A F R I CA O P E R AT I O N S S E RV I C E S S E R I E S 76047 Avoiding Tokenism in Demand for Good Governance (DFGG) Activities Lessons from World Bank- nanced Lending Projects in Zambia ore Operatio C ns S ervices Quality OS DE P Capacity Results C FM ASP EN THE AFRICA REGION | THE WORLD BANK GROUP AFRICA OPERATIONS SERVICES SERIES Avoiding Tokenism in Demand for Good Governance (DFGG) Activities Lessons from World Bank-financed Lending Projects in Zambia Kate Bridges Contact details: Kate Bridges, consultant, Zambia country office in Zambia, e-mail: kbridges@worldbank.org. The author would like to thank Gaël Raballand for his invaluable input during the research, analysis and writing of this paper. Thanks also to Katherine Bain, Hélène Grandvoinnet, Asli Gurkan and Ghazala Mansuri for important comments on initial drafts, and to Marie Sheppard, Indira Ekanayake, Justin Runji and Patricia Palale who provided feedback for the stocktaking review on which this paper is based. AFRICA OPERATIONS SERVICES SERIES Contents Abstract  v Abbreviations and acronyms   vi DFGG terminology explained   vii iii Introduction  1 I. Putting DFGG in Context   3 A. DFGG and the World Bank   3 B. Evolution in development practice   4 C. Value of downwards accountability   5 II. The Zambian Portfolio Findings   7 A. Organizational commitment   7 B. Organizational value system   10 C. Location in a broader governance agenda   11 D. Evaluation  12 E. Importance of context   16 F. Persistent myopia   17 III. Conclusion and Recommendations  19 A. Be experimental   19 B. Be analytical   21 C. Be managerial   23 References 27 Annex 1. Methodology for Assessment  29 AFRICA OPERATIONS SERVICES SERIES Abstract In recent years, the World Bank has made increasing voluble commitments toward greater multi- stakeholder engagement, downwards accountability, transparency, and participation. It is com- mon nowadays to hear Bank promises of improved engagement with non-state actors being referenced at the highest level of management and reiterated within major policy papers. The questions now need to be asked: How smoothly is this high-level discourse translating into prac- tice? And, what challenges is it encountering along the way? v This working paper aims at such an analysis. The paper begins by explaining the Demand for Good Governance (DFGG) concept, identifying its increasing prevalence as a theme in World Bank discourse, locating it in a broader development agenda, and drawing out its implications for World Bank projects. The paper then presents the Zambian lending projects as a case study, drawing out the factors that contribute to DFGG success and failure on the ground. The rel- evance of this Zambian case study is that it demonstrates some of the particular challenges of trying to move DFGG commitments from “paper to practice”. As evidenced, the complexity of the shift should not be underestimated. While acknowledging that there are some hopeful trends, this paper contends that there are key areas that need to be addressed if there is to be a better synchrony between ideals espoused at a theoretical level and on-the-ground practices. This paper concludes with specific recommenda- tions as to how interventions can be more experimental in their philosophy, more analytical in their preparation, and more managerial in their attempts to address internal obstacles. Avoiding Tokenism in Demand for Good Governance (DFGG) Activities Abbreviations and Acronyms ADSP Agricultural Development and Support Project ALPS ActionAid’s Accountability, Learning, and Planning System CEP Copperbelt Environment Project CoST Construction Sector Transparency (Initiative) vi CSO Civil society organization DFGG Demand for Good Governance E-ISR External Implementation Status and Results Reports GAC Governance and anti-corruption IDSP Irrigation Development and Support Project ISR Implementation Status and Results Reports MBP Malaria Booster Project NGO Nongovernmental organization RDA Road Development Association RRMP Roads Rehabilitation and Maintenance Project SEED Support for Economic Expansion and Diversification AFRICA OPERATIONS SERVICES SERIES DFGG terminology explained Transparency involves the provision of clear, honest, and comprehensive information about proj- ect activities to beneficiaries. This openness can be evidenced through public policies, clear mandates, clearly articulated planning and budgetary processes and the presence of information sharing systems. Information is the center of the relationship between a government and the people it governs as well as between the World Bank and the beneficiaries it targets. Information is crucial to trans- vii parency and accountability since “[w]ithout information, the public cannot understand, let alone ask how decisions were made on their behalf” (Chase and Anjum 2008: 11). Participation is the process whereby beneficiaries can exercise an influence over World Bank decisions, and share control over resources and processes that affect their lives. Participation of this sort requires the creation of an environment where the various stakeholders to a project can take meaningful part in the projects that affect them, at all stages of the project cycle—design, implementation and monitoring. In short, it refers to a relationship based on partnership between development agency and beneficiary. Consultation can be defined as “a wider continuous process of participation of all stakeholders in the decisions throughout the formulation and execution of a project leading to sustainable development for the population in the area” (Chase and Anjum 2008: 13). A two-way conversa- tion of this sort is crucial for improved understanding and the informed decisionmaking. Monitoring refers to “the regular collection and analysis of data on specific indicators to assist timely decision making, ensure accountability and provide the basis for learning” (Chase and Anjum 2008: 14). Oversight is more akin to a proactive “watchdog” mechanism whereby the organisation’s pro- cesses are proactively scrutinized in order to detect any instances of corruption or abuse. Combined, monitoring and oversight can take the form of citizen feedback surveys, social audits, and independent watchdog committees, and ombudsmen. Capacity building is defined by the United Nations Development Program (UNDP) as “the pro- cess by which individuals, organizations, institutions and societies develop abilities (individually and collectively) to perform functions, solve problems and set and achieve objectives.” In the case of DFGG, capacity enhancement refers specifically to attempts made to strengthen and enable capacities of multi-stakeholders to participate in project planning and decisionmak- ing. It does not refer to more general capacities. AFRICA OPERATIONS SERVICES SERIES Introduction Given the potential benefits of multi-stakeholder engagement for improved delivery of public services, it has become increasingly clear that a greater commitment to Demand for Good Governance (DFGG) activities is desirable for the World Bank. A number of recent World Bank speeches and papers strongly assert this point.1 What is not quite so well 1 established or discussed however is the significance of the organizational shift that is required or the challenges faced when attempting to translate DFGG commitments from rhetoric into practice. In a speech on citizen empowerment in the Middle East, World Bank President Zoellick (2011) intimates the growing recognition of the need for non-state actor engagement means challenging the traditional World Bank structures and seeking innovative solutions: I suggest it is now time for the World Bank to examine, with its Board and sharehold- ers, whether the Bank needs new capabilities or facilities that could leverage support from countries, foundations, and others to strengthen the capacity of CSOs working on accountability and transparency in service delivery. Perhaps even more significant than the need for structural innovation is the fact that DFGG activities are deeply connected with staff attitude and organizational culture. “DFGG is 80 per- cent political and 20 percent technical,” and it is therefore “as much about changing mentalities, building relationships, and developing capacities as it is about introducing mechanisms and tools.” (Agarwal and Van Wicklin 2011). The relevance of this Zambian case study is that it dem- onstrates some of the particular challenges of trying to move DFGG commitments from concept to practice. As will be evidenced, the complexity of the shift should not be underestimated. This paper is based on an evaluation of the Zambian lending portfolio carried out in early 2011. The desk-based evaluation utilized various World Bank documentation—project appraisal docu- ments, project implementation plans, operational manuals, performance frameworks, environ- mental and social impact assessments, aide memoires, mission documents, project evaluations, midterm reviews, and special studies on governance and accountability—as primary sources of information. The data gathered via documentary analysis was further verified through in-depth interviews with task team leaders and through perceptions of beneficiaries using the beneficiary feedback mechanism, External Implementation Status and Results Reports Plus (E-ISR+). Nine projects were considered in total. Each project was assessed for DFGG mechanisms in the following 4 categories: transparency and information, participation and consultation, monitoring and oversight, and capacity enhancement.2 The established mechanisms were then considered according to a set of 4 criteria: effectiveness, efficiency, inclusiveness, and sustainability. 1 Chase and Anjum (2008), Singh and Sanjay (2011), Zoellick (2011), Agarwal and Van Wicklin (2011), and World Bank (2011. 2 See Annex 1 for further detail on methodology including criteria for selection. Avoiding Tokenism in Demand for Good Governance (DFGG) Activities This paper presents a background and explanation of the DFGG concept, describing its increas- ing prevalence as a theme in World Bank discourse, its location in a broader development agenda as well as its implications for World Bank projects. It further illustrates the Zambian lending projects as case studies, drawing out the factors that contribute to DFGG successes and failure on the ground. On one hand, the implementation of DFGG in Zambian lending projects shows a number of positive trends; newer projects are making a more comprehensive attempt to resource, imple- ment, and monitor DFGG mechanisms than older ones; there is evidence of good practice in number of projects, and there is an increasing and promising inclination to locate DFGG mecha- nisms within a broader governance and anti-corruption agenda. On the other hand, there are also significant challenges. First and foremost, the DFGG mecha- 2 nisms that are implemented receive little, if any, monitoring and evaluation. A significant gap in all of the projects is the absence of systematic third-party monitoring mechanisms. Inclusivity more broadly is problematic, with considerable doubts over the representativeness of a number of civil society intermediaries. As the analysis in this paper illustrates, each of these deficiencies has consider- able impact on the success of the mechanisms and thereby on the effectiveness of the project overall. As the designed mechanisms struggle with moving from concept to practice, a number of them end up being applied in a fairly tokenistic manner. A significant contributor to that struggle is the fact that they have often been designed without sufficient consideration for the broader enabling environment. A key finding in this regard is the extent to which the attitudes of the project man- agement and partners become obstacles to meaningful implementation. It is also problematic that, across the projects, downwards accountability (to lower-level beneficiaries) still receives far less priority than does upwards accountability (to funders) or horizontal accountability (to part- ners). The continued marginalizing of project beneficiaries in particular has direct implications for the sustainability, effectiveness, and efficiency of the project (Box 1). Since it enables benefi- ciary trust, encourages humility in practitioners, and is focused on learning from and responding to the user, downwards accountability should be a direct means of ensuring locally owned and locally appropriate action. The paper concludes with specific recommendations on how interventions can be more experi- mental in their philosophy, more analytical in their preparation, and more managerial in their attempts to address internal obstacles. Box 1. The Lowers and Uppers Downwards accountability is the type of accountability that moves down the aid chain, from those with more power to those with less; it can be described as the accountability of uppers (those who are powerful and dominant in a given context) to lowers (those who are weak and subordinate in a given context). AFRICA OPERATIONS SERVICES SERIES I. Putting DFGG in Context Worldwide good governance is recognized as an essential requisite for sustainable development. The genesis of demand for good governance comes from the growing realization among practitioners and policymakers that engaging with citizens, communities, privates sector, 3 and civil society organizations is essential to improve governance and service delivery (Chase and Anjum 2008). A. DFGG and the World Bank The call for greater engagement with non-state actors and for a more concerted focus on the “demand-side” of governance is an increasingly prevalent theme in World Bank literature and discourse. Within the World Bank this agenda has historically been pursued under a number of varied terms: civic engagement, community-driven development, social accountability, voice, democratic policy reform, electoral systems and processes, human rights law, access to informa- tion, social inclusion, women’s empowerment, civil society capacity development. The variety of terms partly reflects the extent to which the agenda has previously remained somewhat periph- eral and ill defined (Bhargava and others 2011). But now, that agenda seems to be moving center stage. In recent years, the call to engage non- state actors more systematically has been most clearly reflected in the Governance and Anti- Corruption (GAC) Strategy 2007. It directs the Bank to “give explicit consideration, underpinned by improved diagnostic work, to governance shortcomings and corruption in the country,” and to find “ways in which private sector engagement and domestic accountability mechanisms can be used to support and strengthen… implementation and governance outcomes.” Principle #5 of the GAC Strategy 2007 confirms the World Bank’s commitment to specific improvements at the project level:1 Engaging systematically with a broad range of government, business and civil society stakeholders is key to GAC reform and development outcomes—so, consistent with its mandate, the [World Bank Group] will scale up existing good practice in engaging with multiple stakeholders in its operational work, including by strengthening trans- parency, participation and third-party monitoring in its own operations. The progressive transition toward greater multi-stakeholder engagement has been happening in the World Bank for a number of years. The “World Development Report (WDR) 2000: Attacking Poverty” highlights the importance of empowerment in addressing the needs of the poor (World Bank 2000). The “Empowerment and Poverty Reduction Sourcebook” (World Bank 2002) iden- tifies access to information, participation, accountability, and organizational capacity as key elements of empowerment. The “WDR 2004: Making Services Work for the Poor” highlights the role of client power and voice in influencing the accountability relationships that make service 1 http://go.worldbank.org/32PINXVIJ0 Avoiding Tokenism in Demand for Good Governance (DFGG) Activities delivery pro-poor. In 2005 the Bank had its first Social Development Strategy and in 2007 its first Governance and Anti-Corruption Strategy. And in 2011, the World Bank President recognizes citizen participation and social accountability as being at the center of the Bank’s future role in his speech, A New Contract for Social Development (Zoellick 2011). These concerted attempts by the World Bank have helped to more neatly frame related multi- stakeholder activities and focus future interventions under the term demand for good gover- nance. DFGG activities can take place at the project, country, or global level. This review deals specifically with the project level, at which DFGG involves “setting up systems to ensure that beneficiaries have a greater voice in planning and implementation and that the project is down- wardly accountable and responsive to their needs (Agarwal and Van Wicklin 2011).” 4 B. Evolution in development practice The emphasis on responsiveness and downwards accountability is particularly significant since it places this relatively new World Bank agenda within a broader movement that has been ongoing in the development arena for a while. That movement is characterized by the recogni- tion that among NGOs and international development agencies, engagement of and account- ability toward beneficiaries has typically been woefully neglected (Davis 2007, Kilby 2006, Slim 2002, Stockton 2006). There are various reasons for this neglect, but chief among them is the fact that beneficiaries are the least powerful link in the development chain. Accountability toward them has therefore not been necessary for an organization’s survival, even if it has been necessary for its effectiveness. As such, something of a paradox has been operating in develop- ment practice; The central dilemma here is that mission-driven accountabilities are often to people, ‘intended beneficiaries,’ who have little influence and in general no power over the organization, while contractual accountabilities, where the power lies, in general resides with people at the other end of the pipeline, with the funders.2 In agreement with the above quote by Simon Zadek, who places power at the heart of the issue, accountability theorist Austen Davis (2007: 9) insists “all agents can be held to account by those who have power over them”. The logical conclusion of this has been, as Nicholas Stockton (2006: 3) states, “the quality of humanitarian accountability declines in direct proportion to the relative power of the stakeholder.” Consequently, in the development sector, the upwards rela- tionship toward donors has historically been the overwhelmingly myopic focus of mechanisms of NGO accountability. More recently however, as the power of the development sector has grown (both in terms of resources and influence) and as criticism of development practices has become more vocifer- ous, there have been growing calls for improved accountability toward those with less power. As Hugo Slim (2002) points out, the proactive and passionate implementation of downwards accountability within development projects is crucial if development organizations are to avoid charges of hypocrisy. If the World Bank, for example, is to maintain its integrity while insisting governments cannot have successful development without good governance and without the participation of its citizens, then it must necessarily be implementing participative mechanisms within its own projects (Zoellick 2011). 2 Simon Zadek quoted in Bendell (2006: 16). AFRICA OPERATIONS SERVICES SERIES C. Value of downwards accountability Implementing downwards accountability is not only a contributor toward integrity however. For those development professionals who believe that poverty and suffering are exacerbated by unequal power distributions, it is clear that downwards accountability can have inherent value since it aims at redistributing power to the least powerful. Critics insist that since the notion of helping people to help themselves rather than simply imposing solutions is central to the humanitarian and development enterprises (Ellerman and Hirschman 2006), there must be meaningful attempts made to empower lowers in their dealings with NGOs. Otherwise, “to give people help, while denying them a significant part in the action, contributes nothing to the development of the individual.”3 World Bank literature has not tended to focus on the inherent value of downwards accountabil- 5 ity, but it has certainly recognized its instrumental value. Within the World Bank, citizen engage- ment, transparency, and downwards accountability are increasingly being heralded as essential prerequisites for good governance and effective development (World Bank nd). Three major benefits of downwardly accountable and responsive projects are as follows (Agar- wal and Van Wicklin 2011): 1. Better development outcomes, such as improved service delivery, improved program effec- tiveness and public expenditure efficiency, reduced corruption, and improved governance; 2. More effective institutions, processes and systems through user feedback mechanisms, community monitoring, and stronger linkages between local governments and civil society organizations; and 3. Better projects through increased community participation, inclusion, and improved targeting. None of these instrumental benefits are simply theoretical. A recent report by the Transparency Partnership Fund gathers together the evidence of impact around DFGG approaches and con- cludes that DFGG contributes directly to citizen empowerment, enhanced development, and improved governance illustrated in Figure 1 (Bhargava and others 2011). 3 Alinsky as quoted in Ellerman (2001: 15). Avoiding Tokenism in Demand for Good Governance (DFGG) Activities Figure 1: The Impact of DFGG Approaches 6 AFRICA OPERATIONS SERVICES SERIES II. The Zambian Portfolio Findings Each of the Zambian projects was assessed for mechanisms that could be differentiated according to 4 different categories: (a) transparency and information; (b) participation and consultation; (c) monitoring and 7 oversight; and (d) capacities enhancement.1 Table 1 presents examples of the actual mechanisms by which these different elements can be realized (the list of examples is by no means exhaustive). Multi-stakeholder engagement, accountability, consultation, and transparency across the Zam- bian projects portfolio are consistently highlighted as being critical aspects of project design and implementation. But while all of the projects, without exception, do subscribe to the importance of accountability mechanisms, not all of them manage to successfully translate these lofty com- mitments into practical activities. In fact, the diversity across the projects is significant. Table 2 lists the projects under evaluation and gives a rough indication as to the strength of the mechanisms they are being implemented in relation to the 4 DFGG categories. Note that, except in the case of 2 projects, which were evaluated prior to achieving implementation, this measure- ment does not indicate the strength of what was designed but of what was actually implemented. This is important since a number of the projects did in fact design strong mechanisms but then failed to move those designs from “paper to practice”. Although there are apparent struggles with implementation of DFGG mechanisms in Zambian projects, this is less a reason to be disheartened than it is a reason to realize the significance of the shift that is being required by the World Bank’s decision to engage in an increasingly down- wardly accountable manner. Most importantly, it is a reason to assess what is working or not working and why. This following section draws together some of the major contributors to current successes and failures regarding the Zambian portfolio’s DFGG activities. A. Organizational commitment World Bank projects have included activities focused on downwards accountability, participa- tion, and consultation for a number of years. It is only recently however that such activities have 1 There are various ways of categorizing the elements that lie at the core of improved beneficiary engagement activities. All of these tend to involve some combination of consultation, information, influence, and redress. For the purpose of this paper, categories were based on those used by the World Bank in a recent DFGG stocktaking report (Chase and Anjum 2008). Avoiding Tokenism in Demand for Good Governance (DFGG) Activities Table 1: DFGG elements and mechanisms Key Element Examples of Tools and Mechanisms Transparency and information • Community outreach Promoting and creating • Proactive Information disclosure, dissemination and opportunities for access to demystification information and transparency • On-demand information provision norms • Implementation of transparency and information laws • Social contracts • Citizen charters Participation and Consultation • Participatory design 8 Encourage and mediate • Participatory implementation opportunities for multi- • Citizen councils/ Membership on advisory bodies stakeholder participation and • Community mapping/ data gathering input • Consultation workshops • Communities generating indicators • Interface meeting between providers and beneficiaries Monitoring and oversight • Citizen report cards Promote methods and • Expenditure tracking incentives to hold actors to • Civil society oversight committees account by empowering multi- • Citizens jury/ user platforms stakeholders • Service user association/ action group • Participatory monitoring • Third party monitoring • Community scorecards • Social audits • Grievance mechanisms • Legal recourse and redress Enhancing capacities • Training modules Strengthen and enable • Community sensitization capacities of multi-stakeholders • Coalition building to participate in project • Capacity building planning and decision making begun to be emphasized as a central aspect of Bank projects. There are still some indications that the commitment to a new engagement with multi-stakeholders; non-state actors; and ben- eficiaries, in particular, has not yet been fully embraced at the country project level. This lack of commitment toward implementation is naturally proving to be a significant obstacle to the success of DFGG mechanisms. Within project documentation there is a preference for abstract, non-specific language of implicit commitment to transparency and accountability. Some projects make aspirational claims about pursuing “full partnership” or “sustained consultation” with project beneficiaries but fail to spec- ify exactly whose responsibility that will be, how it will be funded, and how performance shall be monitored. The tendency to remain abstract about these particulars simply means they will not end up being meaningfully implemented. Within the Malaria Booster Project, for example, “there is no clear guideline on what, where, and how information is to be shared at different AFRICA OPERATIONS SERVICES SERIES Table 2. Strength of DFGG in Zambian Portfolio KEY Participation and Consultation Transparency and information Good Monitoring and oversight Fair Enhancing capacities Poor Note that the strength indicated is only an average. In the case of some projects they have both very good and very poor mechanisms within a category. Project name 9 Increased Access to Electricity (IAE) Project Malaria Booster Project (MBP) Copperbelt Environment Project (CEP) Public Sector Management Program Support Project (PSMPSP) Agricultural Development Support Program (ADSP) Water Sector Performance Improvement Project (WSPIP) Support for Economic Expansion and Diversification (SEED) Projects not yet implemented  evaluated on strength of design only Irrigation and Development Support Project (IDSP) Road Rehabilitation and Maintenance Phase II (RRMP II) levels of government,” and that consequently “practice in information sharing varies by health facilities and communities at service delivery level.”2 Given the non-specificity of the language, implementation ends up being overly reliant on the individual personalities and priorities of the project management and implementation units. In some projects, changes in individual staff leads to either noticeable improvements or marked deterioration in implementation of the mechanisms. This overt reliance on particular personali- ties ensures that the sustainability of mechanisms remains weak. The abstraction around language is sometimes compounded by a reluctance to make DFGG mechanisms a requirement. Several task team leaders have observed that unless something is explicitly required in project documentation, it is unlikely to be effectively undertaken by imple- menting partners. The suggestion from various task team leaders is that the project environment is often conducive to a more exacting approach, but that DFGG mechanisms have simply not been considered important enough to warrant such treatment. Again, this seems linked to the priority (or lack of such) that DFGG mechanisms are currently being accorded in the Bank’s own internal structures. Task team leaders confirm that they do not have the impression that non-state actor engagement and DFGG mechanisms really are a central concern for the World Bank. According to one team leader: 2 Back to Office Report (BTOR) 2011, Malaria Booster Project. Avoiding Tokenism in Demand for Good Governance (DFGG) Activities If this really is a priority then it needs to be better emphasized within the Bank. No one ever asks me about this. If it’s important then it needs to begin to be reflected in Bank systems, in quality enhancement requirements, and guidelines for the concept note. A further symptom of how organizational commitment may not yet have caught up with rhetoric is the lack of adequate financial and human resourcing for DFGG activities. With regard to human resources, there is still a clear misalignment between the skills set of Zambia’s country office-based staff and the governance agenda of the World Bank Group. There are few staff in the Zambia office who are cognizant of both the rationale and mechanisms of beneficiary accountability. Most task team leaders are in fact unfamiliar with the DFGG concept, which naturally hampers any effort to implement, monitor, and evaluate it objectively. [This was first and foremost reflected in their surprise at being asked about these mechanisms during this review.] By contrast, the task team leader of the newer Irrigation Development and Support Project (IDSP), which has had DFGG 10 input directly from Bank governance experts, demonstrates a far clearer understanding of the con- cept and its implications for the project. Likewise, the Irrigation Development and Support Proj- ect itself demonstrates the most promising DFGG design across all the Zambian projects. Clearly, input from a comprehensive skills set leads directly to considerably more robust design and, presumably, implementation. The fact that several DFGG mechanisms in Zambia are mainly being supported by a short- term funding mechanism—the Governance Partnership Facility—may also be indicative of as yet insufficient organizational commitment. DFGG is after all a necessarily long-term agenda, and mechanisms are likely to need significant time before they bear fruit. To fund it with a short- term mechanism raises issues about both sustainability and effectiveness. It likewise conveys the impression to project staff that it is not yet a key organizational concern (Box 2). B. Organizational value system Closely linked to the notion of organizational commitment is the idea of “organizational culture” or “organizational identity”. The organization displays a recognizable identity or culture when there is considerable agreement, similarity, or overlap among the individual belief systems that outweighs the differences among them. Typically, this means a small set of interrelated values, rather than any one single value. This composite set of values, internally consistent, may be referred to as the organizational value system (Padaki 2000). Within the World Bank, it seems clear that the current value system has not quite caught up with the rhetoric being espoused by key reformers and Bank Management. This is significant because among Bank staff, and project implementers more broadly, it is apparent that personal attitude and beliefs about beneficiary engagement are having a direct impact on the priority that DFGG mechanisms are being accorded in their project. Box 2. Long-term nature of DFGG DFGG is a long-term and deep agenda, involving changes in attitudes, values, behaviours and incentives. It is not well suited to a traditional project-based investment approach, since it requires a long-term perspective, persistence, and sustainable follow-through” (Bhargava and others 2011: 22). AFRICA OPERATIONS SERVICES SERIES The relevance of value systems should come as no major surprise given that accountability theorists have for a long time pointed out that organizational culture and individual attitudes are central to the implementation of meaningful downwards accountability mechanisms (Ebra- him 2005; Chambers 1993; Ellerman 2001; Edwards and Sen 2000). What is surprising is that DFGG-related attitudes and values are not routinely being addressed in the Bank. This fact occurs despite the significance of personal values and attitude having been previously recog- nized by the Bank. It would be a mistake to assume that everyone involved in Bank projects automatically ascribes to the view that a broader citizenry should be given influence over the development activities that affect them. This is not the case, on two levels: the first, in relation to attitudes of Bank staff themselves; and the second, in relation to attitudes of project implementers more broadly. The divergence in application and effectiveness of downwards accountability mechanisms within 11 and between Zambia’s lending projects suggests that attitude on both levels plays a big role. With the External Implementation Status and Results Reports Plus (E-ISR+), for example, it is clear from their attitudes some staff are particularly proactive in assessing and responding to the feedback.3 The IDSP task team leader, for example, took proactive measures to disseminate the findings among project teams and also to ensure that some of the recommendations were actively taken up. For some other staff, the exercise is mainly considered one more task to add to an already long list, and they show little willingness to allow the feedback to alter or inform future imple- mentation. Similarly with broader project implementers, differences in attitudes contribute to significant divergences. The Malaria Booster Project, for example, notes a significant difference in the willingness of health center staff to implement feedback mechanisms for beneficiaries in a more meaningful manner. The continued reluctance to give beneficiary engagement mechanisms much thought is often due to some project teams not having internalized the notion that beneficiaries really should have a central role in all aspects of design and implementation. They do not necessarily under- stand or appreciate either the inherent or instrumental value of such mechanisms. Again, given the value-laden and relationship-driven nature of non-state actor engagement, it is not surprising that individual personalities and attitudes play such a significant role. Develop- ment theorist Robert Chambers points out the need to recognize that sustainable and authentic downward accountability will be achieved not simply by implementation of sectorwide proce- dures but by organizations that are internally accountable to the essential humanity and dignity of individual people (Chambers 1993; 1997). The World Bank’s attempt to ensure greater multi-stakeholder engagement and downwards accountability is a significant attempt at organizational change; as such, it should be assumed that the perspectives of staff will need to be directly addressed. This has not happened. C. Location in a broader governance agenda The newer projects in the Zambian portfolio tend to focus on the engagement of beneficiaries with a specific view to improving governance. The Irrigation Development and Support Project, for example, locates the majority of its beneficiary engagement mechanisms directly in a broader GAC agenda. And the Road Rehabilitation and Maintenance Project (RRMP) Phase II follows a 3 E-ISR+ is a mechanism that facilitates beneficiary feedback on project successes and failures. Avoiding Tokenism in Demand for Good Governance (DFGG) Activities similar path via the Construction Sector Transparency (CoST) Initiative and its corresponding Road Sector Governance Plan (RSGP). The fact that the older projects do not tend to locate their engagement mechanisms in a wider governance and accountability agenda inevitably has an impact on the design and implementation of those mechanisms. The tendency to treat mecha- nisms of consultation and participation as predominantly technical solutions is to ignore the very factors that will determine their success. Older DFGG mechanisms appear more often to be implemented as one-offs, done to a minimum requirement, on an ad hoc basis and with little attempt at institutionalization. The broader view, by contrast, explicitly requires project management to implement the mechanisms in a more sustained manner, on the basis of sound political analysis and with a view to long-term empow- erment of communities. 12 The mechanisms to be implemented in the Irrigation Development and Support Project, for example, are designed specifically to address broader governance risks and to empower small- holder farmers as a means of ensuring effective development. Low capacity of stakeholders is identified and then responded to by delegating capacity enhancement to a contracted service provider for “the full duration of the project”. This responsiveness and emphasis on sustained engagement is testament to a meaningful attempt to engage citizens in a non-tokenistic manner. The fact that such objectives will be closely monitored is also noteworthy. In short, the presence of a clear philosophy and rationale for the DFGG mechanisms in the Irrigation Development and Support Project seems to have led directly to them being well resourced, clearly defined, and broadly monitored. By contrast, in a project like the Malaria Booster Project, which did not benefit from in-depth governance risk analysis—although there are some ad hoc beneficiary feedback mechanisms in place—they vary drastically across project locations and are largely personality driven. The absence of a persistent, uniting rationale for beneficiary feedback that goes beyond minimum safeguard requirements seems to have contributed to it being given a fairly low priority by several implementers. D. Evaluation The most conspicuous trend across all the projects was that those DFGG mechanisms being implemented are generally not being monitored and evaluated. As such, the follow-up to demand-side measures is weak. Supervision missions make no systematic attempt to find out whether beneficiaries are being engaged or whether the existing mechanisms are inclusive, efficient, effective, or sustainable. There are in fact few, if any, financial provisions for them to do so. In the Agriculture Develop- ment and Support Project (ADSP), project information is proactively disseminated to the broader public during the biannual trade fairs and annual agricultural show as well as via a website. It is not evident however who exactly attends these events, receives the leaflets, or reviews the website. In all likelihood, they are a small and atypical element of society. In short, notions of whether DFGG mechanisms are successful or not are subsequently anecdotal, not proven. This does little to address the attitudinal impediments among project staff regarding the necessity of implementing such mechanisms. Robust evaluation studies on the effectiveness of key demand- side measures would make it more likely that task team leaders and clients will adopt measures that have greater operational value. Another major reason to ensure more effective monitoring and evaluation of the projects is to guarantee that good lessons are not being lost. A number of the projects assessed during the review do in fact have in place some mechanisms that seem particularly promising and may AFRICA OPERATIONS SERVICES SERIES be applicable to scale-up across other projects. Examples of some noteworthy mechanisms are shown in Table 3. The tools and mechanisms mentioned in Table 3 show significant potential; but without system- atic monitoring, causality cannot be established. The concern is that any opportunity for lessons being learned or good practices being scaled up is being lost. Moreover, the lack of monitoring contributes to a further absence of incentive on behalf of project teams to implement DFGG meaningfully. As such, it both fuels and compounds any existing attitudinal impediments. A benchmarking review conducted by the Quality Assurance Group (QAG) confirms that this lack of follow-up is a problem with projects Bankwide: Table 3: Examples of Good Practice across projects 13 Key element Examples of tools and mechanisms Transparency and information • Proactive Information disclosure, dissemination, and demystification via the Environmental and Social Promoting and creating Management Framework in various projects opportunities for access to information and transparency • Specific attempts at demystification of project norms information in the CEP • Proactive information dissemination via project newsletter and website in the ADSP • Awarded contracts proactively made available for review by local communities in the RRMP Phase II. Participation and Consultation • Multi-stakeholder and bottom-up implementation in the Malaria Booster Project Encourage and mediate • Robust consultation and participation during opportunities for multi-stakeholder safeguard processes across projects participation and input Monitoring and oversight • External complaints hotline in the RRMP Phase II • Robust participatory decisionmaking process around Promote methods and incentives tourism priorities in the SEED Project to hold actors to account by • Grievance mechanisms as a part of resettlement empowering multi-stakeholders policy framework processes in various projects • Citizen involvement in supervision of RRMP Phase II • Civil society oversight committees in the Malaria Booster Project • Participatory design and implementation around Service Delivery Improvement Facility component in the Public Service Management Support Program. Enhancing capacities • Sustained capacity building of beneficiaries throughout the duration of the Irrigation Strengthen and enable capacities Development and Supprt Project. of multi-stakeholders to participate in project planning and decision making Avoiding Tokenism in Demand for Good Governance (DFGG) Activities While almost half of projects included actions to promote transparency, only one- seventh included arrangements to monitor the effectiveness of these actions. A quar- ter of projects included grievance redress measures, but only half of those were obliged to act on complaints or provide written feedback (World Bank 2009). The lack of monitoring may be symptomatic of a deeper problem regarding organizational com- mitment to DFGG. In any case, its consequences for the effectiveness of DFGG mechanisms are likely to be significant. Although this review cannot claim to have undertaken the level of compre- hensive evaluation that is necessary to definitively establish effectiveness, it is clear from the broad analysis that terms such as “consultation” and “participation” are being applied to activities that have vastly different levels of non-state actor engagement. Consultation in one project, for exam- ple, applies to a process whereby citizens are invited to comment but where there is no assurance that their concerns and ideas will be taken into account. In another project, consultation involves 14 a process whereby citizens have some genuine bargaining influence over project implementa- tion. In yet another project, consultation involves a process whereby non-state actors are placed on rubberstamping advisory committees or boards mainly for the purpose of “educating” them or engineering their support. In short, not all consultations are equal. Some are placatory, some are genuinely participative, and some are even manipulative. In the Copperbelt Environment Project, for example, it was made clear during a recent E-ISR+ evaluation that the mechanisms have not facilitated trust between the beneficiary and the project management, neither have the mecha- nisms enabled beneficiaries to communicate their preferences or have a say in key decisions. The real concern here is that what is sometimes called consultation or participation in the Zam- bia lending projects is actually tokenistic—a perfunctory or symbolic gesture is made toward engaging citizens, but in reality the process is superficial and fails to involve them in any mean- ingful sense. There are concerns that the inclusion of the Zambian National Farmers’ Union (ZNFU) in the ADSP steering committee only goes halfway in representing the broader farming stakeholders, for example. The task team leader observed that while National Farmers’ Union was often presented as the pulse of the farming community, the views it represents are often of a select, powerful few, rather than the broader majority. In short, the DFGG mechanism being used—in this case stakeholder representation on steering committees—appears to be neither effective (in the sense of enabling smallholder farmers to participate in key project decisions) nor inclusive (in the sense of ensuring that intermediaries are representative). Likewise, task team leaders for other projects expressed doubt as to whether the civil society intermediaries present on their project’s steering committee were actually representative of broader communities. The danger is that civil society organizations are being uncritically added to steering committees in a perfunctory attempt to make them look community driven and inclu- sive. But since these organizations are in no sense a perfect conduit for community concerns, their actual representativeness needs to be evaluated rather than presumed. Furthermore, such mechanisms need to be backed up by other activities such as, for example, a solid mechanism for complaints grievance redress so that concerned community members are not limited only to feeding issues through a potentially biased intermediary. The current absence of such parallel feedback channels in Zambian projects is noticeable. Only the newer projects have formal- ized grievance redress mechanisms mentioned in their design documentation. None have yet achieved implementation. It is clear that any tendency toward tokenism will be exacerbated by the fact that the perfor- mance of participatory mechanisms is seldom being closely evaluated. A positive first step would be to devise a clear set of indicators that constitute the level of participation aimed for by the Bank. The second step, of course, would be to ensure that performance of those mechanisms is actually monitored. AFRICA OPERATIONS SERVICES SERIES In this regard, the newly initiated RRMP Phase II is a useful example of both the challenges and opportunities for better procedures. The project’s planned mechanisms for monitoring and oversight will be implemented in tandem with internal governance tools and are based on an initial review that highlighted significant gaps and risks. Included in the mechanisms for citizen oversight are such specifics as an external complaints hotline and representation from local communities on Road Development Association (RDA) project supervision teams. There is also in place an initial set of indicators for monitoring the actual implementation of these mechanisms; this is a crucial step. The task team leader confirms that the real challenge for the RRMP DFGG mechanisms is whether they translate from “paper to practice”. With that chal- lenge in mind, monitoring and evaluation of the mechanisms needs to remain an absolutely critical aspect of their implementation. Not only will monitoring ensure that the incentive to implement the mechanisms is maintained, it will also contribute to helping the Bank assess whether these mechanisms represent models of good practice that could have relevance for 15 other projects. By contrast, the various commitments to sustained consultation and participatory processes during implementation in a number of the other projects do not benefit from such clarity. Neither do they include indicators for monitoring. As a result, the majority of them have simply not achieved implementation. Experience indicates that in the absence of a precise activity plan and clear indicators for monitoring performance, an activity is unlikely to be prioritized. There is a further notable absence in the projects, specifically regarding third-party monitoring. In the Public Sector Reform Project, for example, all monitoring and evaluation remains internal with no sustained non-state actor oversight. The task team leader surmises that the lack of exter- nal pressure has likely been a contributor to the hampering of procurement reforms, and that the absence of civil society engagement is evidently a missed opportunity. Independent evaluation of projects is both necessary and difficult to implement. It is necessary because it provides external verification of project successes and failures and can ensure that beneficiary and non-state actor views are being represented without internal bias. As such, it is a critical demand-side mechanism. It is difficult because CSO monitoring capacity in Zambia is low and financing streams are either scarce or compromised. As the recent report by the Trans- parency Partnership Fund points out, “there is an inherent conflict of interest for a civil society organization to accept funds from a government agency or the World Bank and then engage in independent monitoring of that same agency in a Bank-financed project” (Bhargava and others 2011). To avoid this conflict of interest, there needs to be ideally a hands-off pot of funding that exists only to support DFGG and third-party monitoring. Some of the evaluated projects have participated in the E-ISR+ initiative. E-ISR+ makes an admi- rable attempt to provide unbiased, external verification of project success and failures as well as to assess non-state actor’s satisfaction with their existing levels of engagement in the project. It does, however, continue to suffer from potential conflicts of interest given that the research company is being paid by the Bank to evaluate Bank-financed projects. The E-ISR+ also relies heavily on the characteristics of individual project task team leaders and Bank staff who are required to drive the process and define questions, locations, and stakeholders for evaluation, and who ultimately decide on whether to adopt the recommendations. Even if they are a par- ticularly motivated task team leader, their underlying incentive will be to target the evaluation at “good examples” and obedient, uncritical stakeholders, rather than the problem areas that most require an objective evaluation. Therefore, there is little sense in which the E-ISR+ can be called “independent”. And while it is external on one level (in that an external evaluator carries out the task), it is very much controlled by internal Bank personalities and biases. Avoiding Tokenism in Demand for Good Governance (DFGG) Activities E. Importance of context The issue with implementation of mechanisms is compounded by a propensity for neglect toward the mechanism’s “enabling environment”. In Bank-funded projects in Zambia it is often the case that a promising mechanism is referred to in the design documentation but fails as a result of broader personality, capacity, politics, or other factors. Of course, not all possible eventualities can be accounted for, but a failure to consider the broader enabling environment certainly increases the likelihood of mechanisms being ineffectively implemented. Thus many of the projects—older ones in particular—find that after designing comprehensive mechanisms there is no incentive on the behalf of implementing partners to actually execute them. This sug- gests that the mechanism itself may have been conceived of in isolation from the broader factors that will ensure its effectiveness. Fortunately this tendency has not gone unnoticed, and more 16 recent projects are making diagnosis of the socio-political environment a far more central aspect of project design. A plethora of recent Bank-funded Zambian political economy analyses are also testament to management having prioritized such diagnoses. Such analyses are critical at many levels since the enabling conditions for DFGG vary not only from country to country but also from sector to sector and from project to project. A recent World Bank working paper that uses Zambia as a case study finds that sector-level analysis is particularly important: Lessons from the Zambia case demonstrate that while country-level [political econ- omy] analysis is useful as background information, [political economy] studies should probably focus more at the sector level as the recommendations from these studies tend to be more actionable” (Beuran and others 2011: 14). The Malaria Booster Project has been able to take advantage of a high-level capacity that exists among health-related NGOs to implement a fairly successful multi-stakeholder implementa- tion model. It would be a mistake however to assume that such a model could be transplanted wholesale into the road sector, for example, where such NGO capacity is lacking. There was a similar attempt made to implement community-driven water watch groups in the water sector,4 which failed due to factors in the enabling environment. Project team members confirmed that as a result of insufficient demand-side resources and inadequate enforcing power, these groups are now largely dysfunctional. The implementation of water watch groups appears to have been done without sufficient prior analysis of either the willingness and capacity on the part of the community to continue demanding change or the willingness and capacity of the utility to actu- ally respond to those demands. By contrast, the Irrigation Development and Support Project undertook thorough analysis of demand-side capacity during its preparation phase. As a result of this analysis, it was determined necessary to enlist a service provider to enable “the establishment of water user associations and producer organizations,” and to increase “the knowledge capacity of smallholder farmers to negotiate with their commercial partners in the value chains.”5 That such capacity enhancement is clearly delegated to a contracted service provider for “the full duration of the project” is testa- ment to a meaningful intention to engage citizens in a non-tokenistic manner. Not only is this an example of good and timely context analysis, it is also an example of genuine responsiveness to that analysis. 4 These groups were tasked with monitoring whether the utility was actually providing what it had committed to provide. 5 Project Appraisal Document for IDSP, 2010 AFRICA OPERATIONS SERVICES SERIES The newer projects like Irrigation Development and Supprt Project and RRMP Phase II have only recently been initiated and are yet to be fully evaluated, but the fact that they have used political economy analysis to directly inform their design is promising. Older projects would clearly have benefitted from more assessment of their enabling environment prior to deciding on suitable mechanisms. In particular, the capacity and legitimacy of civil society and implement- ing organizations to implement, monitor, and assess DFGG interventions should be realistically assessed. Thorough analysis guards against the tendency prevalent among donors to view civil society through a rose-tinted, overly idealistic lens (Box 3). Civil society engagement should be done critically and thoughtfully if donors are to avoid simply exacerbating latent inequalities and prejudices. Although commendable, new demand-side projects like E-ISR+ are still being implemented with rather a one-size-fits-all approach. The reality is that different projects and different sectors will 17 require different approaches. Since DFGG is primarily a political rather than technical solution, cookie-cutter approaches will always remain limited in their effectiveness. F. Persistent Myopia There is a continued myopia that exists in the development field and is reflected in the DFGG approach taken by Bank projects: the prioritization of upwards over downwards accountability. While all of the projects assessed during the review insist on comprehensive reporting systems, financial auditing, and monitoring mechanisms that go up the financial chain or across to imple- menting partners, there is still little evidence of prioritizing such mechanisms in a downwards direction. This leads to a critical gap in the overall DFGG approach. Although all the projects do have some form of non-state actor consultation, the trend is to consult predominantly with more senior figures and established organizations rather than with the average citizen or end-level beneficiary. Within the Copperbelt Environment Project the task team leader observed that this specific tendency had directly contributed to project beneficiaries being uninformed and involved only in a tokenistic fashion. The Environmental Management Framework steering committee, for example, is charged with taking into account “the views and interests of all stakeholders, including inhabitants of mine townships and other members of civil society,” and yet its membership includes only a single representative from a nongovernmen- tal organization. The reality therefore is that key project decisions are really being undertaken by ministry, local government, and mining company representatives. The fact that communities expressed anxiety with regard to the project’s cessation is likely indicative of the lack of consulta- tion with regard to key project decisions. Box 3. CSO as Not Ideal Monitor “Civil society is not the all-encompassing movement of popular empowerment and eco- nomic change portrayed in the revelling and exaggerated celebrations of its advocates. It is simply not a democratic deus ex machina equalizing life chances and opportunities; crippled by material limitations and class impairments it constitutes at best a very uncertain substitute to what had previously been the corrupt and class-based patronage of a more profligate state” (Fatton 1995: 72). Avoiding Tokenism in Demand for Good Governance (DFGG) Activities The preference shown for higher-level stakeholders is often related to attitudes, in particular the extent to which the Bank and project staff really believe that the average beneficiary should have an influence on the project. In the case of the Copperbelt Environment Project, it was predomi- nantly a result of the fact that the task team leader was based out of country and was simply not available to ensure continuous communication with beneficiaries. In other projects—particu- larly technical ones such as those in the road sector—it is due to the fact that meaningful engage- ment with these broader stakeholders is particularly challenging and resource heavy, especially where significant capacity building needs to take place before the consultations can happen in an informed manner. But, perhaps the most obvious reason for the preference of more senior stakeholders is that inclusivity and equitable access to the mechanisms has often not been high- lighted as a central priority for project management. The fact that none of the mechanisms are systematically monitored for their levels of inclusiveness is testament to this, as is the almost total absence of third-party monitoring mechanisms. 18 In the Public Service Management Support Program, for example, the main method of project information dissemination to broader non-state actors is via a few annual forums and fairs. Some project details can also be found at online websites, and the PSM component disseminates a regular newsletter. There is however no specific attempt made to ensure that the information reaches as much of the public as possible and no monitoring is done with regard to who actually utilizes these avenues. Given this trend among projects, the Malaria Booster Project is a notable outlier. Its unique combination of bottom-up planning and multi-stakeholder implementation means that non-state actor engagement is by design a central element of the project. It ensures that a number of popu- lar participatory structures, which include representatives from communities and civil society in each catchment’s area, are intricately involved in resource mobilization, planning, and imple- mentation of project plans and facilitating information dissemination to communities on public health issues. The effectiveness of this non-state actor-focused design is apparent in view of the broad support and trust that the project receives from the local communities as well as the extent to which beneficiaries are able to participate in a number of key project decisions. Clearly, if the Zambia Country Office (and the World Bank, more broadly) does value participa- tion and consultation with lowers, they should take more initiative to reflect this in their project structures. The natural power imbalance between uppers and lowers means that if left to its own accord, the needs and interest of lowers will inevitably be “sidelined”. The individual project analyses demonstrate that although newer projects are certainly moving in the right direction, such sidelining is still the general trend. AFRICA OPERATIONS SERVICES SERIES III. Conclusion and Recommendations This review has sought to identify some of the broad gaps and trends that are currently characterizing DFGG mechanisms in the Zambia lending portfolio. The overall message is positive in that it points toward generally progressive improvement in these mechanisms, particularly when 19 comparing older and newer projects. Continued work is required before the practice on the ground matches the aspirations being espoused by Bank Management. The approach moving forward needs to be far more comprehensive, strategic, and sustained than it has been so far. Based on this review’s summary of current trends and in keeping with the nature of DFGG itself, the following recommendations are less absolute rules than they are guiding principles. Recom- mendations fall under three broad headings: 1. Be experimental 2. Be analytical 3. Be managerial A. Be experimental The approach to non-state actor engagement needs to be experimental (Box 4). The centralizing of DFGG activities in projects is still too recent and unexamined for the World Bank to proceed without significant humility as well as considerable stress on the importance of monitoring. The emphasis must be on discovering what works and learning from what does not. It means leaving room for failure as well as success, and taking lessons from each. On a more practical level, this means adapting structures so that they enable both evaluation and flexibility. With regard to evaluation, DFGG mechanisms on a Bank-wide level are seriously under-eval- uated.1 The same criticism is clearly pertinent at the Zambian project level. This needs to be Box 4. Be experimental “There is a case for external action, but the intervention needs to be designed in ways that take account of these constraints, with a particular emphasis on a process of discovery and experimentation of what works” (Devarajan, and others 2011: 30). 1 “DFGG is not yet anchored in a systematic evaluation of its impact on government effectiveness, efficiency, accountability, or corruption” Bhargava and others 2011: 21). Avoiding Tokenism in Demand for Good Governance (DFGG) Activities immediately addressed if DFGG approaches are to gain traction. Evaluation of DFGG requires recognition of a number of key factors: • DFGG monitoring requires a long-term agenda since it involves measuring for changes in attitude, value, behavior, and incentives. This reality needs to be taken into consideration when evaluating pilots in the short term so that indicators are realistic. Project task teams could begin by working with the DFGG staff to develop and implement key performance indicators for the activities within their projects that are aimed at enhancing the non-state actors’ voice. This emphasis on the long term has implications for funding as well; there is a need to secure longer-term funds for evaluation of DFGG mechanisms. Better than simply revising a project supervision budget would be to ensure hands-off funds for this type of monitoring (e.g., a trust fund managed separately and dedicated solely to this role) so that conflict of interest issues can likewise be overcome. 20 • Since DFGG mechanisms may not be particularly suited for many of the current monitoring trends at the Bank, which are often short term and highly quantitative, all impact evalua- tions need to consider mixed methods and innovative approaches. The World Bank can learn from NGOs in the development sector, who have already been experimenting with more innovative and flexible models. • Monitoring and evaluation mechanisms need to be specified from the initial design of an intervention. The evidence shows that as long as monitoring commitments remain in the abstract, they will fail to be meaningfully implemented. • DFGG evaluations need to be mindful of measuring mechanisms only for their instrumen- tal value. As mentioned in the background section on downwards accountability, there is also an inherent value that needs to be considered. Even if DFGG mechanisms have no contribution toward development effectiveness, they may still be fulfilling a core value of development in the sense that they redistribute power, influence, and dignity to the power- less. An awareness of this inherent value means the measurement of success is not tailored solely to indicate improved institutions, better projects, and better development outcomes. A case must be made also for evaluating mechanisms based on the level of empowerment and voice that they enable, regardless of their overall contribution to development effectiveness. An emphasis on experimentation also requires an improved willingness amongst project staff and management to be flexible with project realities as they become apparent. Demand-side mecha- nisms are simply not conducive to one-size-fits-all solutions. When DFGG is treated as a “tick box” rather than as a critically conceived strategy, based on in-depth understanding and analysis of the existing context, implementation is likely to be significantly compromised. It is apparent in the Zambian projects that a few of the DFGG mechanisms, particularly those involved around safeguards (e.g., grievance mechanisms, consultative workshops, etc.), have been implemented more with an interest in meeting the requirements of a checklist than as a result of proper atten- tion given to the enabling environment. The tendency to mandate a single approach across an entire portfolio is counterproductive. Not all environments are actually conducive to DFGG engagement, and certainly they are not all conducive to exactly the same approach. “Ignoring the underlying political economy drivers of accountability may have been a principal factor accounting for cases where results were less than satisfactory.” (Devarajan and others 2011: 31) In this respect, the Bank could learn more from the experience of international NGOs that have sought to pursue downwards accountability and beneficiary responsiveness while avoiding the temptation to become overly bureaucratic. For many theorists, what has been most encourag- ing about recent initiatives is the very radical step that they have taken to move away from AFRICA OPERATIONS SERVICES SERIES accountability systems based solely on externally dictated rules and move toward ones born out of internally nurtured principles. This is the case with something like ActionAid’s Account- ability, Learning, and Planning System (ALPS) where the creation of room for staff judgment and the development of moral sensitivity is placed at the very center of the initiative (Guijt 2004). Such methods are particularly applicable to DFGG initiatives; since “enabling environments” are constantly in flux, it is absolutely critical that experienced staff are given the space to use their judgment accordingly. It is only if such flexibility is allowed that project teams can use timing and pragmatism to their best effect. Though initiatives like ALPS are unlikely to be perfectly suited to the World Bank environment, the underlying message is instructive: DFGG objectives are likely to be best achieved when they are characterized by a preference for principles (mechanisms that shape and drive rather than dictate) over rules (mechanisms that are absolute and inflexible). 21 B. Be analytical Mechanisms are reliant on their enabling environments. Alone, they achieve little and can even be counterproductive. As such, it is insufficient to take a tool and implement it without consider- able analysis having first been conducted of the existing context and characteristics of demand and supply. These initial analyses need to be a key part of any pilot. They are a key way of ensur- ing that a “tick box” approach is avoided. The key point to recognise here is that a DFGG tool or mechanism is unlikely to ever succeed in isolation; its effectiveness is reliant on it building upon certain preconditions and an existing enabling environment. Figure 2 is instructive in this regard since it demonstrates that the mechanism itself is only one factor in the success of DFGG; there also needs to be demand for accountability on the side of citizens, capacity to account on the side of institutions, a conducive political and legal environment and a host of other enabling factors. Figure 2: Necessary Factors for Success of DFGG mechanisms Source: adapted from Agarwal and Van Wicklin 2011 Avoiding Tokenism in Demand for Good Governance (DFGG) Activities Box 5. Questions to consider Has the ministry expressed willingness to engage citizens more in the project? What are the levels of decentralization in your sector? Have verbal and/or written commitments been made to improve non-state actor engagement in certain areas? Are there specific NGOs in your sector who are capable of engaging in third-party monitoring? 22 While diagnosing the socio-political context of their project, staff should therefore be focused on identifying key entry points for mechanisms within their projects (Box 5 for questions to consider). Based on these entry points, project staff will be able to specify the roles and respon- sibilities to be detailed in the project implementation manuals as well as to determine the level or conditionality they can attach to DFGG mechanisms. If it has not already been done, project teams can consult with the DFGG team to identify the specific governance issues and risks. They can then work alongside DFGG staff to select appropriate DFGG tools for addressing the identi- fied risks—that might include political interference, elite capture, corruption, and absenteeism. Project teams can also seek to become more informed by taking advantage of existing DFGG resources (such as online learning modules) that exist within the Bank. For Bank Management, it will be necessary to outline a countrywide DFGG approach that leaves room for customiza- tion at the sector and project level. This should begin with the country strategy and program and include the following (World Bank 2011): • A good understanding of social, political, and institutional dynamics (including historical legacies) and of the history of engagement between state and society; • An assessment of the formal enabling environment for DFGG, including rules and regula- tions as well as practice, which may differ significantly in some settings; • A mapping of actors (including but not limited to civil society groups). Analyzing the environment also means analyzing possible linkages, since mechanisms work better when they involve a number of mixed stakeholders. It would be just as undesirable for DFGG mechanisms to be associated only with end-level beneficiaries or NGOs as it would be for them to only involve higher-end stakeholders and supply-side actors. Instead, a broad mix of stakeholders is required: “experience has shown that high-quality DFGG work should not be limited to NGOs, but should also involve groups such as parliamentary accountability institu- tions, the judiciary, business or professional associations, think tanks, and others” (Bhargava and others 2011: 11). Assessing linkages also means prioritizing mechanisms that involve collaboration with rather than opposition to government. Persuasion, where it is possible, is a more effective means of reform than outright denunciation: “It is particularly important to work not just with civil soci- ety on the one hand and government on the other, but to focus on the interface between the two—creating spaces and mechanisms that link citizens and the state and help to make their actions as constructive, effective, and equitable as possible” (Bhargava and other 2011). As such, Bank Management should facilitate an approach that implements demand-side mechanisms in parallel with supply side (e.g., work at enhancing investigative capacity among media while AFRICA OPERATIONS SERVICES SERIES also encouraging implementation of Freedom of Information legislation). All of this requires significant pre-analysis. Being analytical also means knowing what DFGG interventions are being (or have previously been) undertaken in the relevant country, sector, or project, thus avoiding “reinventing the wheel”. It is no secret that the World Bank is far behind much of the NGO community when it comes to experience of demand-side mechanisms. In Zambia, there are many NGO projects in various sectors that are enhancing the voice of non-state actors. It would be counterproductive for the World Bank to implement pilots without first assessing exactly what exists and partnering with more experienced organizations where appropriate. C. Be managerial 23 The World Bank aims to change the external culture as it relates to norms around transparency, governance, and accountability. This is a laudable pursuit, of course, but the summary of project trends strongly suggests that internal normative change has not yet been sufficiently accom- plished. As observed, downwards accountability still receives far less priority than upwards or horizontal accountability. Given the relative powerlessness of lowers, this is unsurprising. It has in fact been the predominant trend across development projects globally (Ebrahim 2005). Unfor- tunately this myopia means some of the DFGG mechanisms are being implemented ineffectively. There is an apparent tension within the Bank between those who think DFGG should only be considered for its instrumental value and those who think that it is inherently valuable and should be implemented regardless of its outcomes. Such disconnects highlight the fact that the organizational culture, skills, and resources of the Bank need to be carefully managed if they are to provide an environment conducive to an acceleration of DFGG approaches. The following are broad ways in which the World Bank could manage internal dynamics and thus accelerate the internal enabling environment for DFGG: • Managing DFGG resources sufficiently. As previously highlighted, downwards account- ability, unlike upwards, is not essential for organizational survival. Being discretionary, it is often the first thing to be cut when resources (time or financial) are low. If the Bank does highly value downwards accountability, then it needs to back up its claims by dedicating significant time, sustained funds, and experienced staff to building appropriate relation- ships and implementing appropriate mechanisms. Under-resourcing of these mechanisms ultimately reflects under-prioritization. It is critical, for example, that a sustainable funding mechanism be put in place before the Government Partnership Facility2 ends. With regard to human resources, it is also critical that there be a better alignment between the skill set of World Bank staff and the governance agenda. There are few staff in the Zambia office who are sufficiently cognizant of DFGG approaches. Improving skills can begin with project teams ensuring that country office staff are trained in DFGG approaches and tools.3 In the longer term it may mean actively seeking a bigger DFGG team in country offices. It will also be worth considering how certain task team staffing norms affect the logistical ability of the Bank to adequately consult with lowers. For example, the fact that various task team leaders are based in Washington rather than in country and that they are often covering many proj- ects at once clearly contributes toward a marginalizing tendency, whether intentional or 2 The Government Partnership Facility is the multi Donor Trust Fund currently funding DFGG activities in Zambia. 3 Take advantage of existing courses (DFGG core course, ICT for DFGG, DFGG in CAS, DFGG in DPL etc), communities of practice, case studies, how-tos, brown-bag lunches, etc. Avoiding Tokenism in Demand for Good Governance (DFGG) Activities Box 6. Managing attitudes and values “Transforming systems of power is the key to a sustainable future in which all people can live in dignity and fulfilment, but it is impossible unless we ourselves are also transformed” (Edwards and Sen 2000: 23). not. When such task team leaders visit the country on their short-term missions, it is hardly feasible for them to find the necessary time and resources that are required to assess benefi- ciary engagement in their projects. Their natural default is to consult with higher-level staff. 24 • Managing attitudes and values. Focusing on DFGG-related strategies, structures, and systems is necessary, but it is not sufficient. The World Bank needs to implement a clear organizational value system that provides depth, stability, and consistency to their man- agement systems. With downwards accountability it is clear that the soft subject of values has direct implications for the hard facts of performance. The implementation of demand- side governance mechanisms hinges on the extent to which project staff and implementing partners value multi-stakeholder engagement. The implication of such a relational agenda is that project staff need to have the proper attitudes and sensitivities to engage in those relationships appropriately (Box 6). It also means that there needs to be a consistent mes- sage regarding DFGG, its definition, its importance, and its anticipated effects coming from higher-level Bank Management. Currently there is still confusion regarding DFGG, particu- larly with regard to why it is important. In short, downwards accountability begins with people and perceptions, not mechanisms. The Bank needs to have project teams who value and understand demand-side mechanisms and who also value and understand people. Bank task managers, for example, must be persuaded that citizen oversight is both important and worthwhile. Again, part of this will mean regularly dem- onstrating, via evaluation, why DFGG is integral to aid effectiveness. More accountable projects are expected to lead to better development outcomes; more effective institutions, processes, and systems; and overall better projects. But it is not enough to assume these results; they must be monitored and proved to those who are sceptical. It may also mean building an appreciation among staff as to the importance of an inherent value that is often accomplished regardless of overall aid effectiveness. A consistent value framework needs to be constantly maintained, disseminated, and re-assessed. This sustained management is particularly important since DFGG approaches are a marked shift from the Bank’s traditional supply-side engagement. As such, any increased attempts at engage- ment with the demand side are likely to require a significant change in the Bank’s organizational culture (Box 7). At some point, it would be prudent to consider more innovative means of manag- ing staff values. Again, the Bank cannot hope to change external culture relating to transparency and accountability if it is not first managing internal values regarding the same. Recognizing that downwards accountability is intricately linked to personal attitude (Box 8), Bank Management could consider encouraging models of self- and peer-assessment. Managing staff values will likely mean creating space and opportunities whereby staff are encouraged to examine contra- dictions, personal behaviors, and attitudes. In the long term, it will require a more consistent attempt to assess and address the Bank’s organizational identity as a whole. AFRICA OPERATIONS SERVICES SERIES Box 7. Organizational values “[O]rganisational values are too important to be taken for granted. They need to be identified, articulated, and revisited periodically. The compatibility between organisational values and management systems (as they actually work) must constantly be verified. Indeed, exercises in changing organisational structures or management systems must be seen as serving the purpose of reinforcing the organisation’s value system.” The task of assessing organizational values requires the following steps: • Identifying the predominant belief clusters among a critical mass of people in the organisation; 25 • Assessing the extent of consensus among them; and • If necessary, identifying the forces or mechanisms by which prevalent value systems are maintained or may be altered. Source: Padaki (2000) Box 8. Accountability and personal attitude “For it is not they who have been wrong, but us. The first step then is humbling. It is to recognize our ignorance and error. Gradually and none too soon, development professionals are coming to see that the problem is more ‘us’ then ‘them’. It is with ourselves that we have to start” (Chambers 1993: 32). AFRICA OPERATIONS SERVICES SERIES References Agarwal, Sanjay, and A. Van Wicklin III. 2011. “HOW-TO NOTES: How, When and Why to use Demand-side Governance Approaches in Projects”. World Bank. Agarwal, Sanjay, Rasmus Heltberg, and Myrtle Diachocl. 2009. “Scaling-Up Social Accountabil- ity in World Bank Operations.” Social Development Department. World Bank. Washington, D.C. Bendell, J. 2006. “Debating NGO accountability.” NGLS Development Dossiers. United Nations. 27 Beuran, Monica, Gaël Raballand, and Kapil Kapoor. 2011. “Political Economy Studies: Are They Actionable? Some Lessons from Zambia.” Policy Research Working Paper. WPS 5656. Africa Region. World Bank. Washington, D.C. Bhargava, Vinay, Kit Cutler, and Daniel Ritchie. 2011. Stimulating the Demand for Good Gov- ernance: Eight Strategic Recommendations for Intensifying the Role of the World Bank. The Partnership for Transparency Fund. Washington, D.C. Chambers, Robert. 1993. Challenging the Professions: Frontiers for Rural Development. ITDG Publishing. . 1997. “Whose reality counts? Putting the first last.” London: IT. Chase, Robert S., and Anushay Anjum. 2008. “Demand for Good Governance Stocktaking Report: Initiatives Supporting Demand for Good Governance (DFGG) Across World Bank Group Sectors and Regions”. World Bank. Davis, A. 2007. “Concerning Accountability in Humanitarian Action”. Humanitarian Practice Network. Google Scholar. Devarajan, Shantayanan, Stuti Khemani, and Michael Walton. 2011. “Civil Society, Public Action and Accountability in Africa.” Policy Research Working Paper, WPS 5733. Develop- ment Research Group. World Bank. Washington, D.C. Ebrahim, Alnoor. 2005. “Accountability myopia: Losing Sight of Organizational Learning.” Non- profit and Voluntary Sector Quarterly 34, no. 1: 56. Edwards, M., and G. Sen. 2000. “NGOs, Social Change and the Transformation of Human Rela- tionships: A 21st-Century Civic Agenda.” Third World Quarterly 21, no. 4: 605–616. Ellerman, D. 2001.“Helping People Help Themselves: Toward a Theory of Autonomy-compatible Help”. World Bank Group. Ellerman, D., and A. O Hirschman. 2006. Helping People Help Themselves: From the World Bank to an Alternative Philosophy of Development Assistance. University of Michigan Press. Fatton, Jr., R. 1995. “Africa in the Age of Democratization: The Civic Limitations of Civil Society.” African Studies Review (1995): 67–99. Avoiding Tokenism in Demand for Good Governance (DFGG) Activities Guijt, I. 2004. “ALPS in Action: A Review of the Shift in ActionAid towards a New Accountability, Learning and Planning System.” ActionAid International. London. Kilby, P. 2006. “Accountability for Empowerment: Dilemmas Facing Non-governmental Organi- zations.” World Development 34, no. 6: 951–963. Muller, Mike, Robin Simpson, and Meike van Ginneken. “Ways to Improve Water Services by Making Utilities More Accountable to their Users: A Review.” Water Working Notes No. 44225. World Bank. Washington, D.C. www.worldbank.org/water Padaki, Vijay. 2000.“Coming to Grips with Organisational Values.” Development in Practice 10, no. 3: 420. 28 Singh, Janmejay, and Agarwal Sanjay. 2011. DFGG at the Project Level & An Overview of Tools and Methods. Presented at the DFGG & OBA clinic. World Bank, April 2011. Slim, H. 2002. “By What Authority? The Legitimacy and Accountability of Non-governmental Organisations.” The Journal of Humanitarian Assistance 10. Stockton, N. 2006. “The Accountability Alibi.” Humanitarian Exchange 34. Wallace, T., L. Bornstein, and J. Chapman. 2006. The Aid Chain: Coercion and Commitment in Development NGOs. Practical Action Pub. World Bank. 2011. “Strengthening the World Bank’s Engagement on Demand for Good Gover- nance: A companion piece for the GAC-2 Report”. Draft. World Bank Group. . 2009. Governance and Anti-Corruption in Lending Operations: A Benchmarking and Learning Review. Quality Assurance Group . nd. DFGG in Bank Operations. Social Accountability and Demand for Good Gover- nance. Social Development Web page. http://go.worldbank.org/09MJLAICW0 . 2004. World Development Report 2004: Making Services Work for the Poor. Washing- ton, D.C.: World Bank and Oxford University Press. . 2002. The Empowerment and Poverty Reduction: A Sourcebook, PREM, World Bank, Washington, D.C. . 2000. World Development Report 2000: Attacking Poverty. Washington, D.C.: World Bank and Oxford University Press. Zoellick, Robert B. 2011. “The Middle East and North Africa: A New Social Contract for Develop- ment.” The Peterson Institute for International Economics, April 6, 2011. http://go.worldbank. org/LL2NNYC2F0 AFRICA OPERATIONS SERVICES SERIES Annex 1. Methodology for Assessment Criteria for selecting case study projects All projects that are currently active within the current lending portfolio were eligible for review. The projects, including relevant details that were correct as of June 2011 (when the initial review 29 was undertaken), are listed in the following table: Sources of information • World Bank documentation: project appraisal documents, project implementation plans, operational manuals, performance framework, environmental and social impact assess- ments, aide memoires, mission documents, project evaluations, midterm reviews, special studies on governance and accountability; • Task team leaders and task team members; • Beneficiaries and stakeholders (via E-ISR plus). Data collection methods • Collation of World Bank documentation (project appraisal documents, Implementation Sta- tus and Results Reports, briefs, etc.); • Semi-structured interviews with task team leaders; • Desk-based review of previous beneficiary evaluations (E-ISR+). Quality assurance • Triangulation/ Multiple sources were used wherever possible • Multiple methods (interview, desk-based documentary analysis) were utilised to enable cross-checking • Advice and recommendations were successfully sought from task team leaders to ensure that material being covered remained useful to researcher and the Bank. Limitations of review The initial review on which this paper is based was conducted by one consultant, on a part-time basis over a period of roughly 10 weeks. As such, it does have some limitations. The most sig- nificant limitation is the absence of any field component in the review. Sources of information are limited to World Bank documents, interviews with Bank staff, and project evaluations as car- ried out by Bank-financed consultants. The absence of direct contact with beneficiaries or other external stakeholders (e.g., government or civil society organizations) means that the consultant’s ability to triangulate across multiple sources is limited. That said, there was a stakeholder evalu- 30 # Project name Proj. ID Additional financing Funding Team lead Status Approval date Closing date 1 Support for Economic P071407   IBRD/IDA Sheppard, W. Active 29-Jul-04 30-Nov-11 Expansion and Marie Diversification (SEED) 2 Water Sector Performance P10458 Water Sector Performance IBRD/IDA Tavares, Luiz Active 16-Apr-09 N/A management Improvement Project – Claudio Martins Additional Financing P071259 Water Sector Performance IBRD/IDA Active 5-Oct-06 30-Jun-12 Improvement Project 3 Increased Access to P121325 Additional Financing for IBRD/IDA Rezaian, Active 22-Sep-10 N/A Electricity Zambia Increased Access to Abdolreza Electricity Services Project Avoiding Tokenism in Demand for Good Governance (DFGG) Activities P077452 Zm: Increased Access to IBRD/IDA Active 20-May-08 31-Dec-13 Electricity 4 Malaria Booster P120872 Malaria Booster Project IBRD/IDA Vledder, Active 9-Dec-10 N/A Additional financing Monique P096131 Zambia Malaria Booster IBRD/IDA Active 15-Nov-05 31-Jan-13 Project 5 Copperbelt Environment P070962   IBRD/IDA Coombes, Peter Active 20-Mar-03 31-Mar-11 Project John (continued) (continued) # Project name Proj. ID Additional financing Funding Team lead Status Approval date Closing date 6 Road Rehabilitation and P120723 Road Rehabilitation and IBRD/IDA Runji, Justin Active 14-Oct-10 N/A Maintenance Maintenance Second Additional Financing P106596 ROAD REHABILITATION IBRD/IDA Active 16-Oct-09 30-Jun-14 AND MAINTENANCE PROJECT PHASE II P093611 Zambia: Road Rehabilitation IBRD/IDA Active 29-Mar-07 N/A & Maintenance Project – Addtional Financing P071985 Road Rehabilitation and IBRD/IDA Active 9-Mar-04 30-Jun-12 Maintenance Project 7 Agricultural Development P070063   IBRD/IDA Ekanayake, Active 16-May-06 31-Mar-14 Support Program Indira Janaki 8 Public Sector Management P082452   IBRD/IDA Palale, Patricia Active 5-Jan-06   Program Support Project 9 Irrigation and Development P102459   IBRD/IDA Ekanayake,  Not yet 2010 2016 Support Project Indira Janaki active AFRICA OPERATIONS SERVICES SERIES 31 Avoiding Tokenism in Demand for Good Governance (DFGG) Activities ation that had been conducted for 2 successive years by an external research company on a majority of the projects under review and the data gathered during this process acted as a fairly independent verification tool for those projects at least. Given the desk-based nature of the review, the timeline, and the absence of discussion with external stakeholders, there were also some limitations in the consultant’s ability to knowingly recommend project-specific improvements. Comprehensive, project specific recommendations will require additional in-depth field analysis of the enabling environment and focus group dis- cussions with associated stakeholders. 32