embarking on pluralistic reforms (which undermine the regime), one possible solution to the collaborative governance dilemma is to embark on the decentralisation of state governance. Decentralisation without devolution of power is a concrete manifestation of authoritarian intensification where the regime permeates down through local government entities and mass associations to benefit from local presence and new controls, including over local agencies, external expertise, and various associations. The government, therefore, ideally exercises control both vertically, through hierarchical political structures, and horizontally, through local agencies and associations. The inherent dynamic involves a conscious response to increased complexity in environmental management by means of intensified multilevel government controls, essentially capturing collaboration at the roots. A crucial underlying factor is the state's dominant and unchallengeable position that tends to stifle spontaneous, bottom-up environmental action that may otherwise mobilise the public (Ngoc 2017) . As already suggested by Beeson (2010) , recent studies have indicated a spiral of authoritarian self-intensification when confronted with socio-environmental emergencies. Multiple studies have coupled the rise of state environmentalism in real-life authoritarian regimes with increasing use of repressive power, including new digital controls and restrictions on civil society. For instance, studies suggest that the more there is at stake in environmental conditions that may foster social upheaval and destabilisation, the harsher and more exclusively authoritarian are the means of government (Ahlers and Shen 2018: 301) , that a distinct paradigm of recentralised environmental governance has emerged in China under Xi Jinping (Chen and Lees 2018; Lo 2021) , and that environmental governance increasingly forms part of a much larger effort at centralisation and power consolidation in the hands of the state (Kostka and Zhang 2018; Li and Shapiro 2020: 24) : this may in turn potentially increase the power of technocrats. At the same time, instead of coercing citizens into submission a common new model is based on the conscious manipulation of information, including imitations of democracy, efforts to conceal state repression, and the adoption of a rhetoric of performance (Guriev and Treisman 2019) in which environmental and ecological rhetoric are among the key means of demonstrating competitiveness. The qualitative and quantitative data used in this article draw on several interdisciplinary research projects on climate change adaptation, vulnerabilities, water disasters, and environmental management in Vietnam during the period 2009-2021. The research has been donor (Danida) and otherwise foreign funded and conducted in cooperation with Vietnamese research institutions, thus including policy development components. The provinces covered in research include Nghe An, Ha Tinh, Quang Binh, Quang Nam, and Quang Ngai along the central coast, and Lai Cai Province in the North (Fig. 1 ). Access to first-hand qualitative and quantitative data over a substantial period of time allows us to conduct fairly robust triangulations of findings, and the highly centralised nature of the regime makes us believe that the findings, to a large extent, are representative of Vietnam as a whole. In addition, a range of recent literature is consulted. The Among a broad variety of authoritarian regimes Vietnam can be characterised as a single-party regime with authoritarian structures exclusively built around the CPV. This regime type is fairly widespread in Asian countries apart from Vietnam, including China, Burma, Laos, and (arguably) Cambodia. Thus, Vietnam constitutes an exemplary or typical single case of AE within dominant party regimes in Asia, without the analytical findings necessarily being applicable to all centralised political systems (Seawright and Gerring 2008). Within these authoritarian environmental governance structures, we examine the subset of water management, focusing on the key branches of flood response and hydropower construction. However, we believe there is scope for broader generalisations. The two examined branches of water management have high priority for the CPV and display distinct characteristics: hydropower plant construction is primarily an energy decision with environmental consequences, whereas flood responses contain both prevention and mitigation measures in the interface between environmental and disaster management. It should be highlighted that the purpose of the subsequent analysis is not to provide full-fledged case studies of the two areas, but rather to illuminate the inherent features of authoritarian intensification and technocratisation. As a key aspect of water management in the context of recurrent seasonal flooding, hydropower development in the highlands has become the single most contentious issue in recent decades. After 1975, and particularly after the Doi Moi economic reforms of the 1980s, Vietnam's political and economic elite-initiated hydropower planning and construction activities on a massive scale; only the technical barriers were considered. After a few large-scale plants had been built with assistance from the Soviet Union up until 1995, frantic construction activity between 1995 and 2005 reflected both Vietnam's demand for power to feed economic growth and the free hands of large-scale investors: without considerations for environmental and societal costs, hydropower was cheap. After 2005, several new hydropower plants were constructed, and a phase of in-depth development, including capacity enhancement and cascading, was initiated. Altogether, the construction of over 800 hydropower dams, of which 80 are large scale, has almost fully exploited Vietnam's hydropower potential and affected practically every river in the country. Focusing on favourable locations and low investment costs, many projects were initiated by Vietnam Electricity (EVN) without proper approval procedures, reflecting a policy model that entirely side-lined environmental and local community concerns. Moreover, foreign Clean Development Mechanism financing under the Kyoto Protocol, though with sustainability impacts racked by doubt, has added to a "boom and bust" atmosphere in hydropower construction (Smits and Middleton 2014) . Apart from a huge loss of natural forests and biodiversity by submersion, at least 200,000 people have been displaced for hydropower construction. Over 90% of the relocated people are ethnic minorities who already suffer from the highest poverty incidences in the country (Ty 2015) . Moreover, the coercive modernisation efforts of one-party authorities in the region had already resulted in the relocation of ethnic minorities out of forest areas and into settled agriculture on a massive scale (Eyler 2019: 53). As a key example, in upland Quang Nam Province, dams and hydropower installations were constructed in nearly every possible location along the Vu Gia-Thu Bon River within a short period of time, mainly by EVN. The provincial hydropower plan, approved by the Ministry of Industry and Trade (MOIT) and the Quang Nam People's Committee in 2006, included 40 projects (including eight of large scale), but in 2008 projects mounted to over 60. Some sources estimated that there were as many as 110 stations (including ten large scale), with many not even being formally registered; similar issues are reported for other provinces (e.g., VietnamNet 2020). As a consequence, there are few paddy farmers in the province who are not affected somehow. All researched localities showed considerable impacts, including increased or disturbed flooding, water shortages, and sedimentation. For instance, in Dai Loc District, farmers reported that the A Vuong hydropower station on the upstream Thu Bon River caused the water level to rise an extra metre or more during the seasonal flooding, causing traditional flooding measures to be insufficient and leading to the destruction of dikes, crops, and property. In addition, both farmers and NGOs working in the area reported vast sedimentation in parts of the district, in many cases burying fields in sand and mud. In the context of officially approved research, Vietnamese researchers are not permitted to focus on contentious political issues and adverse consequences of state policy. However, at this point many Vietnamese scientists instead criticised the technocratic thinking of their government on the Internet, and several international organisations supported them by raising their concerns over Vietnam's commitment to sustainable and equitable development. For instance, an environmental assessment by the Asian Development Bank concluded that the hydropower plan was detrimental to the provincial economy, with serious adverse consequences for upland ethnic minority groups, natural systems in general, water supply, and mitigation (ADB 2008: 7, 142) . If conflicts over hydropower construction in Quang Nam had fermented earlier, they burst out in public after the 2009 Ketsana typhoon. During fieldwork, the A Vuong hydropower station, allegedly without warning, released 150 million cubic metres of water during the height of the rains, resulting in a catastrophic downpour of masses of water. Farmers and local authorities in lowland Dai Loc District were in an uproar as unprecedented water surges took everyone by surprise, crushed houses and property, and sent people running for their lives (Bruun 2012) . The water caught people in fields and forests and exacted the loss of 163 lives across the province. Interviews revealed that this and other hydropower plants had largely been constructed without involving the affected communities, as local arms of the state saw no obligation to consult or even inform them (e.g., Bruun 2012) . Local people in the affected areas began comparing hydropower to "water bombs" placed on top of their heads (Le 2015: 6) . The coastal city of Hoi An experienced the highest flooding for decades, with a maximum flood level reached after just 3 days as compared to the typical 2 weeks of monsoon rain. The flooding caused immense losses to the unprepared citizens and business owners. The event generated a public outcry and harsh criticism from the city government, with accusations made openly against hydropower management. An unusually open public debate with the Department of Agriculture that turned against the local People's Committee for allowing the release marked the beginning of policy change and ended in the cancellation of several projects on the Vi Gia-Thu Bon River in 2013 (Le 2015) . Similar events across Vietnam in the following years elevated the conflict to the national level (Ty 2015) and induced the central government to issue new directives on strengthened interagency cooperation. Before 2011, when Decree No. 29/2011/ND-CP on public consultation was issued and in practice installed elements of collaborative government, hydropower construction and management were exclusively technocratic and participation by non-state actors was practically excluded. Larger hydropower plants (over 30 MW) were approved by the National Assembly and the MOIT and were installed even before an Environmental Impact Assessment (EIA) had been conducted. Likewise, provincial and local People's Committees were required to comply with the decisions and cooperate on project implementation in a joint Board of Project Management. The vast majority of Vietnam's larger hydropower installations predate Decree No. 29. From those locations where research has been conducted, a fairly consistent picture emerges. Adverse effects include deforestation, flooding issues, droughts, fish death, and increased salinity. However, in particular, the rights and livelihoods of highlanders were ignored. As an example, the Song Tranh 2 hydropower project, also in Quang Nam Province, implied the relocation of more than 1000 ethnic minority households in Bac Tra My District to make way for the reservoir. According to government decrees on resettlement, people must be offered conditions at least equal to those of their original residence. However, the households were resettled in a protected forest area and the majority never received new productive land, thus leaving them without livelihood assets (Le et al. 2016: 64) . Poverty and food insecurity induced new deforestation and the exploitation of protected forest resources, a situation recognised across the highlands. As an illustration, Vietnam's~30 national parks, mostly established on ethnic minority lands, on average comprise 2.5 hydropower plants (Le 2015) . After the 2011 decree, public participation is mandated during the project development and approval stage, but not during the construction and operation stages (Le et al. 2016: 68) . The decree stipulates that EIAs and Strategic Environmental Assessments (SEA) are to be carried out for all installations over 1 MW, and in particular, the EIA must include not only environmental conditions but also socioeconomic profiling and local consultations to ensure the well-being of affected communities. The public consultation process includes a range of concerned social groups, sociopolitical organisations, professionals, and specialists. However, a narrow definition of affected communities is commonly applied in order to lower project costs, to the effect that displaced people are consulted but not those indirectly affected in uplands and lowland agricultural areas. At the same time, there is a lack of public supervision on the EIA and SEA processes, and no demand to make those pertinent documents available to the public. Displaced people are mostly groups of highlanders with low levels of Vietnamese education and language skills, groups that also constitute the poorest segment of society. They have inadequate skills and leverage when negotiating with powerful investors and, at the same time, they lack the freedom of organisation. In the absence of effective public supervision, investors will collaborate directly with local "communities" and "social organisations". These in reality refer to local arms of the state represented by People's Committee Chairmen at district and commune level and leaders of the propaganda-oriented mass organisations. In some instances, "distinguished members of the community", such as revolutionary heroes and old Party leaders, are involved. The involved parties must agree on forms and models of resettlement and sums of compensation. Yet information tends to be sparse since the EIAs and SEAs are not made public, and many unfounded promises are made, while all communication takes place in Vietnamese (Le et al. 2016: 71-75; Le 2015) . After the public consultation processes, the task of the local People's Committees is to carry out the resettlement in collaboration with the affected people, who are given certain choices of housing sites and design. Some sources describe the process as passive participation or manipulated collaboration, as an outcome of the existing structures of government, which simultaneously prevents a broader range of stakeholders from participating (Le et al. 2016) . In a range of cases, ethnic minority households reportedly have returned to the unsubmerged parts of their ancient lands above the reservoirs. As indicated in theory, an approach to environmental management based on the unquestionable statute of the leading role of the party across state and society has been the basis for an overly technocratic approach. Key stakeholders and collaborators are defined as state organisations, hydropower investors, and local People's Committees, who act on behalf of local people across ethnic boundaries. As a consequence, the sums paid as land compensation are at a level predetermined by the Provincial People's Committee, well below market prices. Those who lose their land are not able to individually negotiate prices, while a range of sources report that compensations may diminish or disappear entirely in the process (Ty 2015; Tran et al. 2013; Ha-Duong et al. 2016) . During fieldwork, provincial authorities were found to unanimously support economic growth over environmental concerns and considerations for highland people displaced by hydropower construction (Buch-Hansen 2013; Bruun 2020b). Similarly, hydropower investors and local governments usually collaborate on the selection of resettlement areas before directly involving the people to be relocated, resulting in most resettlement areas reportedly being insufficient and having poor soil quality. Moreover, reports indicate that in cases where local people resist, forced evictions involve arrests, beatings, and the use of hired gangs (Ha-Duong et al. 2016: 58) . However, hydropower dams have created intra-state water management conflicts between various sectors, not least between electricity production interests and the water resource needs of downstream communities depending on agriculture, aquaculture, fishery, transportation, and drinking water. On their side, the vulnerable households lacked the institutional mechanisms to push for a more holistic approach to water management beyond immediate in-kind disaster relief, such as including the various stressors created by multiple government policies. At the same time, the concept of seeking compensation or complaining about mismanagement at the plants was simply not an option for those common villagers interviewed. Yet, these adverse consequences of the hydropower plants continued to be largely unaddressed by the government until 2013. In connection with National Assembly and government reviews of hydropower during 2012-2014 (Le 2015: 6-10), in 2013 National Assembly resolution 62 (62/2013/QH13) most forcefully gave voice to the rapidly mounting criticism of technocratic governance in hydropower construction. Angry and resentful voices were heard from critical scientists, civil society organisations, environment journalists, local authorities, and the general public. A lively Internet debate occurred with participation from many new actors, and several TV documentaries on hydropower impacts were produced and broadcasted. Criticisms reflected in public debate included concerns over loss of forest land, livelihood disruptions, dam failures, controlled and accidental water discharges, changes in flooding regimes including extended flooding, faulty EIAs, environmental corruption, and a long series of small earthquakes around major hydropower installations in the highlands, not least in Quang Nam. Resolution 62 called on ministries and provincial People's Committees to improve the quality of formulating, appraising, approving and managing implementation of hydropower projects, such as by participatory and collaborative means, as well as well to suspend and cancel inefficient and insecure hydropower projects which had negative impacts on river flows, the environment and people's lives. Both the resolution and a general fear of public unrest resulted in the cancellation of over 400 hydropower projects in the national Power Development Plan (PDP7) in 2016. Yet hundreds of other projects persist and will reportedly increase hydropower generation from 17,000 MW in 2016 to 27,800 MW in 2030. The same technocratisation bias found in hydropower construction is also present in flood responses, suggesting that this dynamic is not uniquely related to a particular branch of water management. The Vietnamese collaborative approach to flood responses has a clear focus on gauging the biophysical impacts of floods. There are specific structures and hierarchies in the Vietnamese flood response that confine and standardise initiatives along the more technocratic dimensions of the mainline ministries. A technocratic approach to disaster management prioritises flood responses in the form of investments in dikes, higher bridges, water reservoirs, and more resilient physical infrastructure. As such, flooding is understood as a geophysical problem that can best be managed through technical control (Fortier 2010; Garschagen 2016; Rubin 2015; Nguyen et al. 2021) . Captured collaboration has some advantages in terms of reaction time, penetration, and implementation. Surely, the reach and the swiftness with which flood relief could be distributed is impressive for a lower-middle-income country such as Vietnam. Remarkably, 93% of flood victims surveyed reported receiving support from the local authorities in the wake of a major flood (Survey 2011). The collaborative nature of this type of governance, however, is limited. Our empirical evidence indicated little deliberative collaboration in flood responses and a reliance on technocratic solutions. Both the surveys and interviews suggested that there was little encouragement for local people to collaborate spontaneously to carry out environmental adaptation measures, such as dike construction, pooling of resources, protective forest belts, joint house building, and the like. The governance structures tended to rely on command-and-control systems with limited civil society participation. The vast majority (more than 60%) of the respondents in our 2013 survey could mention technocratic initiatives, such as the provision of better roads, drainage, dikes/dams, and housing. Only a small minority (12%) mentioned flood management initiatives that could be interpreted as rooted in collaborative arrangements at the local level, such as coordinating local meetings, sharing best practices, encouraging village self-help groups, and participating in community rescue plans (Survey 2013). The qualitative interviews also pointed to state-led preparatory activities that were mostly focused on physical infrastructure, such as preparing equipment, stacking sandbags, strengthening dikes and houses, constructing safehouses, relocating people, and deepening the waterways. The point here is not to denounce the importance of information dissemination or investments in physical infrastructure as part of flood response. Rather, it is to uncover the asymmetric collaborative nature of the flooding response where captured collaboration prioritises physical infrastructure at the expense of civil society involvement at the local level. The result was environmental management that was mainly technocratic and focused on climate proofing through the construction of dikes, higher bridges, water reservoirs, resilient infrastructure, and so on. In terms of socio-economic support, short-term relief in the form of rice, noodles, drinking water, clothes, and a little cash was by far the predominant type of support that flood victims received from the state. Longer-term aid to bolster resilience to the recurrent floods received less attention. Only 1% of the surveyed flood victims reported receiving loans and compensation that would allow longer-term investments in floodresilient housing or infrastructure (Survey 2013). In part, this reflects that collaborative arrangements are not built on a legislative framework of citizens' rights but rather captured by the state (Rubin 2013) . This impedes on citizens' abilities to participate in collaborative governance processes and provides the CPV with much discretion in deciding when and what type of support should be provided in situations of major floods. As previously mentioned, Vietnam has undergone several decentralisation reforms during the last decades (Garschagen 2016; Anh 2016). Despite these reforms, central agencies continue to retain control over the process of decision-making in local organs. Rather than fiscal devolution, decentralisation is used to bring governmental agencies as close as possible to the villages in order to consolidate their organisational base. The central government largely continues to set tax rates, determine tax bases, and interfere in subnational spending programmes. One of the chief means of rural consolidation and control is the Fatherland Front, an umbrella group encompassing the many different CPV mass associations, including the Farmers' Union, Women's Union, the War Veterans' Association, and the Youth Unions. In this way, the CPV stretches downwards through the government tiers into most aspects of rural organisation. In that light, decentralisation in Vietnam can be interpreted as being more of an exercise of power than a devolution of power. Both cooperation with and pressure from international donors have had a considerable impact on Vietnam's policy processes in the areas of environmental management and public participation. Since 2000, a comprehensive range of environmental and climate strategies and policy documents has been built with the assistance from bilateral donors, the UN, the World Bank, and the ADB, yet without comparable outcomes (Ortmann 2017; Schirmbeck 2018) . Similarly, a Vietnam Water Resource Strategy Until 2030 was approved in 2020. However, pressures from international donors and various international regimes with respect to collaborative management forms have particularly focussed on protected area forestry ("special use forests"). Specifically, REDD+ and the Forest Carbon Partnership Facility have pushed a shift from state-controlled to community-based approaches. Research on the results remain inconclusive: the success of collaborative management may depend on the further devolution of power to district governments to facilitate horizontal networked collaboration with local communities as much as on a legal recognition of communities as independent from the state apparatus (KimDung et al. 2016; Bruun 2020b) . As compared with collaborative forest management, which mostly concerns degraded, low value forest areas, collaborative water management involves stakes of a much higher order and entails far higher complexity. With key inputs to national power generation and agricultural production, water management presents an entirely different set of policy challenges, which hydropower construction and its many unintended consequences only have exacerbated. As confirmed by fieldwork, rivers are considered common property and the historical basis for Vietnam's irrigationbased civilisation (Le 2015: 17) . Only to be expected, largescale, expert-driven hydropower projects implemented in a top-down bureaucratic fashion with adverse impacts on the environment and society have ignited public protests across Vietnam, both in the resettled upland minority communities and in common lowland communities subjected to altered river flow patterns and water disasters. With no intention to let go of the CPV's constitutional right to lead the country, a range of legal documents since the Grass Roots Democracy Decree in 2008 and the new Constitution in 2013 have sought to console technocratic governance with increased public participation and collaboration. The Grass Roots Decree and other documents encourage communities and individuals to monitor and evaluate investments, contractors, and project construction, and thus, in principle contribute to building collaborative governance. Similarly, the 2013 Law on Water Resources states that all agencies, organisations, and individuals can use water resources on the basis of the equality of rights and obligations (Le 2015: 17) and the 2013 Land Law allows both individuals and NGOs to supervise land management (Ty 2015: 55) . However, as reflected in our surveys and interviews, despite a comprehensive legal framework to support the construction of a law-based state and to permit the public to hold authorities to account, both information dissemination and community consultations have faltered. Foreign donor and domestic reform pressure have put the party-state in a quandary. Ethnic minority deprivation after resettlement and multiple hydropower-enhanced flooding events have started off a series of criticisms from a broad range of actors, including many calls for granting a crucial role to civil society organisations as a means to improve conditions for collaboration, such as to raise the standing of local communities when in competition with large state enterprises for irrigation water or when subject to land acquisitions (e.g., Dong 2016) . Civil society organisations are underdeveloped and struggle with tough political restrictions (The Asia Foundation 2012; Gerard 2014), to the effect that they have small footholds in common Vietnamese villages. For instance, fieldwork revealed that CPV organs have a monopoly on the distribution of emergency aid to the effect that NGOs and private actors might only deliver aid to local governments. Particularly with a focus on assisting communities in water management issues and collaboration with authorities, new specialised NGOs and Internet forums have sprung up, such as the Ethnic Minorities Working Group, the Climate Change Working Group, and the Vietnam Rivers Network; some see a new political space for NGOs in assisting resettled communities through Payments for Ecological Services programmes (TY 2015) . The party's refusal to permit political liberalisation has contributed to increasingly confrontational calls for transparency and accountability; many observers see environmental issues as a major threat to regime legitimacy (e.g., Schirmbeck 2018) . However, an opposing trend to increased environmental organising may be attributed to authoritarian self-intensification in the face of combined socio-environmental emergencies (Bruun 2020a) . As noted in the theoretical section above, the more there is at stake in environmental conditions that may foster social upheaval and destabilisation, the harsher and more exclusively authoritarian the means of government may become (Ahlers and Shen 2018: 301) . Simultaneously with broadened formal access to collaboration, authoritarian legislation, rigid enforcement, media and Internet suppression, and restrictions on NGOs have increased. As a consequence, Vietnam now ranks on par with China and Iran in terms of its level of press freedom and draws heavy criticism from human rights organisations over the persecution and imprisonment of government critics, democratic and environmental activists, religious minority representatives, Internet journalists, independent writers, and artists (e.g., HRW 2020). With formal bans on criticising government policy from 2013 and 2016, and a cybersecurity law criminalising online criticism of government since 2019, critics and non-state actors operate within narrow bounds. Authoritarian intensification is also evident in flood responses, though with less critical and deliberative discussions. One of the strongest dynamics of authoritarian intensification can be seen at the local level of the flood responses with the involvement of local agencies and community associations. The Ordinance on Flood and Storm Control in 1993 stipulated that a Steering Committee for Flood and Storm Control shall be established at all levels throughout the country, from central to provincial, district and commune levels. In 2015, the Central Steering Committee for Natural Disaster Prevention and Control was established, headed by the Minister of Agriculture and Rural Development. The Committee is a coordinating and mobilising entity with the explicit aim of bringing together all resources from ministries, local governmental agencies, organisations, and individuals to deal with consequences from natural disasters (CPV 2015) . In 2021, a new National Strategy on Natural Disaster Prevention and Control through 2030 was approved by the CPV. The strategy is based on several specific outcome targets (halving the fatalities caused by flash floods, for example) and having 100% of "government agencies at all levels, organisations, and households to fully receive information and understand natural disaster prevention skills" (CPV 2021). These regulations, often pushed by international organisations, are indicative of some form of collaborative governance. However, as previously mentioned, most local associations are essentially political mass organisations. The state-led mass organisations have a near monopoly on rural organisation, and any voluntarism in flood management was channelled through these mass organisations. CPV exercises strict control over new civil society movements by demanding that they register under the existing state-led mass organisations. The term "civil society" is nowhere to be found in official documents or public discourse, and there are both formal and practical barriers to bottom-up organisation in disaster responses (Bruun and Olwig 2015; Wischermann et al. 2018; Bruun 2020b) . Garschagen (2016) argues that decentralisation codifications are often adopted only on paper to please the international donor community. Rather, governance continues to exhibit "persisting forces of centralised power and hierarchical, even authoritarian, decision-making procedures" (Garschagen 2016: 49) . This has also been referred to as intraorganisation authoritarianism and is characterised by severe restrictions on the autonomy and self-determination of the agencies and organisations involved in collaborative governance (Wischermann et al. 2018) . In our survey, 95% of respondents answered that members of their household were involved in formal associations (Survey 2011). Only 6% reported being members of organisations other than the major mass organisations tied to the CPV (Survey 2011). Throughout the 2010s, various formal mass associations have been increasingly involved in flood management initiatives, including forming rescue teams, running training exercises, and performing crisis communication in the wake of major floods. These collaborative governance initiatives were carried out in close collaboration with the local commune with the primary objective of effectively implementing CPV policies. Our interviews indicated that getting people to follow government policies was often considered one of the core mandates of the associations, emphasising their propaganda role. While local authorities were indeed actively preparing for floods, they did so as subnational implementation units and not as independent actors in a deliberative collaborative process. Some respondents referred to "making announcements" as the single most important activity for the community associations. Only 2% of our respondents emphasised deliberative collaboration in the form of experience/informationsharing as an activity in collaboration with the government. An overwhelming majority (more than 90%) described local meetings that could best be characterised as forums for state policy announcements and information dissemination. People were, in their own words, "reminded", "informed", "encouraged", and physically "moved". Cumbersome administrative processes hampered any collaborative deliberations. For instance, suggestions from the public needed to be presented in writing and passed through several bureaucratic levels. The broad-based collaboration with the involvement of many local mass organisations was mainly a mechanism for control and implementation rather than actual deliberation of policy options. We refer to this as captured collaborative governance. Further, research suggests that authoritarian intensification crowded out reliance on civil society resources (Rubin 2015) . The majority of the households surveyed indicated a dependence on the state and close family relations during floods. Few emphasised village or civil society as important sources of flood resilience (Survey 2013). As an example, 73% of the respondents highlighted relief from the state as the most important source of flood resilience. Only 4% of the respondents identified neighbours as an important (not even the most important) source for disaster protection. More than 80% of respondents rated aid from relatives and neighbours as irrelevant during disasters (Survey 2013). Instead, the CPV was present in the lower governance tiers and dominated most aspects of the flood response. While collaboration between stakeholders was considerable, the autonomy of the stakeholders, even for local "civil" associations, was undermined by the CPV. Thus, local-level involvement and community participation can be seen as an expression of authoritarian intensification where CPV power permeates communities down to the lowest level. Thus, collaboration builds on asymmetric power relations, with state control of the financial, institutional, and political resources of local stakeholders. This hints at very limited deliberative collaborative governance at the grassroots level. Instead, the collaborative process appears to be captured by the CPV. Using AE as the analytical entry-point, the analysis above identified governance dynamics that we suggested could be characterised by the term captured collaborative governance. At first glance, the term might be considered an oxymoron: If the state controls key governance processes, how can these processes be described as collaborative? Captured collaboration deviates from mainstream understandings focussed on deliberation, policy debate and consensus seeking (e.g. Ansell and Gash 2008; Doberstein 2016; Sorensen and Torfing 2009, 2021) , but may still retain analytical value as a separate governance category. It does not preclude widespread collaboration and interdependence internally and hierarchically (between different sections of the government) as well as externally and horizontally (including mass associations and local organisations). Essentially, however, it determines the nature of these collaborations and interactions, because it defies common distinctions between state and society by referencing the leading role of the party in both. Thus, captured collaboration might spur a high degree of interaction within local communities, including mass associations. However, the nature of this form of collaboration at the same time inhibits spontaneous organising and community-led participation, as much as it retards synergistic cooperation across all actors and stakeholders. This makes captured collaboration distinct from mainstream deliberative collaborative governance, as well as from centralised and hierarchical governance systems that primarily rely on exercising power through vertical structures. While it is generally acknowledged that collaborative dynamics might often occur within the "shadow of hierarchy" (see Whitehead 2003; Zhou and Dai 2021) , with captured collaboration, the state and local elites are not lurking in the shadows but are openly exercising direct control over every actor in the collaboration from local governments to "civil society" organisations. As an aspect of authoritarian power sharing and illustrated in multiple cases above, captured collaboration may easily translate into elite capture of vital resources. Figure 2 illustrates the difference between these governance systems. In centralised governance systems, stakeholders are primarily subject to state control through hierarchical structures that determine the direction and rules for the interaction. Stakeholders largely retain their autonomy and self-determination, but the playing field is formed by the state. Centralised systems are usually associated with authoritarian regimes but are also widespread in pluralistic regimes during crisis management (states of emergency), such as in the initial phases of the recent COVID-19 crisis. As noted by Zhou and Dai (2021: 15) , "hierarchical interventions play important roles in collaborative decisions in many countries regardless of the political system that a country adopts." In the conventional deliberative collaborative governance system, non-state stakeholders are included as equal partners in a horizontal decision-making process (but without the same influence over decisions). This horizontal network is characterised by dialectic interactions between the key actors, illustrated by the circular network of nodes. These collaborative interactions will often involve "trust-based knowledge sharing, joint exploration of problems and solutions, compromises and agreements about joint action, critical scrutiny of the impact of new governance initiatives, and responsive discussions of problems and failures, and thus the need for future adjustments and revisions." (Torfing and Sorensen 2021). The captured collaborative governance system mimics the horizontal system of deliberative governance but restricts the autonomy and self-determination of the participating stakeholders from within. This collaborative dynamic shares similarities with the concept of meta-governance, which focuses explicitly on the practices and procedures that secure governmental influence, command, and control within governance regimes (Mu et al. 2019 ). There are, however, important differences between meta-governance and captured collaboration. Meta-governance can frequently be traced in democracies (indeed one of the earliest proponents of the concept used it to study governance in the UK; see Whitehead 2003) , and the research focuses on the dialectical interactions between self-organisational networks and the hierarchical structures within which they are embedded (Whitehead 2003) . Thus, both horizontal and vertical metagovernance often presupposes the existence of a fragmented political system based on a high degree of autonomy for a plurality of self-governing networks (Sorensen 2006) . Captured collaborative governance, by contrast, highlights the undermining of self-organisation and autonomy in the first place. An authoritarian regime, such as the Vietnamese, can be considered a necessary but not sufficient condition for captured collaborative governance dynamics. Thus, the concept bears resemblance to Chen et al.'s (2021) description of a Regionally Decentralised Authoritarian (RDA) regime in their analysis of the Chinese intergovernmental collaboration of air pollution control. The authors find the RDA regime generally more effective in addressing complexities because the higher level of central intervention allows the state to retain substantial control over its local agents, while shaping collaborations among local governments and steering their behaviour. However, in contrast to the RDA regime's lack of attention to adverse consequences, this study suggests that the complexities of water management further authoritarian intensification and in turn captures collaborative dynamics. This paper analysed the interface between two modes of governance, collaborative governance and AE, as they emerge in the field of water management in Vietnam. We identified two inherently adverse collaborative features, technocratisation and authoritarian intensification, and analysed how the resulting dynamics manifest in hydropower construction and flood response. We introduced the term captured collaborative governance as a distinct type of collaboration that blurs the common distinction between state and society due to extensive party control over key actors across both realms. This type of collaboration inhibits the deliberative features that are normally associated with collaborative governance such as spontaneous organising, community-led participation, trust-based knowledge sharing and critical feedback loops. However, our analysis also revealed pressures for more deliberative governance. In the case of hydropower construction, the mounting criticism from international and national NGOs as well as the affected communities provided an interesting example of how non-state actors using social media could carve out new platforms for policy debate. However, formal collaborative structures remain locked while more deliberative arrangements are still located at the fringes of governance processes. It is an open question whether these novel participation dynamics from "grassroots", "communities" and "social organisations" in hydropower and flood policies can be channelled into broader governance deliberation processes that may eventually spur a genuine paradigm shift. Hitherto, the captured collaborative governance under oneparty dominance has been a highly resilient management system. SEA of the Quang Nam Province hydropower plan for the Vu Gia-Thu Bon River Basin. ICEM Final SEA Rep Breathe easy? Local nuances of authoritarian environmentalism in China's battle against air pollution Vietnam: decentralization amidst fragmentation Collaborative governance in theory and practice Understanding inclusion in collaborative governance: a mixed methods approach Collaborative governance: an international bilbiometric study of the last decade The coming of authoritarian environmentalism Sending the right bill to the right people: climate change, environmental degradation, and social vulnerabilities in central Vietnam Environmental protection in the hands of the state: authoritarian environmentalism and popular perceptions in Vietnam Lost in authoritarian development: have global climate deals and the aid community sacrificed the Vietnamese highland population? Is local community the answer?: The role of "Local Knowledge" and "Community" for disaster prevention and climate adaptation in central Vietnam Paradoxes in adaptation: economic growth and socio-economic differentiation. A case study of Mid-Central Vietnam Local interest group activity and environmental degradation in authoritarian regimes The new, green, urbanization in China: between authoritarian environmentalism and decentralization Which works better? Comparing the environmental outcomes of different forms of intergovernmental collaboration in China's air pollution control en/Pages/tasks-of-the-centralsteering-committee-for-natural-disaster-prevention-and-control. aspx CPV (2021) Decision No. 379/QD-TTg 2021 the National Strategy on natural disaster prevention through 2030 Environmental collaborative governance degree of government, corporation, and public Democratic regression in comparative perspective: scope methods and causes Designing collaborative governance decisionmaking in search of a 'collaborative advantage Concepts for modernization of water management in the Bac Hung Hai irrigation system What makes for good and bad neighbours? An emerging research agenda in the study of Chinese environmental politics Why resource-poor dictators allow freer media: a theory and evidence from panel data Environmental Management An integrative framework for collaborative governance Chapter 1-collaborative governance and collaborative governance regimes Taking a climate chance: a procedural critique of Vietnam's climate change strategy Decentralizing urban disaster risk management in a centralized system? Agendas, actors and contentions in Vietnam ASEAN and civil society activities in 'created spaces': the limits of liberty Authoritarian environmentalism and China's response to climate change Social acceptability of large infrastructure projects in Vietnam HRW (Human Rights Watch) (2020) Vietnam: crackdown on peaceful dissent intensifies Collaborative environmental governance, inter-agency cooperation and local water sustainability in China The Vietnamese state and administrative co-management of nature reserves Command without control: the case of China's environmental target system Tightening the grip: environmental governance under Xi Jinping Vietnam's hydropower policy reform Demand and supply factors affecting participation in hydropower management in the Song Tranh 2 Bureaucracies count: environmental governance through goal-setting and mandate-making in contemporary China China goes green: coercive environmentalism for a troubled planet Ecological civilization, authoritarian environmentalism, and the eco-politics of extractive governance in China Authoritarian environmentalism, just transition, and the tension between environmental protection and social justice in China's forestry reform Assessing and explaining interagency collaboration performance: a comparative case study of local governments in China Grassroots environmental activism in an authoritarian context: the Trees Movement in Vietnam Understanding and assessing flood risk in Vietnam: current status, persisting gaps, and future directions Environmental governance in Vietnam: institutional reforms and failures Collaborative crisis management: a plausibility probe of core assumptions Impediments to climate-induced disaster management: evidence from Quang Nam Province, Central Vietnam Linking Social Capital": evidence from four Vietnamese provinces Case selection techniques in case study research: a menu of qualitative and quantitative options The climate change challenge and the failure of democracy New arenas of engagement at the water governance-climate finance nexus? An analysis of the boom and bust of hydropower CDM projects in Vietnam Metagovernance: the changing role of politicians in processes of democratic governance Making governance networks effective and democratic through metagovernance Accountable government through collaborative governance? Cambridge The Asia Foundation (2012) Civil Society in Vietnam: a comparative study of civil society organisations in Hanoi Hydropower dam development and local communities' livelihoods: a case of Yali Hydropower project Dilemmas of hydropower development in Vietnam: between dam-induced displacement and sustainable development How does collaborative governance evolve? Insights from a medium-n case comparison VietnamNet (2020) Floods, climate change and hydropower In the shadow of hierarchy': meta-governance, policy reform and urban regeneration in the West Midlands Do associations support authoritarian rule? Evidence from Algeria, Mozambique, and Vietnam Within the shadow of hierarchy: the role of hierarchical interventions in environmental collaborative governance Acknowledgements Support was received from Independent Research Council Denmark, Project no. 9038-00138B. Conflict of Interest The authors declare no competing interests.Publisher's note Springer Nature remains neutral with regard tojurisdictional claims in published maps and institutional affiliations.