More aid is flowing to the world’s poorest countries and more of it than ever is passing through multilateral institutions. In December 2007, the World Bank’s International Development Association (IDA) received the largest expansion of donor financing in its history. Over the next three years, it will have access to $41.6 billion.1 Eighty countries, with a combined population of 2.5 billion people, are currently eligible to receive this assistance. Over half of these countries are either prone to or affected by violent conflict. They pose the greatest challenges to sustainable development and are the most difficult contexts for all donors, including multilaterals, to work in.
‘Fragile’ countries are often affected by deep undercurrents of discontent and division. Where these erupt into instability and violence, donors may be forced into a major rethink of their country and region-wide strategies and previous development gains are reversed. This has been shown by recent events in Kenya and those in Ethiopia in late 2005—places not traditionally identified as ‘fragile’ or ‘conflict-affected’. For the 40-50 states labelled as such,2 the Bank is among a number of donors where some staff are arguing for an improved international approach.
In the Bank over recent years, two specialist units have produced conflict guidance for country teams, and have pushed for better institutional approaches and financing in Fragile States. A new operational policy has been agreed and senior managers have begun hosting conferences on fragility as well as internal ‘retreats’ on conflict. In late 2007, as part of the Bank’s latest strategic review, President Robert Zoellick set up a task force on the institution’s response to fragility, and the Bank is in the process of changing its institutional set-up relating to conflict and fragility.
The key question, however, is whether the Bank’s programmatic prescriptions have remained largely the same regardless of the emerging knowledge and the lessons of hard experience. Are the Bank’s internal systems continuing to undermine its overall effectiveness?
Issues for the World Bank
Issue 1: Understanding how fragility affects operations and how operations affect fragility
• Informal institutions, processes and power relations are underplayed in Bank decision-making
Informal ‘institutions’, power relations and ‘systems of meaning’, such as caste or ethnicity, have profound impacts on governance and access to economic opportunity. Yet the Bank, institution wide, underplays the importance of these dynamics and the effects of its activities on them. The Bank’s main analytical tool (the Country Policy and Institutional Assessment—CPIA) centres primarily on formal institutions and policies. Despite staff awareness of power dynamics and local politics, as an institution the Bank’s approach to economic reform and project activities can still be overly technical. It can even appear detached from the conflict realities of the context. This was the case with the second tranche of the Poverty Reduction Support Credit debated in Nepal in late 2005. Targeting deeply sensitive issues such as labour relations and fuel subsidies, the PRSC conditions risked further destabilising a volatile political environment.
• Multidisciplinary analysis is needed for internal decision-making
In recent years, the Bank has sought to improve its knowledge on a wider range of contextual dynamics. For example, with regards to Nigeria, the Bank is equipping itself with a deep assessment of political economy and social issues. Multidisciplinary analysis (including on social impacts) is, however, not yet systematically integrated into internal decision-making processes. Sensitivities amongst member countries mean that necessary changes to the CPIA, which is shared with partner governments, remain some way off. Such sensitivities should not, however, prevent deeper, multidisciplinary analysis from being embedded in the internal working practices of country-, research- and results-assessment teams.
• The Articles of Agreement should not prevent engagement on political economy and social issues
Article V states that the Bank must ensure the ‘proceeds of any financing are used. . . Without regard to political or other non-economic influences or considerations’. The clause affects the Bank’s work in all its operating contexts and is sometimes used as a rationale for pursing ‘development’ as if it can be isolated from politics. There is, however, a fundamental conceptual difference between avoiding the politicisation of IDA and ensuring that IDA commitments are based on sound analysis and are adapted to political economy and social factors. Article V is, in fact, already inconsistently interpreted across the Bank due to a recognition by many staff that insufficient attention to such dynamics leads to superficial and unsustainable outcomes.
Issue 2: Adapting the approach to the context
• Alignment should not be rushed
Even more, perhaps, than bilateral donors, the Bank is guided by a principle of country ‘ownership’. Under its constitution, it is obliged to agree on projects with partner governments. However, even if governance capacity-building remains a vital task, it should also be recognised that the ‘State’ can be a major part of the problem. With the alternative approach of ‘shadow’ alignment remaining ill-defined and their own vested interest in ‘low transaction costs’, the Bank and donor partners need to take care not to rush into formal alignment with government. There are a number of reasons for this including:
- Weaknesses in oversight and accountability systems and in perceptions of a shared ‘political community’ mean that government ownership should not be equated with country ownership. Government may be riven with factional interests and personality politics;
- An agenda agreed with a Ministry of Finance or a National Planning Commission does not, in and of itself, amount to government-wide leadership. Such bodies may not be able to marshal other institutions of the executive, making activities difficult to sustain; and
- Competition within government may prevent progress through sector-wide approaches (swap) particularly where the swap relies on cooperation among a number of ministries (such as for rural infrastructure).
To highlight these challenges is not a call to avoid government systems and leadership. What is vital, however, is again to be clear about political dynamics and to provide close accompaniment to implementation.
• An over-focus on the executive government capacity may undermine progress
Sustainable outcomes depend on patient engagement with all major stakeholders. This is necessary so as to help change the ‘cultures of power’ that have been driving conflict. For example, to deliver improved livelihoods to beneficiaries, the ongoing coffee-sector reform in Burundi needs to taken slowly. As an industry with a long history of dominance by state decisionmakers, time and close engagement by the Bank is needed to build understanding and capacity amongst coffee farmers and their federations—i.e. Beyond the executive. Without it, reform could actually serve to entrench monopolistic control of the coffee sector by a small well connected elite. Although the frequency and expertise of IMF and Bank supervision missions is positively seen, the Bank must continue to be careful to avoid ‘supply-driven’ timelines which may rush the process.
Issue 3: Matching human resources to operational realities
• Bank teams need to include a wider range of disciplines
In recent negotiations over IDA15, major donors such as the UK have pushed the Bank to pay greater attention to social analysis and social impact assessments.3 With quantitative evidence as the primary criteria for decision-making, the Bank tends not to systematically focus attention on the human dimensions of the context (culture, traditions and other ‘systems of meaning’). These often have to be covered by discrete trust funds and/or specific ‘social development’ projects. The provision of in-country conflict sensitivity expertise, too, has largely been on a short-term basis thanks to supplementary bilateral resources—for example from Finland and then Sweden in the Bank’s Sri Lanka office. Yet economic and social analysis should be seen as complementary, allowing quantitative and qualitative approaches to be brought together. The Bank needs to rethink human resource policies in order to strengthen country, review and research teams with this wider range of expertise.
• Work in fragile countries is expert labour intensive
Effective engagement in fragile contexts requires not only knowledge but adequate staff numbers. From some quarters in the Bank (supported by DFID and certain other bilateral donors), there are welcome efforts to increase staffing and improve incentives for staff to work in such countries and reduce staff turn-over. However, senior management and, indirectly, IDA donors sometimes apply pressure for country teams to meet spending targets (measured by the quantity of disbursements) or otherwise face a reduction in the funding available from the administrative budget. Additional obstacles to ‘context-driven’ staff resourcing include the global Bank effort to reduce the overall administrative budget in absolute terms, and the ‘flat’ regional allocation system (which sees, for example, one country office’s administrative budget fall if another office in the same region is allocated a rise—and the rest of the region remains the same). This system defies the lessons of experience in dozens of countries.
Even if every fragile context is different, research (including from within the Bank) suggests that turnaround comes through slow, incremental change which may be subject to the occasional setback. As the achievements and failures of the Bank-managed Multi-country Demobilisation and Reintegration Program (MDRP) in Burundi show, almost every objective—national and local—depends on the skilled and delicate application of expertise and leverage to unblock political obstacles, and tackle long-established systems of power and hierarchy. The level of the administrative budget, globally and for country teams, has not yet been adapted to ensure these outcomes. Stronger staffing for the MDRP, for example, relied on additional finance from leading bilateral partners. Anchored in formal institutions and in policy rather than real practice, the CPIA does not sufficiently reflect underlying ‘cultures of power’ and ‘rules of the game’. In fragile contexts, these are particularly important determinants of the operating environment. Senior management in the Bank should, therefore, not base decision-making on administrative budget allocations primarily on the CPIA. The Post-conflict Performance Indicators, which are substituted for the CPIA when deciding on exceptional post-conflict funding, also need to be improved as a mechanism for determining financial allocations. The categories of ratings for the PCPI need to be further disaggregated and nuanced to reflect the complexity of the relevant operating environment.
• A ‘mission’ culture remains problematic
‘Task Team’ trips from Washington to the field put a considerable burden on national counterparts and country office staff to organise themselves around the visit. In addition, these missions are generally not frequent or long enough to unpack complex political and technical issues, including at local level, and sustain engagement on them. The Bank’s ongoing decentralisation is welcome but the need remains for senior staff to be located closer to the power politics and stakeholders that relate to their project. This would mean also that they would be better able to ensure closer scrutiny of local administrations and implementing contractors—primary determinants of the quality of project implementation.
Issue 4: Defining effectiveness
• Efficiency must not be conflated with effectiveness
The Bank has made significant progress in emphasising issues of ‘uneven development’ and ‘exclusion’. Its recent discourse in Nepal and Sri Lanka, as well as the planned focus of the 2009World Development Report on the geographic distribution of assistance (‘spatial disparities’), exemplify this shift. It is also helpful that certain ‘Corporate Advocacy Priorities’, such as ‘Empowerment, Security and Social Inclusion’, are now available for staff to highlight their attention to the human dimensions of economic development. Substantial Bank investments in community development projects such as the Poverty Alleviation Fund in Nepal are now seeking to address these broader development goals.
However, the current ‘results’ framework still falls short in integrating a long-term change agenda with regards to political economy and social relations, including gender equality. For country assistance as a whole and for projects, ‘results’ remain recorded largely on the basis of technical statistical measurements, with insufficient articulation of spatial, gender or identity group distribution. Instead, the Bank’s focus on ‘economic’ development, combined with external pressure for it to spend still more money, tends to make the speed and absolute quantity of disbursements the main benchmark by which ‘success’ is measured. This is compounded by the ‘scaling-up aid’ agenda. Too much focus is then put on the endogenous, internal systems-related factors of a ‘successful’ project and not enough on exogenous issues in the context. Yet it is the latter which determine the quality of outcomes over time for the country’s peoples.
• It is important to be candid about the quality of data
Due to problems of access or verification, it may not be possible to collect or rely on data. This has been the case, for example, in Nepal and Sri Lanka in recent years. It is a fundamental issue for an institution that focuses heavily on quantitative information. Where data is out of date or insufficient, decision-making needs to recognise this and be adapted accordingly.
• There are ‘right’ and ‘wrong’ kinds of risks
Under the current ‘results’ system, staff may be incentivised to do the following:
• Avoid certain kinds of risk—such as applying a slow incremental approach to the thorniest issues in the most unstable districts. If the Bank shies away from the hardest challenges, this may exacerbate actual or perceived disparities of the kind that are driving conflict in Nepal, for example. It may mean, as with Bujumbura Rurale in Burundi, that a province is not prioritised even where its volatility threatens development progress country-wide. It risks also that gender equality is only ‘tacked-on’ as a concern.
• Take the wrong kinds of risk. Notwithstanding the importance of preventing state bankruptcy and ensuring payment of public sector (and security service) salaries, relatively ‘easy’ risks can include rapid moves towards general budget or sector-wide support. They can also involve multi-million dollar project disbursements to government or quasi-statal recipients with little demonstrable interest in ‘pro-poor’ development.
It may be that the ‘easier’ risks prove unsustainable or unworkable over time, while the hardest ones, despite setbacks, have the most transformative impacts over the long term. This has implications for staff accountability.
Issue 5: Bank and/or donor?
Among board members and internally, there is a core tension as to the Bank’s role. Is it purely a financial institution that provides lending to recipient governments, or a donor that uses money and expertise to help ‘transform’ the development contexts in which it works? With a deepening IDA discourse about providing grants rather than loans in the poorest countries and a strong commitment to ensure ‘pro-poor’ development, many of the Bank’s operations show it to be increasingly playing a ‘change agent’ role. Improved performance in fragile countries will partly depend on how well all its stakeholders come to a shared understanding about this role and its implications.
What needs to change
a) Ensure decision-making is based on political economy and social analysis
The Bank should make multidisciplinary analyses mandatory in internal processes to formulate country strategies and to decide the projects that flow from them. These analyses should cover formal and informal institutions, processes and power relations. They should assess local, national and region-wide issues.
b) Adapt the size and use of the administrative budget
The Bank needs enough staff with the right skills to engage on the range of issues that determine project outcomes. IDA donors and the Bank management should agree that:
- Decisions on the allocation of the administrative budget should be informed on the basis of the Bank’s internal context analysis. Staffing should be determined on the basis of what it takes, in any given setting, to deliver sustainable outcomes rather than on the level of disbursements and the performance of the Bank’s current portfolio.
- Multidisciplinary expertise should be embedded in country, review and quality control teams (such as the Quality Assurance Group) as well as the Development Research group (DRG) of the Development Economics Vice Presidency (DEC). Discrete donor trust funds should not be relied on to achieve this. It should be prioritised in the administrative budget. Headquarters must, however, be careful, however, not simply to add further layers of evaluation to an already quite heavy set of processes. The key will be to ‘streamline’ those processes while getting the balance right.
c) Amend the way ‘results’ are measured
The way results are defined has a profound impact on Bank processes, outputs and outcomes. This is what determines staff incentives. The Bank needs to reassess the 14 country outcome indicators in its Results Measurement System (RMS) and the nine indicators for monitoring IDA’s contribution to country outcomes. These should integrate political economy considerations and pay greater attention to social cohesion and how to empower people (men and women) to access economic opportunity and participate in decision-making. Governance and social development experts should join the experts in DEC in compiling and appraising the data. To institutionalise this change, it is vital that the Poverty Reduction and Economic Management (PREM) Vice Presidency, so dominant in the Bank, is fully engaged in the process.
d) Improve internal and external accountability
To systematise a different approach to ‘results’, there needs to be a shift in the internal accountability within the Bank. Accountability for staff involved in designing projects but whomove on in the early stages of implementation should be applied and run at least for the fulllength of the project. Those who manage the implementation, or who assume the running of theproject after the formulation stage and lead it through all its ups and downs, should be assessedfairly in light of the complex contextual challenges. In addition, given the pivotal role of countrydirectors in the Bank system, their accountability for ensuring conflict-sensitive approaches,mutual coherence among operations as well as attention to the region-wide context, isparticularly important.
Despite its commendable degree of transparency, the Bank needs to dedicate more time and effortto improve external accountability. It must ensure (i) two-way communication with communitiesbeyond urban elites or other individuals with internet access, and (ii) greater involvement ofdeveloping country parliamentarians in discussing and monitoring Bank-funded activities. Thereare, of course, significant infrastructure- and capacity-related obstacles to these goals, but theBank must nonetheless invest more time and resources in reinforcing expectations ofaccountability and responsiveness.
e) Improve collaboration between the Bank and bilateral partners
Bilateral donors need to establish sufficient staff resources and skills in order to engage in a proper appraisal of the Bank’s conflict sensitivity. Bilateral staff numbers in Washington representations (additional to those paid out of the already overstretched administrative budget), in headquarter IFI-related units as well as in the field must be adjusted in order to take on this work and make it mutually reinforcing with their own objectives.
In addition, it may be useful for these bilaterals donors to agree on ‘focus’ countries. This might then involve a pre-identified representative of the donor’s Executive Director team noting that a strategy or a major project for that country is coming up for appraisal by the Board. He/she would then activate his/her own system of officials, as well as other bilateral partners, in order to gather together comments from hqs and field missions. To work, this system would have to build in enough time for the collation of inputs.
It will certainly be problematic for bilateral donors to IDA to resource this kind of additional scrutiny due to budget constraints within bilateral agencies. Also, it would be far preferable if the Bank itself had the systemic capacity and incentives so as to make such scrutiny unnecessary. In the short to medium term, however, investment to ensure mutually reinforcingin-country, and region-wide, work among bilaterals and multilaterals should be a top priority. The benefits of such coherence (including its ‘multiplier effect’ on outcomes) and the necessity for bilateral government to meet their commitments to their taxpayers outweigh the costs of heavier strategy and project processes in the Bank. Over time, it would be expected that the emergence of improved staff incentives in the Bank would allow bilaterals to cut back on their additional staff costs.
Conclusion
In conflict-affected and fragile contexts, governance and economic systems very often favour a small group of competing elites who also determine policy formulation and public expenditure. Such systems are immense obstacles to the kind of ‘open access’ societies on which sustainable poverty reduction depends. Yet, because the Bank’s decision-making remains focused on formal policies and institutions and is structured in line with the orthodoxies of ‘alignment’ and ‘results’, its effectiveness is impaired.
It is vital that the Bank’s performance criteria and administrative budget institutionalise a multidisciplinary approach to the challenges of its operating contexts—in analysis, implementation and evaluation. Without such shifts, the Bank will continue its tendency to take ‘context as the starting point’ only on an ad hoc basis on the initiative of individual senior staff members. The Bank is improving its discourse on these issues but, institution-wide, the quantityof assistance still remains the principal determinant of ‘success’. Yet the hard lessons of experience underline that it is as much the process as the expenditure that matters: the how, not the how much.
