Political agenda-setting for strategic delta planning in the Mekong Delta: converging or diverging agendas of policy actors and the Mekong Delta Plan?

This article uses the lens of the Multiple Streams Approach to explore whether the agendas set by political actors in Vietnam converged with the agenda set in the Mekong Delta Plan (MDP). The MDP presents policy choices for the development of the Vietnamese Mekong Delta. The plan offers economically attractive, climate adaptive and environmentally sustainable paths forward in the face of climate change and economic uncertainties. We collected our data using qualitative techniques, including a literature review and interviews. We found convergence between the MDP’s agenda and political actors’ agendas, though divergences were also detected. Between the delivery of the MDP in 2013 and formal endorsement of its ideas in 2017, the problem stream, policy stream and politics stream were brought together by the actions of “policy entrepreneurs” (scientists and experts). Our findings suggest that agenda-setting and convergence were a crucial step towards endorsement of the strategic delta planning process for the Mekong Delta. Further research could explore issues of power mobilization in enabling or constraining decision-making.


Introduction
Vietnamese delta managers are calling for more responsive policies to cope with the challenges of climate change, upstream hydropower development and economic activities within the Mekong Delta (Smajgl et al. 2015;Anthony et al. 2015). Examples of such policies are "Living with the Floods" and "Turning Disasters into Opportunity" (Vietnamese Government 2017). Up to now, Vietnam has applied a classic approach in delta planning, characterized by short-term master planning and sectoral planning (5-10 years), though without overall coordination and prioritization (Royal HaskoningDVH et al. 2013). Strategic delta planning is conceived as a new way to formulate a development agenda for the delta in order to address both socio-economic and physical-ecological aspects over a longer timeframe (Healey 2004). Previous delta planning research (Seijger et al. 2017a) examined the agenda-setting process that starts the formulation of a strategic delta plan. This entails bringing the topic of strategic planning to the fore in policy and academic circles and determining what are the key issues and problems to be covered by the strategic delta plan. The current study expands on that work, exploring the process of agenda-setting after formulation of a strategic delta plan, in our case the Mekong Delta Plan (MDP). We analysed how the agenda set by this plan has influenced Vietnamese political actors, particularly their policy agendas and priorities, and how this led to its formal endorsement.
Policy studies focusing on Vietnam are rare, particularly those focused on policy agenda convergence in relation to delta planning. Most research in this area has examined specific policies or offered a historical or contemporary view on government polices influential in the Mekong Delta (Hoanh et al. 2003;Hoanh, Suhardiman, and Anh 2014;Biggs et al. 2009). The current study applies the lens of the Multiple Streams Approach to explore the effectiveness of strategic delta planning in influencing the policy agenda. Specifically, it explores the extent to which Vietnam's strategic delta plan has fostered alignment of different policy agendas for long-term development.
The MDP is a climate change adaptation strategy. It defines no-regret measures for sustainable development of the Mekong Delta in the face of contemporary and future challenges. As such, it defines a new development paradigm that, to have real influence, must influence decision-makers and policy agendas. Agendas are considered across different tiers of governance (national, provincial, local and donor) and across different sectors (e.g., those represented by the different line departments). Indeed, sustainable development of the Mekong Delta is a multifaceted objective involving a diversity of actors and policy domains. Influencing agendas across these actors and domains is key to fostering broad endorsement of the delta planning goals.
In the policy sciences, decisions are regarded as policy outputs. The current article assesses the process of formal decision-making or policymaking, by looking at two such outputs, namely Resolution 120 and Decision 593. Both confirm endorsement of the ideas of the MDP at Vietnam's highest political level, that is, by the Vietnamese Prime Minister. Using these outputs, we studied convergence and divergence between the agenda set by the MDP and those adopted by political actors.
Thus, the aim of our study is to explore whether the actors involved indeed moved towards adoption of the agenda set in the MDP. That is, to what degree did the agendas of the various actors converge with, or diverge from, the ideas presented in the MDP. We conducted this assessment using the Multiple Streams Approach. This approach employs three separate, though interconnected, streams: the problem stream, the policy stream and the politics stream. A policy issue receives serious attention and prioritization in agenda-setting and policymaking when these three streams come together through decision makers' efforts (Kingdon 1984). Agenda-setting involves a coupling of the problem and politics streams, while policy implementation involves a coupling of the policy and problem streams. However, implementation can only succeed if the politics stream is consistently favourable (Mu 2018).

Theoretical foundation
2.1. Agenda-setting as a concept Kingdon (1984, 3) defined an agenda as "the list of subjects or problems to which governmental officials, and people outside of government closely associated with those officials, are paying some serious attention at any given time." If an issue is not on an agenda, it cannot be subjected to decision-making processes and will never become a decision (Princen 2009). According to Zahariadis (2016), the term 'agenda' first appeared in the policy science literature in 1971. Meanwhile, 'agenda-setting' is defined as "the process of turning public issues into actionable government policies" (Zahariadis 2016, 6). Cloete, Wissink, and De Coning (2006) suggested that communities' perceived problems enter the political system as inputs from interest groups via the political process. Considering agendas analogously as inputs sheds light on how ideas are brought into the policymaking process. This view helps us to understand how political actors and institutions may perceive problems, construct their own agendas and view a strategic delta plan in relation to their traditional ideas or policies (Seijger et al. 2017a).

The multiple streams approach
The Multiple Streams Approach emerged as an adaptation of the Garbage Can Model by Cohen, March, and Olsen (1972). It was developed by Kingdon (1984) for use in the US political context, further Liu et al. (2010) also applied it to the US's politics at local level. Various international scholars have since adopted it for studies exploring a range of policy contexts, from the EU and Germany to China and Hongkong (Teodorovi c 2008;Baumgartner, Jones, and Mortensen 2014;B eland and Howlett 2016;Chow 2014;John 2003;Mu 2018;Jones et al. 2016;Zahariadis 1999Zahariadis , 2003Zahariadis , 2008Zahariadis and Herweg 2017;Zahariadis 2016;Zohlnh€ ofer, Herweg, and Huß 2016;Baumgartner and Jones 2010;Huitema et al. 2009;Cairney and Jones 2016;Rawat and Morris 2016). Recent work on spatial planning and delta development suggests that strategic planning is becoming the preferred means of setting policy agendas for long-term delta development (Seijger et al. 2017a;Healey 2004;Friedmann 2004;Olesen 2017). Our study was interested in the influence of strategic planning in delta development. Specifically, we explored the extent that Vietnam's strategic delta plan has fostered convergence among different stakeholders' policy agendas for long-term delta development. Kingdon (1984) used several metaphors to illustrate the Multiple Streams Approach, including 'open windows' and 'primeval soup', though without clearly defining the different streams. Derived and developed from this original idea, Zahariadis (1999, 32-34) identified three streams as follows: "[T]he problem stream consists of various conditions that policy makers and citizens want addressed; … ; the policy stream includes a primeval soup of ideas that compete to win acceptance in policy networks; … ; the politics stream consists of three elements: the national mood, pressure group campaigns, and administrative or legislative turnover". Although these streams developed in part independently from each other, once they are coupledthat is, when a problem is recognized simultaneously with an available solution and there is political support in favour of the changea window of opportunity or policy window opens (Guldbrandsson and Fossum 2009;Kingdon 1984).
The current research perceives the problem stream as a process of setting agendas and raising concerns from both inside and outside the political system. The policy stream encompasses those actors that can introduce policy ideas to policymakers and policy networks (e.g., experts, researchers and consultants) after which the ideas vie for acceptance. The politics stream is about politicians and the decisions or policy outputs they generate in formal decision-making processes, including legislative turnover and political attention. Because the Multiple Streams Approach was originally developed for the US political context, it might not be completely applicable to the Vietnamese situation. Indeed, Mu (2018) adapted the approach to the Chinese context. This study adopts those modifications. Following Mu (2018), we interpreted the politics stream as referring to political attention, in addition to legislative turnover, and not including the national mood and pressure group campaigns. Political attention tells us where policymakers turn in their search for solutions, and what public values government pursues (Mu 2018). Furthermore, we label actors from academic institutions as 'policy entrepreneurs' (Huitema and Meijerink 2010) due to their particular roles in the Vietnamese policy process, including consulting and advising the policy makers. They do not take a direct part in decision-making, but rather push the agenda process and try to merge the three streams (Mu 2018).

Data collection and analysis techniques
We adopted a qualitative research technique with three parts: a literature review, indepth interviews and a content analysis. To trace the policy agendas of the past, we relied on Biggs (2004Biggs ( , 2012, Biggs et al. (2009) and Vormoor (2010) as primary sources on the history of Mekong Delta planning and development. We conducted comprehensive interviews with a diverse pool of 15 actors from 13 organizations (Table 1), ranging from the local to the national level. Our interview selection process was designed primarily to identify key policy informants with knowledge and information about agenda-setting in policy on the Mekong Delta. All these informants acted as policymakers to varying extents. Some were practitioners, some worked at academic institutions, and some came from international donor organizations. One interview subject came from the Vietnamese Communist Party. We used the snowball technique to identify potential interview subjects, as we did not expect any single actor in the delta planning process to be able to express all views on the subject of delta planning. The interview subjects were all men and ranged in age from 30 to about 80.
We hypothesized that the different policy agendas for the Mekong Delta would converge in the strategic delta plan. The historical record and the interview results helped us determine whether that was indeed the case. To facilitate the analysis process, we categorized the 15 interview subjects into four groups based on the functions and duties of institutions with which they were affiliated. These groups were labelled as follows: (A) "national governors", (B) "local governors and practitioners", (C) "national experts and advisors" and (D) "international development partners". These roughly represent Kingdon's (1984) policy and politics streams. The policy stream is linked to group C ("policy entrepreneurs") and D (introducing new ideas and proposals); while groups A and B are linked to the politics stream (making decisions). All of the four groups are related to the problem streams. The SWSC (that belongs to group A) is also termed as "policy entrepreneur". The actors did not behave as groups, however, as agenda topics were not always shared between group members. The interviews were digitally recorded, transcribed and analysed.
For our case analysis, to obtain the interview subjects' agendas we applied content analysis (Krippendorff 2004). This involved a three-step coding process (Boeije 2009). The first step was to apply initial coding to the themes discovered within the data. For example, an agenda topic noted by the Southwest Steering Committee (SWSC) interviewee was to promote a regional integration mechanism. The second step was pattern coding to uncover major themes, or patterns, underlying the data segments. We then linked the themes coded from the interviews to the agendas presented in the MDP and linked this data to theory. The patterns indicated the extent to which certain themes recurred across interview subjects (convergence or alignment), or did not recur and different themes prevailed (divergence). In the third step, we triangulated patterns and themes to seek a new depth of understanding of existing knowledge. For this step, we reviewed the interviews and compared them to the results of our previous two coding steps. Finally, we reviewed emergent themes, findings and notes in the light of the Multiple Streams Approach literature.
It must be noted that while the sample is diverse, it is also limited in the sense that not all actors were involved. For example, national-level ministers and the prime ministers and vice prime ministers were left out, and more local-level officials might have been included. In the donor group, no-one was interviewed from the Japan International Cooperation Agency (JICA). However, we did interview international donor representatives, alongside high-ranking government advisors closely associated with national-level ministers. They had good insight into the relevant policy processes.
4. Development of the Mekong Delta from a policy agenda perspective 4.1. Policy agendas through history As the starting point of our analysis, we examined past policy agendas to better understand how the Vietnamese Mekong Delta has developed and the historical context of the scenarios presented in the MDP for the delta's future. In order to do that we applied literature review instead of applying MSA, since in this part we focus mainly on the problem and policy streams without analysing the process of politics that is difficult to assess historically. The earliest policy agenda in Mekong Delta development was exploiting the delta. Under French colonial rule, the government implemented intensive canal digging and initiated a dike-building strategy to control and manage the delta waters (Biggs et al. 2009). This agenda became hugely influential, and extended almost through to the present day.
The second key agenda was food security, which was pursued by closing off the delta with the building of a massive network of dikes, sluice gates, dams and weirs (Vormoor 2010) ( Figure 2). As acid sulphate soils cover two-thirds of the delta, land reclamation was another important agenda, to reclaim 'wastelands' for agriculture. However, as new canals were dredged, civilization penetrated further and deeper into isolated places (Vormoor 2010;Benedikter 2014;Biggs 2012). Thus, another important agenda from the past was rural settlement, by which large populations from northern Vietnam and Cambodia were resettled to rural areas of the Mekong Delta (Sanh, Xuan, and Phong 1998). This appeared on the policy agenda in other terms, too, that is, as colonization of southern 'wastelands' by the northern Vietnamese (Biggs 2004). Prior to the end of the Indochina War, the policy agenda of expansion of agricultural areas gave additional impetus to reclaiming acid sulphate soils. After the nation's reunification, these agendas were narrowed to the agenda of rice cultivation everywhere to ensure food security. This was a reaction to the severe food shortages experienced following the war, especially in northern Vietnam.
After 1975, the key policy agenda became nation-building, including the Mekong Delta. Agricultural diversification was also pursued. This began with the agenda of restoration and excavation of waterways to facilitate agricultural collectivization (Huynh 2015, 45). After 1986, the agenda of political and economic reform came to the fore, particularly Doi Moi. The Doi Moi reform was associated with politicaleconomic decentralization and liberalization, specifically towards household productivity, and ruralagricultural development. The agenda of transforming the Mekong Delta into the country's rice bowl led to the intensified rice strategy, which justified investment in flood-protection dikes, canals and pumping stations (Biggs et al. 2009). To accommodate the agenda of expanding rice double cropping, huge areas of land were excavated for canals and irrigation and water control infrastructure (Vormoor 2010). This did enable Vietnam to become self-sufficient in rice. The economic reforms released farmers from production quotas. The policy agenda here was agricultural diversification and a switch to higher-value crops, such as fruit trees, and aquaculture development. Nonetheless, except in the zones under tidal management, the Mekong Delta's dike infrastructure was designed for rice irrigation (Biggs 2012). Agendas thereafter focused on control of the delta waters to allow multi-cropping, particularly cultivation independent of the natural, seasonal hydrological regime (Vormoor 2010).
We analysed these past policies to derive the entrenched policy agendas that informed the development of the Mekong Delta. Central in these historic agendas was the building of high dikes and accompanying sluice gates, irrigation infrastructure and canalsthough these did not always meet the challenges the Mekong Delta faced. In 2013, the MDP brought a new agenda, offering scenarios for economically attractive development, environmental sustainability and adaptation to climate change. The MDP is analyzed below. It is the only present-day delta development plan included in our study because, to date, there is no more influential agenda for Mekong Delta development.

Agenda set in the Mekong Delta Plan
The MDP arose from an agenda-setting process concerned with economic development and climate adaptability. The three pillars of this agenda can be summed up as follows: economically attractive development, environmental sustainability and capacity to adapt to climate change. In particular, the MDP's primary agenda is to shift the Mekong Delta away from the longstanding food security and rice bowl emphasis. The MDP recognizes that the focus on rice production will not be economically attractive in the future, as population growth is predicted to significantly slow down. Continuation of the agenda of intensive rice production would, in fact, further damage the adaptive capacity of the delta, according to the plan. In contrast, and this is a key argument of the MDP, it is much more economically advantageous to move towards sustainable aquaculture (e.g., polyculture, sustainable brackish aquaculture, mixed fruit-aquaculture systems and value added to agricultural products).
The MDP sets out specific agendas for three distinct delta zones. The first zone, the Upper Delta, is favourable for developing farming systems that can cope with seasonal flooding and increased volumes of flood waters. This agenda seeks to grasp windows of opportunity for river expansion. The building of high dikes and sluice gates is discontinued, though this will diminish opportunities for triple rice cropping. The second zone is the Middle Delta. Here, the MDP foresees responses to the challenges of the region's declining freshwater supplies in the dry season. The third zone is the Sensitive Coastal Zone. Saline intrusion is the primary focus here. However, salinity is no longer treated as a disaster. Rather, a brackish zone is viewed as an environmental resource.
One of the most striking elements of the MDP is its formulation of four possible development scenarios: agro-business industrialization, corridor industrialization, dual node industrialization and food production. Of these, shifting to an agro-business industrialization development model is considered the preferable scenario, as it seems most feasible, viable and sustainable. This scenario takes advantage of natural systems and the richness of the Mekong Delta and promotes a form of agriculture that capitalizes on the biological, socio-economic and physical characteristics of the delta. Agrobusiness is expected to provide an excellent basis for long-term, sustainable economic growth. This scenario has the added advantage of being aligned with a decision taken by the Prime Minister on 17 December 2012 to pursue hi-tech agricultural development up to 2020.
4.3. The policy process leading up to the Mekong Delta Plan, 2010-2017 All of our key informants were involved in different phases of the organization and proceedings of the 2013 Mekong Delta Economic Cooperation Forum (MDEC) and the subsequent Mekong Delta Forums. 1 These are both annual high-level policy events centred on future development of the Mekong Delta. At these meetings, the MDP, and thus its agenda, was a frequent item of discussion. Figure 1 schematizes the roles of the different interview groups and their agendas, which are further detailed in later sections. The early MDP process, from 2010 to 2013, involved consultations with national counterparts and experts, whom we labelled 'policy entrepreneurs'. This was followed by intensive discussions with all stakeholders. After delivery of the MDP, at the 2013 MDEC, an international donor group was engaged, which led to a letter of endorsement from them of the MDP. Subsequent to the MDEC, the donor group, together with the national governors began organizing the annual Mekong Delta Forums, which bring in all the involved actors. This political process propelled two outcomes: the Prime Minister's 2016 endorsement of Decision 593 on Promulgation of the Regulation on Pilot Coordination for Regional Socio-Economic Development in the Mekong Delta (2016-2020) and the 2017 adoption of Resolution 120 on Sustainable and Climate-Resilient Development of the Mekong Delta. This legislation inaugurated the ideas of the MDP into national policy.

Agendas set by political actors and their convergence or divergence with the Mekong Delta Plan
Having categorized the relevant stakeholders into the four groups, this section examines the core agendas of each group in turn (Table 1). Group A represents central policymakers, most of whom were based in Hanoi, except the representative of SWSC, who was regionally based in the Mekong Delta. Group A is made up of political actors with direct influence on the policymaking process. Though currently inactive, SWSC was a powerful entity on Vietnam's political scene at the time of the interviews, being the government Communist Party for the Mekong Delta provinces. Group B represents local government, including departments functioning as practitioners on the ground. Those in this group worked directly with farmers. Their views are thus closest to those of local communities. Group C is the group with the most technical expertise. It is made up of national experts from universities, line ministries' research institutes and policy think tanks, alongside retired high-ranking government officials. All were active in the Mekong Delta planning process. Group C is also labelled 'policy entrepreneurs', as its members operated as mediators of agenda-setting, primarily due to their unique position in the stakeholder arena and their wider and longer-term access to policy discussions and actors. Group D represents international donors. They provided financial assistance and conducted research and projects for the direct benefit of the local communities. In the analyses using the Multiple Streams Approach, groups A and B are linked to the politics stream, as they supported and initiated the coupling of the problem stream with the policy stream. Groups C and D are linked to the policy stream, as its members consulted, advocated, introduced new ideas and gave advice to facilitate achievement of the policy output.

Group A: National governors
Within group A, a retired official from the Ministry of Agriculture and Rural Development (MARD) noted improving coordination between entities (ministries and provinces) and creating a regional coordination mechanism as prominent policy agendas. Another interviewee, who was working as an advisor to MARD at the time of the interview, prioritized addressing problems with upstream countries caused by competition over water resources, although this topic was not included in the MDP. This interviewee also pursued an agenda of solving salinity intrusion problems, as salinity intrusion has become increasingly problematic in recent years.
An interviewee from SWSC mentioned the agenda of supporting and cooperating with the regional coordination mechanism (more recently the term interprovincial collaboration mechanism has become more popular) to adapt the Mekong Delta to climate change. This reflected the subject's former formal role, as SWSC was part of a national committee to respond to climate change. It was mandated to organize at least two annual meetings on the topic, involving all ministries, sectors and the 13 provinces of the Mekong Delta. It thus played a bridging role between the local and central government levels.
Our interviewee from the Ministry of Natural Resources and Environment (MONRE) expressed two major agendas: achieving further cooperation with international agencies and experts to come up with concrete activities and plans and raising awareness among the public and local communities. Coordination with local communities was considered important because the agendas typically set by MONRE for addressing climate change and other problems took the perspective of the entire Mekong Delta, whereas local inhabitants tended to focus on impacts or problems at the scale of their own community or region.
In conclusion, within group A, the policy agendas of the institutes based in Hanoi, being located at a greater distance from the Mekong Delta, were inclined towards institutional development and furthering macro-level priorities, such as solving upstream development problems and fostering cooperation between stakeholders. Meanwhile, institutions located in the Mekong Delta, here represented by SWSC, had a stronger link to the situation on the ground within the Mekong Delta and thus cared more about improving local livelihoods and regional cooperation.

Group B: Local governors and practitioners
One group B interviewee, an official from a local branch of the District People's Committee (DPC) of Tra Vinh province, indicated concern with improving local livelihoods. This interviewee prioritized sustainability and agreed with the need to switch coastal areas to brackish water aquaculture and support polyculture instead of the intensive rice monoculture. Another agenda mentioned was creating reservoirs to supply freshwater to local people, as the already limited groundwater was becoming increasingly depleted.
Most of the action plans proposed by the Department of Agriculture and Rural Development (DARD) (Tra Vinh Province) were inspired by ideas in the MDP. Our DARD respondent noted the boost that the MDP had given to some of the projects already being implemented in the province. It was thus positively impacting the local level, the interviewee said, for example, by supporting sustainability, including the switch to a brackish environment and storing freshwater.
The interviewee from the Department of Natural Resources and Environment (DONRE) said the biggest concern was obtaining financial support to implement new policy and strategy. This interviewee strongly supported the MDP, but disagreed with the goal of increasing the share of integrated shrimp-mangrove farming systems and protected forests. The alternative agenda offered was tailored models towards diversification using a zoning approach.
We spoke with two people from the Department of Planning and Investment (DPI), though both were vague in expressing their agendas, which seemed geared mainly towards developing their own institutional plans and projects. DPI's agendas focused primarily on industrial development, including enhancing added value in agriculture and development of infrastructure, such as roads, waterways and industrial corridors.
In sum, this group of practitioners, being civil servants working directly with local communities, seemingly understood the real needs and demands of the region. Their agendas were mostly in accordance with the ideas presented in the MDP, and they understood the practical reasons underlying them. An exception was DPI (of Tra Vinh Province), which was not directly involved in the planning process or in farmers' livelihoods. Its interests lay more in its own plans and projects, particularly industrial development. Aside from this exception, a relatively convergent trend was found across this group's agendas in relation to the agenda of the MDP.

Group C: National experts and advisors
The national experts and advisors were involved from the early stages in creating a platform for discussion of the Mekong Delta's present and future. They pursued the agenda of determining responses and solutions that are preferably "no-regret". Such responses, thus, should mitigate any potential for devastating impacts on local communities and natural resources. Alternatively, some in this group suggested that the problems facing the Mekong Delta could be viewed as opportunities. For instance, a professor from Can Tho University (CTU) suggested considering salinity in the coastal zone as an advantage for developing a brackish-water based economy. This informant indicated support for a dual zone approach to coastal management combining brackish aquaculture with agriculture. The agenda pursued was improving local livelihoods and alleviating poverty at the local level, as well as improving human resources and agriculture to help local communities. This expert, who had been involved in formulating the MDP, expressed as priorities addressing saline intrusion and the lack of freshwater in dry years, climate change and developing agro-business.
Our second interviewee from CTU had contributed to scenario-building for the MDP. This expert strongly disagreed with the food security agenda of the past, as this was said to have brought huge costs and few benefits to the Mekong Delta. This interviewee prioritized agribusiness and aquaculture development and enhancing the functions of nature, for example, via restoration of the mangrove forests.
The interviewee from Tra Vinh University (TVU) was concerned mainly with how saline intrusion might motivate local people to shift to aquaculture. A related concern was the entrenched tradition of rice production, as this expert considered it would be hard to change. Resolving conflicts from competing interests in water management was another policy agenda mentioned, especially the potential for benefit-sharing between upstream and downstream communities and provinces (e.g., between An Giang and Tra Vinh or Ben Tre).
Our fourth expert was a retired official from the National Assembly's Committee on Science and Technology (NACST) and was still active in the planning process. This person was strongly supportive of sticking to traditional agendas, supporting a food security approach and building more high dikes and sluice gates. The informant thus opposed the new agenda, citing the huge changes it would bring. However, this person did support the idea of restoring the freshwater of the Mekong Delta, especially for provision in the dry season.
Hence, most of the experts' agendas were in harmony with the MDPthe exception being the informant from NASCT, who favoured adhering to the agendas of the past and keeping the Mekong Delta as the nation's rice bowl. This relative harmony among agendas is remarkable, as theory suggests that people perceive things differently, then pursue agendas based on these different perceived interests. That is why in policymaking processes, it is never easy to reach absolute agreement on final choices, as different agendas exist from the start, rooted in the different perceptions. In this group, we see a convergence, or alignment, of agendas with the MDP, although it is not absolute.

Group D: International development partners
In our international development partner group, the World Bank recognized the advantages of the agenda set by the MDP, as the plan is largely aligned with the World Bank's own project in the delta. The World Bank generally pursues a holistic, integrated planning approach instead of the traditional sectoral method of planning. Adaptive delta management is considered a valuable means towards sustainable exploitation of water resources, development of agriculture and protection of the environment, while adapting to climate change. The World Bank prioritizes no-regret measures that take into account climate change and development uncertainties. For example, rice is preferred for regions for which it is perfectly suited. In other regions where more profitable livelihood options exist, such as polyculture and organic shrimp production, those should be pursued. Most World Bank projects reflect strategic choices for controlled flooding in the upper delta and sustainable aquaculture in the estuary zones. The World Bank also focuses on livelihood models. The organization is innovating in integrated mangrove-shrimp and rice-shrimp models for coastal zones, and looking at different kinds of flood-based agriculture for the upper delta. For example, floating rice or lotus-based systems might be considered appropriate in some places, while two rice crops with a differing third crop could be favoured elsewhere. The World Bank observed that a lot of data and information on the Mekong Delta is split between different research agencies and ministries without protocols for data sharing or collaboration. One of its agendas is improving that situation.
The interviewee from the International Union for Conservation of Nature (IUCN) expressed his key agenda as addressing problems of rice intensification and flood risk. However, IUCN cannot accomplish this alone. That is why, according to our interviewee, there was a need for cooperation between various development partners, such as the Dutch government, the World Bank and GIZ (Deutsche Gesellschaft f€ ur Internationale Zusammenarbeit GmbH). This interviewee expressed deep concern about losses in capture fisheries and the loss of absorptive capacity upstream, linked to increased salinity intrusion in the dry season.
The GIZ representative expressed the organization's priority agenda as integrated coastal management, saying this was a major focus of their projects. In collaborating on the MDP, GIZ's main agenda was zoning. In particular, this related to what the MDP proposed regarding adaptive rice production in the upper delta and brackish aquaculture in coastal zones. Inclusion of these specifics was necessary for them to lend collaborative support to the plan. GIZ's second priority was formulation of an overarching plan or strategy for the Mekong Delta that was officially adopted by the government. GIZ is also working hard to improve regional coordination.
Ultimately, we found an absolute alignment of the strategies and interests across this donor group, and with the MDP. Nevertheless, coordination was still considered necessary, not only with respect to agenda-setting, but also in practical actions, as these actors usually focus on their own projects. One interviewee in this group suggested that the development partners had to realize that an integration process would be required to take the next step.
5. Discussion on processes steering convergence and divergence 5.1. Evidence on convergence and divergence The interview data provided strong evidence of convergence between the agendas of political actors and the agenda set in the MDP. Further evidence of this convergence is provided by the Prime Minister's formal endorsement of Decision 593 in 2016 and Resolution 120 in 2017. That legislation contains many of the ideas of the MDP. Indeed, Decision 593 echoes the MDP's call for a sub-regional planning approach. Another link to the MDP agenda is Decision 593's call to strengthen regional planning and collaboration between provinces (Vietnamese Government 2016). In the past, people did not talk about such mechanisms. In Resolution 120, the main directive points are turning challenges into opportunities; adapting to climate change and sea level rise and pursuing nature-based development; switching from intensive rice cultivation to diversified agriculture and agro-business; considering salinity as a resource for development of a brackish zone economy; and promoting interprovincial collaboration (Vietnamese Government 2017). Those ideas are perfectly aligned with the MDP agenda. What processes explain this convergence? We explored this question by studying the interconnections between the problem, policy and politics streams, and the roles of policy entrepreneurs in bringing these streams together. The formal endorsement of Decision 593 and Resolution 120 is considered proof of convergence of the agendas of political actors and the MDP.
As noted, groups A and B are linked to the politics stream, and group C and D to the policy stream. An analysis using the Multiple Streams Approach lens suggests that these streams operated both independently, each setting their own agendas and priorities, and in an integrated fashion, via the consultations and group discussions culminating in the consecutive Mekong Delta Forums. We identified the problem stream by way of monitorable indicators, such as salinity concentration, drought frequency, flood intensity and crop failure; focusing events, such as salinity peaks, that draw attention to problematic conditions; and feedback from previous programmes, which can highlight what works and what may not. As such, a disaster in the form of an extreme drought with salinity intrusion that irregularly occurs could become a focusing event to draw politicians' attention and create a sense of urgency for steps towards the MDP agenda.
The policy stream includes a variety of policy agendas that vie to win acceptance within policy networks in Vietnam. The number of agendas is large, so only a few can receive serious consideration. We observed a diversity of agendas that competed to win acceptance and become policy in the Mekong Delta. Since policymaking in Vietnam is mostly sector-driven or project-driven, clear boundaries between sectors remained, as well as overlaps in project proposals between sectors and ministries (Conway 2004).
We applied the politics stream to the Vietnamese political context only in regard to legislative turnover and political attention, as ideas must be formalized in official documents (e.g., a resolution, decision or decree) before they become policy. The MDP provided a vision to guide political actors in composing their proposals; but its ideas were not formalized in national policy until the adoption of Decision 593 and Resolution 120, respectively, in 2016 and 2017.

Policy windows revealed by the multiple streams approach lens
Resolution 120 was endorsed at a critical point when the three streams came together. Occurrence of a focusing event, in our case a 20-year peak of saline intrusion measured in 2016, engaged all the relevant political actors. Endorsement of Resolution 120 was also facilitated by the consecutive Mekong Delta Forums, especially the 2017 event, which was chaired by the Prime Minister.
Throughout the development history of the Mekong Delta, the agendas of the stakeholders in the political process have changed. For instance, the agendas of the past were more conscious of the need to ensure food security and enable intensive rice production. Meanwhile, since the arrival of the MDP, each of the stakeholders has composed and presented newer agendas to the central government, in the form of action plans and project proposals, that sometimes overlap or conflict, both with each other and with the MDP. According to the Multiple Streams Approach, this process forms part of the problem stream. Agenda-setting for the MDP climaxed at the 2013 MDEC and at the Mekong Delta Forums that have been organized annually since 2013. These propelled a process of intensive consultations involving MARD, MONRE, the Prime Minister and focus groups (see Figure 2). All these actors represent the politics stream.
There were also protracted engagements with the entrenched political agendas of the past. Here, SWSC emerged as a key policy entrepreneur, as this organization responded positively to the MDP agenda. The sudden rise of SWSC in support of the MDP agenda thus significantly contributed to the further endorsement of the plan. The MDEC and the later Mekong Development Forums also engineered important political breakthroughs that opened windows of opportunity for considering alternative development agendas. Due to the presence and involvement of all the various actors at these events, a political process got underway that greatly facilitated convergence between the agendas of stakeholders and that of the MDP. The group of national experts (C) acted as indispensable policy entrepreneurs. They played crucial roles in promoting the MDP's agenda: consulting, advising, giving feedback on formulating the MDP's agendas; delivering training courses; taking action for the MDEC; and taking action with donors and with the national governors. They engaged with many actors. Once national experts supported the MDP agenda, they became important in influencing the agendas of others.
At each step, from the MDEC and delivery of the initial MDP, to the first Mekong Delta Forum in 2013 and endorsement of Resolution 120 and the latest Mekong Delta Forum in 2017, the agenda of the MDP increasingly converged with the agendas of stakeholders and became more 'Vietnamese' (see Figure 1 on convergence over time and Tables 2 and 3 for convergence across actors). The stakeholders operated at different steps of this political process. The efforts and actions of policy entrepreneurs led to consolidation in agenda-setting across all stakeholders along the way. This eventually culminated in endorsement of Resolution 120 by the Prime Minister at the point in time when the three streams came together. Consolidation of agenda-setting thus took time and effort. Rawat and Morris (2016) argued that, although agenda-setting may be perceived as non-incremental, specification of alternative policy can be both non-incremental and incremental. This was indeed found to be true in our case. The agenda set by strategic delta planning underwent a long process of soft implementation, as the many minds and agendas involved did not automatically converge with the new strategic agenda set in the MDP (Seijger et al. 2017b).
However, looking back, although a clear convergence of goals and agendas can be observed between the stakeholders and the MDP, this was certainly not uniform across all stakeholders. Nuances and divergences were also found in all groups. Group A, for example, set agendas mainly at the macro level, such as cooperation between national and international entities and solving problems caused by upstream developments. Group B focused more on improving local livelihoods, while group C was most interested in no-regret solutions for the entire Mekong Delta and embracing a brackish-water based economy. Group D displayed elements of all these agendas. The divergences can be traced to institutional political affiliation, differences in age and selfinterest, and more importantly, to geographical locationthat is, whether the actor was based inside or outside the delta. The agendas of the 'insiders' (SWSC, DARD, DONRE and DPC) were more in line with the MDP agenda concerning the practical needs of local communities, for example, creating better livelihoods, improving the water supply situation and willingness to embrace a brackish environment. The 'outsiders' (MONRE, MARD) were concerned with institutional agendas (e.g., cooperation at the national level, solving upstream conflicts and raising awareness). Divergences were also identified between local and national groups. The former, practitioners, were more problem-oriented and set agendas that reflected the real needs of farmers. They were situated closer to the emerging problems of sustainability, economic development and the need to adapt to the changing environment. Their agendas and concerns were thus more closely affiliated with the problem analysis and solution strategy embedded in the MDP agenda. Meanwhile, the latter were more concerned with the institutional and governance aspects of the agenda than with the details of the problems and solutions. For example, they sought enhanced cooperation between entities and pursued ways to cope with climate change at the delta scale. Another type of divergence was found in the mindset of one national expert. The interviewee from NACST objected to significant changes in the development model for the Mekong Delta, and thus could not fully support the MDP.
The conclusion we draw from this analysis is that agenda-setting is a long-term incremental process that requires sustained stakeholder engagement across a wide spectrum of the policy and politics arena. In the case of the MDP, engagement and cooperation of policy entrepreneurs who were strategically positioned within this arena proved pivotal. We hypothesized that the different policy agendas for the Mekong Delta would converge in the strategic delta plan. Our evidence confirms this hypothesis. Although it took time, the various actors' different policy agendas for Mekong Delta development did converge with the agenda put forward in the strategic delta plan.

Conclusion
The MDP arrived on the Vietnamese political scene as a paradigm shift. It introduced a whole new development agenda for the Mekong Delta. The MDP agenda brought emergent issues to the fore, thus representing a break from the agendas of the past, particularly the longstanding key agenda of food security through intensive rice production. The MDP proposed a better way to achieve food security and sustainable rice production; that is, through flood-adapted farming systems and a switch to brackish-water based economic development, while viewing the saline zone as an environmental resource. This was proposed to replace the past strategy of building high dikes and sluice gates, irrigation infrastructure and canal projects. Many interviewees considered the agendas of the past to be ineffective today, and saw the Mekong Delta as in need of new development directions. Many of them, especially those from SWSC, CTU, TVU, donors and local practitioners, saw their own concerns and agendas reflected in the MDP.
After delivery of the MDP, in 2013, formal recognition of the ideas presented in the plan by the Vietnamese Prime Minister followed in 2016 and 2017. Divergence of agendas was narrowed over time by the actions of policy entrepreneurs, particularly, through SWSC's organizing of the annual Mekong Delta Forums and the intensive discussions and consultations held among national experts and academics and their active cooperation in these events. This finding is important, as it demonstrates that the strategic agenda set by the MDP eventually did influence the agendas of a wider pool of actors, ranging from the local to the national and international levels. The MDP strategic agenda can also be seen as a success in terms of decision-making. Insofar as strategic delta planning and agenda-setting are concerned, the MDP has certainly made inroads in aligning development policy agendas, although this result was not uniform or across the board. As such, it is not yet a finished process.
This convergence of agendas, however, was not achieved overnight. Having a sound problem analysis and a good story, or strategy, is not the only thing that matters. A dedicated political process is also required in which different actors, operating in different arenas of decision-making and acting at different stages of the process, are engaged in pursuit of the agenda, trying to merge the three streams.
While this analysis provides a deeper understanding of political agenda-setting following from strategic delta planning, it also presents some limitations. First, the sample size was limited and did not cover all actors involved in the Mekong Delta planning. Second, the interviewees did not represent all the involved institutions. However, the interview subjects were highly placed in the political system of Vietnam. They, moreover, played important roles in the MDP formulation process and in its uptake in policy processes, such as Decision 593, Resolution 120 and the annual Mekong Delta Forums. They can, therefore, be considered knowledgeable informants.
Our analysis confirmed the usefulness of applying the Multiple Streams Approach in the Vietnamese context, albeit with several modifications. In particular, we viewed the politics stream as focused on legislative turnover and political and institutional attention, taking into account the newness of the strategic delta planning approach in Vietnam. Additionally, the Multiple Streams Approach revealed the crucial role of agenda-setting as a step towards achieving consensus among actors. Given the importance of such consensus, especially regarding support for the strategic plan, the political process of agenda-setting merits further attention in future studies on strategic delta planning or strategic spatial planning. Further research is also recommended on the role of mobilizing power and resources, and the links between these and the sorts of agendas being prioritized for delta development. They could prove particularly influential in delta decision-making processes.

Note
The fourth MDF is a special Conference on Sustainable and Climate-Resilient Development of the Mekong Delta of Vietnam (This conference was a special meeting among high-level participants, chaired by the Prime Minister held in Can Tho city on 26-27 September 2017). It can also be considered the fourth Mekong Delta Forum, which has been held every year since 2013 (the first was MDEC).