Abstract
In donor aid programs, the claims of causal links between inputs and outputs are crucial to establish the effectiveness of aid. As a result, there has been ample research on the degree and direction of correlation and causality between aid and poverty reduction. While evaluating end outcomes has its merits, this article aims to assess how some donor aid programs come under early criticism and/or are dropped or modified in order to ensure their continuation. This article examines a tender award dispute in a forestry program in Nepal, and traces the causal contribution of actors and actions that obstructed and changed the decisions related to the program implementation plan of this multi-million-dollar initiative. The article employs the process tracing method to search, collect, and assess evidence on the tender dispute case of the selected program. By tracing the logical sequence of evidence, this study establishes causality, linking the hypothesized causes and their effects to explain how the tender award decision that was announced by the lead donor and endorsed by the Government of Nepal was eventually nullified.
Keywords
Introduction
All development programs have the ultimate goal to improve the quality of life of a targeted population. One of the key concerns for the stakeholders, therefore, is whether the program has achieved its desired and targeted results. As a result, there has been ample research on the degree and direction of correlation and causality between aid and its impacts both in the development sector as well as in academia.
While measuring results has its merits, surprisingly little attention has been paid to analyzing key events in a development project’s lifespan that alter the projects’ plans, design, and/or implementation strategies. It must be noted, however, that increasingly donor organizations and aid agencies have been promoting flexible and adaptive approaches for project management to account for unplanned yet inevitable events in a project timeline (Asian Development Bank, 2012; World Bank, 2017). The Asian Development Bank (2012) points out that without an openness and a systematic approach to adjustments due to unforeseen difficulties, projects, especially those operating in fragile states, are likely to be delayed or even cancelled. Despite this increasing emphasis on the projects being adaptive and responsive to changes due to unforeseen events, there has been inadequate attention to understanding the causal mechanisms in these key events and how they unfold in specific contexts.
In the social sciences, there is an increasing interest in the process tracing (PT) method to investigate within-case inferences in single case study or unique interventions. This is because the PT method has been proposed as a markedly suitable approach to improve rigor and transparency in causal attribution and claims. It offers a Bayesian approach to evidence search, collection, and assessment. However, the application of this method is still limited (Busetti and Dente, 2017). This study has identified a tender dispute event in a forestry program in Nepal as a fitting candidate for the application of the PT method and to test its usefulness in collecting and assessing evidence for causal claims.
This study recognizes that there is a great deal of information in these key events that alters approved project designs and plans; and that parsing out the causal connections of actions and events provides valuable information to open the black box of causality, thereby helping improve future performance. By applying the PT method to the tender decision controversy in a forestry project in Nepal, this article examines the actors and their activities that led to a project deadlock and the eventual revisions made on the decisions to move forward.
Brief information on the forestry sector in Nepal
Nepal’s forestry sector operates in a complex environment with many stakeholders that include but are not limited to the formal bureaucracy, international development agencies, local forest user groups, professional associations, federations of user groups, NGOs, and the private sector. Scholars argue that the forestry sector in Nepal serves as a key site where struggles for resource governance are manifested and determined through contestations between techno-bureaucratic control, community rights for resources, the interests of international development partners, and ongoing socio-political struggles (Nightingale, 2017; Ojha, 2006, Timsina and Gotame, 2014).
In recent times, forestry in Nepal has witnessed increasing calls for wider stakeholder engagement in decision making processes. This is attributed to the presence of vibrant non-state actors alongside a powerful state (Luintel, 2006; Ojha, 2006; Timsina and Gotame, 2014), and further to the influence of an emerging global agenda of multi-stakeholder collaboration for effective resource governance (European Commission, 2007, 2011).
In line with this emerging trend, the Government of Nepal with support from its long-time development partners, the Governments of Finland, Switzerland, and the UK, initiated a policy-level experiment in January 2012. A new forestry program was planned that was to be steered by multiple forestry stakeholders from planning to the implementation stage. This program was named the Multi-Stakeholder Forestry Program (MSFP).
The Multi-Stakeholder Forestry Program
The MSFP was by far the most ambitious forestry program in Nepal, with a financial commitment of US$150 million for 10 years. The overarching aim of the program was donor-aid harmonization, institutionalization of multiple stakeholder collaboration in decision-making processes, and governance reform in the forest sector. The stakeholders included central government, local governments, civil society groups, and the private sector – representing their constituents in a national steering committee to steer the program and to contribute to policy making.
Right from its conceptualization, the MSFP was closely aligned with the international development agendas as guided by the Millennium Development Goals and the aid dissemination modalities under the 2005 Paris Declaration for Aid Effectiveness.
This is reflected in most aspects of the program’s design, outputs, outcomes, and implementation procedures. Some key examples include the program’s main focus on contributing to the national goals of poverty reduction through forestry intervention contributing directly to Millennium Development Goals 1 and 7 (MSFP, 2012). Additionally, the focus on harmonizing donors’ support echoes the fundamental principle of the Paris Aid Declaration that stresses donor-aid harmonization to reduce transaction costs and to ensure aid effectiveness (MSFP, 2012; OECD, n.d.). One key program implementation innovation was the formation of the first of its kind multi-stakeholder steering committee to provide strategic direction to the program and to contribute to inclusive policy process. This echoes the OECD transformative aid agenda to create country ownership and co-responsibility among national stakeholders (European Commission, 2011).
Despite adopting the acclaimed approach of stakeholder collaboration for resource governance and a generous financial aid commitment, surprisingly the program terminated early without transitioning to the second phase. This article, however, will not discuss the reasons for its early termination. This study is focused on explaining causal mechanisms in a specific event that caused a huge setback in the program’s early phase. This event was about a controversy related to the tender award decision that delayed the program implementation by nine months.
The event under investigation: Tender award decision
In July 2011, the lead donor, the Swiss Agency for Development and Cooperation (SDC), 1 on behalf of all three donors, released a Call for Proposals from interested and eligible national implementing agencies for implementing the first phase of the program in 18 selected districts of Nepal. On March 15, 2012, the tender award decision was announced. A national non-profit, Rupantaran Nepal, was selected as the awardee for implementing the first phase of the MSFP. Soon after, the decision was opposed and disputed by several stakeholder groups. A four month-long controversy ensued, attracting wider stakeholder debate and coverage by the news media.
Two contesting groups formed for and against the decision. The first group consisted of representatives from the three donors and key officials involved as members in the selection committee representing the Ministry of Forests. This group is referred to as the tender-defending group in this article.
The second group consisted of non-profit organizations engaged in forestry development and activism. In this case, this group was represented by the Federation of Community Forestry User Groups Nepal (FECOFUN), the Federation of NGOs (NGO Federation), and professional organizations including Nepal Forester’s Association (NFA). This group is referred to as the tender-opposing group in this article.
Claims of the tender-opposing and tender-defending groups
The opposition group objected to the tender award decision, accusing the selection committee of a biased selection process. They claimed suspicious alignment of the selection committee with the awarded NGO, a lack of transparency in the process, and unfair advantage to Rupantaran Nepal, the NGO that was ultimately awarded the bid. Some quotes in the daily newspaper by officials from various opposing parties illustrate this further: ‘…the lack of transparency and irregularities in the bid’s selection process of MSFP has left us surprised’ (General Secretary, FECOFUN, in Pandit, 2012a); ‘The tender decision was a result of a pre-planned design that came without holding consultations with partners concerned, including the civil society’ (President, NFA, in Shahi, 2012b).
The defenders refuted the accusations, stating that the decision process followed the standard procurement procedures of the Embassy of Switzerland in Nepal (the lead donor) and adding that the selection committee comprised the donors and representatives from the Ministry of Forests: ‘Tenders have been called (per notice of 8th July 2011) and information related to the selection process and criteria for the assessment was provided along with the tender document. The selection will follow the procurement procedures of the Embassy’ (Embassy of Switzerland in Nepal, 2012).
After four months, the SDC cancelled the tender outcome in July 2012 and announced a new tender process.
Study objective
Through application of the PT method to the tender decision controversy in a forestry program in Nepal, this article aims to establish that key events, especially those that obstruct an approved project plan and design, provide a wealth of information such as stakeholder interests and negotiation strategies. And discounting deep investigation on these key events will result in lost opportunities to learn and to improve upon future processes. The PT method promises to enhance rigor in within-case inferences and single case studies. The general objective of this research is to apply the PT method to add rigor and transparency for drawing inferences and causal connections to investigate a key event in a large multi-stakeholder development project in Nepal.
There exists a general understanding that the tender decision in the MSFP was revoked due to stakeholder dispute. However, how this outcome came about – for instance, how the stakeholders organized and/or negotiated to contest – is not assessed in depth.
The specific purpose of this article is to employ the process-tracing method in order to search, collect, and assess evidence on how the tender dispute resulted in its cancellation. By tracing the logical sequence of evidence, this study aims to establish causality, linking the hypothesized causes and effects to explain how the tender award decision that was announced by the lead donor and endorsed by the Government of Nepal was eventually nullified.
Theoretical underpinnings
An important methodological entry point in studying dispute resolution processes related to resource governance is to acknowledge that there exists potential for the domination of some parties by others and that opening spaces for democratic processes offers a platform to challenge such power imbalances between actors. In the case of forestry in Nepal, the advent of community-based forestry in the 1980s and the national political movements of 1990 and 2006 opened spaces for civil society groups to demand space for engagement in decision-making processes (Luintel, 2006; Ojha, 2006; Ojha et al., 2007). As a result, collaboration between state and non-state actors for inclusive governance has started to feature prominently in policy discourses.
The tender decision dispute in the MSFP offers a distinct case to understand how actors in arguably weaker power positions can still block and/or severely delay implementation plans in collaborative settings characterized by differential power sharing arrangements. Existing research suggests that collaboration can be cultivated in power asymmetry and can function well if it is accepted socially as reasonable (Ran and Qi, 2018). Purdy (2012) adds that collaborative governance in itself is an arena for social construction of power. Ran and Qi (2018) argue that the shared power that is socially constructed and culturally patterned may be more beneficial in specific collaborative arrangements.
Following this, the article starts with a premise that the aim in collaborative settings is not an equal share of power, but to arrive at a situation or an understanding where the power imbalance is acceptable to all stakeholders. Here, power is conceptualized broadly not just to include coercive power but rather as a resource that is relational, productive, and transferable (Ahlborg and Nightingale, 2018), indicating that power can be expanded, diminished, or transferred over time (Purdy, 2012).
Theories on advocacy networks suggest that actors use strategies to frame issues in order to make them comprehensive to target audiences, to attract wider public attention, and to encourage action (Falleti and Lynch, 2009). This is considered as a strategy to broaden legitimacy. For instance, Purdy (2012) refers to the power of discursive legitimacy that is drawn from raising issues that are linked to societally important ideals. Parties who exercise this strategy generate their power from the status of the values or the logic they represent.
Scholars of network politics agree that a popular strategy used by networks to generate attention to new issues is to leverage media as essential partners to help set agendas, and provoke further media attention, debates, hearings, and meetings (e.g. Shanahan et al., 2017; Stone, 1989). Another strategy is that of searching for a more receptive political venue (Baumgartner and Jones, 1991).
According to Stone (1989), in order to campaign for an issue, the actors must convert it into a causal story of establishing who bears the responsibility and the burden of policy change. Falleti and Lynch (2009) add that in addition to the causal story, the causal chain in that story needs to be sufficiently short and clear to make a convincing case about responsibility and guilt. Shanahan et al. (2017) claim that not only will advocacy groups work on self-presentation (identifying themselves as the losing group and others as the winning group), but the losing group will necessarily engage in strategies that aim to expand the scope of conflict such that they will project as if the costs are diffused and benefits are concentrated to a few. Conversely, the winning groups will try to contain the issue to the status quo (e.g. concentrating costs and diffusing benefits).
The MSFP tender dispute captures the differential power positions of actors in forestry governance in Nepal. It offers a fitting case to analyze how actors in varying power positions strategize to negotiate for a decision in their favor.
Methodology
Process tracing: A brief introduction
Process tracing (PT) is a tool of qualitative analysis for drawing within-case inferences in single case studies (Beach and Pederson, 2013; Bennett and George, 2005; Collier, 2011). While there is no single framework to conduct PT case studies, the key variant that is standard to all PT analysis is the identification of the causal mechanisms in the process that lead to certain outcomes. Causal mechanisms are referred to as an individually necessary element of a whole that connects causes to outcomes. Beach and Pederson (2013) argue that causal mechanisms are embedded in a deterministic understanding of causality, as opposed to a probabilistic approach.
The key difference between quantitative statistical studies and the qualitative PT method is that in the latter the researcher is interested not in the magnitude of the causal effect but in whether the hypothesized causal mechanism was present or absent in the evidence collected and how a causal mechanism contributed to producing the outcome.
Key methodological tools for conducting PT-based studies are theory-testing, theory-building, and case-centric. All three follow the Bayesian logic of inference to determine the presence and/or absence of causal mechanisms. This article uses the case-centric methodology to determine the underlying causal mechanism that accounted for the outcome.
2
Beach and Pederson (2013) recommend a three-step procedure to make causal inferences about observed effects using the PT method: -Generate a clear hypothesis of how an outcome was caused. -Predict pieces of evidence (observable manifestations) that we should expect to find for each part of the mechanism if the mechanism is present in the case. -Pass the empirical tests: this refers to the probative value of the piece of evidence based on certainty and uniqueness. Certainty illustrates what evidence must appear in the empirical record if the causal mechanism is present. High certainty indicates that the evidence should be there if the hypothesis is true, and low certainty indicates that the evidence is not important. Uniqueness illustrates what evidence must be found that cannot be explained with an alternative hypothesis. High uniqueness indicates that the evidence is specific to the hypothesis and unexplained by alternative hypotheses, and low uniqueness means that there are other explanations for finding the piece of evidence (Busetti and Dente, 2017; Mahoney, 2012). Literature has focused on four tests based on a belief in hypothesis or a Bayesian logic, namely the straw in the wind, hoop, smoking gun, and doubly decisive tests (Beach and Pedersen, 2013; Bennett and Checkel, 2012; Collier, 2011). These tests vary in terms of their inferential leverage. For instance, the hoop test is necessary but not sufficient for the validity of a given hypothesis. This enhances the disconfirmatory power of the evidence. The smoking gun test is sufficient but not necessary for the validity of a given hypothesis. This enhances the confirmatory power of evidence, as finding such evidence greatly increases confidence in the hypothesis, but not finding it has only limited disconfirmatory value.
In summary, our ability to update our confidence of the causal mechanism depends on combining two sources of reasonings: the existence of theoretical literature to link the cause and outcome (referred to as the theoretical prior 3 ), and the probative value of the observed evidence based on certainty and uniqueness. This reasoning is used in the analysis of the selected case in the discussion section of this article.
Data collection
This research employed a qualitative method for data collection, analysis, and interpretation of the first tender award and cancellation process for the implementation of the MSFP in Nepal. The first part of data collection and analysis consisted of an in-depth review of program-related documents. This included the common program document, program implementation plan, briefing notes, program inception report, and other relevant program materials. This was followed by a systematic review of newspaper coverage on the program from March 2012 to December 2012. These dates were selected as the announcement of the bid winner was made on March 15, 2012, after which the controversy started. The tender process was cancelled on July 27, 2012. Two English-language national dailies were selected, with the highest and second highest circulation rates for English newspapers in Nepal. Four Nepali newspapers were also reviewed – their selection was based on suggestions from program staff members who reflected on which newspapers best covered the tender issue. The national newspapers covered the tender controversy 20 times in the four months following the award decision. All newspaper coverage in Nepali was translated into English.
Coding was done manually as a way of organizing data and scanning for preliminary ideas and themes. This initial coding was followed by a focused coding strategy (Lofland et al., 2006) as informed by the analytical constructs of the process tracing method. In other words, the focused coding entailed coding data into categories of observable manifestations (evidence) for the four tests, straw in the wind, hoop, smoking gun and doubly decisive, of the process tracing method. I acknowledge that given the qualitative nature of data analysis, the underlying meanings of the text, as interpreted by the researcher, are inherently subjective.
Data analysis
Following Beach and Pederson’s (2013) recommendation of the three-step procedure to make causal inferences using the PT method, I start by generating a hypothesis of how the outcome (cancellation of tender award) occurred. The hypothesis illustrates the predicted activities of the entities in each part of the process.
Step 1: Generating the hypothesis
H1: The tender-opposing group opposed the tender decision, claiming favoritism and lack of transparency. Further, they partnered with media to draw wider attention to this case. The negative press linked the case to transparency issues in foreign aid. This raised reputational risks for the donors and the Government of Nepal, resulting in the cancellation of the tender award.
The hypothesis on entities and their activities is illustrated in Table 1 as a four-part process leading to the cancellation of the tender decision.
Conceptualization of a four-part tender award dispute.
Step 2: Identifying observable manifestations
In this step, I predict the evidence (observable manifestations) that is expected to be found for each part of the mechanism if the causal mechanism is present in the case.
For part one of the hypothesis above (Table 1), official statements by the tender-opposing group reflecting negatively on the bidding process could serve as the potential evidence. For part two, the potential evidence could include self-framing by the tender-opposing group as actors struggling to protect the interests of society from the powerful groups that have historically resorted to abuse of power and exploitation of state resources for personal benefits. News media statements with accusations and references of the abuse of power could provide valuable evidence to support this claim. Similarly, for part three, the news media coverage portraying the tender-defending group in a negative light could serve as evidence. Lastly, for part four, internal memos and reports as a response to the negative press coverage by the tender-defending group could serve as a useful source of evidence.
Step 3: Passing the empirical tests
The first step is to determine the prior confidence in H1, which is followed by tracing sequential evidence to test whether this prior confidence can be updated to strengthen the confidence in H1.
H1 proposes that the tender-opposing group claimed favoritism and lack of transparency in the decision process, and partnered with media to draw negative press on the issue, thus garnering wider public support. This eventually compelled the tender-awarding group to cancel the decision.
Existing literature determines the prior confidence in H1 as moderate based on two major ideas. The first is that even though the main actors in the tender-opposing group, the civil society organization, are considered as vibrant watchdogs effective in advocacy, awareness-raising for development, and acting as a bridge to link policies and practices (Ojha, 2006; Timsina and Gotame, 2014), they are not completely trusted by large segments of society. This is because a vast majority of Nepali non-profits are criticized for their association with political parties as well as for shifting political alliances (Rankin et al., 2018). In addition, they suffer from an image as donor-driven organizations that are supported largely by financial and technical contribution from international sources and are more accountable towards fulfilling the donors’ interests rather than those of the beneficiaries (Ulvila and Hossain, 2002). Secondly, the hypothesis considers the media as a key actor in this case. Since this is a case of central-level actors, focus on the capital city of Kathmandu suggests that there exists good newspaper readership, drawing from the city’s 70.62% literacy rate (compared to the national average of less than 40%) (Adhikari, 2005). Further, Timsina and Gotame (2014) highlight that the media has become increasingly important through enhanced investigation and reporting on issues related to forest governance, forest mismanagement, and best practices for forest management. Therefore, media coverage cannot be ignored as it is generally perceived to enhance government transparency and create an informed civil society.
This contextual knowledge about civil society organizations and media perception in Nepal lends some weight to the plausibility of H1, but not significantly. Therefore, a cautious estimate would set the prior confidence in the hypothesis as moderate or moderately high. Next, I test the sequential evidence to check if it allows me to update the confidence in H1.
Discussion
Analyzing the strength of the evidence to update the hypothesis based on the test results
In this section, I discuss the evidentiary impact on the confidence of H1 based on the collected evidence, description of a certainty and uniqueness evaluation, discussion of whether specific Bayesian tests were confirmed, and how changed confidence in H1 from the prior confidence that was characterized as moderate or moderately high.
Table 2 illustrates what evidence was used in each of the four tests (i.e. straw in the wind, hoop, smoking gun, and doubly decisive) and whether they passed the test or not.
H1 proposes that the tender-opposing group claimed favoritism and lack of transparency in the bid process. And, that they further leveraged the media to broaden their legitimacy. This constituted part 1 and part 2 of the hypothesis H1 (see Table 1).
Evidence to test part 1 of H1 (lack of transparency and favoritism in the process) includes direct quotes from the daily newspapers highlighting transparency issues and abuse of authority. Examples of evidence supporting the claims of the tender-opposing group include: ‘The activities of the donors participating in the program are of serious concern. The program is a result of a pre-planned design that came without holding consultations with partners concerned, including civil society’ (Debesh Mani Tripathi, President of Nepal Foresters’ Association, in Shahi, 2012b); ‘the lack of transparency and irregularities in the bid selection process has left us surprised’ (Ganesh Karki, General Secretary, Federation of Community Forest Users Group, in Pandit, 2012a).
This type of evidence is best described as support for a straw in the wind test. A straw in the wind test suggests that the evidence provided is of low uniqueness and low certainty. This is the weakest of the four tests and is neither necessary nor sufficient to confirm H1. However, passing multiple straw in the wind tests can add up to important evidence (Punton and Welle, 2015). As described in Table 2, there are multiple traced examples of evidence (E.1.1., E.2.1, E.2.2, and E.2.3 in Table 2) that pass this test, providing some limited support for the lack of transparency and favoritism in the process.
Part 2 of H1 proposes that the tender-opposing group framed itself as the losing group and leveraged media to expand the scope of the conflict by diffusing costs to the wider society (Jones and McBeth, 2010). In this case, the news coverage helped broaden the scope of the dispute by linking the case to larger issues of corruption, foreign aid mismanagement, and the democratic rights of information access, all wider political issues in Nepal. Newspaper headlines serve as examples of this framing in the media: ‘Donor assistance under the grip of forest mafia’ (Nepal Samacharpatra, in Panta and Sapkota, 2012a); ‘Multimillion-dollar forestry program courts controversy’ (The Kathmandu Post, in Shahi, 2012a); ‘Corruption in foreign aid’ ( Nagarik Daily, in Subedi, 2012).
In the four-month timespan after the decision notice, the donor group did not release any press releases as a response to the negative press coverage to explain about the selection process to the wider public. This suggests that the tender-defending coalition tried to contain the issue by framing it as regular complaints that would be dealt under the standard operating procedure and procurement policy. This group also highlighted the use of a fair process for dealing with complaints, trying to convince opponents and the public that any costs or concerns would be addressed through a transparent process.
In Part 3, H1 proposes that the media presented the tender decision mostly in a negative light. Evidence to support this part of the hypothesis comes from newspaper coverage of the tender offer and withdrawal, which characterized the tender process as a case of abuse of power and corruption, as cited in the headlines above. The national newspapers covered the tender controversy 20 times in four months following the award decision, and most of the coverage presented the tender-defending group in a negative light. This evidence passes the hoop test (E.3.1 and E.3.2 in Table 2). A hoop test provides high certainty that the hypothesis is correct and is necessary to confirm that a causal mechanism is present (Mahoney, 2012); it also helps rule out alternative hypotheses (Beach and Pederson, 2013).
The other evidence (E.3.2 in Table 2) qualifies as smoking gun evidence for this case. A smoking gun test is related to evidence with a high level of uniqueness, increasing the confidence that the hypothesis is true. The media attention described above led to the Parliament’s Finance and Labor Relations Committee’s investigation into this case. The committee directed the concerned Ministry to submit all documents related to the program. Evidence for this test includes newspaper coverage that highlighted the issue escalation to the parliamentary committee. 'The parliament’s Finance and Labour Relations Committee has instructed the Ministry of Finance and Ministry of Forests to furnish all the documents related to the flawed decision to award the Multi Stakeholder Forestry Programme (MSFP) to Rupantaran Nepal for investigations' (Nepal Samacharpatra, in Panta and Sapkota, 2012b); ‘We have ordered the concerned Ministries to submit all documents related to the tender process. We will investigate this. If we find any flaws, we will instruct the parties concerned to scrap the decision’ (Sushila Kandanga, Chairperson, Parliamentary Committee on Natural Resources, in Pandit, 2012b).
This evidence notably adds to the confidence in H1 by showing that the negative press generated enough public interest to require a high-level commission to investigate the case. Getting drawn into a controversy of this scale was perhaps unanticipated by the tender-defending group.
In Part 4, H1 hypothesizes that the tender-defending group foresaw reputational risks and a decline in stakeholder support, so they cancelled the decision. An internal donor report of the Department for International Development (DFID) provides evidence in its discussion of turmoil within donor issues in an annual report: ‘MSFP has been subject to negative press coverage in both the Nepali and English media over the past few months, including “donor bashing” on aid modalities, with potential reputational risk to DFID’ (DFID, 2013)
On July 27, 2012, a press release was issued by the Embassy of Switzerland, acting as the lead donor on behalf of all donors, announcing the tender cancellation and informing that the next tender would be designed in a more multi-stakeholder decision process:
After today’s meeting of the Ad-hoc Steering Committee of the MSFP, comprising representatives from the Ministry of forests and the three donors (UK, Switzerland and Finland), the Embassy of Switzerland, Swiss Agency for Development and Cooperation (SDC) decided to cancel the tender to implement MSFP in 18 districts of Nepal originally announced on 8 July 2011…A fresh, redesigned tender will be called soon. A working group has been constituted to that effect. (Embassy of Switzerland in Nepal, 2012)
Evidence (E.4.1. and E.4.2 in Table 2) passes the hoop test, providing additional support for the causal mechanism described in H1.
Testing H1 against the evidence traced
Collected evidence, certainty and uniqueness evaluation, and test results.
From Table 2, we see that H1 passed two straw in the wind tests, two hoop tests, and one smoking gun test. In this respect, the confirmatory power of H1 is enhanced.
Conclusion
It is worth noting that the process tracing (PT) method pushes researchers to increase rigor in reasoning and evidence search to make causal connections. It is possible that there is a diverse set of causal paths that led to the tender decision cancellation. However, what process tracing enables us to do is to open the black box of causality by assessing the power of different forms of evidence submitted to a range of tests (Beach and Pederson, 2013). This method provides transparency about how evidence is assessed as well as clarity in the weight of the evidence that can be used to confirm or update confidence in the hypothesis.
In this case, we saw that the hypothesis was corroborated by the evidence collected in all four parts of H1. The weight of the evidence was assessed through informal Bayesian reasoning and satisfied the hypothesized causal mechanism. Here we assessed the causal contribution of actors and actions that resulted in the tender cancellation. Understanding the causality of this outcome was important as it obstructed and changed the course of action initially designed for this multi-million-dollar donor aid-funded program. In addition, it also helped us understand the power dynamics in collaborative settings and how actors negotiate and strategize to reach an agreeable power-sharing arrangement.
By using the method of process tracing and applying the Bayesian logic to an evidence search and analysis of the tender dispute case, this article establishes how the stakeholder groups, arguably in a weaker power position, were able to organize, partner, and work together to negotiate and overturn a decision made by the stakeholder group in a more dominant power position. The PT method can add rigor and transparency for drawing within-case inferences in single cases, which cannot be done easily using traditional qualitative analysis methods.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the The Oregon Sasakawa Young Leaders' Fellowship Fund (Sylff) Graduate Fellowships for International Research 2017-2018.
