Barriers and enablers to climate change adaptation in hierarchical governance systems: the case of Vietnam

ABSTRACT

Governments fulfil important roles in increasing the adaptive capacity of local communities to respond to climate change impacts, particularly in developing countries. Existing studies on how governments enable and constrain the ways in which local level communities learn and build their adaptive capacity, however, generally adopt network or market-oriented types of governance. However, the most vulnerable regions to climate change impact in the world are generally governed through hierarchical policy systems. This research aims to understand how the hierarchical policy system in Vietnam creates enables and/or constrains the policy capacity of policy actors to contribute to effective climate change adaptation. We conducted interviews (n = 26) with key actors at multiple levels of government. Our findings show the importance of clear legal institutions, available financing for implementing policies, and the training of governmental staff, particularly at district and commune levels where the policy capacities are generally too low to deal with climate change impacts. We conclude that any efforts to support local actors (i.e. smallholder farmers) should include investments in policy capacity to ensure uptake and upscaling of adaptation actions more broadly.

KEYWORDS:

• Social learning
• hierarchical state
• constraints
• enablers
• adaptive capacity
• climate change adaptation

1. Introduction

Communities across the globe are adapting to the impacts of climate change. Particularly in developing countries, there is an important role ascribed to governments to create an enabling policy system that helps the most vulnerable social groups and regions to start adapting (Bardsley, 2015; Ensor & Harvey, 2015). Many examples of governments building this capacity have been recorded, including increasing knowledge exchange, creating training and educational programmes, empowering the poorest and most vulnerable groups, enhancing social networks, and connecting administrative levels and scales (Armitage & Plummer, 2010). However, planning for climate change adaptation (CCA) in developing countries is considered highly complex as this takes place in a setting of multiple socio-economic and political challenges including poverty, gender inequality, illiteracy, corruption, rapid urbanization, food insecurity, and extensive extraction of natural resources for economic development (Mohabbat & Shahriar, 2015). Moreover, over time several discourses have emerged in most developing countries to reduce social vulnerability, including through development aid, disaster risk reduction, sustainable development, and more recently CCA efforts, creating a complex and fragmented policy landscape (O'brien, Eriksen, Nygaard, & Schjolden, 2007; Vij et al., 2017). Although the role of the state has been questioned in these contexts for not having the policy capacity to ensure timely implementation – hence the mushrooming of NGO initiatives and internationally funded projects to help start local communities to adapt – there is still a potentially important role for the state. After all, governments have unique constitutional tasks and responsibilities as well as vast policy instruments at their disposal – most explicitly the rule of law – that make them indispensable for ensuring timely CCA across scales (Araos, Ford, Berrang-Ford, Biesbroek, & Moser, 2017; Jordan et al., 2015). Moreover, the state is (un)willingly creating constraining conditions that affect the capacity to adapt to climate change, such as creating conflicting tools and guidelines or under-reacting to climatic changes (Peters, Jordan, & Tosun, 2017). Paradoxically, the state is also instrumental in removing such barriers (Biesbroek, Klostermann, Termeer, & Kabat, 2013), for example to formulate vertical coordination mechanisms to create a social learning environment that allows local stakeholders to build adaptive capacity. Amundsen, Berglund, and Westskog (2010), for instance, find that without national authorities giving clear signals through designing and facilitating national adaptation policies, lower administrative levels will find it more difficult to develop and implement effective adaptation policies.

Although several studies exist that investigate the capacity of government to enable adaptation actions, they have focused on network-oriented or market-oriented policy systems that exist in predominantly in high-income countries (Armitage, Berkes, Dale, Kocho-Schellenberg, & Patton, 2011; Nilsson, Gerger Swartling, & Eckerberg, 2012). However, the most vulnerable regions in the world are often governed through governance models that are more hierarchical in nature. Studies on adaptation in developing countries often suggest improvements based on network- and market-orientated principles that do not necessarily fit with the existing hierarchical policy system context.

This research therefore aims to understand how in a hierarchical policy system the policy capacity at different levels of government enable or constrain policy actors to help local actors to adapt to climate change impacts. To address this question, we focus on Vietnam – a highly vulnerable country with a hierarchical autocratic policy system. More specifically, we focus on the agricultural sector which is an important driver for the Vietnamese economy but highly vulnerable to current and projected climate impacts. Over the past decades, the Vietnamese government has already implemented several CCA activities, policies, and strategies (Hoang et al., 2014; Knaepen, 2014) in a formalized and top-down manner to ensure timely actions (Rubin, 2014).

2. Examining policy capacity in hierarchical multilevel government settings

We characterize CCA as a societal issue that cross-cuts different spatial, temporal, and administrative levels and scales. Moreover, adaptation is understood here as those intentional actions to reduce social vulnerability taking into consideration the long-term climate impacts (Dupuis & Biesbroek, 2013; O'brien et al., 2007). In its broadest sense, adaptation requires coordinated responses across multiple levels of government to prevent fragmented and maladaptive actions (Araos et al., 2017), hence referring here to multilevel government as a system of continuous negotiation of nested governments at several territorial tiers: national, regional, and local (Hooghe, 1996). Central to the notion of multilevel government is the recognition that there are different ways in which power diffuses between the actors and institutions and that different levels may have different tasks and responsibilities when it concerns adaptation (Araos et al., 2017). For example, in most contexts, local governments are responsible for delivering basic services to their citizenry whereas the national government is responsible for creating an enabling environment by making basic resources available, for example. Studies find that in governing CCA, cross-level and cross-scale imbalances and contradictions are inevitable (Bauer & Steurer, 2014; Fröhlich & Knieling, 2013). Multilevel government studies therefore seek to identify the level at which the problem is manifested and the level at which the climate change impacts are being managed (Termeer, Dewulf, & Lieshout, 2010), which are not necessarily in sync.

In several developing countries, and particularly in Vietnam, the policy systems are characterized as ‘bureaucratic hierarchies’, ‘administrative states’, or a ‘mono-centric modes of governance’ (Schreurs, 2010). In these systems, the focus is on the centre of political power and authority – the state – that sets the agenda of societal problems, decides upon policy goals and means, and implements its policies at lower administrative levels in a top-down manner (Painter & Peters, 2010). They are characterized by the specialization of functions, objective qualifications, civil servants who follow a fixed set of rules, and a hierarchy of authority. Steering and controlling are key concepts within this system where the state does the steering rather than the rowing (Osborne & Gaebler, 1992). The top-down institutional structures strongly influence the patterns of interaction and coordination within and between different jurisdictional levels and scales (McNeeley, 2012; Rubin, 2014).

Like any other system, hierarchical systems are operationalized by a set of shared rules, norms, and practices that shape the implementation preferences of bureaucratic actors (Painter & Peters, 2010). These policy system characteristics are difficult to change even in the face of major challenges such as climate change as the political interest in retaining the status quo is often much stronger than efforts of changing existing institutions. Although many hierarchical systems are changing, for example due to globalization, such change usually goes very slow. Several studies show that the ability of the state to respond to climate change is both enabled and constrained by the characteristics of the policy system (Craft & Howlett, 2012; Wellstead & Stedman, 2015)

2.1. Understanding policy capacity in hierarchical system settings

Many authors have listed different types of possible constraints to adaptation, including cognitive, political, social, and institutional constraints (Biesbroek et al., 2013; Oberlack, 2016; Walker, Adger, & Russel, 2015). Especially in low-income countries, socio-economic factors, resource constraints, societal hierarchies, and psychological factors have been widely identified as major constraints (Bayard, Jolly, & Shannon, 2007; Deressa, Hassan, Ringler, Alemu, & Yesuf, 2009). Other frequently reported barriers include unfamiliarity of the local people with available data on climate change, lack of local expertise, lack of a clear role for local governments, and lack of focus on adaptation at the national level leading to a lack of attention to CCA at the local level (Biesbroek et al., 2013; Hughes, Gleeson, Legge, & Lin, 2015).

We adopt the policy capacity framework (Wu, Howlett, & Ramesh, 2017) to analyse more deeply the policy capacity in Vietnam and how this enables and constrains adaptation actions. Policy capacity is understood as the ability of the government to make intelligent collective decisions, to mobilize necessary resources, to produce robust evidence-based policy, to weave together different organizations and interests, to coordinate policy-making across and external to government, and to implement as well as formulate meaningful policy (Hughes et al., 2015; Oliphant & Howlett, 2010; Wellstead, Stedman, & Howlett, 2011). In the context of this paper, we operationalize policy capacity as: (1) the institutional characteristics of the policy and governance system; (2) the resources required and available in order to design and implement CCA, and (3) the policy analytic capacity of policy actors involved in CCA (Craft, Howlett, Crawford, & McNutt, 2013). We briefly discuss the three main elements of this policy capacity framework and how they enable or constrain adaptation actions across levels and scales.

The first element of policy capacity is the institutional characteristics of the system. Institutions are macro-level arrangements that are contrived of formal and informal rules, norms, cultures, and beliefs that influence the ways in which government is structured, how power and authority is being distributed within government and between government and society, and how the state interacts with society. Institutional arrangements have specific strengths (e.g. coercive power, clear division of tasks, and responsibilities) and weaknesses (e.g. procedural restriction, fragmentation) that influence the ways in which governments (can) respond to climate change impacts (Doelle, Henschel, Smith, Tollefson, & Wellstead, 2012). Institutional characteristics are known to both empower and constrain societal and governmental actors. CCA studies in southeast Asia, for example, show that institutional fragmentation is considered a critical reason for governments not being able to respond to climate change impacts as there are many institutional gaps, coordination issues, bureaucratic separatism, conflicting responsibilities, and objectives, as Lebel, Manuta, and Garden (2011) show for Thailand. Studies in Vietnam show that formal institutions have constrained collaborative governance processes by obstructing the participation of actors across levels and areas of expertise and weakening the roles and functions of local authorities in adapting to climate change (Ho, Cottrell, Valentine, & Woodley, 2012).

The second element of policy capacity is governing resources. The government has specific governing resources at their disposal such as financial, knowledge, legal, organizational, political, social, or human-related resources (Koch, Vogel, & Patel, 2007). Resources are the means through which governments can achieve particular policy goals (Henstra, 2016). As such, resources are critical to measure policy capacity and are considered crucial inputs for designing, recommending, and implementing policy decisions (Wellstead & Stedman, 2015). Several studies show that lacking of governing resources will limit local actions to CCA (Brown, Nkem, Sonwa, & Bele, 2010). Arguably, the most frequently reported constraint in low-income countries is the lack of governmental funding for designing and implementing adaptation measures as well as for hiring appropriate personal resources (Antwi-Agyei, Dougill, & Stringer, 2015). Moreover, limited access to domestic and international financial resources was found to negatively influence analysis and implementation of adaptation, for instance in Ghana (Antwi-Agyei et al., 2015) and Bangladesh (Ahammad, 2011).

The third element for studying policy capacity is related to so-called policy analytical capacity of civil servants. Policy analytical capacity refers to the capacities of actors, policy-makers, and policy workers such as their competences, skills, attitudes, and knowledge (Howlett, 2015; Lodge & Wegrich, 2014). Howlett (2015) argues that the capacities of individual policy actors are important as they determine, for example, the amount of research a government can conduct or access, the efficiency and effectiveness of policy implementation, the possibilities to create specific trainings for public and societal actors, and to provide recommendations to local people. These determinants are crucial as they allow governments to communicate about climate change policy and, vice versa, to incorporate insights from society into governmental decision-making (Howlett, 2009; Tiernan, 2011). However, several studies show that low skills and abilities of public sector actors seriously affect overall policy analytical capacity. The lack of knowledge and skills of governmental actors about climate change impacts and adaptation, for example, has proven to be a major constraint in Bangladesh to integrate adaptation (Ahammad, 2011). Low degrees of analytic capacity of policy actors explain the variation in how, within the same country or region, governmental responses of governments can vary greatly (Wellstead et al., 2011).

3. CCA in Vietnam

We adopt the policy capacity framework to understand the exiting CCA policy efforts in Vietnam. Vietnam frequently is characterized as a strong developmental state (Gainsborough, 2009) and autocratic one-party system where the Communist Party of Vietnam holds the monopoly of the political process and is the absolute leader of society (Desbarats, 1987; Minh Chau, 1997). The Party implements its leadership through the state via the principle of democratic centralism which means unconditional implementation of the decisions taken at higher levels by the lower levels. However, recently there have been many socio-political reforms opening up the country for international organizations and non-government stakeholders (Painter, 2003) thereby increasing democratic principles and practices. In recent years, administrative reforms have changed the instruments of coordination and transformed diffusion of power (Garschagen, 2015). When it comes to CCA, the government is responsible for building, steering, and implementing all policies, measures, and strategies in a rather formalistic and technocratic way, although it depends on the approval of the Party (Clemens, Rijke, Pathirana, Evers, & Quan, 2016). Figure 1 presents the hierarchical multilevel system for CCA in Vietnam.

Figure 1. The Vietnamese government organization structure based on the mandates of agencies and main CCA policy in Vietnam (adapted from MONRE, 2008).

At the national level, different ministries collaborate on CCA and each ministry has the authority to make decisions that should be implemented at lower levels of government. The Ministry of Natural Resource and Environment (MONRE) is responsible for and coordinates all CCA-related activities, for example through the National Target Program to Respond to Climate Change (MONRE, 2008) and National Strategy on Climate Change (MONRE, 2011). When it comes to long-term CCA efforts in the agricultural sector, however, the Ministry of Agriculture and Rural Development (MARD) is the main ministry responsible. This ministry has developed several policies, most importantly the Action Plan Framework for Adaptation to Climate Change in the Agriculture and Rural Development for the period 2008–2020 (MARD, 2008), and a Climate Change Adaptation Vision to 2050 (MARD, 2011). In addition to MONRE and MARD, there are four main ministries with specific adaptation responsibilities: Ministry of Planning and Investment; Department of Public Finance of the Ministry of Finance; Department of Student Affairs of the Ministry of Education and Training; and the Ministry of Information and Communication. Although the policies have emerged as result of experienced climate change impacts that are highly diversified across Vietnam (Christoplos, Ngoan, Sen, Huong, & Lindegaard, 2017), the implementation of these policies is top-down and centralized. This means that the national level can force the Peoples’ Committee at local levels to implement certain CCA policies and measures. At the provincial and district levels, the institutions for CCA are structured the same as at the national level (see Figure 1). There is an important role for the Provincial Peoples’ Committee as this level is central in translating district (and sometimes commune) level requests, as well as converting proposals from sectoral departments and academic institutions into concrete project proposals that can be presented to international organizations. At the commune level, the lowest management level in the policy system, there are no institutions with an explicit responsibility for CCA issues. At the lower administrative levels (i.e. province, district, and commune), the Peoples’ Committees approve and sign all legal documents which are used for day-to-day management and governance activities. The coordination between national level and other levels takes place via two routes: (1) top-down steering from national level to the local level through decisions and (2) decrees and resolutions and feedback from local level to national level through annual meetings, workshops, and reports.

4. Methodology

For this research, we used a qualitative nested case study approach (Keessen et al., 2016) to gain an in-depth understanding of the policy capacity of the four different administrative levels in Vietnam (national, provincial, district, and commune levels). Vietnam is very vulnerable to climate change impacts (Bruun, 2012; Maplecroft, 2011), specifically agriculture and food security are under serious pressure. To focus our analysis, we analysed particularly the experiences of policy actors’ adaptation efforts in the agricultural sector. In Vietnam, agriculture is a major economic force, contributing to more than 16.32% of the GDP of the nation and providing employment for 42.4% of the working population (GOS, 2016). Of the 63 subnational provinces, we selected Thua Thien Hue province in the centre of Vietnam as it is considered one of the most vulnerable regions to extreme climate events (Beckman, 2010). The province is already recognizing the impacts of climate change and has started to invest in CCA. The third and fourth tiers of analysis are the Quang Dien district and Quang Loi commune, as they are thought to be one of the most climate vulnerable areas in Thua Thien Hue by the government with high poverty rate and farmers’ livelihood strongly depending on income from agricultural production and governmental support to adapt (TTH Provincial People Committee, 2014).

4.1. Data collection methods

We use interview and document analyses as main data sources. In the summer of 2016, we conducted 26 in-depth interviews with policy actors from different administrative levels: 4 at national level, 9 at provincial level, 5 at district level, and 4 at commune level. These were the policy actors most responsible for adaptation of the agricultural sector at the specific level. We also interviewed two households at farm level, one research organization and one non-government organization. Each interview lasted between 60 and 90 minutes during which interviewees were asked about specific themes: their current understanding and knowledge of climate change, their tasks, responsibilities and policy actions on adaptation, the other actors’ involved, horizontal and vertical coordination between each level and between sectors, and the three dimensions of the policy capacity framework. The interviewed policy actors were also asked to critically reflect on the key factors they perceived as being enabling and constraining the implementation of CCA at the local level. The interviews were recorded when allowed. In several instances, however, the interviewees did not agree to be recorded in which case detailed notes were taken by the two interviewers. The interviews were transcribed shortly after each interview. We contacted several interviewees for follow-up questions. In addition, secondary data were collected including legal documents and governmental reports from all the four levels for the time period 2008–2016. This data allowed us to prepare the interviews and include specific questions that could not be distilled from the literature.

4.2. Data analysis methods

The data were analysed through several steps. First, the content of documents was analysed by identifying the main events, policies and (political) decisions to reconstruct the adaptation policy development at each of the four levels under analysis (2008–2016). Second, the interview data were analysed using open coding to identify which enabling and constraining conditions were mentioned by the interviewees. We clustered and prioritized these using the three elements of policy capacity: institutions, resources, and individual policy actors. We used quotes to provide some examples from interviews clarify the findings. The quotes were selected that directly spoke to the key questions being asked in this study. The most frequently reported barriers were discussed and cross-checked with the secondary data. The findings of this analysis are reported in Section 5.

4.3. Limitations

This study has several methodological limitations. Firstly, the findings of this study were limited by the limited number of nested cases we studied. For analytical and pragmatic reasons, we focused on one province, district, and commune rather than including multiple cases at each level. This will have consequences for the upscaling and generalizability of our findings, as we discuss in the final sections. Second, interviewing government officials in hierarchical systems are known to be a challenge for several reasons, including trustworthiness of the findings (Rubin, 2014). This is particularly the case when it comes to politically sensitive issues. Whilst this could be a limitation in our study, we combined multiple interviews and document analysis to ensure the robustness of our findings. Finally, we mostly studied government officials, which is central to our analytical framework, but including more non-governmental actors could have enriched allowed us to understand how other actors perceive the policy capacity of governments.

5. Results

Our analysis shows several key conditions and processes across the different levels of government that were experienced as enabling and constraining We have clustered, analysed, and presented these through the three elements of policy capacity: institutions, resources, and individual policy actors.

5.1. Policy capacity: institutions

5.1.1. Key enabling institutional characteristics

First, interviewees at national and province level argued that the existence of the NTPRCC and NSCC is helpful as it clarifies roles, tasks, and responsibilities of policy and societal actors. The fact that adaptation is not the sole responsibility of one ministry but rather shared across multiple ministries is considered to be strength, as it ensures that mainstreaming adaptation across vulnerable sectors is ensured. Consequently, at the national level, several policy networks and collaborations have emerged to share information and encourage policy learning for implementing across different sectors. Second, interviewees argued that because adaptation at subnational levels is mainstreamed in annual and five year social-economic development plans (SEDPs) there is increasing institutionalized support for staff at province, district, and commune levels. This allows each level to gradually increase their governing resources and analytical capacity of their policy actors. Interviewees also reported that the hierarchical and structured approach ensured some degree of consistency and coherence between sectors (horizontal) and across all levels (vertical) in how to frame and address climate change impacts.

However, since adaptation is not fully institutionalized and mainstreamed across all levels, much of the adaptation taking place is still rather fragmented. The interviewees at the district and commune level did not consider this a key constraint, but rather argued that this creates plenty of opportunities for learning, participation, and experimentation by smallholder farmers.

At district level, there are not yet specific CCA policies for agricultural development and adaptation. However, we are already organizing trainings, building plot demonstrations, and conducting agricultural experiments in local communities. Mainstreaming of CCA in these activities increases the learning process and farmers’ capacity to implement adaptation measures to respond to climate change impacts. Advanced climate models and new agricultural techniques are made available to us by the scientific and technological sub-departments which have increased farmers’ knowledge and their capacity to adapt to climate change impacts. (Interviewee of DARD at district level)

5.1.2. Key constraining institutional characteristics

However, interviewees also identified a number of key institutional constraints. First of all, much of the institutionalization of adaptation is still at the early stages. Although a lot has been realized at the national and provincial level, these are not yet dispersed to district and commune levels which mean that there is no legal basis for lower levels of government to start acting and making adaptation a central policy issue. Although several policies and plans for CCA in the agricultural sector have been adopted at the national level, these too have not yet been adopted and integrated at the lower levels. Consequently, all interviewees at the lower levels argued that their roles are unclear and their tasks are very still limited. Thus, all decisions for building and implementing CCA strategies in the agricultural sector depend on the discretion of the Peoples’ Committee at provincial and district level as they control the financial distribution for each department.

Because of the unclear roles and the lacking of a legal mandate, the question of who is held accountable for adaptation remains unclear, particularly at the lower levels. As one interviewee at the district level aptly notes:

Although mainstreaming was mentioned in the provincial CCA policies, this strategy is not the main responsibility of the agricultural department. Given that these departments do not force us to engage with adaptation, we focus on the several other things that we [are legally required] to do. (Interviewee of DARD at district level)

The second main institutional constraint was related to the top-down steering of implementing adaptation. Interviewees noted that it leads to mismatches between what the central government tells the lower levels to do and what the lower levels want to do given their specific contexts and vulnerabilities. So far, all existing policies, plans, or guidelines are covering the whole country whereas large parts of the implementation are expected to be carried out at commune levels. This creates difficulties for lower levels if they want to adjust their adaptation measures to fit the appropriate climate change impacts. As one interviewee notes:

All CCA strategies and measures have to follow the higher level provisions, so it is very difficulties for our departments to adjust or change these measures and make them more suitable to our specific context and needs. Departments such as DONRE and DARD at provincial level often select the techniques used in projects; smallholder farmers or commune authority do not have any opportunity to choose projects and measures they are interested in. (Interviewee of DARD at district level)

Lower levels of government were also not involved in the design of these new adaptation policies, which means tensions are emerging between the new top-down designed adaptation policies, and the existing policies and practices at the lower levels. This is also affecting the policy actors (at lower levels) who mentioned that it is unclear how to coordinate across scales as there currently are no formal and informal mechanisms that allows them to ‘do-not-harm’ or to engage in positive coordination. During several projects, policy actors experienced several conflicts and contradictive actions that resulted in ineffective implementation of local CCA. This also reduced the possibility of reflecting on their experiences and to share lessons learned. The only mechanism through which feedback across levels is organized is through formal reporting, but the effectiveness of this mechanism in terms of policy orientated learning is questioned by local and provincial interviewees.

The hierarchical system has also resulted in ‘silofication’ where departments stick to their legally determined tasks and responsibilities and hardly share information or coordinate actions. As two interviewees note:

Our department works following the principle: who has capacity and can find financial resources will implement CCA strategies. There is no need to coordinate as long as you do your job and I’ll do mine. (Interviewees of DONRE provincial level)

As a result, the sharing and updating of information about CCA between the agricultural and other sectors have not yet occurred, particularly not at provincial and district level. The lack of horizontal and vertical coordination is hampering the policy-oriented learning process among departments and staff in agricultural sectors. This means that upscaling of other initiatives is complicated:

At province and district level there is recognition that small pilot projects, often led by NGOs, can result in valuable lessons for adaptation. But the lack of a mechanism for coordinating and sharing these lessons prevent up-scaled programming by the government. […] The lack of horizontal coordination has lead in some cases to overlap in pilot projects between different departments. (Interviewee of DAEC at district level)

5.2. Policy capacity: governing resources

5.2.1. Key enabling governing resources

Interviewees noted there are quite some financial resources available to start adapting to climate change. In recent years, the financial commitments by donor agencies to invest science, technology, and society have increased and climate change has become an important component of international funding. In total, the budget allocated for CCA is 0.1% GDP of Vietnam, with two-third of the funding from the Vietnamese government, and one-third by international donors (MPI, 2015). In 2013, 79% of the total climate finance budget is allocated to adaptation of the agricultural and rural development sector. Although this budget is still relatively small, these investments have grown rapidly in recent years (MPI, 2015). Interviewees argued that the large bureaucracy in Vietnam means that there are many civil servants working on agricultural issues. However, the number of people working on adaptation (rather than disaster risk reduction or sustainable development) is still relatively low, but slowly increasing.

5.2.2. Key constraining governing resources

Still policy actors considered the lack of financial resources a main governing resource constraint. Interviewees noted that there is a significant imbalance in the financial distribution across levels and sectors. The vast majority of financial resources is allocated to improving and building infrastructure at the national and provincial level, such as improving dykes. Investments in national infrastructure, for example, covered 93% of the total funding available for adaptation in the agricultural sector. In 2013, of the remaining budget, 4% was allocated for developing science and technology, 2% for enhancing governmental capacities, and a meagre 1% for developing adaptive capacity of communities and smallholder farmers (MPI, 2015). Interviewees reported that the unequal distribution and emphasis on technical measures prevented creating a social learning environment for smallholder farmers but also constrained civil servants for improving their policy analytical capacity.

In addition, whilst some funding for building adaptive capacity and stimulating social learning was available, this funding generally went to small scale pilot projects that were largely driven by donor funding. Currently, there is no systematic or continuous financial flow to ensure implementation of CCA in the long run. Interviewees reported that the existing policies and plans designed by the national government are far more ambitious than the available governing resources at the different levels. The disbalance in financial distribution is also reflected in the limited human resources for implementing CCA at local levels in terms of quantity and quality of staff. The limited number of staff that is knowledgeable about CCA is illustrated by one of the interviewees from DONRE:

We work to mainstream CCA in all sectors for the Thua Thien Hue province. However, our department has only three staff members who are knowledgeable about climate change and none of them are knowledgeable about adaptation of the agricultural sector. Similarly, there is only one staff member for CCA at district and commune level but he is not knowledgeable about agriculture. (Interviewee of DONRE at provincial level)

This can partly be attributed to institutional constraints as the tasks and responsibilities of the civil servants are not legally determined. This means it is up to the discretion of the Peoples’ Committee to determine if, when and how governing resources are invested for adaptation:

The Peoples’ Committee at each level manages the human resource and finance of these departments. Moreover, agricultural departments at lower levels also are managed by agricultural departments at higher levels where emphasis is placed on management and steering of specialized technical agricultural activities. During the implementation of CCA strategies, there are overlapping and inefficient uses of human and financial resources. (Interviewee of DARD at district level)

5.3. Policy capacity: individual policy actors

5.3.1. Key enablers of individual policy actors

Although financial resources are limited, interviewees referred to several training courses to increase the policy analytical capacity of individual actors’ working on climate change. These trainings are specifically designed for the implementation of the NTPRCC and NSCC. Most training follows the train-the-trainer principle, assuming that knowledge and expertise on CCA diffuses in the bureaucracy. Policy actors working on CCA at national level generally have high levels of education and are knowledgeable about CCA. There are also several opportunities to access updated climate change and CCA information in agriculture through national and international knowledge exchanges (e.g. conferences). Similarly, at the provincial level, several of the staff of DONRE and DARD are knowledgeable about climate change. These actors play a crucial role to enhance awareness about climate change for farming communities in their province.

5.3.2. Key constraints of individual policy actors

However, all interviewees agreed that policy analytical capacity at district and commune level is very limited particularly due to the limited knowledge and skills about CCA in the agricultural sector. This can be attributed to some extent to the type of Vietnamese bureaucracy where generalists generally prevail over specialists and where civil servants rotate throughout the bureaucracy frequently. This means the very few civil servants knowledgeable about adaptation also change jobs and since there are no institutional mechanisms in place, this critically influences the organizational expertise and lowers the motivation of civil servants to become very knowledgeable about CCA. This is illustrated by one interviewee:

I am trained in environmental inspections and I have worked for DONRE. From 2013 to 2014 I was moved to another department to support the provincial adaptation strategy development team. After that I returned to my old position. Since the beginning of 2016, however, DONRE has established the CCA department and I have worked as staff for this department for a while. However, I do not know whether I will continue to work here or if I will be moved to another department the future. (Interviewee of DONRE at provincial level)

A second key constraint is that knowledge and information about CCA is lacking and no substantive trainings are offered for individual policy actors at local levels. Even if they were offered, interviewees mentioned that they would attend, but only when it was mandatory. Consequently, the knowledge level at district and commune level remains relatively low with very limited comprehensive knowledge on the consequences of climate change. This again comes back to the institutional characteristics, as mentioned by one of the interviewees:

Some staff participated in training courses or CCA pilot projects. They have the opportunities to access knowledge and information but they did not share it to other staff because they do not consider this to be their responsibility. (Interviewees DAEC at provincial and district levels)

Interviewees also mentioned that the train-the-trainer principles did not work simply because the person trained did not have the time, nor made it a priority to share the knowledge to others in the department. Whilst CCA is considered an important task by all civil servants interviewed, it is clear that individual civil servants already have a lot of tasks that are mandatory. As CCA is currently an integrating rather than a legal mandatory task, it is not considered the primary task of civil servants working at district and community level. The motivation to learn and acquire new knowledge about CCA and to improve skills to support the social learning of smallholder farmers is not considered a priority.

In our community there is only one staff in charge of agricultural, extensional, and economic development for the whole commune. Therefore, I do not have enough time to learn or update new climate change and CCA knowledge. Although I work closely with smallholder farmers, I know very little about CCA in agriculture. I think that this is also true for the leader of community, staff of the agricultural cooperative, and other people in the community. (Interviewee agricultural staff at commune level)

6. Discussion

Climate change impacts in Vietnam are increasingly forcing the communist Party in Vietnam to encourage government to invest in CCA. Over the past years, the Vietnamese government has substantiated their portfolio of responses to manage the impacts of climate change by explicitly considering long-term and planned CCA in addition to other policy initiatives including disaster risk reduction, development aid, and sustainable development. Our findings show that like any government system, the hierarchical multilevel setting of Vietnam creates several enabling and constraining conditions that influence if, when, and how adaptation takes place at local levels. Although often critiqued in the literature, our findings suggest that the hierarchical policy system can offer some benefits, especially when it comes to vertical coordination and mainstreaming of CCA across departments (Araos et al., 2017). Vietnam is rapidly developing its CCA policies and measures, particularly for agriculture, and several successful examples of mainstreaming adaptation have been reported in the literature. The mainstreaming strategy seems well suited for such a governance context since it places CCA on a broader development pathway and can be implemented at all levels (Ward, Pauw, Van Buuren, & Marfai, 2013).

However, as this study also shows, this potential is not fully used and policy actors in the agricultural sector experienced various key constraints, including unclear institutions in terms of structure, roles, and accountability mechanisms, a lack of clarity with regards to the coordination within and across levels, a lack of human, financial, and legal governing resources, and limited individual policy actors’ understanding of climate change and its impacts. These are mostly manifested at district and community level, but the very nature of the hierarchical system means that many of the constraints are shared across the different levels of the government system. Moreover, the political power structure remains highly centralistic with centralized planning and management paradigms persist (Garschagen, 2016), leaving less room at local levels to manage the specific vulnerabilities the agricultural sector faces.

The nature and design of institutional arrangements determine both the manoeuvring space and policy capacity of civil servants working on adaptation in the agricultural sector. Dedicated institutions are therefore portrayed as playing a crucial role in CCA processes (Dovers & Hezri, 2010). Our study shows more generally that in Vietnam the institutionalized collaboration between national and local levels for implementing CCA currently is problematic. Especially, considering that there is no mechanism amongst the agencies across sectors and scales to coordinate adaptation efforts. Consequently, some responsibilities may be either taken up by more than one agency or overlooked altogether because each agency assumes another one is responsible. In hierarchical governance systems, this tends to lead to policy paralysis where no policy actor is feeling responsible to take action (Peters et al., 2017). Moreover, the lack of dedicated institutions increases the risk of maladaptation. Whilst this institutional void is sometimes filled by non-governmental organizations, this remains a patchy and ad-hoc answer to structural and systemic institutional problems that might hamper future progress on adaptation.

Our findings contribute to the ongoing debates about the pros and cons of institutional hierarchies (McNeeley, 2012) or complex and inflexible institutional frameworks (Craig, 2010) which are said to reduce adaptive capacity and exacerbate vulnerability to climate change (Ahammad, 2011). Studies show that the bureaucratic characteristics in hierarchical systems constrain the ability of managers and staff to develop plans and projects that cross agency jurisdictions (Rutherford, 2005; Thomas, 2003). Recent changes of institutions and organizational structures towards more flexible and robust systems allow for more efficient CCA governance, as can be observed in countries like the Netherlands (Ward et al., 2013), Finland (Juhola & Westerhoff, 2011), or Canada (Burch, 2010). Authors have argued that a flexible institutional arrangement and organizational structure is needed to deal with the complex contextual conditions and a combination of a top-down and bottom-up approach. Whilst we agree with the general line of reasoning, studies in public administration show that routinized practices and traditions are not easily changed, let alone transformed (Termeer, Dewulf, & Biesbroek, 2017). Although the policy system in Vietnam is clearly opening up, there is still a long way to go before fundamental changes in the institutional system can be expected. This means that governance arrangements specifically for adaptation in the agricultural sector need to be aligned to the administrative traditions of Vietnam to prevent institutional misfits that lead to inefficiencies and ineffective policy processes on adaptation.

One of the most important policy approaches that plays a crucial role in Vietnam is the mainstreaming strategy adopted in the so-called SEDPs. Reported benefits from the SEDPs at local levels include the increased coherence among policies and the reduced chances of duplication and policies that contradict each other (Rauken, Mydske, & Winsvold, 2015). These benefits of the mainstreaming are recognized by Vietnamese government (Knaepen, 2014), but how to mainstream effectively still needs to be determined (Runhaar, Wilk, Persson, Uittenbroek, & Wamsler, 2017). This requires reforms of the CCA organizational structure and associated institutions to help accelerate and deepen the development of the policy capacity needed to implement CCA measures for the agricultural sector at local levels.

The two other components of policy capacity – governing resources and policy analytical capacity – are directly influenced by the limitations of the institutional setting. We find that it is not necessarily the amount of money that is a constraint, but rather the way in which the money is allocated for developing and implementing CCA at local levels. Here, the political willingness of the Party is an important factor for the distribution of resources. In line with the work of Burch, we argue that finding more financial resources is not more important than facilitating the effective use of existing resources (Burch, 2010). In addition, the analytical capacity of civil servants plays decisive roles in performing key functions in policy processes (Brown et al., 2010; Sietz, Boschütz, & Klein, 2011). This capacity was found to be limited at the local levels as no clear institutional arrangements have been made to enable civil servants to build such capacity. Existing training programmes to adapt agricultural production processes to climate change have yet to allow for learning, transferring, and co-creating knowledge about CCA by policy actors and local farmers. This is in line with an earlier study that demonstrates that smallholder farmers in this region perceived their key constraints predominantly as a lack of information on climate change characteristics and CCA strategies provided by government (Phuong, Biesbroek, & Wals, 2017).

Understanding the interactions of the three dimensions of policy capacity that enable and constrain policy actors at different levels of government is important to consider when designing policy interventions to overcome some of these constraints in the agricultural sector in particular and other sectors more generally. It requires comprehensive assessment that aims to understand the intricacy of the different causes that constrain CCA (Biesbroek et al., 2015). A key recommendation is, therefore, to enhance local policy capacities in Vietnam by establishing a clear legal mandate for adaptation that makes it a primary concern instead of a criterion to be considered in SEDPs. Addressing this main cause would already address many related barriers, including those related to accountability and coordination. Educating and enhancing knowledge and skills for agricultural staff at local levels on the long-term climate impacts and possible adaptation actions will be crucial to further advance local adaptation, but this will require political willingness and institutional mandates. Finally, it is important to consider the hierarchical context in which these processes take place. Solutions such as governance system transformations are not likely to emerge anytime soon nor will they be driven by climate change. Designing new institutions, building governance resources, and increasing policy analytical capacity should therefore consider the existing hierarchical governing system.

7. Conclusion

This paper investigated how in Vietnam, the policy capacity at different levels of government enables or constrains the CCA at local levels. We conclude that in Vietnam’s agricultural sector, the current hierarchical multilevel setting enables implementing CCA at national level but it creates several interlinked constraints at district and commune levels. Whilst several constraints have been reported by policy actors, we find that the institutional setting and lacking legal mandate are crucial to explain the limited progress on CCA in Vietnam across different administrative levels. Our findings suggest more generally that overcoming institutional, resource, and policy analytical capacity constraints will be vital to further prepare Vietnam to the impacts of climate change. This requires to some extent acceptance that the hierarchical policy system in Vietnam has both advantages and disadvantages for governing CCA.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

G. Robbert Biesbroek http://orcid.org/0000-0002-2906-1419

Notes on contributors

Le Thi Hong Phuong is the lecturer and researcher at the Faculty of Extension and Rural Development, University of Agriculture and Forestry, Hue University, Vietnam. She obtained her PhD at Wageningen University in 2017. Her current research includes in designing, supporting and evaluating learning processes that can (re)connect people and planet, and help overcome global systemic dysfunction (particularly to respond to climate change) and poverty issues in Asia [Email: [email protected]].

Robbert Biesbroek is Assistant Professor at the Public Administration and Policy group, Wageningen University & Research, the Netherlands. He obtained his PhD at the Wageningen University in 2014. His current research includes causal mechanisms of complex decision making, dynamics of policy (dis)integration of crosscutting societal issues, tracking policy change, and the political and bureaucratic responses to climate change adaptation [Email: [email protected]].

Arjen Wals is a Professor in Transformative Learning for Socio-Ecological Sustainability at Wageningen University in The Netherlands and a Guest Professor at the University of Gothenburg in Sweden. He obtained his PhD at the University of Michigan in 1991. His main interest lies in designing, supporting and evaluating learning processes that can (re)connect people and planet, and help overcome global systemic dysfunction [Email: [email protected]].

Additional information

Funding

This work was supported by The Nuffic Project [grant number NICHE/VNM/105].