Room Document 7 DRAFT Synthesis of Guidance and Key Policy Messages on Aid, Accountability and Democratic Governance OECD-DAC-GOVNET

Programme on Improving Support to Domestic Accountability

This work represents a collective effort in the OECD-DAC Governance Network in collaboration with partners in developing country accountability institutions e.g., parliaments, civil society organisations, political parties, and the media - to explore citizen-state relations and better understand the impact of aid on domestic accountability. It provides direction on how to take more strategic approaches to support for accountability systems (budgeting/service delivery in sectors) and provides principles for democratic governance assistance in areas of electoral assistance and support for parliaments, the media, political parties, etc. It is based on country case studies in Mali, Mozambique, Peru and Uganda, a range of literature and donor innovations in this field since mid2009 and a series of high-level dialogues on trends in support to accountability. May 2011 The Management Group on Aid and Accountability will follow through to further develop and then finalise the document and related plans for implementation, following discussion and review in GOVNET on 7 June 2011. GOVNET members should aim to draw out key messages from this work also reflecting on their own internal efforts - for the High Level Forum on Aid Effectiveness in Busan, end-2011 (see also RD 8).

MEETING of the DAC NETWORK ON GOVERNANCE OECD, Paris 6-8 June 2011
1

Draft Synthesis of Guidance on Aid, Accountability and Democratic Governance May 2011

Key Policy Messages

Over the last two decades, support to strengthen domestic accountability has been a growing component of donor engagement in developing countries, with donors spending over $10 billion annually on governance support and accountability a common component of this. But it is currently at a crossroads. There is increasing recognition that much of this support has made inaccurate assumptions about the nature of democratic and institutional transitions. This has translated into commitments to Western models and technical approaches that use blueprints, templates and best practice approaches. More recently, this has been challenged by calls to work with the grain of societies and to develop country specific strategies which represent the best fit (Centre for Future State 2010). This requires changed roles for external assistance, with greater emphasis on facilitating or convening locally driven reform processes. Support to domestic accountability now needs to go beyond business as usual approaches. This will involve engaging with some challenging realities. Greater recognition is needed of the extent to which patterns of aid dependence or other international drivers can contribute to weakened domestic accountability. Taking context seriously needs to sit at the heart of programme design and much greater attention needs to be paid to the incentives and power dynamics of all actors involved in accountability reform processes. How should external actors navigate this complex landscape more effectively? First and foremost, external actors should not weaken relationships between states and their citizens. Total levels of aid and the way that aid is given can shape the wider governance and accountability context, particularly in aid dependent countries. Donors need to take care to recognise the impacts of where and how aid is given, and take action where accountability to donors outbalances accountability to citizens. The Paris Declaration committed to increase levels of programme based aid to improve country ownership and donor alignment. In reality, ad hoc portfolio approaches have emerged, which combine project and programme based aid. Wherever possible, aid should work towards being programmatic, in that it should work in support of national strategies rather than donor priorities, regardless of the specific modality chosen and more strategic use of aid portfolios, based on assessments of the context, should be prioritised. Increasingly some donors are bringing accountability benchmarks into budget support debates. DFID, for example, has made this a specific commitment, with an 2

amount equivalent to 5% of budget support to be spent on accountability support in each country. Mutual accountability should be taken much more seriously, with donors realising commitments to improve their own accountability to aid recipient countries. Governments should be open about what they expect from donors and in return, donors should be open about how their own internal processes and political considerations will impact on how aid is delivered. While donor assistance seeks to strengthen systems of domestic accountability through external support, those systems should eventually operate without any outside help. Ultimately, the purpose of donor assistance should be to remove the need for continuing support. This will require exit strategies for any support programme. Greater coordination among a range of external actors is needed. Any donor accountability support or engagement should be grounded in complementary diplomatic policies that seek to nurture or reinforce commitments.

Second, donors should stop doing what does not work. Too often, support to domestic accountability has been characterised by the use of ideal models which work badly in the contexts in which they have been applied, and technocratic best practice solutions which take insufficient account of political realities. Donor engagement needs to accept different starting points in each country and work with the grain of local institutions and reformers, rather than importing models. This will require much greater realism, both about the reform space for accountability in each country and the longer timeframes involved in realising transformational institutional reform (World Bank 2011). Because change happens slowly, support should be based on multi-year commitments. The assurance of a long-term presence is likely to enhance relationships with local partners and increase the chances of genuine impact. Donor support needs to move away from supply driven, top down forms of assistance, often targeted only at formal accountability institutions. The way organisations are funded is important - the projectisation of accountability support has resulted in local organisations and groups effectively competing for funding rather than cooperating to support change processes. Instead, support should incentivise cooperation and collaboration. To be fit for purpose to deliver more politically aware accountability support, agencies need to address their own staffing rewards, so that staff are incentivised to understand country contexts, through training, time spent in one country and performance measures.

Third, the next step is to understand that domestic accountability is dynamic and functions as a system, which brings together a range of actors and institutions. Unless attention is paid to the way in which the system functions, support which targets a single actor or institution can unbalance that system as a whole.

To date, donors have focused overwhelmingly on support to a select number of accountability actors (civil society, audit institutions and increasingly parliaments and judiciaries). This has meant very little coordination of policy or programming across different actors and institutions and little understanding of how support to one shapes the operating environment for others. Adopting a systems approach should start at the analysis and design stage of programming. For example, the design of parliamentary support needs to work from the position of the parliament within the overall system of domestic accountability, as well as examining the parliaments internal procedures, resources and operations. The use of political economy approaches and tools can be particularly useful. They allow attention to be paid to power dynamics and incentive structures. But these must be operationally relevant forms of analysis, which address particular issues, sectors or problems. Wherever possible, analysis should be shared by donors and this should provide the underpinning of any subsequent division of labour discussions. Adopting an accountability systems approach does not necessarily mean ever more holistic programming (for example, bringing together all aspects of accountability support). But it may mean more support to build links across actors or groups and supporting coalitions of reformers rather than particular individuals or groups in isolation. For some forms of support, such as electoral assistance, it will be particularly important to be as comprehensive as possible, so that support works across the many domestic institutions and sectors that are involved in an electoral process. Viewing accountability as a system allows for the recognition that it is dynamic and changes over time. Examples such as Twaweza, which makes use of both new and old technologies to expand citizens ability to access government information and hold leaders accountable in East Africa signal how accountability systems are evolving in many countries, and how they are shaped by technological transformations, among many other factors. More could be done to promote citizens access to media and mobile technologies as well as citizens media literacy. This includes access to media products and infrastructure as well as the ability to make sense of information. Information flows are therefore crucial and can shape the incentives of different actors to respond or be held accountable. Transparency is rightly recognised as an important building block for accountability. But it cannot be viewed in isolation and it is not, on its own, sufficient to realise accountability. In particular, greater support is needed to strengthen how information is analysed and acted upon examples of Parliamentary Budget Offices are a good example of how this capacity can be supported for parliamentarians.

Fourth, support to domestic accountability can get most traction when it is anchored in specific sectors. Budget monitoring initiatives in health and education or the monitoring of global standards in oil, gas and mining have often gone furthest in realising concrete accountability reforms. Working in a particular sector reinforces a systems approach as it provides a clear rationale for bringing together a range of actors (inside and outside the state). It 4

requires moving beyond narrow supply versus demand divides towards engaging with coalitions of reformers around a specific development issue or challenge. A sectoral approach allows for greater recognition of the need to engage formal and informal institutions. Accountability initiatives which work only with formal institutions may exclude some of the most important players. For example, traditional chiefs can be significant actors in the provision of security or justice or in relation to key social services, with implications for the workings of accountability systems. While direct support can be challenging, there are options to work with intermediary organisations, to provide training or capacity development, or to support initiatives which work with hybrid institutions. Any accountability support in a sector needs to be much more effective at linking national and local level accountability initiatives and processes. For example, too often budget monitoring initiatives remain isolated to specific districts or regions and are not integrated into national level reform processes. Instead, opportunities to bridge local and national accountability gaps need to be identified.

Fifth, support which tackles some of the deeper drivers of state society relations, such as taxation and its relationship to citizenship, should increase. That tax can play a key role in building a social contract between citizens and the state is well recognised, as when citizens invest in the state, it can increase the incentives for responsiveness. Two factors in particular determine the impact of taxation on the relationship between government and citizens - the degree to which a state relies on tax versus non-tax sources of revenue and how the government taxes citizens (Moore 2007). Donor support has been slow to connect tax reform to accountability support. .At present, only around 0.1% of donor assistance supports taxation (OECD DAC statistics). Resurgent interest in tax issues represents recognition of the centrality of the tax system to accountability processes but also growing interest in domestic resource mobilisation to progress towards an eventual exit from aid dependence. Support can cover various types of technical assistance. More innovative examples of support include facilitation of national level dialogue (with government, CSOs, parliamentarians and others) on levels of taxation and domestic resource mobilisation. Aid modalities choices present a range of options for inclusion of tax issues. For example, forms of budget support (general and sector) can incorporate assessments of tax performance as part of their performance indicators.

Finally, measuring results and evaluating the impact of accountability support is key and does not have a strong track record. Donors face increasing pressures to demonstrate results. This is necessary and has been a significant gap. Pressures to record and monitor impact should work with, not against, achieving improved and sustainable development outcomes and strengthened accountability institutions in the longer term. Donors own accountability to their taxpayers should reinforce accountability to countries in which they provide aid commitments to 5

improved transparency offer a good example of the mutual benefits of improved accountability to different stakeholders. Ongoing risk monitoring and management needs to be much more central within all forms of accountability support. For example, elections assistance providers must respond to the recent rise in electoral and post-electoral violence by giving greater attention to electoral risk and the long-term causes of political violence that ignite in election processes. Approaches which make explicit (and test) the theories of change which underlie programmes for accountability are particularly helpful. There is no silver bullet for evaluation methods instead, a combination is needed, which can bring together forms of both quantitative and qualitative measures where appropriate. Crucially, evaluation should be seen as part of a learning process for strengthening accountability engagement in the future.

Table of Contents
1 2 Introduction.................................................................................................................................... 8 From donor accountability to domestic accountability ............................................................ 9 2.1 2.2 3 Towards portfolio aid approaches.................................................................................... 10 Realising mutual accountability ........................................................................................ 12

Moving away from supply driven assistance.......................................................................... 14 3.1 The use of political economy analysis ............................................................................ 16

Working with domestic accountability systems ...................................................................... 18 4.1 4.2 Adopting a systems approach .......................................................................................... 21 Facilitating information flows ........................................................................................... 22

5 6 7 8

The benefits of a sectoral or issue driven approach ............................................................. 23 Engaging with the drivers of domestic accountability ........................................................... 27 Measuring results ....................................................................................................................... 30 Key lessons and recommendations ........................................................................................ 34 8.1 8.2 8.3 Implications for programme design ................................................................................. 34 Implications for programme implementation .................................................................. 35 Implications for monitoring and evaluation ..................................................................... 37

Annex 1: Principles for Improving Donor Assistance to Support Domestic Accountability ..... 38 1) 2) 3) Principles on International Elections Assistance ............................................................... 40 Principles for political party support ..................................................................................... 42 Principles for media assistance ........................................................................................... 46

4) Principles for support to parliaments .................................................................................... 49 Annex 2: Country case studies ........................................................................................................ 52 Annex 3: Political economy frameworks ......................................................................................... 52 References .......................................................................................................................................... 53

Introduction

The OECDs Development Assistance Committee (DAC) mandated its Network on Governance (GOVNET)1 to explore ways to improve donor support to domestic accountability in a programme from 2009-2012. As part of this, GOVNET has commissioned research to explore the realities of aid and domestic accountability, involving country case studies in Mali, Mozambique, Peru and Uganda, which involved interview based fieldwork supplemented by reviews of relevant policy and academic literature.2 In addition, it held a series of meetings and gathered expert advice to develop a set of principles for international engagement with domestic accountability actors. This guidance note synthesises the findings from this work and sets out key recommendations. It uses a definition of accountability which refers to the relationship between decision-makers and those affected by decisions, involving answerability, enforceability and transparency. Answerability occurs where decision makers justify their actions and respond to demands; enforceability is the sanctions which may be used if decision makers do not meet their commitments or fail to meet certain standards; and transparency refers to information regarding the commitments made by decision makers and whether and how these commitments have been met (Hudson and GOVNET 2009; Domingo et al 2009). There is a particular focus on two forms of accountability; first and foremost, domestic accountability and, to some extent, mutual accountability, as well as the links between the two. Conventional approaches to support to domestic accountability have undoubtedly led to a number of beneficial impacts, but two key challenges have been identified by a growing body of evidence (see Hudson/GOVNET 2009). Firstly, not enough attention has been paid to the specific challenges and opportunities in different contexts, including a lack of understanding of the wider political and economic context in which accountability processes take place. Secondly, the majority of donor support has focused on building the capacity of specific actors (civil society, parliaments) largely in isolation from their relationships and inter-actions with other actors and institutions (Ibid.). Progress towards more accountable governance, however, requires building links between actors and constructing strong constituencies or coalitions for change that involve civil society, the media, parliaments, political parties and a range of other institutions on both the demand and supply-side. Thus, this guidance note presents an accountability systems approach to allow for greater attention to these dynamics in different contexts, and to ensure greater attention to the relationships and networks between actors and institutions. At the same, it acknowledges that the majority of donor support in this area still tends to be designed according to support for specific actors or institutions. Therefore, it compliments a systems approach with a set of core principles for support to specific institutions.

1 2

www.oecd.org/dac/governance A summary of the research methodology and information on each case study can be found in Annex 1

2 From donor accountability to domestic accountability

Aid and other forms of development assistance can disrupt the relationships between citizens and their governments in developing countries. This has been an important recognition in recent years, not least in the commitments made at Paris and Accra. There is clear evidence that total levels of aid and the way that aid is given can shape the wider governance and accountability context, particularly in countries which receive high levels of aid This means that donors need to take care to recognise the impacts of where and how aid is given, and to take action where accountability to donors outbalances accountability to citizens. The need to deliver aid in ways which allow developing countries to own their development efforts, and to be held to account for those efforts, was a core component of the Paris Declaration. This led to commitments to improve ownership, particularly by linking aid to agreed priorities set out in national development strategies and plans. Concepts of ownership are now being re-examined. Too often, commitments to ownership have been undermined by the ongoing influence of different external actors. Ultimately effective development processes, and processes of institutional reform leading to greater accountability, need to be endogenous and locally driven. This is fundamentally a political process and it requires external actors to play supporting roles, in facilitating or convening key players. Operating in this way requires a reorientation of mindsets and aid models, from technical approaches to more political outlooks that recognise the power dynamics and incentives that can drive reform. Too often, this has not been openly acknowledged and has meant that accountability support has struggled to engage with the political dynamics and systems in which they operate. Addressing this means taking ownership seriously in relation to accountability support and designing support so that it works with locally driven reform processes rather than promotes external models. This is discussed in more detail in the following sections. More fundamentally, external actors should not obstruct relationships between states and their citizens through their aid relationships. Donor accountability should not be stronger than domestic accountability as this potentially further weakens internal accountability relationships. Ultimately, donor assistance can seek to strengthen systems of domestic accountability through external support, but it should be explicitly recognised that the ultimate goal is for those systems to eventually operate without any outside help. This means that clear exit strategies need to be developed, based on key milestones or measures of progress, and particular care should be paid to issues of sustainability to ensure support achieves long term institutional transformation. This is not only applicable to donor agencies and development practitioners. Any donor accountability support or engagement should be grounded in complementary diplomatic policies that seek to nurture or reinforce commitments. This has been particularly noted in relation to support to more obviously political actors, such as political parties, but should be seen as general rule of thumb for all forms of accountability engagement. This may be easier for agencies which sit as part of their ministries of foreign affairs, but even in these circumstances, strategies need to be designed to ensure that development and diplomatic engagement is mutually reinforcing. 9

2.1

Towards strategic aid portfolios

The way aid is given can significantly shape its accountability impacts. While the Paris Declaration committed to increase levels of programme based aid, in reality a variety of programme and project based approaches have continued. This has led to portfolio approaches which often combine a range of different modalities in each country. In some countries, these portfolio approaches have worked well, with the use of targeted project support (for example, to address a particular area of weakness within the accountability system) complementing programme based approaches3. Ensuring that different aid modalities are complementary, in relation to accountability and more broadly, however, will require much greater transparency and more substantive collaboration between donors. Wherever possible, aid should work towards being programmatic, in that it should work in support of national strategies rather than donor priorities, regardless of the specific modality chosen. Some donors are bringing accountability benchmarks into budget support discussions and performance indicators. DFID, for example, has made this a specific commitment, with an equivalent of 5% of the total amount spent on budget support committed to accountability support in each country. These approaches have yet to be tested, and are still being translated into operational plans, but they are likely to rise in number as bilateral donors, in particular, cite governance concerns in relation to the use of budget support. More broadly, leveraging improvements in domestic accountability through conditions in programme based aid should be adopted with care. In each country, it should be based on analysis of the political incentives for reform, whether there are existing government commitments to reform and whether there are champions to take reform forward. Lessons should be learnt from political economy analysis of reform processes (see Grindle 2002; 2004). The use of conditionalities linked to governance or accountability reform should be treated with care. Recent experience in Mozambique in relation to General Budget Support is instructive in this respect (see Box).
Box: The use of General Budget Support in Mozambique (OECD 2011d) In 2010, donors temporarily suspended General Budget Support in response to governance concerns. The government responded by making a series of commitments in a political matrix. Donors subsequently renewed GBS funding. The results of this crisis are mixed, however, and raise important lessons for future government-donor dialogue. Analysts have warned that threatening to cut GBS harmed trust between government and donors and that GBS proved too blunt an instrument for incentivising improvements in domestic accountability. Instead, they suggest that donors need to mix aid modalities and conditionalities to influence governments performance on governance and accountability.

While project aid is likely to continue as a common modality, and can help to address important financing gaps, they can pose significant challenges to the budget process itself and to national strategies for development. Where aid is provided off budget and information about that aid is not made easily available, it undermines the scope for domestic accountability. This means that large areas of health or education spending, for example, can effectively operate outside the oversight of key domestic actors such as the government,
3

See OECD Mozambique study, Annex 2 and OECD (2011d).

10

parliamentarians and civil society groups. In addition, the use of separate reporting structures and processes can create additional burdens for government and nongovernment actors, where reporting lines flow from the implementing agency back to the donor only, resulting in little wider dissemination on aid commitments and disbursements as examples from the health sector in Uganda illustrate (OECD 2011a). Assessing the merits of portfolio approaches reveals the need for much greater nuance in defining on- and off-budget aid modalities. Simplistic dichotomies in reality fail to capture the wide range of modalities in use. The Collaborative Africa Budget Reform Initiative (CABRI) usefully presents a more refined definition, which outlines seven different dimensions of onbudget aid (see Box).
Box: Dimensions of capture in on budget aid (CABRI 2009) Seven dimensions of on budget aid: On plan: Programme and project spending is integrated into spending agency planning documentation On budget: Programme and project aid (and its intended use) are reported in budget documentation On Parliament: Aid is included in the revenue and appropriations approved by Parliament On Treasury: Aid is disbursed into the main revenue funds of government, and managed through government systems On accounting: Aid is recorded and accounted for in the governments accounting system On audit: Aid is audited by the governments audit system On reporting: Aid is included in ex-post reports by government.

In practice the degree of capture varies widely across modalities (and even within the practices of individual donors), but it is useful to adopt this wider range of distinctions, as aid which is labelled on budget can in fact be on plan but not approved by parliament, audited by the governments audit system or included in ex-post reporting. It also allows for recognition that donors are working in more programmatic ways even where aid is not formally classified as on budget. Recent Global Fund experience in Uganda is illustrative in this respect. Following a corruption scandal in 2005, the Global Fund developed a set of arrangements with the Ugandan Ministries of Health and Finance to allow for some use of country systems, albeit with elements of earmarking (see Box). These arrangements effectively allow Global Fund resources to work within government priorities and systems. This impacts on the wider accountability system too, as Global Fund assistance is now audited by the Office of the Auditor General.4 It suggests that a more open acceptance of the intermediate steps between budget support and project support could have important implications for accountability. Further research is needed into the implications for domestic accountability of the finer distinctions offered by CABRI.

Progress here is partial, as Global Fund audits are not included in the consolidated annual audit report to the Public Accounts Committee; however, stakeholders recognised the added value of government access to Global Fund financial information.

11

Box: Long-term Institutional Arrangements (OECD 2011a) The Long Term Institutional Arrangements (LTIAs) were developed on the principles of the Sector Wide Approach (SWAp), which stipulates all funding for the sector supports a single sector policy, strategic agenda and expenditure programme, under government leadership, adopting common implementation, monitoring and evaluation mechanisms across the sector and progressing towards relying on government procedures to disburse and account for all funds. The LTIAs recognise that budget support remains the preferred mechanism for the Government of Uganda. Global Fund, support is channelled through the budget as general budget support but ring-fenced in the Poverty Action Fund (PAF) mode. MoFPED remains the principal recipient for all health grants, while a Civil Society Fund may be used for productive CSO involvement. Operational budgeting and planning under the LTIA is guided by the 2001 Budget Act, with decisions on resource allocation based on the work of the technical working groups and budget working groups, with participation of civil society 2.2 private sector.mutual accountability and Realising

In theory mutual accountability, a key component of the Paris Declaration, should be a key vehicle to address imbalances in accountability. It aims to shift accountability from a largely one way focus to a two way or mutual process. In practice, however, it has proved one of the most challenging commitments to realise. Accountability to donors far outweighs the accountability of donors, and this skews partner countries own accountabilities away from their citizens and towards donor requirements. This can be seen in patterns of implementation of different aid modalities and of mutual accountability mechanisms. In theory, budget support modalities or programme based aid approaches provide additional incentives for realising mutual accountability, as they are based around frameworks of indicators for both government and donor performance. But in practice, donor requirements for disbursal have meant that their conditions hold greater weight. Indicators for government performance are often treated differently by donor agencies, with some relying on analyses of aggregate performance, others on a select number of indicators, and others on additional benchmarks, undermining the usefulness of these frameworks as single, comprehensive sets of indicators against which governments will be held to account. In addition, indicators for donor performance are often less comprehensive and there are very limited enforcement mechanisms to address where these indicators are not met. Project based aid approaches typically offer reduced scope for mutual accountability requirements. In principle, even where funds are channelled outside government systems there should still be scope for governments to jointly evaluate or assess projects. But in practice, donors usually conduct these evaluations on their own, with limited involvement of domestic stakeholders. Away from specific aid modalities, the links between mutual accountability mechanisms and domestic accountability have remained weak, although there are areas of progress. Mutual accountability has been largely conceived as a technical agenda for agreeing shared agendas and measures of performance between governments. But it has ensured spaces for dialogue and debate which have opened up greater opportunities for a range of national stakeholders to participate, including in sector working groups which have increasingly included parliamentarians and/or civil society representatives (Steer at al 2009). Moreover, strengthened domestic accountability, such as parliamentary oversight of national development strategies and results monitoring, can improve the climate for partnership with donors. However, these links between domestic and mutual accountability have not been capitalised on. Instead, there are growing examples of parallel agendas and monitoring systems, with potentially conflicting lines of accountability (Ibid.). Examples from 12

Mozambique reveal the challenges this poses (see Box). Thus careful consideration is needed, to determine how mutual accountability can be made complementary to domestic accountability and where possible how mutual accountability mechanisms can link into domestic accountability processes.
Box: Policy dialogue in Mozambique (OECD 2011d) Donors are perceived as dominating the space for political and policy dialogue in Mozambique, hindering the participation of parliamentarians and civil society in domestic accountability. Some analysts argue that the space for this dialogue is usurped by donors increasing role in sectorbased working groups and GBS joint reviews. Although civil society groups, for example, are part of the Joint Review process, they can lack the capacity and resources to have significant influence. In a country with a strong ruling party and a centralised state, fora for dialogue around aid could potentially serve as a platform to allow civil society and parliamentarians to challenge policies and hold government to account. This is reportedly starting to happen; for example, donors have begun to support civil society in preparing for Joint Reviews.

This should be accompanied by firmer commitments to mutual accountability overall. Governments should be open about what they expect from donors and in return, donors should be open about how their own internal processes and political considerations will impact on how aid is delivered. Lessons from Rwanda are relevant here, as the government has set up a system for monitoring and rating the performance of multilateral and bilateral donors (see Box). Opportunities should also be sought out to deepen the involvement of a range of domestic accountability actors in substantive ways in relevant mutual accountability mechanisms. In 2008, the Government of Rwanda, along with donors, developed the Donor Performance Assessment Framework (DPAF) which provides a joint tool for the monitoring of donor performance against national-level and international commitments (on the volume and quality of aid). The DPAF is presented in aggregate form (comprising all donors) and disaggregated by donor to allow for comparison, individual reflection on performance, and peer pressure. The DPAF for 2009/2010 indicated improvements at the aggregate level in a number of indicators, but most indicators did not meet targets set for FY 2009/2010 or the Paris Declaration
Box: The Donor Performance Assessment Framework, Rwanda (Government of Rwanda 2010) Targets for 2010. In 2010/2011, all of the official Development Partners in Rwanda agreed to use country systems, including treasury, reporting, procurement and auditing systems, and to communicate monthly projections of disbursements.

Committing to make aid information easier to access, use and understand for governments, parliaments and citizens, in their own and developing countries, is also important for realising both mutual and domestic accountability. There is evidence of significant improvements in aid transparency, as more accurate data is available as a result of budget support frameworks and more ad hoc reforms. This seems to be particularly the case at the sectoral level, where sector working groups have often established regular interactions between donors and a range of domestic stakeholders. In Uganda, for example, the creation of a simplified reporting spreadsheet for donors in the Local Development Partners Group has helped to collate, in one document, donor commitments and actual disbursements on a quarterly basis, searchable by sector (OECD 13

2011a). This echoes trends in other countries towards the sharing of aid information and attempts to collate it in usable ways, for example through Aid Management Policies or Platforms (see Box). Furthermore, signatories to the International Aid Transparency Initiative (IATI) have agreed to make information about aid spending easier to access, use and understand. These signatories should now fulfil these commitments, including providing more accurate, timely and comprehensive information on aid commitments and disbursements (see Box).
Box: Improving aid information Aid Management Platforms

Aid Management Platforms are Web-based applications that allow governments to better manage and coordinate development assistance. Software developed by the Development Gateway Foundation provides a virtual workspace where governments and donors can share aid information from planning through implementation and then analyse this by donor, sector, status, region, timing and other factors
International Aid Transparency Initiative (IATI) IATI is developing international standards for the way donors report information about aid spending. To date, 19 donors are signatories. The IATI standards will include: Agreement on what information will be publish and how detailed this should be A system for categorising different types of aid spending /commitments A common electronic format making it easier to share information A code of conduct on what information donors will publish and how frequently, how users may expect to access that information, and how donors will be held accountable for compliance.

While this represents significant improvement, there are remaining challenges where donors do not provide information in ways which are compatible with how governments record information (Moon and Williamson 2010). Moreover, there has been little focus on support to domestic actors in terms of strengthening their ability to interpret and use aid data. This would be useful as part of broader support to enable stakeholders to interpret and analyse all sources of government revenue (but would only apply to forms of aid which are on budget).

Moving away from supply driven assistance

Where support to domestic accountability has been provided in the past, it has too often been characterised by the use of ideal models which work badly in the contexts in which they have been applied, and represent technocratic best practice solutions which take insufficient account of political realities. For example, the majority of support to parliaments has involved forms of technical assistance and capacity development. This has included technical assistance to draft bills, tailored analysis, support to strengthen Parliaments representation, legislative and control functions. In general, these generic forms of capacity development or technical assistance have not effectively engaged with the wider political context. At times, they have struggled to link support programmes to the realities of the wider political context or to the informal rules 14

of the game. In Mali, for example, parliamentary support has done little to engage with informal accountability systems and traditions of consensual politics which challenge some of the proscribed roles for parliament including in the formal budget process (see Box).

Box: Parliamentary strengthening, Mali (OECD 2011b) Past experiences of colonialism and dictatorship, combined with ethnic and regional diversities, have contributed to a political culture in Mali which emphasises decision by consensus. This works in contrast to models which see parliament as a counter-weight to the executive. This is reflected in the dominance of cousinage relations, in which patterns of political interaction are mediated through reference to familial relations, or the use of so-called joking relationships which allow conflict or tensions to be voiced within in humorous rather than confrontational ways. These all potentially challenge models which posit political competition as a part of decision making (for example as many Westminster multi-party systems do). Moreover, it creates strong informal accountability relationships within and between parties which can be overlooked. Few aspects of donor support in Mali currently engage substantively with these realities.

Similarly, civil society support has tended to focus on well established civil society organisations or Non Governmental Organisations (NGOs). This often translates into funding for Northern based international NGOs or to local level NGO partners (often urban based). This can be limiting, where there are challenges in the perceived legitimacy, representativeness, independence, credibility and sustainability of some of these organisations (Rocha Menocal and Sharma 2008: 26). It has meant that support has overlooked the wide range of actors and organisations which in reality contribute to civil society in many developing countries, including trades unions, social movements and religious groups (see box).
Box: Results of a Citizens Voice and Accountability Evaluation (drawn from Rocha Menocal and Sharma 2008: 26-27) A review of Citizens Voice and Accountability Initiatives (looking across the support of five donors) found a number of examples of non-traditional civil society groups which were effective partners for accountability support. In Indonesia, for example, Islamic mass-based organisations helped to open doors usually closed to secular NGOs. In Bangladesh, social movements such as Samata used entitlement to government land and water bodies to push for more responsive state institutions.

The need to recognise the wider structural and institutional contexts in which different domestic accountability actors sit is applicable across the range of support which is provided for domestic accountability. Political party support, for example, is increasingly recognised as part of a range of accountability support. But parties themselves remain bound to the political systems in which they operate and reflect processes of state formation and political and institutional development (Wild et al 2011). Some of the structural features which shape party development include the nature of the electoral system, informal rules of the political game and wider geo-political histories (Ibid.). This has contributed to a number of widely documented weaknesses, including poor links to citizens, weakly institutionalised structures, fragmented opposition and a lack of robust legal and regulatory frameworks (Carothers 2006). However, standard approaches to political party support often focus on technical assistance which does not little to engage with these deeper, structural challenges (Ibid.). 15

This has led to a disjuncture between the types of assistance provided and the full nature of the underlying accountability problems (Ibid; see Annex 1). Addressing this means moving away from supply driven, top down forms of assistance, often targeted only at formal accountability institutions. Instead, donor engagement needs to accept different starting points in each country and do much more to work with the grain of local institutions and reformers, rather than importing external models. It needs to grapple with the realities of the incentives and power dynamics which shape the structural conditions for accountability in most countries.

3.1 The use of political economy analysis

Forms of political economy analysis can be particularly helpful in tailoring support to different contexts and in ensuring that support addresses a range of wider structural and power dynamics. A growing number of multilateral and bilateral donor agencies are investing in developing their own political economy tools5 or commissioning political economy analysis to inform their programming. According to OECD DAC definitions, political economy is: concerned with the interaction of political and economic processes in a society: the distribution of power and wealth between different groups and individuals, and the processes that create, sustain and transform these relationships over time.6 At its core, it seeks to understand the power dynamics and incentives of key actors. Reflection on political economy issues can help to avoid the use of standard models or blueprints that adopt technocratic approaches and work to achieve best practice solutions at times ill-suited to the contexts in which they are implemented. Political economy analysis has been criticised where it has not been seen to be operationally grounded or has not been seen to offer insights relevant to programming realities. In recent years, problem-driven forms of political economy analysis have been developed to address this challenge. As suggested by the name, problem-driven analysis begins with the identification of a particular problem, opportunity or vulnerability (often arising from specific operational challenges). This process helps to narrow the scope of the analysis, and to drill down to understanding those institutional arrangements that are most relevant and how those arrangements lead to the particular problem identified. Specific and feasible options for reform can then be identified, which either seek to work within existing spaces for reform or which attempt to increase that space. These feasible options for reform are often likely to be second-best options, rather than first-best textbook solutions (Fritz et al., 2009). A number of donors, including the UK, the Netherlands, the European Union and the World Bank have begun to develop their own tools, frameworks and approaches to political economy analysis. The World Bank, for instance, has developed a framework for problem driven governance and political economy analysis (see Table 1). This usefully assesses the existing institutional and governance capacities for a given issue and then assesses the political economy drivers which underlie this, to identify why things are they way they are and what feasible options for reform might be. See Annex 2 for a selection of other frameworks and approaches.
5

See for example: World Bank (Problem Driven Framework); EC (Sectoral analysis framework), UNDP, DFID (How To Note) 6 http://www.oecd.org/document/8/0,3746,en_2649_34565_37957768_1_1_1_1,00.html; Collinson 2003

16

Table 1: Problem-Driven Governance and Political Economy Analysis

Vulnerabilities/ challenges? Institutional & governance arrangements & capacities Evidence of poor outcomes to which GPE weaknesses appear to contribute What are the associated institutional set-up and governance arrangements? E.g.: repeated failure to adopt sector reforms, poor sector outcomes. Mapping of relevant branches of government, ministries, agencies, and SOEs and their interaction. Existing laws and regulations. Policy processes (formal rules and de facto). Analysis of stakeholders, incentives, rents/ rent distribution, historical legacies, and prior reform experience

Problem Driven GPE Analysis

Political economy drivers

Why are things this way? Why are policies or Institutional arrangements not being improved?

Source: World Bank, 2010 Political economy approaches emphasise more realistic, incremental action to support accountability, grounded in countries political realities. In particular, they highlight the need to pay close attention to the interaction between formal and informal rules and institutions, as key components for understanding how accountability really works in a given context. Where political economy insights have been examined, they have revealed the potential for different ways of working to support accountability. Emerging political economy insights on social accountability, for instance, illustrate some significant implications for changing practice. Conventional approaches have often sought to respond to weaknesses in the supply side of state accountability through increased attention to the demand side, including forms of local democratisation and citizen-empowerment or social accountability initiatives. In the case of basic service delivery, for instance, the need to strengthen the direct accountability of service providers to users has been promoted, through a range of social accountability initiatives (World Bank 2004). Yet recent political economy grounded research suggests that establishing formal mechanisms for participation or strengthening (through providing resources and skills building) of the demand side cannot on its own support the forms of bottom up accountability envisaged under some of the standard approaches (Centre for Future State 2010; Rocha Menocal and Sharma 2008; Booth 2010). Instead, it can be most helpful to identify bridging channels that bring together citizens and the state, and address incentives for bottom up and top down reform. The work of the Centre for the Future State, for instance, has emphasised that support to a particular set of actors (such as CSOs) is not particularly effective and instead, support should be directed to broad based alliances which bring together a range of actors with common interests in reform (and which cross public-private divides) (Centre for Future State 2010: 45). Therefore, accountability support needs to be informed by a realistic approach to power dynamics and incentives that shape pressures for more or less accountability. Political economy approaches also reveal that the way organisations are funded is important, as it shapes their incentives and motivations. The projectisation of accountability 17

aid has meant that in many countries accountability actors, such as CSOs, effectively compete for funding rather than cooperating to support change processes. In some countries, this leads to a number of accountability initiatives operating in the same district or region with overlapping remits but which are largely uncoordinated, rather than coordination to increase the coverage and scope of these activities (OECD 2011a; OECD 2011b). Adopting political economy insights, as the above analysis suggests, will require much greater realism, both about the reform space for accountability in each country and the longer timeframes involved in realising transformational institutional reform (World Bank 2011). Because change happens slowly, support should be based on multi-year commitments. The assurance of a long-term presence is also likely to enhance relationships with local partners and increase the chances of genuine impact. While there is growing recognition of the utility of political economy analysis, there are remaining challenges for donor agencies in terms of their own incentives and organisational cultures which can mitigate against the uptake of these findings. Ostrom et al (2001), in one of the most in-depth studies of the institutional incentives of a donor agency (in this case SIDA), highlighted how information asymmetries, rapid staff turnover and pressures to disburse all created incentives which mitigated against attempts to foster strong understanding of the context in which aid was delivered. Aid spending imperatives can also mitigate against more realistic and politically feasible ways of working (which can take longer and may suggest lower levels of funding over longer time periods). To be fit for purpose to deliver more politically aware accountability support, donor agencies need to address their staffing rewards and incentives, so that staff are incentivised to understand country contexts and adopt nuanced programming approaches, including greater use of innovation and piloting where standard approaches have been found wanting. Improving staff handover procedures and increasing the amount of time spent in one country, to encourage the accumulation of local knowledge and its retention in the institutional memory, would support this. Attention should also be paid to the types of skills needed to operate in more politically savvy ways, both in the recruitment and training of staff (Wild and Foresti 2011).

4 Working with domestic accountability systems

The majority of support to domestic accountability is structured around support to specific actors or institutions. In the main, it has been focused on support to state responsiveness on the supply side, such as audit institutions or parliaments and to the demand side, principally through support to CSOs. Some cross cutting support has been provided, commonly linked to specific, time bound events, such as electoral assistance. Many providers of external assistance to accountability are likely to continue to provide funding in separate funding lines to these different groups, organisations and institutions. Where this is the case, lessons should be taken on board for how to achieve more effective practice. As a first step, a set of principles for support to parliaments, political parties, the media and electoral processes is set out in Annex 1. These principles make clear that the design of support to any of these organisations must start with an assessment of the wider system or context and must take care to consider the implications of support for other actors or institutions within that system. They provide a guide for how to achieve more effective, and more politically aware, programming in these areas. The common themes from these principles are summarised in the box below. 18

Box: Support to domestic accountability actors General principles which apply to all support: Take context seriously and align support programmes with wider accountability objectives Establish realistic and long term objectives Ensure local ownership and work with incentives for reform Pay attention to gender issues and inequalities

Principles on International Elections Assistance: Be alert to electoral risk the long-term causes of political violence Ground electoral assistance in complementary diplomatic policies but dont instrumentalise it Recognize the role of regional organisations Be as comprehensive as possible

Principles for Political Party Support: Be aware of but not paralysed by the sensitivities of party aid Build on the interconnections between party aid and other elements of political aid. Dont confuse party diplomacy with party aid Dont assume common goals between providers and recipients

Principles for Media Assistance: Incorporate media indicators and audits into governance diagnostics and needs analysis Cooperate with media development CSOs and determine media objectives and outcomes, not methodologies Support independent, sustainable, and capable local media in developing countries Support systematic research on the effects of media and information access on domestic accountability Learn about and harness new technologies

Principles for Support to Parliaments: Focus on institutional change leading to behavioural change Understand parliaments incentive structures Dont ignore political parties Identify and address the causes of underlying parliamentary weakness

Taking on board the principles set out in Annex 1 will be an important first step. However, they also reveal the limitations of working in a purely siloed approach. Much support to date has not effectively grappled with the wider political context in which it is implemented. There has been very little coordination of policy or programming across these different actors and institutions. More worryingly, there has been little understanding of how support to one can shape the operating environment for others. This means that conventional modes of accountability support does not adequately capture the extent to which domestic accountability is dynamic and functions as a system. These systems bring together a wide range of actors and institutions alongside information flows and patterns of incentives. 19

This was a clear finding from recent OECD DAC research in Peru, Mali, Mozambique and Uganda, which look a closer look at the workings of accountability around specific entry points, namely budget processes and service delivery7. Looked at from this perspective, it was clear that in practice accountability processes brought together a range of actors and groups within a given accountability system. For example, a supreme audit institution needs a level of capability and resourcing to prepare its reports, but it also draws on information produced by government systems; its reports are used by relevant Parliamentary committees; they are in turn used by CSOs to disseminate more widely; and they should be taken on board by Ministries of Finance or other relevant bodies. Similarly, parliamentary support is often designed and implemented separately from support to political parties. However, one of the strongest influences on behaviour in parliament will be political parties and this is commonly excluded from conventional parliamentary support models. These country case studies offered some interesting insights in this respect. In the case of both budget processes and service delivery in Uganda, accountability systems seem to be unbalanced with actors poorly embedded within that system. There is wide recognition of progress in terms of the strengthening of financial systems and processes, and the growing capabilities of some key actors such as Office of the Auditor General, the Public Accounts Committee in Parliament, the Ministry of Finance and some CSOs. However there are remaining weaknesses, where the recommendations of these are not acted on by others, where particular actors struggle to fulfil their oversight duties (such as Parliament) and in the growing divergence between improvements to formal systems at the national level and the realities of service delivery and budget execution at local levels (OECD 2011a). In Peru, given ongoing decentralisation processes, it was found that intervening with many actors but in a few targeted regions offered the best opportunities for promoting multi-actor systems of accountability while at the same time supporting decentralization (OECD 2011c). This was accompanied by calls to move beyond one actor support and to focus on the links between state entities and citizens (Ibid.). Similarly in Mali, accountability was again found to work as a system wherein the actions of one actor impacted the functions and capacities of others. For example, the Bureau du Vrificateur Gnrals ability to audit Government directly linked to the quality of information it was able to collect from line ministries; likewise, civil society and parliament were better able to call government to account if they had access to the Bureau du Vrificateur Gnrals reports (OECD 2011b). In education, the successful functioning of local school committees was found to require effective multi-stakeholder processes, bringing local councillors together with citizens and decentralised education administrators (Ibid.). Thus the relationships and mutual accountability between these various actors suggested that a system approach to strengthening accountability would better align the demand with the supply sides of accountability in Mali (Ibid.). In Mozambique, donors have continued to support individual institutions through capacity development focused on specific functions rather than building accountability relationships between institutions. Stakeholders argued that donors needed to do more to consolidate relationships and networks where they already existed. For example, the Poverty Observatories, which bring together government, civil society and international partners, could be strengthened to become more than consultative bodies and to offer a substantive channel for government to respond to, and answer, citizens concerns (OECD 2011d). This builds on and reinforces a growing of body of research examining the interrelationships between accountability actors, rather than focusing on particular actors in isolation (Centre
7

See Annex 3 for an overview of this research and relevant resources.

20

for Future State 2010; Rocha Menocal and Sharma 2008; Joshi 2010). This has been supported by recent studies of local level accountability, which have also found that the effectiveness of local domestic accountability is largely determined by the relations and interplay between political and social actors (VNG International/African Studies Centre 2011; Booth 2010). In the future, support to accountability should therefore work towards adopting an accountability systems approach, which recognises that support needs to engage with these wider systems of accountability, and with coalitions of actors within them, rather than narrowly focusing on specific groups or institutions.

4.1

Adopting a systems approach

Adopting a systems approach should start first and foremost at the analysis and design stage of programming. For example, the design of parliamentary support needs to work from the position of the parliament within the overall system of domestic accountability, as well as examining the parliaments internal procedures, resources and operations. This is also the case for support to civil society, political parties, audit institutions or to cross cutting issues such as service delivery, elections or budget processes. A systems wide political economy analysis may be an important first step and will help to identify the wider constraints and opportunities within a given system. Where possible, donors should seek to carry out joint analysis or to share analysis. This may help reduce fragmentation in terms of subsequent division of labour for accountability support. A systems approach can be interpreted in multiple ways. At its core, it requires a systems wide analysis of accountability to underpin any intervention. Support may then be targeted at a specific organisation or group but will be done so in ways which connect with the wider incentives and dynamics at play, taking care to address the impacts of this support on other actors too. The principles set out in Annex 1 suggest how this might be done. In addition, it may involve reshaping programmes of support themselves. One key finding from both political economy insights and reviews of accountability support is that greater efforts are needed to strengthen the links across actors and to support coalitions of reformers rather than particular individuals or groups in isolation (Centre for Future State 2010; Rocha Menocal and Sharma 2008; McGee et al 2010). There are a limited number of examples of a more systems wide approach being taken, and building this evidence base further should be a real priority. In Uganda, programmes such as USAIDs Linkages initiative explicitly seek to support the links between actors, including at local levels; in Mali a number of programmes seek to strengthen decentralisation processes by bringing actors together (see Box).

21

Box: Examples of accountability support to multiple actors, Uganda and Mali (OECD 2011a; 2011b) In Uganda, the Linkages programme runs from builds on previous USAID funded parliamentary support and runs from 2007-2010 (at the time of writing, negotiations are underway regarding its renewal). It aims to strengthen democratic linkages within and among the Ugandan Parliament, selected local governments and CSOs, as well as to build their capacities to enhance accountability and improve service delivery (Tsekpo and Hudson 2010). In practice, Linkages provides support to a number of parliamentary committees, as well support to hold policy forums and some technical support to strengthen the shadow cabinet. Alongside this, it funds CSOs to run Budget Conferences at district level, to strengthen participatory involvement in budget processes, and provides some support to leading CSOs to provide training and capacity development to other, smaller CSOs. In this way, it seems to support links between and within different stakeholder groups. In Mali, the Programme dAppui aux Collectivites Territoriales (PACT), funded by GTZ, seeks to strengthen the capacities of communes in order to improve their performance and build synergy between actors promoting social and economic development. This is a twelve year programme launched in 2002 and due to end in 2013. The Programme de Gouvernance Partage (PGP), funded by USAID, focuses on strengthening local democracy by working with citizens, civil society organisations and focused authorities. A first five-year phase began in 2003, transition, such as Where support is communalon a particular event or moment of political with a second phase to it will 2014. With a geographic focus in the Northern regions, the PGP has a that support elections, run to be particularly important to be as comprehensive as possible, sopeace building element and seeks to improve the credibility, accountability and are involved in a works across the many domestic institutions and sectors that legitimacy of the state given and to strengthen electoral process. economic development to improve stability. It includes training programmes to improve the efficiency, accountability and transparency of local government, engagement with civic Facilitating information flows 4.2 organisations and the expansion of the medias coverage of decentralisation.

Information flows are crucial to the functioning of domestic accountability systems and can shape the incentives of different actors to respond or be held to account. Using a systems approach can help promote greater transparency but also recognises that it cannot be viewed in isolation and that transparency is not, on its own, sufficient to realise accountability. As Carlitz notes (s)imply placing more budget information in the public domain will not have an impact unless citizens can understand it, and have the legal and institutional channels to use it (2010: 3). A systems approach can reveal where greater support is needed to strengthen how information is analysed and acted upon. Examples such as Parliamentary Budget Offices, which provide technical support to parliamentarians and parliamentary committees to interpret and use budget information are helpful examples of how support of this can kind can be targeted (see OECD 2011a for examples from Uganda). Acknowledging that domestic accountability systems are inherently dynamic opens up opportunities to recognise and respond to moments of transition or transformation. Social media and mobile technologies are increasingly shaping how people interact with politics and accountability around the world. New information and communication technologies (ICT) have added channels and platforms for citizens to hold their governments to account. Support to accountability therefore needs to recognise that new technologies and mobile applications change the rules of the game completely and constantly. Examples such as Twaweza, which makes use of both new and old technologies to expand citizens ability to access government information and hold leaders accountable in East Africa signal how accountability systems are evolving in many countries, and how they are shaped by technological transformations, among many other factors. More could be done to promote 22

citizens access to media and mobile technologies as well as citizens media literacy and safety. This includes access to media products and infrastructure as well as the ability to make sense of information and to use it in appropriate ways. Looking at domestic accountability systems also reveals the need to engage with different dimensions of accountability. To date, most attention has been focused on strengthening answerability and transparency. Per u is p ar t icu lar ly in st r u ct ive in t h is r esp ect , as ef f o r t s t o im p r o ve t r an sp ar en cy h ave b een leg ally en sh r in ed . Th e Transparency and Access to Public Information Law (2002) stipulates that all information generated by state entities is public (with small exceptions for national security and some confidential information), extending from national to local levels. All public agencies are required to establish an online transparency portal, which includes information on budgets, spending, purchases, plans and activities, and citizens can request access to any information not available online. A number of donors have supported these reforms in recent years (see Box).
Box: Donor support to transparency, Peru (OECD 2011c) A number of donors in Peru have been active in their support for the full implementation of Transparency and Access to Public Information law. This has included: Support to individual state entities to improve compliance with legal provisions (including AECID support to Congress to improve information published online; IDB and UNDP support to the Presidency of the Council of Ministers to increase egovernment implementation). Sectoral support (such as USAID support to the Ministry of Health to implement transparency commitments). Some limited support at local levels, such as addressing internet access and local governments capacity/resources to implement requirements (including USAIDs Commun@s Rural Connectivity for Municipal and Regional Transparency, which has established multimedia kiosks with internet access in 120 municipal offices and trained public officials, community members and journalists in how to use these kiosks).

There has been far less donor engagement with issues of enforceability however, both in terms of administrative enforcement and judicial/legal enforcement. Mechanisms such as judicial review have often been neglected in accountability programming. This is out of step with changing patterns of judicial activism in many developing countries, which have created new jurisprudence on social and economic rights, often in line with new constitutions or expanded bills of rights, as examples from India, South Africa, Colombia, Argentina and elsewhere show. Similarly, conventional approaches have overlooked the emergence of legal empowerment through changing patterns in legal mobilisation around rights claims and public interest litigation. Much greater analysis is needed for what determines legal voice from below and what determines courts responsiveness and the effectiveness of legal recourse, linked to an exploration of the options for external support to facilitate these processes.

5 The benefits of a sectoral or issue driven approach

23

Support to domestic accountability may get most traction when it is anchored in specific sectors. Working in a particular sector reinforces a systems approach as it provides a clear rationale for bringing together a range of actors (inside and outside the state) around a specific development issue or challenge. The delivery of basic services, such as health or education, can shape citizen-state relations as citizens come into contact with the state especially in its local forms most obviously through their use of services provided by the state (Hudson/GOVNET 2009, Blankenberg, 2007; Eldon and Gunby, 2009). In practice, service delivery chains can involve a range of providers and actors both inside and outside the state. There may a wide range of ways in which accountability for service delivery is facilitated. This includes central and local government capacity to respond to citizens; support for greater transparency so that information is put into the public domain which was not previously available; and citizen-led or social accountability initiatives (such as the monitoring of services, citizen score cards and so on). There are some examples of work with conventional accountability actors (parliamentarians, political parties) at the sectoral level and growing recognition of the roles they can play in relation to effective service delivery. In Peru, for instance, support has been provided to political party platforms in relation to health issues such as child nutrition. One notable example is USAIDs support to Acuerdo de Partidos Politicos en Salud (Political Party Agreement in Health), implemented by a local organisation, Abt Associates (see Box). This assistance is seen as having helped create consensus among political parties about important health reforms. While it has not been linked to parliamentary assistance which is an existing gap and nor has it provided ways of linking to citizens engagement with these issues, it seems to represent a useful example of linking party support to core sectoral concerns and issues.
Box USAID support to party platforms, Peru (OECD 2011c) USAID has provided support to parties through the PRAES project and Politicas en Salur (implemented by Abt Associates). This supports the development of parties platforms in health with a focus on using health information and data to craft policies and create consensus, including through the development of the Political Party Agreement in Health. This support reportedly seeks to respond to the lack of party positions and cross party engagement in health policy making in Peru. USAIDs work has sought not to achieve particular health policies but rather to improve parties capacities to analyse health priorities and develop policy platforms around health (as well as putting health issues higher on the public and political agenda). It also gave space to other advocacy organisation to put forward policy proposals. This has reportedly been successful in influencing the platforms and policies adopted by parties, although a lack of enforcement remains challenging, with little monitoring to ensure that platforms are then implemented by participating parties.

However, analysis of local service delivery in countries like Benin, Uganda and Tanzania have found that local councillors and staff may be more concerned with upward accountability to political parties and line ministries than downward accountability, in part due to political networks which link local political representatives to national political patrons (VNG International/Centre for African States 2011a; VNG International/Centre for African States 2011b). 24

Adopting a sectoral approach may therefore allow for finer grain analysis of the key incentives and dynamics at play and may allow for greater recognition of the need to engage formal and informal actors and institutions. This is because accountability initiatives at the sector level which work only with formal institutions may exclude some of the most important players. For example, traditional chiefs can be significant actors in the provision of basic services or of security and justice, with implications for accountability systems. In Mali, traditional chiefs have allegedly not bought into decentralisation processes and are unsatisfied with their diminishing role over key services. This has meant that in practice citizens still tend to consult with chiefs before participating in formal accountability processes (OECD 2011b). In this sense, traditional networks and accountability relationships can work in tension with formal domestic accountability mechanisms, with important implications for support in this area. While direct support can be challenging, there are options to work with intermediary organisations, to provide training or capacity development, or to support initiatives which work with hybrid institutions (Kelsall 2011). Alternatively, it may be useful to adopt an issue driven approach, which focuses on a specific operational challenge, vulnerability or weakness in terms of accountability. Accountability support in the past has at times been seen as rather abstract and lacking a focus on specific issues. OECD research into accountability systems, for example, has revealed the utility of examining how accountability works around various entry point issues, including service delivery but also budget processes and processes of decentralisation (see Annex 3). Turning to budget processes, while support to Public Finance Management (PFM) reforms are not commonly included within conventional approaches to domestic accountability there can be real scope for crossover and complementarity. Reforms which aim to strengthen the credibility of the budget, such as the strengthening of commitment control mechanisms or the roll-out of output-based budgeting software, are often motivated by the need to strengthen country systems and processes required for effective PFM and are not necessarily designed specifically as support to domestic accountability. Regardless of the expressed intent of these programmes, they can however perform important roles in relation to strengthening wider definitions of accountability. Improving the credibility of the budget can go a long way towards improving pure financial accountability (i.e. the control of corruption) but it can also render more meaningful other areas of engagement (for example, debates at either the national or local level about resource allocation). In Uganda, a particularly significant programme of support has been the Financial Management and Accountability Programme (FINMAP) which channels support primarily to the Ministry of Finance, Planning and Economic Development (MoFPED) (see Box). Key reforms under FINMAP which have impacted on domestic accountability include support to oversight committees in Parliament, support to the Office of the Auditor General (OAG), support to strengthen budget transparency and openness, and a number of reforms that increase the credibility of the budget and thus the relevance of the formal budget process. FINMAP is seen as successful in part because of productive working relations between MoFPED and the donors involved.

25

Box: The Financial Management and Accountability Programme, Uganda (OECD 2011a) The Financial Management and Accountability Programme (FINMAP) was designed to consolidate and build upon the achievements of previous donor supported financial management reforms, including the Economic and Financial Management Programme (EFMPI and II) (with World Bank and bilateral support) and the Financial Accountability Programme (FAP) (with UK DFID support) and to ensure sustainability of those reforms. FINMAP is comprised of six main components: i) Economic Planning, ii) Budgeting Systems, iii) Financial Management Systems in MDALGs, iv) Oversight, v) LG Financial Management Systems, and vi) Management Support. Most of these components have links to improving transparency, answerability and enforceability in relation to the use of public resources. The main donors to FINMAP include World Bank, UK, Irish Aid, Norway, Sweden and the EU. At the time of writing, progress under FINMAP was being reviewed and a follow-up programme of support designed.

However, there is some concern that while improvements in capacity (in the basics of financial management) are necessary for improvements in accountability, on their own they are not sufficient for accountable governance to emerge. This is particularly the case with issues of non-financial accountability, which are beyond the remit of many of these institutions and of PFM reform programmes. Care must be taken not to overload PFM reform programmes with additional domestic accountability requirements, but rather that support to specific accountability initiatives or actors should coordinate with any large scale PFM reform programmes, so that they are mutually reinforcing. Finally, a sectoral or issue based approach needs to effectively link national and local level accountability initiatives and processes. Too often accountability initiatives remain isolated to specific districts or regions and are not integrated into national level reform processes. Instead, opportunities to bridge local and national accountability gaps need to be identified. Patterns of decentralisation, for example, can shape accountability relationships. A major (explicit or implicit) motivation behind decentralisation is often the commitment to increasing the accountability and responsiveness of the state. In practice, however, the evidence is mixed as to whether many of the models adopted for decentralisation (and promoted by donors) have contributed to improved outcomes or to strengthened accountability (Crook and Manor, 1998; Crook and Sverrisson, 2003). There is a growing realisation that the causal chains linking decentralisation with greater accountability can be interrupted by a number of factors, including various political and bureaucratic incentives (including those provided by donors themselves). For instance, in Mali, local communes are supposed to consult citizens (or their representatives) in defining local development plans and budgets. The management of local public services is to be done with the participation of citizens, through local associations elected by communities. However, there are no specific roles for citizens in this and poor oversight has contributed to a lack of sanctions where these processes are not followed. This is exacerbated by poor capacity (institutional, operational, and organisational) of district officials (OECD 2011b). Similarly, examples from Peru demonstrate the need to look not just at formal mechanisms and processes but at how they have actually been implemented. Perus legal framework establishes clear responsibilities and roles between citizens and the state, including in relation to citizens participation in policy making. The 2002 decentralisation laws provided for the creation of autonomous administrative units at regional, provincial and municipal levels and mandated that these regional and local governments hold a public forum at least 26

twice a year to report on and justify their actions. They were required to generate development plans through participatory processes and participatory budgeting. The implementation of these processes, however, has been undermined by a reported lack of political will at local levels (where reforms are seen as imposed from the centre), geographical distances from Lima (which mean that oversight bodies have limited effective oversight at local levels) and a lack of implementation capacity or adequate resources to comply (OECD 2011c). A lack of enforcement mechanisms means that citizens inputs and suggestions are not followed up, contributing to a sense of participation fatigue, with citizen participation decreasing over time (Ibid.). These examples suggest a common lack of attention paid to downstream issues of implementation, including issues of accountability and the incentives for frontline service providers, particularly in the context of decentralisation. Analysing sector budget support specifically, Williamson and Dom (2010) term this the missing middle in service delivery, which overlooks the process for management of frontline service providers, the actual delivery of services, human resources management, and the accountability for service provision. Addressing the missing middle between accountability mechanisms requires improving connections between district level actors and national level decision makers (Ibid.). While there are examples of accountability support which attempts to link these different levels, this is not a comprehensive component of much accountability support and there are particular challenges in terms of addressing competing incentives at local and national levels.

Engaging with the drivers of domestic accountability

Thinking about domestic accountability systems in these more holistic ways helps to identify particular gaps or neglected areas for engagement. In particular, it opens up new opportunities to tackle some of the deeper drivers of state-society relations, such as taxation and its relationship to citizenship. Reflection on the nature of state-citizen relations has highlighted that variation in sources of state revenue can have an important influence on the relationship between citizens and the state. Research suggests the need to move beyond a consideration of national aggregates for tax revenues (for example, as a % of GDP) to consider both the relative contributions of tax and non-tax sources of government revenue and that the way in which that revenue is raised (visibility, size of the tax base, distribution of the tax burden, degree of explicit coercion, perceived uses of that revenue) (Moore 2007). Historical examples reveal the implications of changes to taxation for state-citizen relations. In Uganda, for example, the abolition of the Graduated Tax (a poll tax) appears to have contributed negatively to state-society relations (see Box).
Box: The Graduated Tax, Uganda (Wild and Harris 2011) The Graduated Tax was the only significant form of direct taxation Uganda and a significant source of local tax revenue. All able bodied men over the age of 18 and all able bodied women in gainful employment were eligible to be taxed (LGFC 2001). In 2001, opposition party politicians argued the abolition of this tax. This was taken up by the ruling party and the tax was abolished in the run up to multi party elections in 2006. As a result, local governments are now heavily reliant on intra-governmental transfers from central government, weakening their responsiveness to citizens. Those who do still pay tax (market traders, those in formal sector employment such as 27 civil servants) commonly seek to protect their own interests rather than those of the wider population. In practice, low expectations of government have been reinforced by these weak fiscal links.

In contrast, historic examples from Ethiopia show how tax reform can contribute to wider administrative and political reforms (see Box).
Box: Tax reform in Ethiopia (Prichard/GOVNET 2010) Emperor Haile Selassie ruled Ethiopia between 1930-1974, a period marked by attempts to expand the strength of the central state against the regional landowning artistocracy. Taxation formed a core component of this approach, in building the fiscal capacity of the state and representing the replacement of feudal rule with modern government. Tax collection contributed to the creation of a salaried civil service and helped establish a centralised, bureaucratic government. Peasants were given tax receipts for land tax payments which came to act as a guarantee of peasant property rights. In 1967, local Tax Assessment Committees were established which led to more extensive central government presence in the regions.

Donor support has been slow to connect tax reform to governance support. Yet historical evidence suggests that taxation and domestic revenue generation alongside revenuebargaining between states and organised citizens can be crucial for state formation and the development of accountability relationships (Brutigam et al 2008:1). Where states can access other sources of revenue, this can undermine the development of more legitimate and sustainable tax-based social and fiscal contracts between citizens and the state. This has been commonly cited in relation to the resource curse experiences of countries with significant natural resources and it is increasingly applied to analysis of aid flows. Resurgent interest in tax issues therefore represents both the recognition of the centrality of the tax system to governance processes and growing interest in domestic resource mobilisation to progress towards an eventual exit from aid dependence (Prichard/GOVNET 2010). At present, only around 0.1% of donor support supports taxation (OECD DAC statistics). This is commonly in the form of short- or long-term technical assistance; training and learning opportunities; provision of equipment; and investment in infrastructure (see Box).
Box: Technical assistance to tax systems (Bolnick et al 2011) Common forms of assistance include: Technical assistance: Using experts who commit to the long term, to act as mentors and support longer term institutional change or in the short term address specific projects or problems through specific assignments. Training and learning: Strengthening capacity and human resources. This can also involve peer networking or twinning arrangements; opportunities to attend specific courses. To be effective, this needs to be supported by wider incentives and institutional cultures. Provision of equipment: such as IT equipment. Infrastructure investment: This may include funding the construction of one-stop customs posts to facilitate trade

Examples of donor support to civil society campaigns on taxation, or to bringing together a range of actors to debate and discuss these issues, such as those highlighted in the Box below, represent innovative attempts to address some of the deeper drivers of taxation and accountability in many contexts. 28

Box: Support for taxation, Mali and Uganda (OECD 2011a; 2011b) In Mali, many individuals often support their extensive families and contribute to community activities and religious establishments, which mean that direct taxation by government can be seen as a heavy burden, and an unwelcome one where citizens do not perceive much in return. In this context, local authorities have found it very challenging to raise taxes. In response to this, a number of donors, including Denmark, France and Germany, are funding pilot projects to raise citizens awareness of direct taxation and to enable local authorities to explain the importance of taxes to their constituents. In return, citizens have requested improved financial transparency, highlighting a potentially important entry point for the exchange of accountability between state and citizen. Informal accountability relationships could also play an important role in strengthening the social and fiscal contracts between elected officials and citizens. For example, a recent study notes that local authorities should look to collaborate more closely with traditional chiefs, as those with influence over local populations who may help with the collection of local taxes (VGN International 2010). This needs to be approached carefully and cautiously, but it suggests new avenues for donor support to accountability. In Uganda, the AfDB has recently funded a series of country analyses, of which Uganda is one, looking at tax policies and revenue generation. They hope to facilitate in-country debates on issues of tax policy and tax reform, where there is local demand for this. The Netherlands Embassy in Uganda has taken a stronger position, and is providing support to CSO campaigning on taxation issues. The consideration of taxation issues by these actors seems to be a useful attempt to go deeper into some of the core drivers of accountability weaknesses.

As recent OECD research shows, government ownership and leadership is crucial, as aid can only effectively support government programmes to improve the tax systems and these need to be aligned to domestic political incentives (Prichard/GOVNET 2010). How revenue is raised and collected is as important as the total amount, so attention needs to paid to issues of efficiency and equity. The choice of aid modalities also presents a range of options for inclusion of tax issues. Forms of budget support (general and sector) can incorporate assessments of tax performance. Fora for policy review (such as budget support frameworks) can highlight tax performance and create incentives for improving the tax system, including through introducing variable tranche mechanisms linked to revenue targets (Bosnick et al 2011). Sector Budget Support for PFM can also create a direct link between budget funding and PFM performance, including tax performance (Ibid.). Basket financing, with multi-donor pooled funding, can be useful for a specific tax programme or to address specific or concrete areas of reform. According to a recent study, tax officials preferred basket funding where at least three donors are interested in supporting a common tax programme (Ibid.). Multi donor trust funds can also support tax systems through a coordinated platforms or donors can jointly fund projectised assistance. A large amount of aid in this area is still provided through individual bilateral projects, however, which can lead to fragmentation and higher transaction costs (Ibid.). Therefore, there are broad areas of synergy between accountability agendas and agendas for tax reform. Linkages between taxation and accountability involve support to institutions and organizations outside the revenue system, including the justice system, Parliament, and civil society. Accountability support programmes should give special weight to activities that address these synergies (Prichard/GOVNET 2010). 29

In considering deeper drivers of accountability, care must also be taken to identify and address international drivers which affect the incentives and behaviour of political actors at the national level. In recent years, analysis has focused on the impact of increasing global demand for natural resources, which has incentivised political elites in poorer countries to extract rents from the export of these commodities. This has provided unearned incomes which have undermined incentives to develop accountable public institutions (Moore et al 2009). Aid can similarly be seen as providing unearned income in some contexts. In addition, attention to international drivers highlights the roles of illegal arms trade or criminal networks (for example in narcotics or human trafficking) which generate high profits and rents and can shape the incentives of police and politicians. This has been further facilitated by large transnational capital movements and the use of tax havens. There are growing calls for donor countries to respond to these international drivers. Initiatives such as the Extractive Industries Transparency Initiative (EITI) launched in 2002 are some of the most high profile responses. EITI operates as a voluntary, multi-stakeholder attempt to promote transparency in revenues paid to governments by extractive industry companies. In Mozambique, for example, the Government, with donor support, has established an ad-hoc' multi-stakeholder group made up of Government representatives, companies and civil society to increase the transparency of payments by companies to the government, as well the latters revenues from the extractive sector (OECD 2011d). Efforts to address capital migration include the Stolen Asset Recovery Initiative (StAR); the International Centre for Asset Recovery (ICAR): G7 Finance Ministers initiative towards a common set of (voluntary) principles on integrity, transparency and propriety in global business and financial transaction and recent declarations of intent to crack down on tax havens from a number of donors (Moore et al 2009). To date, there has been limited coordination between aid programmes and initiatives which can address international drivers; identifying synergies between these can be particularly helpful and requires linking development assistance much more effectively with diplomatic engagement (alongside trade and economic relationships).

Measuring results

In the current climate of fiscal austerity, many donors are under increasing pressure to demonstrate the results of their development spending. This focus on results is not a new one. Donor agencies have periodically sought to find new ways of measuring the impacts of their interventions and have worked with those within aid recipient countries to better track development trajectories and aid. Frameworks such as the Millennium Development Goals and commitments at Paris and Accra to aid effectiveness are motivated to some degree by commitments to improve results, and to better measure them. Some of the current rhetoric on results has led to worries that a more narrow results agenda may now be pursued, which could marginalise support to accountability and governance, as these are areas which are intrinsically challenging to measure in quantifiable ways. Rather than being viewed as a threat, the prioritisation of results should be seen as an opportunity to reflect on more robust measures and to strengthen monitoring and evaluation going forward. Evidence suggests that monitoring and evaluation approaches for accountability support have not been particularly effective at measuring longer term outcomes and impact. As McGee et al highlight: many initiatives are not underpinned by a clear articulation of exactly what outcome or impact is sought.... or of how the actions and inputs contemplated are expected to generate that outcome or impact. That is, the 30

assumptions underlying the causal chain, from inputs to outcomes and impact are absent, vague or too implicit (2010: 9-10). Incorporating political economy insights into accountability programming should contribute to developing realistic objectives, something which is a key component of more rigorous results measurement. This implies paying much greater attention to objectives, assumptions and programme design at the outset, rather than solely at the end of a programme. Being explicit about the theory of chain behind a programme or area of intervention can be particularly helpful in testing the starting assumptions of any support (see Box).
Box: Theory based evaluation Programme implementation necessarily involves a theory about what might cause change, even though that theory may not be explicit (Pawson and Tilley 1997). A theory of chain approach therefore seeks to make these theories explicit. This involves developing clear hypotheses about how, and for whom, programmes might work. The implementation of the programme, and any subsequent evaluation, then tests those hypotheses. This requires data collection about the processes of implementation and impacts but also measuring key dimensions of the wider context (which may impact on programme outcomes) and monitoring specific mechanisms which may create change. These data collection processes (interviews, focus groups, questionnaires and so on) should be constructed to help refute or refine theories about how and for whom the programme works.

Where theories of change are not made explicit, or where they are based on assumptions that do not reflect the wider context, the effectiveness of any support is likely to be undermined. It will also make measuring progress and final impact more challenging. McGee et al (2010) have highlighted a number of methodological challenges for measuring the results of transparency and accountability initiatives (see Box). Overall, they warn against the use of a generalisable model of evaluation which can be applied across all programmes and all contexts; instead there is a prevailing need to adapt tools for monitoring and evaluation to specific contexts and challenges faced, taking account of the wider enabling environment and feasible options for change.
Box: Constraints on measuring accountability initiatives (drawn from McGee et al 2010: 26): Common limitations to amount and quality of evidence available (lack of comparators or counterfactuals); Lack of use of theories of change and the predominance of untested assumptions; Tensions between correlation and causation (where many factors likely to be beyond control of interventions); Lack of appropriate baselines and indicators (or relevant data).

Adopting an accountability systems approach should further enable both more realistic and more rigorous approaches to measuring results, where they are based on assessments of the wider contexts and the enabling environment for reform. This is particularly important in relation to measuring the impact of aid on the scope for domestic accountability. As McGee et al highlight in the Box above, as a wide number of factors and dynamics shape accountability processes, it can be challenging to identify causation and some factions are 31

likely to be beyond the control of a given intervention. Bringing this into a systems analysis is likely to be key. This means that the measurement of risks should be prioritised within evaluation frameworks. Measuring economic or financial risks is generally easier to assess and identify; political risks remain more uncertain. Building risk assessment (and using political economy tools where appropriate) throughout programme delivery can help to address this (Wild and Foresti 2011). Embedding risk assessment within evaluation requires frameworks and approaches which allow for the observation of change over time (rather than a static evaluation at the end of a programme of support) (Ibid.). In practice, the limitations referred to above, as well as the implications of adopting a systems wide approach means that it is likely to be most helpful to combine different methods and techniques to build up a picture of the impact a programme to support accountability. This could involve the use of counterfactuals and user/citizen perception data, combined with political economy analysis, and more standard tools such as logframes. The Table below summarises some of the potential methods available, and their relative strengths and weaknesses8. Method of evaluation Quantitative: surveys Weaknesses Cannot explain level of implementation achieved or why impact occurred (or did not occur); May ignore what is not easily measurable; sampling can be challenging in some contexts. Rankings and indices Provides at-a-glance Cannot explain why certain comparative information; Can scores were achieved or facilitate peer pressure for low account for wider contexts. scorers. Experimentation Can isolate the impacts of an Cannot produce findings which approaches: (e.g. individual intervention (i.e. can are generalisable; may not Random Control Trials) measure before/after, capture complex interactions of with/without) multiple variables Qualitative: case Allows for purposive sampling Cannot draw general or studies/stakeholder (can focus on examples of representative conclusions (and interviews success or failure); Can often limited by lack of sufficient identify variety of factors and number of multiple cases). dynamics which contributed to a given outcome and account for wider context. User/perception Can highlight perceptions of Can be highly subjective, not surveys or participatory those who should benefit from representative. approaches a given intervention; can build ownership of overall process. Outcome mapping Can measure changes in Does not necessarily behaviour and in the demonstrate relationships between actors; causation/contribution of a can trace emerging change; specific intervention; does not actively engages stakeholders produce generalisable findings.
8

Strengths Seen as more objective; can draw on generalisable findings based on representative sample.

This is adapted from an excellent table in McGee et al 2010: 33.

32

Most Change

Significant Can capture complexity and Does not produce generalisable unexpected changes; actively findings, can be resource engages stakeholders in intensive and time consuming. identifying and prioritising impacts.

33

Key lessons and recommendations

There is now a growing body of evidence of what is more effective, and what is least effective, in support to accountability in developing countries. To date, support to domestic accountability actors has often been provided through forms of top down and supply driven capacity development and technical assistance which have tended to emphasise best practice or templates that assume particular causal chains, and have not worked with the grain of the societies they are delivered in. These have often not achieved their desired impacts and donor support has invariably been provided to individual actors (usually those on the supply or demand side). What is now needed is to join the dots between isolated support to different actors and to recognise that accountability in fact works as a system, which involves a wide range of actors and a number of links (of varying strength) between those actors. Adopting an accountability systems perspective would allow for greater reflection on the range of relationships that exist between actors rather than focusing on particular actors (parliaments, civil society) in isolation. This would mean moving beyond supply versus demand debates to examining the concrete mechanisms through which actors are brought together to be held to account and to hold to account. Looking at accountability systems as whole should facilitate the incorporation of political economy analysis and perspectives to a far greater degree than has been possible to date. It would facilitate analysis of a range of actors and institutions, their incentives and the reform space possible for accountability in each context. This has implications for different phases of the policy and programming cycle.

8.1

Implications for programme design

Adopting a systems wide perspective is most important at the design stage of any programme of support to accountability. Where possible, this should be grounded in forms of political economy analysis. Adopting a systems wide, political economy perspective may mean drawing on pre-existing contextual analysis or commissioning analysis where it is not available, as well as using available evaluations of past accountability support. At this stage, donor coordination and cooperation is important, including the commissioning of joint analysis (although this needs to be handled sensitively and be grounded in shared objectives). Donor coordination at design stages can ensure that subsequent divisions of labour support greater coherence and address fragmentation, which has been a common challenge for past accountability support. It may be useful to conduct a mapping of the wider accountability system (including all actors and institutions) as a first step. Then the feasibility and requirements of a range of programming options can be assessed. From the outset, consideration is needed of the interaction between formal institutions and processes and informal ones. Too often, donor support has been targeted at formal accountability institutions, for example around the formal budget process, which have not addressed some of the informal realities, for instance which determine how resources are really allocated. Similarly, models of political accountability have often been based on assumptions of adversarial relationships between different branches of government, or of latent citizen demand, which may not fit well with realities of political cultures that prioritise consensus and cooperation. 34

Donors are rightly wary of engaging with informal processes and institutions, as these processes can be challenging for external actors to fully understand. Moreover, instances of patronage and clientelism challenge many of the models of accountability donors have supported to date. Engagement does not have to imply endorsing all of these practices and institutions, but it will mean tailoring the design of programming to adjust to different contexts. Key recommendations: Draw on available context analysis (and where not available, consider commissioning analysis) as well as evaluations Consider joint analysis or coordination to develop shared understandings of the accountability system, and use this to address any fragmentation challenges Map formal and informal actors, institutions and processes which shape accountability. Understand the relationship between these features and elements of programme design Identify entry points (service delivery, budget processes) to ground accountability work (including analytical work) Adopt an accountability systems approach, which looks at multiple actors and the mechanisms which bring them together, rather than focusing on individual actors alone Assess both the wider context and the feasibility of a range of options.

8.2

Implications for programme implementation

Engagement with domestic accountability actors requires robust understanding of the wider context or accountability system in which they are located. This can then allow for identification of the more feasible options for support, and for those which account for both the incentives and motivations of an individual actor but can also account for (and monitor) the impacts on other actors. Support may still be targeted at individual actors, groups or institutions. Where this is the case, it should aim to follow the principles set out in Annex 1. Alternatively, support can be implemented within an accountability systems approach. This may involve piloting support which explicitly targets the relationships between actors. There is growing evidence that social accountability support, for example, only gains traction when it is linked to formal mechanisms. Reflection on the mechanisms and linkages between actors should become much more central to decisions as to how to implement accountability systems support. Some isolated examples such as USAIDs Linkages programme in Uganda provide illustrations as to how this might be done effectively. Working across actors in this way implies different ways of working, with a greater focus on identifying a range of domestic actors engaged in change practices and the use of facilitating or convening roles to support coalitions. This may imply greater investments in human resources, to ensure donor staff have the necessary skills and time to engage effectively in these ways. An accountability systems approach may imply working with actors not always associated with accountability reforms, including the private sector and political parties. In many countries, these are actors which may need to become more accountable themselves but which can also be important contributors to coalitions to hold others (donors, 35

governments) to account. Working with platforms which bring these actors into dialogue with others can be useful where direct engagement would be sensitive. Moreover, there are areas which have been overlooked but would benefit from greater investment. Donor engagement on accountability has been slow to connect to tax reform But resurgent interest in tax issues represents both the recognition of the centrality of the tax system to accountability processes and growing interest in domestic resource mobilisation to progress towards an eventual exit from aid dependence. There are some innovative examples of support which aims to facilitate national level dialogue (with government, CSOs, parliamentarians and others) on domestic resource mobilisation in this respect. The choice of aid modalities also presents a range of options for inclusion of tax issues. Transparency improvements are important but attention is now needed to how information is used and acted upon. Important progress has been made in improving both the transparency of aid and of government revenues in general. Greater focus is now needed on how information can be used and acted upon by different groups. Examples of parliamentary Budget Offices reveal the utility of mechanisms to help interpret and analyse available information; similar forms of support are also needed for other actors (civil society, media). Signatories to the Paris Declaration should step up efforts to meet their key commitments, many of which can have important knock-on effects for domestic accountability. Mutual accountability should be taken much more seriously, with donors realising commitments to improve their own accountability to partner countries. Governments should make their expectations of donors clear and donors should be explicit about their own political restrictions and about how their internal processes impact on how aid is delivered. The Paris Principles committed to increase levels of programme based aid to improve country ownership and donor alignment. In reality, ad hoc portfolio approaches have emerged, which combine project and programme based aid. More strategic use of aid portfolios should be prioritised. Wherever possible, aid should work towards being programmatic, in that it should work in support of national strategies rather than donor priorities, regardless of the specific modality chosen. Finally, greater coordination among a range of external actors is needed. Any donor accountability support or engagement should be grounded in complementary diplomatic policies that seek to nurture or reinforce commitments. In particular, attempts to address international drivers including initiatives such as EITI or attempts to address capital flight should work in synergy with domestic accountability initiatives. Key recommendations: An accountability systems approach does necessarily imply ever larger, more holistic accountability programmes; it can mean thorough analysis of the accountability system and then narrow support to address particular weaknesses (including particular actors) Adopting an accountability systems approach implies providing support in ways which help to build relationships, bring together coalitions and support dynamic change processes. It may require different ways of working, including brokering, facilitating and supporting reform processes A systems wide approach may mean working with unconventional actors, such as the private sector, political parties, and on new issues (taxation) 36

Donors can impact on the scope for domestic accountability in how they provide aid. Programme based aid approaches seek to strengthen country systems but do not guarantee support for domestic accountability, unless they include specific agreements in this respect. A mixture of programme based and project based aid remains a reality for many developing countries. Achieving more strategic aid portfolios should therefore be a priority. Greater links should be made between attempts to tackle international drivers and domestic accountability initiatives.

8.3

Implications for monitoring and evaluation

The need to demonstrate results is rising on international development agendas, including in the field of accountability support. To date, there has been a lack of robust measurements for monitoring and evaluation efforts of donor support to accountability. Those metrics in use have not been particularly effective at measuring longer term outcomes and impact. Many initiatives are not underpinned by a clearly articulated theory of change and assumptions and risks are not monitored very effectively. Adopting an accountability systems approach should allow for approaches to measurement which include assessments of the wider context and the enabling environment for accountability reforms. They can help develop realistic theories of change, which consider the incentives and relationships between actors. It will be most helpful to combine different methods and techniques in order to build up a picture of the medium term and longer term outcomes of a programme. This should include a combination of quantitative rankings or surveys with qualitative perception surveys or case studies. Recommendations: Develop a theory of change, underpinned by realities in a given country, for each intervention, which makes explicit assumptions about how and for whom a programme will work These assumptions should be reassessed at milestones throughout the programme and revised where needed Build in assessments of the wider context and risk management throughout the programme cycle (for monitoring and evaluation) Combine evaluation methods to capture medium term and longer outcomes.

37

Annex 1: Principles for Improving Donor Assistance to Support Domestic Accountability

This annex is designed to support the main body of the report by providing principles for donor assistance in key areas of activity designed to support systems of domestic accountability, namely: elections, the media, political parties and parliaments. Electoral assistance is the field that has received most international attention because of the pivotal role of voting in providing the public with a political voice and calling the executive to account for their actions. Although democracy involves more than holding free and fair elections, they are the fundamental building block for effective accountability. Effective accountability also depends on a functioning public sphere - where citizens come together (even virtually), share information, and deliberate on public issues. This depends on a vibrant media which provides information, highlights key issues, facilitates public debate, and acts as a watchdog for the public interest, holding state and non-state actors accountable. The very nature of the media means that it interacts with accountability systems at all levels, and increasingly international assistance is seeking to buttress this vital role. By contrast, political parties have been regarded more cautiously by donor agencies, because of perceptions of political interference in the domestic politics of another sovereign nation. Yet, political parties are an equally vital part of the public sphere, providing the principal vehicle for the articulation and representation of public concerns within the political systems of accountability. The quality of the party system is intrinsically linked with the quality of accountability. Party aid is inevitably political, but it has a legitimate place in foreign assistance in pursuit of democratic and development goals. Finally, parliaments are a key institution in securing executive accountability. In theory, while governments are directly accountable to voters at elections, in between elections it is the duty of parliamentarians to hold ministers and their departments to account on the public's behalf. Parliaments derive much of their authority from the fact that a number of accountability institutions usually report to them - ranging from the supreme audit institution, the ombudsman and the electoral commission, through to utility regulators, inspectorates and agencies. In other words, parliaments should sit at the centre of a web of domestic accountability and are potentially vital allies for donors in securing accountability and improving quality of public services. The principles below are a distillation of key lessons within each field of expertise. Although they reflect the specific dynamics, priorities and challenges within that particular field, several themes recur throughout the papers.

38

Take context seriously: Assistance to any domestic accountability actor needs to work from a deeper understanding of local conditions, in a way which recognises the interconnections between institutions, sectors and actors. Aligning support programmes with wider accountability objectives: Reflecting the interdependence of systems of accountability, projects should be tied to other accountability support efforts. A programme in a particular area - such as elections should actively complement any assistance efforts in other areas, such as parties, parliaments or media or at least should be designed in ways which recognise and adjust to the impacts on other actors. Establishing realistic objectives: It should be recognised at the outset that improving accountability is likely to be slow and incremental. Projects need to be based on a realistic assessment of what is feasible, which will often mean having limited objectives rather than seeking to overhaul the entire system Ensure local ownership. Achieving meaningful change means changing patterns of behaviour as much as increasing resources, reforming institutions or creating new laws. However, behavioural change cannot be enforced from outside, it must be emerge from within. Projects should ensure that local partners are responsible for its success and that support aligns with incentives for reform. Support must be long-term. Because change happens slowly, support must be based on a multi-year commitment. The assurance of a long-term presence is likely to enhance the relationship with the local partners and increase the chances of genuine impact.

The challenge now is to turn these principles into action, so that they inform project design and delivery in each of the fields. We hope that the accompanying notes provide the basis for such development.

39

1) Principles on International Elections Assistance

Elections are only one element of democracy, yet they are an irreplaceable onewithout regular, genuine elections democracy does not function. Elections give form to citizens political voice, constituting both a fundamental root of political accountability and an orderly process for successions and alternations of power. Through substantial support for elections in many countries attempting democratic transitions over the past several decades, the international donor community has helped improve numerous electoral processes. By identifying lessons from these experiences and incorporating some of those lessons into improved methods and practices, international elections assistance is positively evolving. Nevertheless, given the complexities, difficulties, and risks of electoral processes in many developing countries elections assistance continues to face numerous challenges, especially in post-conflict contexts and in fragile states with sharp sociopolitical divisions. Marking a renewed commitment to make elections assistance as effective and useful as possible, the major funders and implementers of international elections assistance agree to the following strategic and operational principles: Take the local context seriously through careful, comprehensive assessments. Elections assistance efforts should be grounded in incisive political economy analyses that identify and examine the determinant power dynamics and political constraints that shape the electoral environment, as well as the specific roles that elections are likely to play in particular settings. Be alert to electoral risk. Elections assistance providers must respond to the recent rise in electoral and post-electoral violence in developing countries by giving greater attention to electoral risk and long-term causes of political violence that could ignite in election processes. These concerns should be reflected in the design and implementation of elections assistance interventions. Don't misuse electoral aid and promote transparency. Elections assistance should be employed based on the highest standards of impartiality and only to promote free and fair elections, not to advance other donor policy goals, such as burnishing the legitimacy of favored partner governments or building friendly relationships with governments. Assistance should be provided on a transparent basis: information on who is providing funding, and assistance should be readily available. Cost-effectiveness should be ensured so that state expenditure is in line with efforts to meet the Millennium Development Goals. Ground electoral assistance in complementary diplomatic policies. Elections assistance should be grounded in complementary diplomatic policies that seek to nurture or reinforce the commitment on the part of partner governments to follow accepted electoral norms. Recognize the role of regional organisations. Elections assistance providers should take full account of the valuable role that regional organisations can play both 40

in election monitoring and election assistance and seek the greatest possible complementarity with such organisations. Embrace a full concept of ownership. Aid to strengthen an electoral process should be owned not only by the relevant partner government but also by the broader political society in question. Elections assistance providers should embrace an interpretation of local ownership that takes account of this political imperative. Build on donor coordination. Elections assistance providers should build on the progress they have made in creating cooperative mechanisms for elections assistance by assessing the record of such mechanisms and seeking ways to broaden and deepen communication, cooperation, and coordination among all relevant aid providers, including around joint funding mechanisms. Be as comprehensive as possible. Designing elections assistance to be comprehensive horizontally across the many domestic institutions and sectors that are involved in an electoral process will ensure better synergies and overall coherence. Elections assistance and observation should be well coordinated, as observation plays a key role in effective electoral support. Think and act across the electoral cycle. Elections assistance should be designed, planned, and implemented in a long-term fashion across the full length of electoral cycle and if possible across multiple cycles, avoiding the common tendency to focus primarily on activities relating to elections day. Donor support should encourage sustainability to ensure that local capacity is built as quickly as reasonably possible. Push for integration. Elections assistance should be actively integrated into the wider domain of democracy support, especially assistance for political party development, legislative strengthening, media assistance, and civic education programs. Emphasise citizens understanding and engagement. Efforts to help citizens understand the utility and significance of elections as one part of a broader set of accountability mechanisms should be an integral element of elections assistance. Experience shows that it is important to support consultative approaches to help election stakeholders to be jointly responsible and to build their confidence, and that of the wider public, in election management bodies and other institutions involved. Add the local to the national. Aid providers should complement their traditional focus on national elections with greater attention to strengthening local elections. Make the connection with accountability. Elections assistance providers should actively connect their activities with the wider set of donor programs supporting accountable governance at all levels. Don't neglect gender. Although progress has been made to widen and deepen the role of women in politics in many developing countries, elections assistance providers should do more to incorporate a full gender dimension in elections assistance, especially concerning candidate selection and voter participation. 41

Respond more consistently to flawed elections. Donor governments committed to advancing free and fair elections should strive toward greater consistency in responding to flawed elections. Keep learning about impact, and act on it. Building on the important learning efforts undertaken in recent years, elections assistance providers should carry out deep-reaching evaluations of the impact of elections assistance in varied contexts and incorporate the learning from those evaluations into assistance practice.

2) Principles for Political Party Support

If they function well, political parties can play a number of fundamental roles in democratic politics, including: aggregating citizens views and interests; providing structured political choices to citizens; engaging citizens in the democratic process; training and socializing political leaders; developing policies and taking responsibility for implementing them; and facilitating coordination within legislatures and between branches of government. In addition, political parties can help advance governmental accountability. Opposition parties have a direct interest in monitoring the actions and checking the power of ruling parties and putting forward viable policy alternatives. Parties are also held accountable for their performance by voters. Compared to individual politicians, parties tend to have longer time horizons and a stake in maintaining a long-term reputation. Any individual politician who ignores the electorate or abuses his power can face pressure from within his or her party to reign in their behaviour. Parties can initiate pro-development policies which reflect the interests of key social sectors and can gain public legitimacy for these policies through electoral competition. They can then ensure the necessary coordination within government to implement these policies. Two striking facts stand out about political parties in developing countries: first, parties are exceptionally unpopularon the whole they are the least respected public institution in most countries; and second, the complaints that citizens have about parties are remarkably similar across different countries and regions. Taken together these complaints form what has been called the standard lament about parties in new and struggling democracies (Carothers 2006). Parties are often seen as corrupt, self-interested organisations dominated by power-hungry elites; hold to ideological positions only opportunistically; become genuinely active only at election time; and are ill-prepared for governing and do a bad job of it when given the opportunity. The causes of the standard deficiencies of parties in developing countries are complex and multiple. They include: compressed transitions: the relatively rapid movement from authoritarianism to multiparty politics characteristic of democracys Third Wave left parties in these countries with little time to develop a broad grassroots base, instead they were thrown immediately into electoral competition and forced to become electorallyfocused; weak rule of law: the weak rule of law characteristic of many developing countries provides an inadequate framework for regulating the financial and other activities of parties; 42

poverty and inequality: widespread poverty and inequalities in many developing countries contribute to the rise and endurance of neo-patrimonial, clientelistic politics; anti-party legacies: in many new or struggling democracies citizens come to democratization with a deeply anti-political outlook based on their previous experience with authoritarianism; and presidential systems: presidential systems can encourage top-down leadercentric parties and weak parliaments.

Despite the ongoing expansion of political party assistance, however, significant parts of the development aid community remain wary of political party assistance, uncertain about the links between political party development and socio-economic development, concerned about the political sensitivities of such assistance, and worried that working with political parties will entail engagement with corrupt, tawdry politicians. Moreover, frustration and disappointment are characteristic of many party aid efforts. Many parties absorb significant amounts of party aid for many years without showing important signs of positive change. Expectations about what party aid can accomplish are often too high. The very modest results of most party assistance can be ascribed to two main factors. First, political parties are very difficult organizations to assist. Many parties are highly resistant to reforms. They are leader-driven vehicles, serving the interests of a narrow party elite willing to use any methods to gain and maintain power. For the leadership of these parties, the reform agenda that international party aid providers bring is largely unappealing and even threatening. Almost every element of that agendawhether greater financial transparency, more internal democracy, greater inclusion of women, more role for youth or more rational management systemsrepresents a potential reduction of the power of the entrenched party elite. Moreover, the rational assumptions that international party aid providers bring to the task concerning the basic aims of parties with which they workthat parties seek to represent citizens interests, to elaborate a well-conceived, technocratic political platform, to compete hard but fairly, to emphasize substance over personalism, and so forthoften are not shared by the party elites in question, who see their own parties in very different terms, as vehicles useful for advancing particularistic interests of the elites themselves. Therefore, the provision of technical assistance, which is by far the largest element of party assistance, does little to ameliorate parties shortcomings. Instead, as noted above, the principal causes of the weak state of parties in most of these countries are much deeper structural conditions. Identifying principles for political party support is not simple. Different aid actors are taking divergent approaches and there remains a weak base of understanding of the results of such efforts. What might seem like obvious principles are often untenable. For example, it might be tempting to suggest that party aid should strive to be nonpartisan but this would not. work for those party aid organizations that utilize fraternal approaches. Or it might be suggested that party assistance should not entail direct financial transfers to recipient parties, yet some of the newer entrants to the party aid domain have been utilizing direct grants to parties. With these caveats in mind, the following principles have been developed: 43

Base party aid on a sophisticated understanding of the political economy of the relevant parties and party systems. Given the wide range of party types, roles, and systems, it is imperative that party aid providers develop deep knowledge of the actual nature and function of parties within their national contexts before undertaking party assistance. Recognise the value of effective political parties not just for democracy but also for development. Political parties play potentially crucial roles in articulating policy alternatives, helping spark public engagement in and legitimacy for prodevelopment policies, and establishing governmental accountability. Be aware of but not paralyzed by the sensitivities of party aid. Party aid is inevitably politically sensitive given its reach into core political processes and institutions. At the same time, however, it has a legitimate place in foreign assistance if pursued openly in genuine pursuit of democratic and developmental goals. Build on the interconnections between party aid and other elements of political aid. Political party work connects naturally to other forms of assistance related to strengthening democratic processesincluding work on legislatures, elections, civic advocacy, and local government performance. Dont confuse party diplomacy with party aid. Western political parties sometimes engage abroad to build political alliances for the sake of building coalitions in multilateral organizations or enhancing bilateral diplomatic relations. Such party-toparty diplomacy is legitimate, but is significantly different from party assistance.

Dont assume common goals between providers and recipients. Party aid providers must be careful not to take recipient parties at face value in terms of their political role and goals. Achieving alignment between the goals of party aid providers and party aid recipients is crucial to success. Party aid providers should pay attention to actors within political parties who may share their goals more closely than party leaders. Stress cooperation rather than competition among party aid providers. As party aid multiplies the need for party aid providers to communicate with each other and avoid overlap or competition increases proportionately. New entrants to the party aid domain should take special care to ground their work in a thorough analysis of what other aid actors are already doing in the same countries. Embrace transparency. Crucial to managing the political sensitivities inherent in political party aid is operating with transparency in carrying out such assistance. Continue emphasizing gender issues. Fostering greater inclusion of women in political parties has been an element of many party aid programs. Encouraging progress has been made in this area in some countries, more so than with respect to many other aspects of party change. But this focus must be sustained and even deepened. 44

Pursue realistic, incremental goals. Given the uncertain and often troubled state of political parties even in established democracies, political party aid must not be based on the pursuit of ideal models but instead on very modest, realistic, and incremental goals based on in-depth studies of the local political environment. Keep strengthening evaluations, but dont overemphasise numbers. Many issues regarding whether and how party assistance works remain insufficiently examined empirically. Party aid organisations should continue to deepen their evaluation efforts and support research and other learning exercises. At the same time, however, funding organisations should recognise that any efforts to reduce political party development to strict quantitative indicators are likely to be unhelpful. Recognise the long-term challenge, but focus on tangible outcomes. Problematic features of political parties and party systems are not amenable to quick fixes and party aid is most effective when pursued on a long-term basis. Nevertheless, party aid programs should define tangible medium term outcomes that define the path of such longer-term engagement.

45

3) Principles for Media Assistance

For much of modern democratic history, media has been considered one of the most powerful and central forces for accountability. It receives special protection within most democratic constitutions expressly because an informed citizenry and a fourth estate capable of acting as a check on executive power are considered to be critical to good governance. Good governance depends on a functioning national public sphere. The public sphere represents the space between government and citizens, where citizens come together (even virtually), share information, and deliberate on public issues. Media provides news and information to the public, brings issues on the public agenda and facilitates public debate and discussion. It serves as watchdog for the public interest and holds state and non-state actors accountable. Media is crucial for good governance: it creates the conditions for inclusive policy dialogue, as well as providing a platform for broad-based participation in actual policy processes. Within the context of aid effectiveness and democratic governance agendas, few question the importance of a free, professional and plural media in contributing to good governancei. Nevertheless, few within the development community accord the media the same importance as other national accountability institutions. The media is mentioned (once) within the Accra Agenda for Action on Aid Effectivenessii within the context of helping to contribute to mutual accountability and several surveys suggest the issue languishes low in terms of governance priorities within development agencies. Funding is increasingly being allocated to media work, but is often not institutionalised or integrated into an overarching policy structure,iii and may therefore not be efficiently utilised. Although there is a substantial amount of research on the role of the media for domestic accountability, this research remains scattered and inconsistent. Studies use vastly different definitions and measurements of accountability and interpret results inconsistently. Anecdotes of successful media interventions outnumber rigorous studies with strong empirical measures. Social media and mobile technologies are increasingly shaping the way people interact with politics and represent an increasingly important accountability mechanism. New information and communication technologies (ICT) have added channels and platforms for citizens to hold their government accountable. A more limited body of evidence exists on ICT (mostly focusing on European and American contexts), but given the pace of change, and the rate of internet/mobile proliferation in many developing economies, building an evidence base remains very much a work in progress. However, a growing list of initiatives illustrates the possibilities and potentials of using social media and mobile technologies to increase domestic accountability.iv Donors need to be aware that new technologies and mobile applications change the rules of the game completely and constantly.

46

In order to achieve more effective and strategic media assistance, the following principles should provide the framework for support: Incorporate media assistance into larger frameworks of development aid. Access to information is crucial for domestic accountability. Media institutions in particular provide tools and channels for accountability that can complement and enhance other accountability mechanisms, but also add new instruments that may be at least as powerful and efficient as accountability measures that are more commonly considered by donors. Weak and/or highly constrained media undermine domestic accountability. The risk of not considering and supporting media as part of broader accountability programmes is significant. Incorporate media indicators and audits into governance diagnostics and needs analysis. The state of the media is inseparable from the state of governance in general. For instance, the UNESCO standard media development indicators can usefully be incorporated into governance needs assessments to more effectively guide interventions aimed as improving media as an accountability mechanism. Cooperate with media development CSOs and determine media objectives and outcomes, not methodologies. Given a lack of specific expertise on media development within the majority of donor organisations and local media beneficiaries, there is a strong argument for developing media support strategies and specific interventions in partnership with media development CSOs. While there is a clear need to ensure that media strategies complement overarching accountability objectives, there is a strong argument for providing CSO implementers with substantial scope and the ability to propose creative solutions as opposed to highly prescriptive requirements. Focus on building public demand for inclusive policy dialogue. There is clear potential for media support that enables and fosters country level policy dialogue, moreover research incorporated into such support can assist in building a body of evidence and understanding of effective strategies for stimulating policy dialogue. Support independent, sustainable, and capable local media in developing countries. Local media in developing countries often enjoy significant reach and audience interest, but lack the resources, skills and support to better understand the needs of populations and effectively hold government to account. In supporting these organisations to better enact their watchdog role, donors can effectively enhance non-media accountability interventions, build peoples demand for domestic accountability, and strengthen local media as an accountability institution. Foster ownership as a central component of support. The nature of productive relationships between media and audiences is one that engenders a sense of ownership. Where people see media acting on their behalf and critically enabling them to engage directly with issues and politicians there exists a clear sense of trust and ownership of media programmes. Promote citizen access to media and mobile technologies as well as citizens media literacy. Media can be an effective accountability mechanism only if citizens are able to utilize them. This includes access to media products and infrastructure as well as the ability to make sense of information.

47

Encourage links between media institutions and the rest of civil society. Media and civil society organizations together can form a formidable coalition for accountability and good governance. Donors should consider joining support for several accountability mechanisms, including media support, in appropriate situations. Support systematic research on the effects of media and information access on domestic accountability. As outlined in this discussion paper, empirical evidence on media effects on domestic accountability is available, but not integrated into a larger theoretical framework. Research, including monitoring and evaluation, should be part of any media support project, but should also be supported in its own right to advance our understanding of the role of media for domestic accountability in different political, economic, and social contexts. Learn about and harness new technologies. Internet and mobile-focused support is not appropriate in all contexts. Needs analyses must properly assess media and communications environments to determine the most appropriate media platforms for supporting accountability. Where interventions do focus on new technologies, research should be incorporated to build a body of policy-relevant evidence to guide subsequent support.

48

4) Principles for Support to Parliaments

Parliaments perform a vital role in any system of representative democracy, but they play an especially important role in emerging democracies - not only in improving the quality of governance by ensuring transparency and accountability, but also playing a critical role in shaping the publics expectations and attitudes to democracy. Parliaments are the single most important institution in overseeing government activity, scrutinising legislation and representing the publics concerns to those in power. Their performance in holding government to account and engagement with voters will help to establish the norms and values in the early years of a democratic culture. Although traditionally a small part of international support programmes, donors have paid greater attention to the role of parliaments in the last decade or so. Most support programmes usually seek to improve the effectiveness of the institution in one of their three key functions: Legislation - Assessment of the legislative function will be concerned with how well parliament scrutinises and amends bills, or simply acts as a rubber-stamp for the Executive; Oversight - Parliamentary oversight is the main means by which government is held to account, parliaments should ensure government departments are run efficiently and that ministers are regularly called to account for their actions, policies and spending; and Representation - Parliament ultimately derives its legitimacy from its ability to represent and articulate public concern and programmes tend to concentrate on the representativeness of parliament (that is how its make-up reflects wider society) and the extent to which MPs consult and engage with voters. In short, a parliaments role is to provide a check on the activity of government. The role might be thought of as providing government by explanation. That is, highlighting issues of concern and ensuring that government is able to justify its actions to the public, or where that policy is deficient, forcing a change. It is in this last area where parliaments have the potential to be most effective in strengthening systems of domestic accountability. Parliaments derive much of their authority from the fact that a number of accountability institutions usually report to them - ranging from the supreme audit institution, the ombudsman and the electoral commission, through to utility regulators, inspectorates and agencies. Such institutions provide a wealth of information on the performance of government in specific policy areas, and provide the evidence on which parliament can hold ministers, and ministries to account. In other words, parliaments should sit at the centre of a web of domestic accountability. Parliaments are therefore potentially vital allies for donor agencies in improving domestic accountability. Yet in many parts of the world legislatures have fallen far short of public (and donor) expectations. In emerging democracies, parliaments are frequently ineffective against a powerful executive, and have little public legitimacy and authority. International support to parliaments has a poor record in improving parliamentary performance. In the last decade donors have placed greater emphasis on parliamentary assistance and there has been a much greater degree of co-ordination and lesson-learning amongst donor agencies. However, traditional approaches to parliamentary support have 49

been essentially technical, relying on superficial analyses of the problems facing parliaments and rarely understanding the political, social and economic context within which they operate. The following principles are neither exhaustive nor comprehensive, but provide a starting point for guidance on parliamentary support projects: Integrate objectives with wider efforts to support accountability: Support to parliamentary institutions should be integrated with wider efforts to support domestic accountability. Support programmes should seek to increase the extent to which parliaments engage without outside institutions (such as the supreme audit institution), and ensure that other programmes designed to strengthen other mechanisms of accountability feed into, and strengthen the parliament.

Focus on institutional change leading to behavioural change: Ultimately, the effectiveness of a parliament will be determined by the behaviour of individuals within it. The purpose of a support programme should ultimately be to change that behaviour so that parliamentarians understand their role in holding government to account, have the resources and capacity to use the relevant procedures effectively, but also have the incentive to perform their accountability function. Understand parliaments incentive structures: Many support programmes assume that all parliamentarians would like a stronger parliament and that donor assistance will inevitably be welcomed. This is rarely the case, and support programmes need to understand the various incentive structures within a parliament, how they are currently shaping political behaviour and how they might be used to generate cross-party backing for the initiative. Dont ignore political parties. One of the strongest influences on behaviour in parliament will be the political parties. Fears of political interference often discourage donors from engaging directly with parties. But, to encourage a less partisan role, programmes will need to understand and work with the political parties in parliament. Programmes should provide them with the opportunities and incentives to engage on a cross-party basis, without compromising donor neutrality. Identify and address the causes of underlying parliamentary weakness. Programmes must be clear about the underlying causes of underperformance. It may be immediately apparent that the parliament is poor at financial oversight, but support projects need to assess whether this is to do with the parliaments constitutional position, its procedures, resources, experience or political complexion. Most often, it is a combination of several factors. Ensure parliamentarians own the problems - and their solution. Local ownership is particularly significant in trying to foster political and behavioural change. Given the complexity of getting change through a parliament, projects need a) a widespread concern that parliament is underperforming, b) cross-party agreement of the reasons for that weakness, and c) some internal consensus that the projects objectives are the best way to address those problems. As such, parliamentary support projects need to be 50

developed in partnership with key interlocutors within the institution, often politicians and staff. Take gender seriously in relation to parliamentary performance. The underrepresentation of women in political decision-making structures has implications at many levels. Support should seek to increase the number of women elected to national parliaments and to improve the impact of parliaments in developing policies that take into account their effect on women and men. Design projects around outcomes rather than activities. Programmes should maintain a clear sense of what they are designed to achieve. Frequently process and outcomes are conflated, with donors measuring activities instead of the impact they were originally designed to have. An outcome-driven approach requires a much greater degree of flexibility in the design and delivery of programmes, with co-ordinated interventions at different parts of the parliament, designed to achieve the same end. Set realistic objectives and a realistic timescale. The conditions for achieving parliamentary change will vary between institutions, but donor-supported programmes need to work from the understanding that in most parliaments change will be haphazard and unpredictable, and that the interests of MPs will change over time. Support should not assume linear development and should recognise that political change happens slowly. It should also recognise that the scope for political change is often limited, and projects which seek discrete objectives will frequently be more effective than institutionwide reform. Match indicators to expected outcomes. Once indicators are in place they tend to determine subsequent project activity meaning that with the wrong indicators, projects do the wrong things. Indicators should be based on a thorough analysis which involves stakeholder perceptions of performance. This sort of monitoring and evaluation needs to be built in at project design stage, and should be a regular and on-going feature of parliamentary support programmes. Recognise that the timing of any project will be a key determinant in its success. The best point to establish new ways of working is immediately after an election. At this point there is likely to be a large number of new MPs, the committees will have a new complexion and the government ministries they monitor are also likely to have changed personnel. Induction programmes should aim to establish new patterns of working and reinforce key principles.

51

Annex 2: Country case studies

Information on case studies TBC

Annex 3: Political economy frameworks

TBC

52

References
Bjorkman, M and Svensson, J (2009) Power to the People: Evidence from a Randomised Field Experiment on Community-based Monitoring in Uganda, Quarterly Journal of Economics, Vol 124 No2, pp735-69 Blankenberg, F. (2007) Taking responsibility and demanding rights: Accountability in service delivery, Dutch Ministry of Foreign Affairs Booth, D and Golooba Mutebi, F (2009) Aiding Economic Growth in Africa The Political Economy of Roads Reform in Uganda, ODI Working Paper 307 Booth (2010) APPP Working Paper 13, Towards a theory of local governance and public goods' provision in sub-Saharan Africa, Africa Power and Politics Programme/ODI Bolnick, B with Hadler, S, Nathan Associates Inc (2011) Appropriate Aid Modalities for Supporting Tax Systems Brutigam, D., Fjeldstad, O. H. and Moore, M. (2008) Taxation and State Building in Developing Countries: Capacity and Consent, Cambridge University Press: Cambridge Carlitz, R (2010) Review of Impact and Effectiveness of Transparency and Accountability Initiatives: Annex 2 Budget Processes, Prepared for the Transparency and Accountability Initiative Workshop, Oct 14-15, Institute for Development Studies Collinson, S (2003) http://www.odi.org.uk/resources/download/241.pdf Claasen M. and Alpn-Lardis, C. (Eds) (2010) Social Accountability in Africa: Practitioners Experiences and Lessons. Cape Town: IDASA/ANSA Africa. Carothers, T (2006) Confronting the Weakest Link: Aiding Political Parties in New Democracies, New York: Carnegie Endowment for International Peace Centre for the Future State (2010) An Upside Down View of Governance, Brighton: Institute of Development Studies Collaborative Africa Budget Reform Initiative (2009) Putting Aid on Budget Using Country Budget Systems. Good Practice Note. Pretoria: CABRI Crook, Richard C. and Alan S. Sverrisson (2003) Does Decentralization Contribute to Poverty Reduction? Surveying the Evidence in P.P. Houtzager and M. Moore (eds.) Changing Paths: International Development and the New Politics of Inclusion. Ann Arbor: University of Michigan Press: 233-259. Crook, Richard Charles and James Manor (1998) Democracy and Decentralisation in South Asia and West Africa: Participation, Accountability and Performance. Cambridge: Cambridge University Press. Domingo, Pilar, Wild, Leni, Hudson, Alan, Wathne, Cecilie (2009) Domestic and Mutual Accountability for Aid: Building stronger synergies - Literature Review and Conceptual Framework, ODI 53

Eaton, K., Kaiser, K. and Smoke, P. (2010) The Political Economy of Decentralization Reforms. Washington DC: The World Bank. Edwards, M (2003) Civil Society Eldon, J. and Gunby, D. (2009) States in development: State building and service delivery, Draft Report for DFID Fritz, V., K. Kaiser and B. Levy (2009) Problem Driven Governance and Political Economy Analysis: Good Practice Framework. Washington DC: World Bank Fox, J. 2007, The uncertain relationship between transparency and accountability, Development in Practice 17. 4-5: 663-671 Goldfrank, B (2006) Lessons from Latin American Experience in Participatory Budgeting, Presentation at the Latin American Studies Association Meeting, San Juan, Puerto Rico Golooba-Mutebi, F (2005) When Popular Participation Wont Improve Service Provision: Primary Health Care in Uganda, Development Policy Review, 23 (2): 165-182 Government of Rwanda (2010) Kigali Statement of Action (KSA) of Ninth Biennial Government of Rwanda and Development Partners Meeting, 4-5th November 2010, Kigali, Rwanda, Endorsed on 5th November, please see here for more information: http://www.aidtransparency.net/news/rwanda-enhancing-transparency-and-mutualaccountability-through-donor-performance-assessment-framework-dpaf Grindle, Merilee (2010) Good Enough Governance Revisited, Development Policy Review 29 (s1): pp s199-s221 Grindle, Merilee (2002) Despite the Odds: The Political Economy of Social Sector Reform in Latin America in Abel, C, Lewis, CM (eds.) Exclusion and Engagement: Social Policy in Latin America, Institute of Latin America Studies: London Grindle, M (2004) Despite the Odds: The Contentious Politics of Education Reform, Princeton University Press: Princeton, New Jersey Harris, Daniel and Wild, Leni with Kizza, Diana (2010) Aid and accountability in Uganda: Lessons from the health sector, ODI Hubbard, P. (2007) Putting the Power of Transparency in Context: Informations Role in Reducing Corruption in Ugandas Education Sector. Working Paper Number 136. Washington, DC: Center for Global Development. Available at http://www.cgdev.org/content/publications/detail/15050/ Hudson, A, GOVNET (2009) Background Paper for the Launch of the Workstream on Aid and Domestic Accountability, OECD DAC GOVNET. Paris: OECD Humphreys, M, Weinstein, J (2010) Policing Politicians: Citizen Empowerment and Political Accountability in Uganda, Columbia University, www.columbia.edu/~mh2245/papers1/scorecard2010.pdf

54

International Budget Partnership (2010a) Open Budget Index 2010: Uganda Country Report. Washington, DC: International Budget Partnership - Center on Budget and Policy Priorities. Available online at: www.internationalbudget.org Irish Aid (2008) Good governance, aid modalities and poverty reduction, From better practice to better theory, The Advisory Board for Irish Aid Joshi, A (2010) Review of Impact and Effectiveness of Transparency and Accountability Initiatives: Annex 1 Service Delivery, Prepared for the Transparency and Accountability Initiative Workshop, Oct 14-15, Institute for Development Studies Kelsall, T (2011) Going with the Grain in African Development? Development Policy Review, Vol 29: S1, S223-S259 Killick, T. (2004) The democratic deficit and the politics of Ghanas budgetary system Kuteesa, F, Tumusiime-Mutebile, E, Whitworth, A, Williamson, T (ed.) (2009) Ugandas Economic Reforms: Insider Accounts. Oxford: Oxford University Press Mainwaring, S. 2003. Introduction: Democratic Accountability in Latin America. In S. Mainwaring and C. Welna eds. Democratic Accountability in Latin America. Oxford: Oxford University Press. McGee, Rosemary & John Gaventa with contributions from Greg Barrett, Richard Calland, Ruth Carlitz, Anuradha Joshi and Andrs Meja Acosta (2010) Review of Impact and Effectiveness of Transparency and Accountability Initiatives: Synthesis Report, Prepared for the Transparency and Accountability Initiative Workshop, October 14 15, 2010, Institute of Development Studies McGee, R (2010) Review of Impact and Effectiveness of Transparency and Accountability Initiatives, Annex 5: Aid Transparency, Prepared for the Transparency and Accountability Initiative Workshop, Oct 14-15, Institute for Development Studies Moon, S, Williamson, T (2009) Greater Aid Transparency: Crucial for Aid Effectiveness, Project Briefing 35. London: ODI Moore, M. (2007) How Does Taxation Affect the Quality of Governance? IDS Working Paper 280. Brighton: IDS Moore, M., A. Schmidt and S. Unsworth (2009) Assuring Our Common Future in a Globalised World: The Global Context of Conflict and State Fragility, DFID background paper for the 2009 White Paper Securing our Common Future, London: Department for International Development Mulley, Sarah (2010) Transparency and Accountability Initiative, New Frontiers: Aid and Donor Finance, ippr OECD DAC (2011a) Donor support to domestic accountability: Budget processes and service delivery in Uganda OECD DAC (2011b) Donor support to domestic accountability: Budget processes and service delivery in Mali

55

OECD DAC (2011c) Donor support to domestic accountability: Budget processes and service delivery in Peru OECD DAC (2011c) Donor support to domestic accountability: Budget processes and service delivery in Mozambique Ostrom, E., C. Gibson, S. Shivakumar and K. Andersson (2001) Aid, Incentives, and Sustainability: An Institutional Analysis of Development Cooperation, Sida Studies in Evaluation 02/01, Stockholm: SIDA Pawson, R. and N. Tilley (1997) Realistic Evaluation, London: Sage Power, G (2008) Donor support to parliaments and political parties: An analysis prepared for DANIDA, London: Global Partners and Associates Prichard/GOVNET (2010) Citizen-State Relations - Improving governance through tax reform, OECD: Paris Rakner, Lise, Luke Mukubvu, Naomi Ngwira, Kimberly Smiddy and Aaron Schneider (2004) The budget as theatre - the formal and informal institutional makins of the budget process in Malawi. Final report Bergen: Chr. Michelsen Institute Robinson, M (2006) Budget analysis and policy advocacy: The role of non-governmental public action, IDS working paper 279, Institute of Development Studies Rocha Menocal, A and Sharma, B (2008) Joint Evaluation of Citizens Voice and Accountability: Synthesis Report, London: DFID Steer L, C Wathne and R Driscoll. (2009). Mutual accountability at the country level: a concept and emerging good practice paper. Centre for Aid and Public Expenditure. Overseas Development Institute, London Sundet, G (2008) Following the money: do Public Expenditure Tracking Surveys matter? U4 ISSUE 2008:8, Anti-Corruption Resource Centre, CMI Tsekpo, A, Hudson, A (2009) Parliamentary Strengthening and the Paris Principles: Uganda case study, London: ODI VNG International/African Studies Centre (2011a) Domestic Accountability at the Local Level Inventory Study Country Case Study Report (Tanzania) VNG International/African Studies Centre (2011b) Domestic Accountability at the Local Level Inventory Study Country Case Study Report (Benin)

Williamson, T (2008) Putting Aid On Budget: A Case Study of Uganda, A Study for the Collaborative Africa Budget Reform Initiative (CABRI) and the Strategic Partnership with Africa (SPA), Mokoro Ltd Williamson, T, Dom, C (2010) Sector Budget Support in Practice Synthesis Report, London: ODI/Mokoro. 56

Wild, Leni, Foresti, Marta, Domingo, Pilar (2011) International assistance to political party and party system development: Synthesis Report, ODI Wild, Leni, Foresti, Marta (2011) Putting Politics into Practice reference TBC (Pogo report) Wild, Leni, Harris, Daniel (2011) Beyond the ballot box: building bridges between citizens and the state, ODI Opinion, February http://www.odi.org.uk/resources/download/5487.pdf Wild, L. and Golooba-Mutebi, F. (2010) Review of international assistance to political party and party system development - Case study report: Uganda. London: ODI World Bank (2004) World Development Report: Services Work for Poor People, World Bank: Washington DC World Bank (2011) World Development Report: Conflict, Security and Development, World Bank: Washington DC

NOTES
i

See Governance and the Media: A survey of policy opinion, BBC World Service Trust, 2009. The word media was inserted, within the context of mutual accountability, at the last moment a few weeks before Accra as a result of advocacy by the BBC World Service Trust. iii It is notable that the most recent and perhaps most useful published analysis of European spending on support to media was commissioned from the National Endowment for Democracy in the United States. iv The Technology for Transparency Network initiative catalogues accountability projects that use mobile technology and social media as accountability tools (http://transparency.globalvoicesonline.org/).
ii

57