Influence of different discourses on the outcomes of development policies and projects : Insights from water project implementation in Nepal

Powered by TCPDF (www.tcpdf.org) This material is protected by copyright and other intellectual property rights, and duplication or sale of all or part of any of the repository collections is not permitted, except that material may be duplicated by you for your research use or educational purposes in electronic or print form. You must obtain permission for any other use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone who is not an authorised user. Haapala, Juho; Keskinen, Marko


| INTRODUCTION
Politicians and decision-makers are interested in improving the ways in which their decisions are converted to social change. Finding appropriate approaches to effective policy implementation remains a major challenge to various development organizations and institutions (Crosby, 1996;DPF, 2015;Makinde, 2005;OECD, 2015;Rondinelli, 2013, pp. 5-6;United Nations Economic and Social Council, 2014;WWAP, 2015).
International development cooperation provides various instruments for policy implementation. Policy implementation and the approaches to development cooperation are thus fundamentally intertwined (Carothers & De Gramont, 2013;Duncan & Williams, 2012). The original, standard device for policy implementation in this context is a development project (De Haan, 2009;Rondinelli, 2013, p. 5). The conjecture is that project organization provides applicable, context-specific intervention tools, especially at operational levels (Rondinelli, 2013, p. 5;Sjöblom, Löfgren, & Godenhjelm, 2013).
Other instruments of development, such as sectoral, programmatic, and microcredit approaches, also become often "projectified" at the operational levels (Rondinelli, 2013, p. 6;Sjöblom et al., 2013, p. 3). From one perspective, development project organizations then function as simple implementation and innovation platforms for the policies. According to this view, external policies direct the ways in which development projects are allowed to operate (Ferguson, 1994;Rondinelli, 2013, pp. 5-6;Rossi, 2004). From a broader viewpoint, project organizations are not only mechanical devices of development, but their operation also depends on local, internal values and participation mechanisms (Cleaver, 2012;Haapala, Rautanen, White, Keskinen, & Varis, 2016;Tortajada, 2014). They promote embedded societal goals and reflect the characteristics of the surrounding society (Cleaver, 2012;Haapala et al., 2016;Tortajada, 2014).
Various tensions between directive policies, project implementation and local realities can be further characterized in the project context.
The first regards the fact that the policy discourses that direct development projects come top-down from variable national (originating either from recipient or donor governments) and international sources. Therefore, they may to a degree be discontinuous from each other (Rossi, 2004). The clearest example is the fuzzy interplay between the international, donor and national recipient policies and objectives regarding the discourses of development interventions (Hänninen, 2014).
Second, projects inevitably must make political, value-laden prioritizations in their every-day operations. However, project management is paradoxically mandated to negotiate and perform only as a politically neutral, implementing actor. Moreover, the directive policies do not necessarily reproduce effective, complementary ensembles at the operational level (Molenaers, Gagiano, & Renard, 2015). Some discourses, for instance, may become incompatible with each other in the context of a project, while others may become operationally marginalized under the pressure of more dominant discourses (Ferguson & Harman, 2015;Rossi, 2004).
Third, policies that work in one local context do not necessarily work in others (OECD, 2015, p. 3;Rondinelli, 2013, p. 16), making project management and application of directive policies always context specific. Aid delivery to the grassroots is a complex and delicate process (Rondinelli, 2013, p. i) that is embedded in local and global power-relations (Abdulai & Hulme, 2015;Cleaver, 2012). Top-down implementation across many cultural spaces and vertical administrative levels may pose inconsistencies between upper-level policies and local politics and realities. Development projects operate under pressure from both directions (Haapala et al., 2016;Haapala & White, 2015).
Fourth, an inevitable tension exists between project design, and local customs and realities (Cleaver, 2012;Haapala et al., 2016). Projects cannot therefore always produce the intended outcomes at the local level by following the project design (Haapala et al., 2016). Moreover, the ability of project management to control the local outset of a project is often limited (Rondinelli, 2013, pp. 17-18) as the local realities and the operational environment greatly influence project outcomes. These four aspects emphasize the central role of successful implementation conducted through project organizations.
Despite the challenges indicated above, little academic research has been conducted on the policy implementation process within development project organizations. To our knowledge, it remains largely unstudied how project management absorbs and addresses the wide variety of development policies and practices, as well as accounts for local realities, within its internal project operation. Equally understudied are the ways in which these policies are transformed into politically informed development outcomes that make the policy have impact.
This article presents a case study that deciphers the ways in which such an analysis can be conducted. The method developed for and applied in this study portrays the ways in which various policy discourses appear and interrelate within project operations. A rural water resources management project operating in Nepal is used as a demonstrative case study.

| Case study context
The case study water project operates in remote, rural Nepal. Fourfifths of the Nepalese population live in rural areas (Central Bureau of Statistics, 2012). Average income per capita (GNI PPP) is 2,500 US dollars (USD), ranking 189th out of 210 listed countries (World Bank, 2016a). In 2010, almost half of the population lived on less than 3.10 USD a day, and 15% less than 1.90 USD a day (World Bank, 2016b).
The least developed areas are located in the most remote and typically also water-scarce rural regions of the country. These localities are characterized by extreme remoteness, rugged and steep terrain, difficult access to villages and markets, and challenging climatic conditions, including severe droughts. Village water supply is often nonexistent, resulting in extensive water-fetching times for local women. Livelihoods rely largely on simple irrigated and rain-fed agriculture on steep mountain slopes. People suffer from malnourishment, inadequate hygiene, and lack of access to basic health care and any advanced education, among various other development challenges.
One important way to address these challenges is through sustainable water resources management and water infrastructure development. Such activities contribute to very basic prerequisites of a better life, such as access to safe drinking water, improved nourishment and livelihoods, enhanced personal health and hygiene, and access to electricity. Before the devastating earthquake in 2015, the government of Nepal targeted achieving universal coverage of basic water supply and sanitation services to its citizens by 2017. The estimated countrywide water supply coverage in Nepal was 84%, but only about onequarter of the water supply schemes were fully functional (UNICEF, 2014, p. 20). Lack of long-term sustainability has long remained a prominent challenge for water sector stakeholders (Bhandari & Grant, 2007).  Possible tensions in the implementation of international development projects. Projects operate in fundamentally political surroundings, between external, directive policies and local realities. Four different types of tensions, described in the text, are marked with black circles. Source: the authors.
2010, Phase II from 2010 to 2016, and Phase III since 2017. The study was conducted during the second phase. RVWRMP is currently the largest rural water-sector project in the country, serving hundreds of thousands of beneficiaries.
The scope of the project is in broad terms water resources development. In operation, it translates to construction of community water supply, micro hydropower and irrigation schemes, dry toilets, green houses and home gardens; and to establishment of community institutions, groups and cooperatives; training and education provided to community people, among many other activities. Human rights, gender equality and social inclusion values are central to all project practices in the field. The broader policy context of the project is described in section 3.1, and the key project activities and discourses are explained more in detail in Appendix 1, as parts of the analysis.

| Characteristics of project personnel and key informants
The project personnel consist of 250 trained individuals. They can be divided roughly into District personnel and the core team. The District personnel constitute most of the project staff, 220-230 individuals. This study focuses on the core team as they hold key positions in terms of shaping the project practices and recreating the project documents and procedures. They guide the operations of the project, and communicate with government and donor representatives, and other key stakeholders. The specialists have a long history of working together as a collective. They are fairly fluent in English, and relatively well aware of international discourses, national legislation and policies, donor and recipient government interests, the political environment, sectoral discourses and best practice. They therefore work at the focal point in which the external, directive policy discourses become organized into functional implementation practices within the project. We therefore postulate that the hypothesized gaps between the informant insights and actual operations remain negligible, and that the informants have a concrete link to the implementation reality and project discourses.

| Analysis method 2.3.1 | Research approach
The analysis method depicts interactions and hierarchies between the identified key project discourses as perceived by the expert key informants. It relies heavily on their viewpoints, participation, knowledge and experience. The method is therefore characterized by participatory, constructivist research. The study has a phenomenographic research orientation (e.g., Svensson, 1997). The orientation scrutinizes the ways in which examinees describe a certain phenomenon (Marton, 1981(Marton, , 1988. It is generally applied for analyzing the differences and similarities in people's ways of understanding and experiencing the world. Furthermore, the analysis is based on an understanding of the study context. It is founded on our field presence, observations and understanding of the specific context that allows constant triangulation with the insights of key informants through the analysis. This contextual knowledge provides an informed identification of project discourses for the analysis, and interpretations of the results. The research is founded on an extensive 5-month presence at the PSU and in the field, including observations, formal and informal interactions with the project personnel and the project's beneficiaries, and participation in several meetings and events. As a key event for this analysis, the first author arranged a project discourse workshop that introduced the main exercise for the personnel (see analysis phase 3), and a feedback session that reported the analysis results to the project personnel (phase 5). The triangulation also involves a literature review on both official, published project documents and internal, unpublished field reports.
The first phase identifies the policy context of the project. The national, international and donor policies, and sectoral practices that steer the project are specified based on policy and project documents and observations in cooperation with project personnel. This analysis phase is based mainly on documentary evidence.
The second phase identifies the key internal project discourses.
The identification is based on the lead author's observations on project management and field operation as well as on daily interactions with project personnel, participation in the day-to-day project operations and meetings, and on a literature review of the project documents over the 5-month period. The criteria for identification of key discourses is the prevalence of the stated concepts in day-to-day activities, in communication between the personnel and beneficiaries, and in project materials.
The third phase consists of a project discourse workshop that documents the key informants' insights about the identified project discourses. Here the informants individually choose a few of the identified project discourses for further scrutiny as per their own interests and field of expertise. They indicate their views about the strongest unilateral and reciprocal influences between each chosen "focus discourse" and the other identified discourses. This is done by drawing arrows to the response sheet (illustrated in the upper part of Figure 5).
While drawing, the informants simultaneously explain the rationale behind their views to the facilitating researcher, as a check. The exercise is blinded; the informants respond to the exercise individually, without knowledge of others' responses. The final outcome of the analysis is realized after compiling all the responses. This allowed us to minimize the researcher's influence on the exercise results. The Project core team's perspective on project management: translating external policies and model practices into project discourses and operation. Source: the authors. Five phases of analysis; see main text for detailed explanations. Source: the authors.
analysis results can thus be seen as the individual informants' "collective mind-set" of the discursive interrelations of the project.
The fourth phase visualizes the analysis results according to the responses compiled. It reveals the interrelations between the identified project discourses, emerging hierarchies between them, and possible clusters of interconnected discourses. The results emerge directly from the exercise compiled responses, without further interpretation by the researcher. "Influence ratio" has a crucial role in phase 4. It deciphers the ability of a discourse to influence other discourses (or being influenced by them), forming a relative hierarchy between each one.
Visually, the ratio describes the proportion of departing to arriving influence for each project discourse in the visualization of the results.
In the fifth phase, the study results are again reflected with the PSU specialists in a common meeting. Triangulated with field observations and a review of project modality documents, it increases the reliability and validity of the method and results. The feedback ensures that the researcher's interpretation of the specific case study context is valid and the conduct of the method and analysis is acceptable for the key informants.   The main policies, mutual agreements and sectoral practices that direct the internal project discourses. Source: the authors.

| Identification of key project discourses (Phase 2)
We concluded that the project's operations and related discourses could be summarized by a selection of nine key discourses, presented in random order in Table 1. Appendix 1 describes how the identified key project discourses link with the observed procedures, field activities and project documents. These key discourses remained at the core of the operational works, were often used in the day-to-day communication within the project, and were also evident in the project documents and materials. Many of the identified discourses were combinations of multiple concepts to better meet with the broader working areas of the project, to capture their observed interrelatedness in practice and to increase the simplicity of the analysis.
The water resources management discourses used by the project, namely IWRM and Multiple Use Water Services, were implemented through rural development-related discourses during day-to-day operations. They were hardly ever mentioned by the personnel in their dayto-day communication, nor in project implementation practices (see section above). Therefore, the analysis considered IWRM and Multiple Use Water Services as implicit discourses rather than independent, explicit key discourses that were chosen to provide closer scrutiny.

| Compilation of exercise results and their visualization (phases 3-4)
Phases 3 and 4 considered the ways in which the identified project discourses appeared and interrelated within the project according to the project's personnel. At this stage, the identified project discourses were placed in a hierarchy as per the compiled responses of key informants. The specialist informants, working together to address common operative problems, evidently had a quite codirectional collective mindset and well-informed understanding of the project discourses.
Compilation of the responses indicated that the nine project discourses formed three distinct groups, based on their relative ability to influence other discourses. This hierarchy was formed by determining the influence ratio for each project discourse (Table 2): From the rightmost column of Table 2 we see that the "foundational discourses" (Group 1) had the greatest influence. This relatively large number implied that these discourses greatly influenced the other discourses but were not greatly influenced by them. The PSU specialists' explanations during the workshop, and their feedback on the results (see the fifth phase), indicated that these discourses evidently set the stage for the operational project activities by providing the project personnel with underlying, procedural frames and collective values for project operations and practices. These interactions thus provided foundations for the other discourses.
The operational discourse group (Group 2) comprised five discourses. The corresponding influence ratios were balanced around a value 1 (Table 2). This implied that these discourses were both influenced by some discourses (here mainly the foundational discourses), and that they correspondingly influenced some others (especially "Poverty Reduction"). Again, the specialists' explanations and feedback indicated that these discourses actually represented the standard operational practices of the project, providing the practical means to achieve the desired project outcomes (especially that of poverty reduction).
"Poverty Reduction" made up the third discourse group. It differed from the other project discourses as it was heavily influenced by them but had no significant influence on the others. Correspondingly, the key informants perceived "Poverty Reduction" as the eventual objective of the project operations to be achieved by operating through subordinate project discourses. This matched the stated purpose of RVWRMP II, namely "to achieve improved well-being and reduced poverty through sustainable management of their water resources" (RVWRMP, 2011a). The idea of the distinct nature of "Poverty Reduction" versus the other project discourses came up in many discussions with project informants, as well as in the feedback session (phase 5).
To visualize the relationship between project discourses and to take the analysis a step further, the set of diagrams in Figure 6 were The nine key project discourses identified. Appendix 1 provides detailed discourse descriptions.

No
Name of project discourse  in the lower part of Figure 6 as per their relative ability to influence the other discourses, based on the influence ratio.
The arrows departing from "Participation and Community Management" and "Institutional Capacity Building" (both from the foundational discourse group) indicate that these discourses had the strongest unilateral influences on the others. In particular, they affected three other discourses (that were part of the operational group), namely "Health and Hygiene", "Livelihoods Development" and "Sustainable Scheme O&M." The also had an influence on "Poverty Reduction." This supported the informants' view that the foundational discourses set the stage for the operational project activities by providing the project personnel with the underlying, collective procedural frames and values. Figure 6 also exposes an additional layer of interaction between the project discourses. "Discourse clusters," shown as triangles of reciprocal interactions in Figure 6, demonstrate that some discourses were explicitly strongly interconnected with each other. One discourse cluster is formed around the foundational group ("Participation and Community Management," "Institutional Capacity Building," and "GESI and HRBA"), while two partially overlapping clusters are found around the functional discourses ("Environmental Protection and Climate Change Adaptation," "Food Security and Nutrition," "Health and Hygiene," and "Livelihoods Development"). Such interconnected clusters of discourses were formed within the identified discourse groups, Top: compiled key informant responses on the response sheet. Bottom: emerging discourse hierarchy as per the responses. Source: the authors.
whereas the most remarkable unilateral influences remained mainly intermediate to the three groups. This further highlighted the different perceived roles of the discourse groups and the clusters within project management and operations. These ideas are further interpreted from an organization management viewpoint in Section 4.1.

| Reflective feedback (phase 5)
The lead author arranged a workshop for the PSU specialists so as to present the method of analysis, process and results. The feedback ensured that the specialists considered the research design to be acceptable and the method sound.
The specialists considered the results to be rational and reliable,

| Reflections on case study analysis
The Introduction describes various tensions that project management encounter and must consider between the policies, internal discourse and realities at grass roots level. The developed method can help organizations to address these tensions within the project operations.
Organization management can then purposefully utilize awareness about the evident discourse hierarchy and clusters during project planning, team formation, monitoring and supervising the operations, and in finding operational synergies between the discourses and corresponding field activities. Project management is able to direct the operations by harmonizing and recreating the internal discourse hierarchies to better meet external policy requirements and local realities, or to improve the internal operability and efficiency of the organization.
Our analysis has several implications for project management. The first advantage of the developed method is the identification of discourse hierarchies: it can examine whether the actual project discourse hierarchy differs from the stated project aims and objectives. In the case analyzed, for instance, it is important to note that the project personnel conceived three distinctly different types of project discourses.
This translates into the project discourses having distinct roles, and that they reflect well the stated project objectives (Section 3.4 in particular). This awareness could be used in project design, planning and monitoring of operations.
Second, the analysis depicted specific discourse clusters. The clusters imply that some project discourses have partially overlapping roles and mutual synergies. The clusters were in our case formed within the discourse groups, rather than between them. This indicates the distinct role of each discourse group, and reflects the existing synergies and interlinkages among the discourses in the same cluster. This could mean that the field activities related to the discourses in the same cluster may benefit from each other. This awareness could be used in the formation of field teams and planning of activities.
The study revealed possibilities for further research.

| Comparisons with related methods
The method developed here partially resembles Q-methodology (Herrington & Coogan, 2011;Zabala, 2014) and the idea of fuzzy cognitive mapping (e.g., Gray et al., 2015;Groumpos, 2010;Kosko, 1986;Stach et al., 2010). These methods share the same phenomenographic starting point, and the first two analysis phases resemble the statement identification phases of the Q-method. However, unlike the Qmethod, the described method does not aim to grasp the variety of viewpoints that exist, nor to sum them in a synthesis. Rather it aims to understand the ways in which the discourses interact within a certain setting. The methodology described in this paper does not use distinct statistical analysis, but relies on the construed expert knowledge of key informants.
The introduced method and the fuzzy cognitive mapping methodology are similar in that they both look at the interconnectivity of certain concepts and encourage the use of expert knowledge and participation (cf. Gray et al., 2015;Stach et al., 2010). In the developed methodology, the fuzziness comes from the different aspects and opinions within the collective mindset of the core project personnel about project functions. However, it not aim to produce a functional causal system or a model about the discourse system functions. We consider the qualitative content of the identified influences as being far too complex and multidimensional to be evaluated or modeled reliably and specifically as a whole. Another difference is the developed method's strong emphasis on contextual understanding and field observations. We believe that this makes it more suitable to analyses in expert organization and project management contexts, allowing continuous triangulation with operational reality. However, the third and fourth analysis phases, in particular, could be described as a participatory application of fuzzy cognitive mapping methodology principles.

| CONCLUSIONS
The aim of this study was the development and description of a participatory research method for unfolding the ways in which policy discourses interrelate within project organizations. Our internal discourse analysis helps decision-makers and project managers to understand the ways in which the various steering policies mould and interact within the implementing organization. This is important to provide more effective policy impact and improved contextual understanding of implementation processes.
The conducted case study analysis demonstrated the ways in which development project personnel transformed the externally imposed policy discourses into an internal, operative project discourse hierarchy. The method visualized the ways in which the key project discourses interacted with each other within project operations, leading to distinct hierarchical groups and clusters of discourses. The clusters indicated partially overlapping roles and strong mutual synergies.
The internal discourse hierarchy steered the observed project practice.
The method has several concrete contributions to improving internal project management: the analysis enables visualizations of positions and interrelations of the key discourses within organizations in a transparent way. It enables comparisons between the discourses, and between stated objectives versus the observed discourse hierarchy, and also helps to recognize mutual synergies or inconsistencies among the discourses within organization practice.
Finally, the participatory analysis methods, including workshops and interviews, help the project personnel to reflect the different roles and relations that different discourses have in their own work, and how they correspond to other staff members and project documents.
At the same time, the possible differences in interpretations between project personnel highlight the potential ambiguity related to the project's stated objectives and actual practices. This paper discusses the ways in which organizations could utilize the analysis in planning, team formation, monitoring and supervision.
This will allow implementing managers to meet with external policy requirements, fit with local realities, and improve policy and project outcomes.

Gender Equality and Social Inclusion, and Human Rights Based
Approach (GESI & HRBA). GESI referred to the Gender Equality and Social Inclusion Strategy and Action Plan (RVWRMP, 2011a(RVWRMP, , pp. 38-39, 2015 with emphasis on HRBA (RVWRMP, 2015). It was based on the recognition that "inclusive targeting" is required for disadvantaged groups, such as women, the poor and people with disabilities to gain equitable access to resources and opportunities (RVWRMP, 2015), and to achieve the "improved quality of life" project target (RVWRMP, 2011a). Related project activities included empowering disadvantaged groups to involve and to take lead roles in project activities; giving preference to the disadvantaged in all training opportunities, awareness-raising campaigns, rallies, door-to-door visits, community meetings and interactions, and collaboration with local administration. HRBA referred to "the norms, principles, standards and goals of the international human rights system into development plans and processes" (RVWRMP, 2015, p. 5). It was closely linked to national and international policy context, discussed in Section 3.1, especially with the Finnish guidelines to development cooperation (MFAF, 2012). GESI and HRBA were mainly used among project specialists, but less so in the field. 6. "Environmental Protection and Climate Change Adaptation." The first part of the concept referred to the recognition that sustainable human wellbeing is founded on the natural environment.
Activities related to environmental protection involved microlevel watershed management, including various types of source protection activities, low-impact infrastructure and technology, construction of water preservation infrastructure, and soil management and erosion control pilots. Enhancing environmental conditions was also explicitly emphasized in the overall project objective (RVWRMP, 2011a, p. 27). The "Climate Change Adaptation" part of the discourse referred to activities that can be considered to contribute to adaptation to climate change by reducing vulnerability and increase resilience to disasters and adverse impacts to climate change. The project goals included empowering people to manage their own water supply. Related activities involved the application of IWRM, promoting efficient use of water, prioritizing multiuse systems, environmental and soil protection activities, and low-input agriculture. Operational tools included the Step-by-step Manual (RVWRMP, 2012b) and Water Safety Plans. 7. "Health and Hygiene." Health referred to the direct project activities regarding overall physical, mental and social human wellbeing. It was manifested in the water project context most closely through food security, nutrition, and hygiene and sanitation issues. Hygiene, in turn, referred to sanitation-and hygienerelated behavior, infrastructure developments and thematic community groups. Related operational practices involved various training at multiple levels, Open Defecation Free and total sanitation campaigns, Water Use Master Plans (RVWRMP, 2011b), water safety planning and implementation, other institutional capacity-building activities such as establishment of community groups, and construction of dry toilets for every household in project localities. The main guiding document in the project context was the Nepal National Sanitation and Hygiene Master Plan (GoN, 2011). The project document referred to health in the project objective as "enhanced quality of life" and in the project purpose as improved "well-being" (RVWRMP, 2011a, p. 27).
Health issues were also among the expected results of the project (RVWRMP, 2011a, p. 29). In practice, health and hygiene improvements were the most often discussed themes in the field, considered an approachable theme by both project personnel and the beneficiaries.
8. "Food Security and Nutrition" referred to the project activities directed towards the availability of, access to and use of nutritious food for healthy life. In the project context, food insecurity and malnutrition were clear problems in multiple areas. These problems originated particularly from an inadequate variety of cultivated plants, remoteness and challenging environmental and climatic conditions, plus gender-related discrimination and various food taboos (see Haapala & White, 2015). These led to problems such as insufficient calorie intake and suboptimal feeding patters.
Related project activities included awareness campaigns, GESI activities, promotion of home gardening, dietary demonstration from locally available foods, IWRM, activities related to health, sanitation and hygiene, Open Defecation Free campaigns, training of farmers and irrigation infrastructure. The project objectives (RVWRMP, 2011a) indirectly referred to food security and nutrition, and the document stated "improved and sustainable nutrition, food security" (p. 29) as an expected result.