Evaluation of DFID’s work on electoral support through UNDP

1 Introduction

1.1 The Independent Commission for Aid Impact (ICAI) is the independent body responsible for scrutinising UK aid. We focus on maximising the effectiveness of the UK aid budget for intended beneficiaries and on delivering value for money for UK taxpayers. We carry out independent reviews of aid programmes and of issues affecting the delivery of UK aid. We publish transparent, impartial and objective reports to provide evidence and clear recommendations to support UK Government decision-making and to strengthen the accountability of the aid programme. Our reports are written to be accessible to a general readership and we use a simple ‘traffic light’ system to report our judgement on each programme or topic we review.

1.2 We have decided to conduct a review of the management of UK electoral assistance through the United Nations Development Programme (UNDP), a major actor in the delivery of electoral assistance internationally and the implementing partner for most of the UK’s electoral assistance. In carrying out this review, we have agreed to co-ordinate with UNDP. This will be a collaborative process, rather than a joint evaluation, with the two organisations sharing their research but drawing fully independent conclusions and each publishing its own report. Working collaboratively will reduce the burden on partners while broadening the reach of both evaluations.

1.3 This inception report sets out the evaluation questions, methodology and a work plan for the delivery of the evaluation. It is, however, intended that the methodology and work plan be flexible enough to allow for new issues and questions that emerge over the course of the evaluation.

2 Background

2.1 One of the goals of UK development assistance is to support strengthening democracy. Although the electoral system is only one of many institutions required to make a functioning democracy, elections are a necessary part of the democratic process. Free and fair elections call for a complex set of rules and organisational capacities, to enable genuine political competition, manage the different processes involved and ensure that voters are able to participate effectively.

2.2 Over the period 2004-09, the Department for International Development (DFID) provided election-related support in 25 countries at a total estimated cost of £121 million. This was one part of DFID’s wider support to political systems (including parliaments, political parties and accountability) which totalled approximately £234 million over the same period or 8% of the governance portfolio.¹ In the current Spending Review period (2011-15), DFID will support elections in at least 13 countries.²

2.3 Most UK electoral assistance has been directed to supporting countries undertaking specific elections. Disbursement tends to be at its largest within the year preceding an election and over the election period itself. Electoral assistance has tended then to significantly diminish or cease until needed for a subsequent election. Experience shows that this pattern of event-driven electoral assistance largely failed to build sustainable capacity within national electoral systems.³ In recent years, a consensus has emerged internationally on the need for an electoral cycle approach – that is, for sustained support throughout the entire electoral cycle, focussing on the development of robust systems and sustainable capacity. In December 2010, DFID and the Foreign and Commonwealth Office (FCO) adopted the electoral cycle approach in a joint guidance note on electoral assistance.⁴ It remains to be seen how much delivery will change away from a pattern of finance that is more narrowly episodic.

2.4 The UK usually chooses to provide the bulk of its electoral assistance with other donors through a common or basket fund. Most of these electoral basket funds are managed by UNDP, whose global mandate as a UN agency helps it to engage in what can be a sensitive political arena. In a few instances, basket funds are managed by other bilateral donors or by non-governmental organisations (NGOs).

2.5 In most cases, Electoral Management Bodies (often Electoral Commissions) are the largest beneficiaries of such basket funds. Typically, these organisations require significant assistance as they are not sufficiently resourced domestically and are often orphans within the national administrative structures.⁵ A range of activities is typically funded out of the basket funds, such as compilation of the electoral roll, boundary delimitation, procuring election materials, voter education, providing technical assistance to build capacity where needed, funding domestic observation and the delivery of the process of the election itself, including the collection, tabulation and reporting of votes. UNDP officials and contracted consultants will engage with the Electoral Management Body in these activities. Other areas that may also be funded through UNDP basket funding include strengthening security around elections, supporting the media and enabling international observation.

2.6 As part of its electoral assistance funding, the UK usually supports the UN in its political role. The focus of such activity is to ensure compliance with international standards for the conduct of elections and, crucially, respect for election results. The most senior UN official (the UN Resident Co-ordinator) tends also to be the head of the UNDP office, the UNDP Resident Representative. It therefore typically falls to the UNDP Resident Representative to present the views of the international community to leaders in-country and maintain the channels for political dialogue, particularly if finalising the political transition after elections requires international mediation and negotiation. In such cases (for instance, where results have been contested or the elections marred by violence) the UN system will generally be a key mechanism for reaching a solution. The UK will often fund such mediation activities and engage actively with the UN (and specifically UNDP) throughout the process.

2.7 The UK also provides a range of smaller-scale electoral assistance through other channels, such as support to local NGOs or international bodies for election observation or voter education campaigns. UK and international organisations, such as the International Foundation for Electoral Systems (IFES), are beneficiaries of electoral assistance. On occasion, DFID also finances the work of the EU and Commonwealth Secretariat to undertake election observation missions at country level.

3 Purpose

3.1 To assess whether DFID’s funding for electoral support through UNDP is being managed so as to be effective and deliver value for money.

4 Relationship to other evaluations and studies

4.1 The Multilateral Aid Review,⁶ undertaken by DFID in 2010 and published in 2011, underlined that UNDP is and will remain an important partner for DFID. It noted that, in 2008-09 alone, DFID had channelled £264 million of Official Development Assistance (ODA) through UNDP (£55 million in core funding, £86 million via multi-donor trust funds, £98 million in direct support of projects at country level and £25 million via system-wide funds and thematic funds). The review’s key conclusions are set out below. These general findings from the Multilateral Aid Review will form an important back-drop to this evaluation.

4.4 A preliminary review of this and other material sees patterns emerging. UNDP is the only practicable provider of large-scale support for elections in most contexts, principally as a result of its international mandate. DFID, therefore, often has little or no choice of partner organisation for delivering electoral assistance. UNDP’s functions are typically:

• financial management of large-scale funding in support of elections, often in partnership with domestic electoral commissions;
• delivering technical assistance to build capacity in-country (among both government and civil society);
• procurement of election materials and equipment;
• co-ordination of external assistance; and
• political engagement with government on the conduct of the elections.

4.5 Initial findings indicate that the last role can dominate, making a focus on delivery often problematic (both between UNDP and its operational partners in government such as Election Commissions and between DFID and UNDP).

4.6 In 2011, UNDP’s Evaluation Office chose to undertake an evaluation of its work in elections. Simultaneously, the UNDP’s Bureau for Development Policy also chose to undertake a review of the organisation’s work on the same topic. We and the UNDP Evaluation Office have agreed, where possible, to co-ordinate our work during the period of our evaluations. This co-ordination seeks to be mutually beneficial, enhancing our understanding of the effectiveness and value for money provided by UNDP to DFID and informing the UNDP Evaluation Office with information gleaned from our work.

4.7 UNDP will undertake both a portfolio review of UNDP’s electoral support since the 1990s and country case studies. The UNDP’s case studies have been chosen to cover a range of country contexts, including immediate post-conflict, conflict/transition and development, as well as different types of electoral assistance (i.e. event-based or electoral cycle approach). Three of the case studies, which will be undertaken in-country, are countries where DFID has provided significant funding; Afghanistan, Bangladesh and Mozambique. Other countries under consideration by UNDP are Bolivia, Chad, Guinea/Conakry, Guyana, Haiti, Kyrgyzstan, Lebanon, Mexico and possibly Yemen (a country to which DFID has provided money for elections).

5 Methodology

5.1 Our evaluation will consider UNDP’s management of UK electoral support, including:

• whether the choice of delivery options and partners is delivering effective assistance and good value for money;
• whether DFID is providing effective management and oversight of UNDP’s electoral assistance; and
• whether UK electoral support is anchored in a credible political engagement strategy and a broader strategy of support for political system development.

5.2 These topics will be framed using ICAI’s standard evaluation framework. Our approach seeks to consider the relationship between DFID and UNDP from the perspective of the delivery of election support in-country, whilst also gathering perspectives from within DFID and UNDP.

5.3 We have agreed that we will engage in a dialogue with the UNDP Evaluation Office to develop approaches and learn lessons from each other’s work. We will share findings on all country studies. In particular, we will shadow the UNDP’s evaluation of work in Bangladesh and in Afghanistan. Findings, however, will be produced and reported independently by each organisation.

Evaluation Framework

The evaluation framework for this review is set out in the table below. This has as its basis the standard ICAI guiding criteria and evaluation framework, which are focussed on four areas: objectives, delivery, impact and learning. It also incorporates other pertinent questions we want to investigate in this review. The questions which are highlighted in bold, including those from our Terms of Reference (ToR), are those on which we will focus in particular.

Using questions aimed at the case-study level, the evaluation framework focusses on the achievements of DFID’s elections-related investments that have been channelled through UNDP. We will collate our case study findings along with cross-cutting information (such as the Governance Advisers review described in more detail in 5.5, element 4 below) to set aggregate traffic light ratings. We will consider in particular whether UK assistance through UNDP is delivering according to the UK’s policy on funding elections as part of a broader commitment to democracy; and whether the UK is supporting a holistic approach throughout the entire electoral cycle.

5.5 The evaluation has the following elements. These will be undertaken in parallel.

1. a mapping of DFID-supported electoral assistance 2001-2011;
2. four country case studies (two involving in-country studies; two desk reviews shadowing UNDP’s own evaluation);
3. a review of data from the UNDP evaluations;
4. dialogue with DFID internal staff to gather their assessment of UNDP as a provider of electoral assistance; and
5. a review of DFID-UNDP management arrangements for oversight of elections.

Element 1: Mapping of UK Electoral Assistance

A mapping of UK electoral assistance according to country, expenditure and delivery method/partner, conducted through analysis of information on DFID’s management information systems (Aries and Quest) and interviews with DFID staff in London.

This will require:

a) a query to be run to generate a single spreadsheet with high-level data, by DFID project code; and

b) a specific investigation of each project code using project data from Quest and/or Aries to identify funds flow, disbursement patterns and the link between expenditure and performance reporting. In particular, we will seek to identify the terms of conditions set out in any Memoranda of Understanding. This will enable information to be provided in detail on who receives funds from DFID for elections, the cycle of disbursement and the nature of the accountability for expenditure (i.e. the terms on which the finance is provided through UNDP).

Element 2: Country Case Studies

Four country case studies (two involving country visits; two desk studies shadowing UNDP’s evaluations) of work in countries where UK electoral assistance is provided through a UNDP-managed basket fund, as well as other channels.

For the two country visits to Burundi and Malawi, we will interview representatives from UNDP, DFID, civil society, the election management bodies, other funders and politicians, using semi-structured interviews guided by the evaluation framework. These visits will be undertaken by a two-person team and will take place between 27 November and 8 December 2011. We will use information from DFID and elsewhere to inform these visits.

Burundi (Country Visit)

DFID has supported elections in Burundi at least since 2004. For the last elections, held in 2010, UNDP managed a US$46 million basket fund. DFID contributed £1 million to it. The presidential and parliamentary elections of July 2010 were accompanied by some violent incidents. This followed allegations of widespread fraud in the May 2010 communal elections. Although international observers considered these to be unfounded,¹⁰ all six opposition candidates dropped out of the presidential election in June 2010, leaving incumbent President Pierre Nkurunziza as the sole candidate. Only two major opposition parties participated in the legislative elections in July. The ruling party won over 80% of parliament seats.

As part of its Bilateral Aid Review in 2010-11,¹¹ DFID decided that its bilateral aid programme to Burundi offered poor value for money compared with other, larger country programmes. DFID said that a large scale-up would have been required to show a significant impact and therefore demonstrate better value for money. DFID believed that achieving this in the short term would have been difficult given capacity constraints in-country. For this and other reasons, DFID decided to close down the bilateral programme and to allocate these resources to larger existing programmes, where it was felt that better value for money and effectiveness could be achieved. DFID maintained some financial support through other channels, in particular a specific regional programme focussing on economic integration. The International Development Committee (IDC) recently carried out an inquiry into this decision, concluding that the bilateral aid programme should be reinstated.¹²

Malawi (Country Visit)

DFID has been supporting electoral processes in Malawi since at least 2000. The DFID support to Malawi’s 2009 presidential and parliamentary elections began three years prior to the elections in 2007. Support to the electoral process was managed through a contribution to a UNDP-administered basket fund. Following the 2009 elections, DFID in collaboration with the Foreign and Commonwealth Office provided support to civic and voter education activities targeted towards the 2010 local government elections. Malawi has faced allegations of electoral fraud. The 2009 elections were peaceful and well-managed but with elements judged as below international standards, particularly abuse of incumbency. Since the election, governance in Malawi has deteriorated markedly with suppression of demonstrations, intimidation of civil society organisations and new laws empowering the government to restrict political opposition. This led to DFID suspending budget support to Malawi in July 2011. The difficult political environment makes for an interesting case study as to how effectively UNDP and DFID are able to defend the independence of the electoral body.

For the two case studies where we are shadowing the UNDP evaluations (Bangladesh and Afghanistan), we will undertake a high-level assessment of DFID’s internal project documentation and lesson learning, as well as a review of open-source third party documentation. UNDP has readily undertaken to share its preliminary and final conclusions with us prior to publication. We have also agreed to engage with the team leaders of each of the UNDP evaluations to exchange information on emerging findings.

Afghanistan (Desk study shadowing UNDP’s evaluation)

Undertaking elections in states where active stabilisation is underway, such as Afghanistan, is a significant challenge. DFID’s internal documentation notes donor fatigue after the 2004-05 elections, which resulted in missed opportunities to provide sustained support for Afghanistan’s electoral bodies. Subsequent to this period, however, new management bodies, the Independent Electoral Commission and the Electoral Complaints Commission, were established, becoming a focus for international support and finance. The majority of external funding for both the 2004-05 and 2010-11 elections has been channelled through a UNDP-managed multi-donor trust fund with activities implemented through the associated programme, UNDP Enhancing Legal and Electoral Capacity for Tomorrow (ELECT).¹³ UNDP’s own reporting shows a total expenditure of US$378 million, provided by 22 bilateral funders, the EU, UNDP and the United Nations Democracy Fund (UNDEF). DFID’s reporting shows it contributed £25.77 million (US$36.96 million) for elections over this period, primarily through this fund. Given the unique circumstances in Afghanistan where the international engagement is governed by a United Nations Security Council Resolution (UNSCR 1917), co-ordination mechanisms for electoral assistance are specified as the role of the United Nations Assistance Mission in Afghanistan and the United Nations Special Representative of the Secretary General. As in other contexts, the UK government has also funded work with civil society bodies in Afghanistan, outside the work of UNDP’s programme. In Afghanistan, for instance, the UK Conflict Pool¹⁴ supported the work of the Asian Network for Free Elections, implemented through the Asia Foundation. This was in collaboration with the United States Agency for International Development (USAID) and the Australian Agency for International Development (AUSAID). UNDP will be concluding its evaluation of Afghanistan by the second week in January 2012.

Bangladesh (Desk study shadowing UNDP’s evaluation)

The UK has supported elections in Bangladesh since the country’s birth in 1971. Elections during the last decade have been uncertain and associated with violence. In 2006, elections were suspended and delayed for two years after the Caretaker Government stayed beyond its mandated three-month period. The principal constitutional responsibility of the Caretaker Government was to ensure free and fair elections. With the military’s backing, the Caretaker Government ran until December 2008 when, after significant investments from the international community, the country held substantially free and fair elections. The bulk of DFID’s assistance for the elections was channelled through UNDP which, as well as managing the multi-donor trust fund, also implemented a series of electoral reform programmes including the Support to the Electoral Process Project (US$4.16 million), Preparation of Electoral Rolls with Photographs (US$78.7 million), the Transparent Ballot Boxes Project (US$6.2 million) and construction of Server Stations for the Electoral Database at a cost of US$47.2 million.

UNDP’s recent country evaluation notes: ‘The Support to Electoral Process in Bangladesh Project was ingenious and flexible in responding to emerging needs in a changed political environment. It directly assisted the implementation of an ambitious roadmap of the Election Commission for holding a free and fair election on the basis of a new voter list and new rules within two years. The preoccupation of the project, however, with the election emergency impeded UNDP’s support for the institutional development of the Election Training Institute and of the Election Commission and its secretariat.’¹⁵

Element 3: Review of Data from the UNDP evaluation

A review of data and analysis from the UNDP evaluation. UNDP will provide us with raw and emerging findings to inform our deliberations. These will relate to all its country studies. Where appropriate, we will also participate in discussions with the UNDP team around emerging conclusions.

Element 4: A review of DFID internal staff to gather their assessment of UNDP as a provider of electoral assistance

A written survey of the DFID Governance Advisers responsible for managing the last episode of electoral assistance in the highest-spending 25 countries to solicit their views on the strengths and weaknesses of different delivery options and partners. We have agreed with DFID that findings (anonymised and at synthesis level only) from this element of the evaluation will be shared with DFID to assist with its ongoing development of the relationship with UNDP. We have also agreed to share this information, in the same manner, with UNDP.

Element 5: Review of DFID – UNDP oversight of elections

This will include interviews with DFID’s United Nations and Commonwealth Department in London and interviews with UNDP management personnel in New York (carried out by staff from KPMG New York).

We will carry out a review of available documentation on lesson learning regarding UK electoral assistance, including individual project reviews and the case studies commissioned by DFID and FCO in 2008.

6 Roles and responsibilities

6.1 KPMG will provide oversight of this review under the overall leadership of ICAI Project Director. The team will consist of the following members:

He is a Director of Agulhas Applied Knowledge. He specialises in aid effectiveness, governance and institutional development. He was originally a health service manager in the UK and has worked on health service reform projects throughout Africa and Asia. He is a member of the core management team for ICAI’s implementation and led the Bangladesh climate change evaluation. He has particular knowledge of the subcontinent, having in the past been a governance adviser for DFID in Bangladesh in the late 1990s.

Team member 1 (Agulhas)

With over 15 years in policy analysis, he has worked for a variety of clients on a range of high-level policy issues including implementation of the Paris Declaration, aid effectiveness and fragile states. He is an authority in international law and human rights and has written widely on post-conflict reconstruction, state-building and the restitution of property. He has more recently specialised in aid effectiveness and governance processes at all levels, including policy development, programme design and evaluation.

Team member 2 (Agulhas)

She is a consultant with Agulhas Applied Knowledge and a former programme manager in the DFID Somalia team. As a consultant for the past three years, she has undertaken reviews of the aid architecture in Kenya and Kyrgyzstan. She is currently working on the assessment of governance reforms in five African states. She is based in Mombasa and has lived and worked in East Africa for most of the last decade. She also undertook the ICAI evaluation of Bangladesh climate change programmes.

Team member 3 (KPMG)

She works in KPMG’s Public Sector Audit Department specialising in external audit and internal audit. Her client base is varied, ranging from central government, local government, health services, trade unions and housing associations. She has over five years’ professional experience, including the provision of legal representation at an international level and the provision of pro bono services to a range of high-profile not-for-profit organisations. She also took part in ICAI’s review of its approach to effectiveness and value for money.

Team member 4 (KPMG)

She is an advisor in KPMG’s Management Consulting Public Sector group, focussing on organisational financial management. She has over ten years’ experience in auditing and advising public sector and government clients. Her main role will be to analyse UNDP data sources and figures to support the findings of the report.

Team member 5 (KPMG)

She is a senior manager in KPMG’s International Development Assistance Services (IDAS) practice and is located in the firm’s New York office. She manages KPMG’s United Nations Desk and, in this capacity, develops strategic relationships and opportunities within the UN system and provides project management oversight of key engagements. She has ten years of international professional experience working at the intersection of international development, foreign assistance policy and global philanthropy. Her main role will be to conduct the UNDP head office data collection, first person interviews and financial analysis to support the findings of the report.

7 Management and reporting

7.1 A first draft report will be produced for review by the Secretariat and Commissioners by the end the week commencing 23 January 2012, followed by revision and review prior to completion and sign off by the end of February 2012.

8 Expected outputs and timeframe

8.1 The following timetable is indicative. The phases are not strictly sequential, so a degree of overlap is possible.

9 Risk assessment

10 How will this ICAI review make a difference?

10.1 It is unlikely that support for elections will cease to be a key part of the UK’s and DFID’s international efforts in the future. Similarly, the role of the United Nations in assisting governments to meet international norms and standards for law and democracy is also unlikely to change fundamentally. DFID and UNDP are, in a sense, tied together in this work for the foreseeable future. This work will provide independent scrutiny for both partners, assisting in the identification of issues for action that should lead to improvements in impact and delivery.

10.2 Certainly two (and, depending on definitions, all four) of the case study countries can be classified as fragile states. With the increasing focus of UK interest in fragile states, where by definition democracy has yet to be embedded, this work will also contribute specifically to improving the delivery and impact of UK assistance for elections in difficult environments.

10.3 In addition to the specific contribution to each institution’s work on elections, this evaluation will provide a contribution to the wider debate on how DFID engages with UNDP. This is a focus of interest for DFID, in the light of the recommendations of its Multilateral Aid Review.

Footnotes

1. Governance Portfolio Review Summary, DFID, July 2011, http://www.dfid.gov.uk/Documents/publications1/governance-portfolio.pdf.
2. Business Plan 2011-2015, DFID, May 2011, www.dfid.gov.uk/Documents/DFID-business-plan.pdf.
3. Draft Principles of Electoral Assistance, OECD Development Assistance Committee, March 2010 (updated June 2010), http://www.oecd.org/dataoecd/43/17/45881241.pdf. Joint Workshop on Effective Electoral Assistance: Participants’ Manual, European Commission, UNDP and International IDEA, June 2009, http://www.ec-undp-electoralassistance.org/index.php?option=com_docman&task=doc_download&gid=201&Itemid=.
4. How To Note on Electoral Assistance, DFID and FCO, December 2010, http://www.dfid.gov.uk/Documents/publications1/elections/how-to-on-elect-asst-summ.pdf.
5. Where they are weakly resourced, Electoral Management Bodies may be perceived to have less influence or impact and therefore be subject to less political pressure. External interference in Electoral Management Bodies, however, is dependent upon many factors.
6. Multilateral Aid Review: United Nations Development Programme (including the Bureau for Crisis Prevention and Recovery), DFID, 2011, http://www.dfid.gov.uk/Documents/publications1/mar/undp-inc-bcpr.pdf.
7. Roger Wilson & Bhavna Sharma, Review of UK Electoral Assistance in the context of lessons emerging from best practice in international experience, commissioned by the Politics and State Team, DFID Policy Division and the Human Rights, Democracy and Governance Group, FCO, December 2008, http://www.gsdrc.org/docs/open/Review_UKElectoralAssistance_2008.pdf.
8. Electoral Assistance and Politics: Lessons for International Support, DFID, 2010, http://www.dfid.gov.uk/Documents/publications1/elect-asst-pol-less-int-supp.pdf.
9. Roger Wilson & Bhavna Sharma, Review of UK Electoral Assistance in the context of lessons emerging from best practice in international experience, commissioned by the Politics and State Team, DFID Policy Division and the Human Rights, Democracy and Governance Group, FCO, December 2008, http://www.gsdrc.org/docs/open/Review_UKElectoralAssistance_2008.pdf.
10. Rapport Final, Elections Communales, Présidentielle, Législatives, Sénatoriales et Collinaires 2010: Mission d’Observation Électorale de l’Union Européenne, http://www.eueom.eu/files/pressreleases/other/final-report-burundi-2010_fr.pdf.
11. Bilateral Aid Review: Technical Report, DFID, March 2011, www.dfid.gov.uk/Documents/MAR/FINAL_BAR%20TECHNICAL%20REPORT.pdf.
12. The Closure of DFID’s Bilateral Aid Programme in Burundi, International Development Committee, 20 October 2011, HC 1134, http://www.publications.parliament.uk/pa/cm201012/cmselect/cmintdev/1134/1134vw.pdf.
13. UNDP ELECT project information, http://www.undp.org.af/whoweare/undpinafghanistan/Projects/dcse/prj_elect.htm.
14. The Conflict Pool is a funding mechanism for conflict prevention activities, managed jointly by the Foreign and Commonwealth Office, DFID and the Ministry of Defence.
15. Assessment of Development Results, Evaluation of UNDP Contribution Bangladesh, UNDP Evaluations Office, March 2011, http://erc.undp.org/evaluationadmin/downloaddocument.html?docid=4843.