DFID’s partnerships with civil society organisations

ICAI score

DFID values civil society organisations (CSOs), but its funding and partnership practices do not fully support the long-term health of the civil society sector. Weak management practices have led to delays and uncertainty. Accountability mechanisms are strong, but DFID pays insufficient attention to the global decline in civic space, the sustainability of the results of CSO work and the uptake of learning and successful innovation.


DFID sees civil society organisations (CSOs) as key partners in delivering humanitarian and development results. It sees a healthy civil society as a goal in itself, as CSOs hold governments to account, and include and give voice to marginalised groups. DFID has reduced the number of its centrally managed funding instruments and introduced stringent requirements, thereby improving CSO transparency and accountability. However, these strict requirements, in combination with time-consuming and costly application processes and a short-term project-based funding model, limit the ability of CSO partners to ensure their relevance, health and adaptability at a time when civic space in many countries continues to decline. DFID does not offer its country offices a framework within which to analyse and respond to this decline, and the wider UK government lacks clear objectives on how to achieve the UK’s goal of protecting and expanding civic space.

Weak process management and poor management of the review period’s many disruptive events, within DFID and most of its funding intermediaries, led to unreliable and unpredictable donor behaviour at all stages of the process leading up to contracts and funding agreements. This caused delays that affected the results of CSOs’ work. Still, these results were positive: in general, DFID-funded CSOs delivered effective development and humanitarian interventions. CSO partners saw DFID as a supportive donor once project implementation had begun. DFID provided useful practical support to protect CSOs, but its overall influencing and programming work in the field of civic space has been slow and limited.

DFID has recognised, but not yet filled, important knowledge gaps on how different funding mechanisms affect the results of CSO work for poor people. We saw evidence of active learning in a range of other areas, with DFID incorporating lessons from previous experience and from its engagement with CSOs and other stakeholders. DFID and its funding intermediaries have funded and facilitated a range of relevant learning systems, processes and events. These were generally well designed and proportionate to the nature of the projects being funded. However, we found less evidence of the results of these learning interventions being applied to programming. While DFID has several funding windows for innovation, its overall funding approach does not generally facilitate innovation, partly because of the strong focus on compliance processes. DFID does not systematically identify and promote innovation successes.


Executive summary

Civil society organisations, or CSOs, are not-for-profit entities, separate from states and the private sector, that operate in the public domain. CSOs covered in this review pursue development or humanitarian objectives, and often focus on improving conditions for the poorest and most vulnerable in society. They are diverse in their size, longevity, regional reach, sectoral focus, working methods and organisational set-up.

Over the past decade, between 15% and 25% of the Department for International Development (DFID)’s bilateral spending has been used to fund the work of CSOs. DFID spending on CSOs peaked in 2014-15, when it reached nearly £1.4 billion, or 25% of DFID’s bilateral spending. It declined to nearly £1.3 billion, or 20%, in 2016-17, largely because of the termination of one of DFID’s centrally managed funding instruments. In addition, a significant but unreported share of DFID funding to multilateral agencies is sub-contracted to CSOs.

DFID’s current objectives for and approach to working with civil society are set out in its Civil Society Partnership Review (CSPR), commissioned in May 2015 and published in November 2016. The CSPR underlines the important role CSOs play in achieving the Sustainable Development Goals, both through delivering services to the poorest and most vulnerable, and by empowering people and holding decision-makers to account. Among other commitments on DFID’s engagement with CSOs, the CSPR commits DFID to working with other UK government departments to address the problem of declining civic space, and the growing threat to the freedom of CSOs to organise, participate, communicate, and influence the political and social structures around them.

The CSPR introduced significant changes to how DFID funds and partners with CSOs, with fewer central funding streams, stronger transparency and accountability requirements and a more competitive, outcome-focused funding model. ICAI decided to conduct a performance review of DFID’s partnership approach to CSOs because of the scale of funding for CSOs and the changes to how DFID works with civil society. Our scope includes DFID funding for local, national and international CSOs that is managed centrally and from DFID country offices, for both development and humanitarian purposes. It also includes DFID’s work to help build the capacity of CSOs. In addition, the review assesses the efforts of DFID and other UK government departments to counteract the decline in global civic space.

We look in particular at three large centrally managed funding instruments that the CSPR focuses on – UK Aid Direct, UK Aid Match and UK Aid Connect – and at DFID’s work with civil society in two case study countries where it has invested in civil society work – Bangladesh and Ethiopia. As a performance review, the focus is on the effectiveness of DFID’s approach and wider UK government efforts, taking into account the evolving national and international contexts for these partnerships.

Relevance: How well does DFID’s approach to partnership with CSOs reflect its CSO objectives and commitments?

DFID’s views and commitments regarding the role and importance of CSOs are coherent, consistent and persuasive, and broadly shared among donors and practitioners in the development and humanitarian assistance community. DFID considers CSOs to have the potential to deliver good value for money through service delivery, particularly in fragile contexts and humanitarian emergencies. It views CSOs as uniquely well placed, through their grassroots networks, local knowledge and legitimacy, to reach marginalised communities, and therefore as central partners in furthering the UK aid commitment to ‘leave no one behind’. DFID’s support to CSOs is also a goal in itself, underpinned by the conviction that a strong civil society fosters diverse, open societies, reduces corruption and holds decision-makers to account. Finally, DFID sees British CSOs in particular as “crucial partners and allies” in its “efforts to modernise, update and improve the effectiveness of the global development system”. However, we found that DFID’s new approach to central funding for CSOs is not well suited to furthering these views and commitments.

DFID has consolidated its centrally disbursed funding for CSOs into a few large, competitive funding instruments with stringent conditions and requirements on CSO grantees that are individually sensible but collectively onerous. The high investment cost required to develop proposals for them, combined with a low success rate for winning grants, pose challenges for CSOs.

In 2016, DFID ended almost all its ‘unrestricted’ or ‘core’ funding that CSOs could use at their discretion to pursue DFID-agreed priorities. From this year onwards, central and in-country funding was almost exclusively project-based and CSOs were often treated as implementers of mostly shorter-term projects, with funding tightly conditioned by results frameworks. This was meant to increase competition, transparency and accountability, and to an extent this worked. However, it also reduced CSOs’ opportunities for self-driven capacity development, longer-term strategic thinking and adaptation to the evolving contexts in which they operated. Though DFID’s central and in-country funding streams included CSO capacity-building elements, these were often overly focused on the requirements and standards of DFID as a donor, and too little on CSOs’ own needs – how to strengthen their organisation, fulfil their mandate and achieve their objectives. In all, DFID’s practice during the review period was not fully aligned with the department’s commitments to foster a healthy, vibrant civil society as a goal in itself.

Following Oxfam’s sexual abuse scandal and Save the Children’s sexual harassment scandal in early 2018, DFID introduced a range of compliance requirements to safeguard against the abuse, exploitation and victimisation of vulnerable people by aid workers, and of aid workers themselves. DFID’s initial reaction was appropriately urgent, but also led to delays in other important work. Meanwhile, overlapping DFID requests led to extra work on the side of CSOs, some of which was duplicative. However, DFID’s compliance-based safeguarding approach evolved during our review period. It included some CSO capacity development support, a pre-grant ‘enhanced due diligence’ process which included safeguarding issues, and a requirement for CSOs to verifiably adhere to standards that ensure protection of vulnerable people, and to minimise the risk of exploitation and abuse.

DFID and the wider UK government have committed to tackling the global decline of civic space through international influencing activities and supporting CSOs in country contexts. However, there is currently no clear approach to achieve this. While DFID has identified the need to provide a guiding framework on how to analyse and respond to closing civic space in its focus countries, it has not yet begun to develop this.

We have given an amber-red score for the relevance of DFID’s partnership approach to CSOs as well as to the influencing efforts of the wider UK government aimed at strengthening civic space.

Effectiveness: How well does DFID’s funding for CSOs and related influencing work contribute to better development results and a more effective civil society sector?

In our sample of projects, we found that DFID funding to support CSOs’ delivery of development and humanitarian interventions generally had positive results. DFID-funded CSO projects have had direct life-changing or even life-saving impacts for beneficiaries, but we also saw projects that did not deliver the anticipated results or lacked the scale or timelines to tackle the complex, interrelated problems they were seeking to address. Across DFID’s CSO portfolio, results were often affected by delayed funding awards and short project cycles.

Once funding was secured and projects launched, CSOs reported that they found DFID and most of its funding intermediaries to be a supportive, if somewhat inflexible, donor. In the stages before project implementation, however, DFID’s process management was weak. Delays – often long ones – between funding calls, funding decisions and funding disbursement were the norm rather than the exception. Timelines and delays were often poorly communicated to CSOs. This led to uncertainty, late starts and gaps in project delivery, as well as a level of disillusionment among CSOs about the nature of their partnerships with DFID.

We found in both Bangladesh and Ethiopia that DFID country offices have provided useful practical support to CSOs on civic space in specific situations and settings. But generally, DFID’s influencing and programming work in this area has been limited and unguided by a clear DFID-wide approach. One possible exception in the future may be in relation to ‘de-risking’ (banks withdrawing their services from fragile areas to avoid inadvertently violating international counter-terrorism rules, which makes it difficult for CSOs to fund their humanitarian and development work in fragile countries and regions). In October 2018, DFID announced that de-risking was a “critical financial development and humanitarian issue” globally, and noted that the UK was well placed to influence international practice in this area, considering its role as a global financial capital and a leading international development actor. We will continue to follow the developments in this area.

DFID’s current approach to funding does not empower CSOs to achieve the best possible project results. Weak process management has caused frequent and lengthy delays and disruptions to plans, which have impacted the effective delivery of DFID-funded CSO projects. We have therefore given an amber-red score for the effectiveness of DFID’s CSO partnership approach to achieve better development results and a stronger civil society sector.

Learning and innovation: How well does DFID promote learning and innovation in its partnerships with CSOs?

At the start of our review period, DFID was planning to commission research and evaluations aimed at filling some of the knowledge gaps on optimising its new CSO funding mechanisms. Knowledge gaps in this area included: how the value for money of centrally funded work compares to work funded from country offices, what types of funding and which types and sizes of CSOs produce the best results in which contexts and for what purposes, and what the advantages and disadvantages of ‘open’ versus thematic funding rounds are. In the end, delays meant that no research or evaluations on these issues were conducted during the period covered by this review (May 2015 until December 2018). We did, however, see DFID-commissioned research in a range of other fields related to its CSO work, and examples of DFID incorporating lessons from its previous experience and from its engagement with specialist CSOs and other stakeholders on issues such as disability inclusion in aid delivery.

DFID also facilitated or funded a range of systems, processes and events that were appropriately designed to facilitate learning across CSOs. We did not find much evidence of the results of this learning being directly applied to programming, but saw signs that they may have contributed to gradual shifts in programming practice in new areas such as disability inclusion and accountability of development and humanitarian CSOs towards people affected by crisis and poverty.

We saw a few innovative funding mechanisms and programme approaches that have the potential for transformative impact, such as a CSO-led quick-disbursement fund for low-visibility emergencies and a six-year project to identify and scale up successful CSO innovations in fields such as refugee education. However, DFID and its funding intermediaries are not well set up to identify such successful innovations, or to maximise their visibility and buy-in. This may be different for the newly launched UK Aid Connect, which aims to develop innovative solutions to complex development problems, and has been designed to invest heavily in real-time learning.

Overall, the lack of efforts to fill knowledge gaps on the impact of different funding types before setting up new central funding instruments, combined with insufficient sharing and uptake of learning and innovations, has led to an amber-red score for DFID’s activities in this area.

Recommendations

We offer the following recommendations to strengthen the long-term effectiveness of DFID’s CSO partnerships.

Recommendation 1

DFID should fill gaps in the knowledge needed to optimise the design of its central funding instruments.

Recommendation 2

Throughout DFID’s central and in-country portfolios, the process towards funding agreements should be more efficient, predictable, reliable and transparent, and should allow CSOs sufficient time to develop proposals.

Recommendation 3

Throughout its central and in-country portfolios, DFID should have a stronger focus on the long-term results of its CSO-implemented programmes, the localisation of development and humanitarian efforts, and its CSO partners’ long-term capacity to deliver relevant results in evolving contexts.

Recommendation 4

DFID should do more to encourage CSO-led innovation, and to recognise and promote the uptake of innovation successes.

Recommendation 5

DFID should provide a guiding framework for country offices on how to analyse and respond to closing civic space within a national context, and work with other UK government departments to agree a joint approach to addressing the decline of civic space at the international level.

Introduction

The term ‘civil society’ refers to the way in which citizens are linked by common interests and collective activity. Part of civil society takes the shape of civil society organisations (CSOs), which “include all non-market and non-state organizations in which people organize themselves to pursue shared interests in the public domain”.

CSOs are pursuing development and humanitarian objectives at multiple levels, from small village-based organisations to international non-governmental organisations with the capacity to deliver humanitarian and development aid across multiple countries. Their work supports the full range of Sustainable Development Goals. In 2008, the OECD concluded that CSOs are “particularly effective at reaching the poor and socially excluded, mobilizing community efforts, speaking up for human rights and gender equality, and helping to empower particular constituencies. Their strength lies not in their representation of society as a whole, but in their very diversity and capacity for innovation, and in the different perspectives that they bring to the issues when engaging in policy dialogue.”

The UK government has been funding CSOs since long before the establishment of the Department for International Development (DFID) in 1997. Some themes, such as health, education and humanitarian action, have been constantly present in DFID’s partnerships with CSOs, while others have come and gone. The volume and modalities of CSO funding have also changed over time. In 2008-09, DFID’s bilateral civil society portfolio amounted to £515 million, or 15% of DFID’s bilateral spend. DFID’s spending through CSOs increased year on year until 2014-15, when it reached £1,375 million or 25% of its bilateral spend. It then declined to £1,268 million, or 20% of bilateral spend, in 2016-17. In addition, a significant but unreported share of DFID funding to multilateral agencies is sub-contracted to CSOs.

DFID’s current approach to working with civil society is set out in its November 2016 Civil Society Partnership Review (CSPR), which it commissioned in May 2015 as part of a suite of strategic reviews. Our aim in this report is to assess how well DFID’s partnerships with and funding for CSOs, and the UK government’s broader influencing work on promoting civil society, are achieving the objectives set out in the CSPR (see Box 3). The CSPR ushered in significant changes to how DFID funds CSOs at a time when CSOs were also adapting to fast-changing operational contexts. This performance review focuses on the effectiveness of DFID’s approach, taking into account the evolving contexts in which these partnerships are situated.

The review covers the period from May 2015, when the CSPR was commissioned, to December 2018. We also include some elements of DFID’s work that began before then, to assess results that require longer timelines. Our scope includes DFID’s CSO funding and capacity-building efforts managed by central DFID teams, DFID in-country offices and through multilateral agencies. It includes funding for both development and humanitarian efforts, and for local, national and international CSOs. We look in particular at three large centrally managed funding instruments (UK Aid Direct, UK Aid Match and UK Aid Connect) and at DFID’s work with national civil society in two case study countries, Bangladesh and Ethiopia.

In addition, the review assesses the efforts of DFID and other UK government departments to maintain and expand civic space, both globally and in DFID focus countries. Civic space is the set of rules and practices that jointly shape the extent to which people are “able to organise, participate and communicate without hindrance” and, in doing so, “are able to claim their rights and influence the political and social structures around them”.

We stress that the subject of this review is DFID and the quality of its CSO partnerships, rather than the CSO partners themselves. While we examined the effectiveness of a sample of individual projects, we did this to explore whether DFID’s funding practices were enhancing their effectiveness.

The review did not assess DFID’s funding for research institutes, people’s movements or business and trade associations. It did not include DFID’s volunteering grants, the Girls’ Education Challenge, or its engagement with CSOs on policy issues. Our review questions and sub-questions are set out in Table 1.

Table 1: Our review questions

Methodology

Our methodology was designed to assess the relevance, effectiveness and learning of DFID’s relationship with and funding of civil society organisations (CSOs), a well as DFID’s and the wider UK government’s work to ensure supportive operating environment for CSOs. We made our assessment by reference to the objectives and commitments outlined in the Civil Society Partnership Review, taking into account the dynamic political context in the UK, in DFID focus countries and globally.

The core of our methodology was a review of three centrally managed funding instruments – UK Aid Direct, UK Aid Match and UK Aid Connect – and of DFID’s work with national civil society in Bangladesh and Ethiopia. We conducted 28 case studies of funding agreements with individual CSOs and 12 case studies of CSO funding proposals that were rejected (a list of the case studies is included in Annex 1). These case studies covered both humanitarian and development interventions, and were selected in two rounds. In the first round, we sampled from small, medium and large CSOs. Within each of the three centrally managed funding instruments, we purposefully sampled from CSOs that were from the UK and from DFID priority countries in equal numbers. Across these size and country-of-origin categories, we sampled grants and contracts for development and humanitarian projects that were awarded both to long-standing CSO partners and to more recent partners. The second selection round aimed to fill gaps in evidence by selecting an additional four grants that were not part of these three centrally managed instruments or two country portfolios.

The resulting sample reflected the diversity of DFID’s CSO funding. The grants and contracts ranged from £164,000 to £223 million. They were used to work towards different Sustainable Development Goals, with implementation periods of between 45 days and five years. The funds were awarded both directly and via intermediaries, on the basis of unsolicited and solicited proposals, framework agreements, competitive funding rounds and tenders that were open both to CSOs and to other bidders.

In the course of our assessment, we conducted key stakeholder interviews and group discussions with DFID staff, representatives of funding intermediaries and recipients, other donors, host governments, external specialists and people affected by poverty and disaster (with a focus, within the last group, on representatives of groups that are often left behind). We visited project sites in Cox’s Bazar District and Khulna District in Bangladesh, and in Asosa Zone and Wolayita Zone in Ethiopia. We triangulated our findings by comparing them with findings as reported in recent literature on civil society in the humanitarian and development sectors, and with evidence and findings from other relevant ICAI reviews.

The methodology is explained in full in our Approach Paper. Both our methodology and this report were independently peer-reviewed.

Figure 1: Components of our methodology

Background

Civil society organisations (CSOs) are not-for-profit entities, separate from the state and the private sector. Beyond these definitional commonalities, CSOs are diverse in their size, longevity, regional reach, sectoral focus, working methods and organisational set-up. The ones that are within our scope pursue developmental or humanitarian objectives.

DFID’s funding to and through CSOs

Figure 2: Centrally managed instruments

In 2016-17, DFID channelled 20% of its bilateral spend, or £1,268 million, to or through CSOs. Of DFID’s bilateral CSO spend, 62%, or £783 million, was managed by DFID’s country offices. In-country contracts and grants ranged from a five-year £223 million direct grant to BRAC (a large Bangladeshi CSO), to grants of less than £1,000 for local CSOs, channelled through civil society funding instruments that were managed by intermediaries such as companies, UN agencies and international or national CSOs.

The remaining 38% of DFID’s bilateral CSO spend in 2016-17 was managed centrally – down from almost 50% in the preceding years. The reduction was caused by the termination of the centrally managed Programme Partnership Arrangements (PPAs) that had provided £669 million to 41 organisations in the period from 2011 until 2016. This marked an important shift in DFID’s CSO funding model, from the provision of unrestricted funding (which can be used at the discretion of the recipient, usually towards DFID-agreed priorities) to a limited number of trusted CSO partners, towards a more open, competitive and outcome-focused approach. The policy decision to terminate the PPAs was followed by the commissioning of the Civil Society Partnership Review (CSPR) in 2015. The CSPR lists four funding instruments as central to this new funding model. Figure 2 details the three centrally managed instruments this review is concerned with: UK Aid Direct, UK Aid Connect and UK Aid Match. Figure 3 provides examples of grant selection rounds within UK Aid Direct.

Figure 3: Examples of grant selection rounds within UK Aid Direct

DFID’s objectives in funding CSOs

At the start of our review period (May 2015), five objectives underpinned DFID’s work with civil society:

  • Provide goods and services to the poor, particularly the hard to reach.
  • Help people in developing countries to hold governments and others to account and influence decisions that affect poor people’s lives.
  • Support poor people to do things for themselves.
  • Build support for development by encouraging UK citizens [to engage].
  • Build and maintain the capacity and space for an active civil society.

Then, with the publication of the CSPR in November 2016, DFID announced eight objectives for its CSO work (see Box 3). These largely overlapped with the previous five, but no longer featured the objective of supporting poor people to do things for themselves. They also included three new elements:

  • Developing and sharing evidence and knowledge of what works to achieve the UK Aid Strategy.
  • Maximising the impact of DFID funding on the lives of poor people by supporting the most cost-effective interventions that will make the biggest difference to the largest number of people.
  • Building diversity within civil society.

DFID is seeking to achieve its objectives in a fast-changing and unpredictable environment

Since the publication of the CSPR, there have been two UK general elections and DFID has had three secretaries of state. Following the 2016 Brexit referendum, and the triggering of Article 50 in March 2017, DFID warned that UK CSOs could face discrimination when attempting to access EU funding. In 2018, Oxfam experienced a sexual abuse scandal and Save the Children a sexual harassment scandal. All these events and shocks delayed funding processes, and some of them affected the nature of DFID’s partnerships with CSOs.

Longer-term trends have also impacted on DFID’s partnerships with CSOs. Most important among these is a general decline in the political and regulatory operating environment for civil society (hereafter “civic space”). In many countries, CSOs are less free to operate now than they were a decade ago. Political rights and civil liberties are under threat in countries on all continents, and have declined in more than half of DFID’s priority countries over the past decade (see Figure 4). In 2019, Freedom House classified only three of DFID’s priority countries as ‘free’ (see Figure 5) and CIVICUS reported that, in 2017, there were serious systemic problems with civic space in 109 countries. CSOs have faced an increase in administrative hurdles, arrests, confiscation of equipment, forced closure and violence. In many countries, new laws have restricted CSOs’ access to international funding. DFID considers protecting civic space to be important as an objective in itself.


We will … robustly defend the rights of civil society in a dangerous and uncertain world. Around the world, civil society is facing unprecedented pressure, from violent attacks to attempts to close down the space for democratic dialogue and debate. The UK Government, as part of its commitment to freedom of thought, association and expression, will stand alongside civil society against these encroachments.

Civil Society Partnership Review, foreword by the secretary of state, DFID, November 2016

The practical response of CSOs and their donors has taken many forms, including withdrawal, relocation, self-censorship or a refocus on politically un-contentious activities such as service delivery, but also protest, advocacy and the provision and use of more secure means of communication to reduce risks to CSO staff.

Figure 4: Changes in associational and organisational rights in DFID’s priority countries and ‘development partnership’ countries between 2008 and 2018

Figure 5: Freedom Rating of DFID priority countries and ‘development partnership’ countries in 2019

A second long-term trend is the shifting thinking about what good humanitarian interventions and cooperation look like. Donors have recognised that their funding practices may inadvertently restrict the capacity of CSOs to serve beneficiary populations. At the May 2016 World Humanitarian Summit, DFID and other humanitarian donors and agencies signed up to the Grand Bargain, a set of commitments designed “to get more means into the hands of people in need and to improve the effectiveness and efficiency of humanitarian action”. The commitments include expanding humanitarian actors’ manoeuvrability by harmonising and simplifying reporting requirements and by providing more un-earmarked and multi-year funding to ensure greater predictability and continuity in humanitarian response. There is also a commitment to providing more funding to local and national responders, and to including people receiving aid in making the decisions that affect their lives – an objective known as ‘localisation’. The goal is to make vulnerable communities more resilient by increasing their capacity to help themselves. For the time being, there is no equivalent formal commitment in the field of development work, but it is widely accepted that localisation is important in this sector as well.

Findings

Unless specified otherwise, our findings apply to DFID’s work with all its civil society organisation (CSO) partners. This includes local and national CSOs in DFID focus countries, UK CSOs and CSOs from other countries.

Relevance: How well does DFID’s approach to partnership with CSOs reflect its CSO objectives and commitments?

DFID’s views on the role and potential value of CSOs are clear and consistent

We found that DFID’s views on the role and potential value of CSOs were clearly and consistently presented throughout policy documents and in our interviews with DFID staff. The most commonly mentioned comparative advantages of CSOs, and the most common reasons given for funding them, are as follows:

  • Service delivery: CSOs have the potential to deliver good value for money through service delivery in fields related to the Sustainable Development Goals. Some respondents mentioned that CSOs often build useful alliances with other providers, including governments, and that they are particularly well positioned to provide services in contexts that are fragile or in conflict. In circumstances where host governments are not yet able or willing to provide essential education, health and other services to all groups within their borders, DFID often funds CSOs to fill some of the gaps.
  • Leaving no one behind: CSOs can help DFID achieve its commitment to leave no one behind in ways other delivery partners cannot. CSOs often focus their efforts on marginalised communities, and their grassroots networks, local knowledge and legitimacy make them well placed to serve hard-to-reach groups. Some respondents emphasised CSOs’ track record of supporting women and girls in particular. DFID also draws on specialist CSOs’ knowledge and experience of inclusive programming to improve its own practices, for instance in the field of disability inclusion. Almost all of DFID’s CSO funding streams promote either inclusive programme approaches or direct funding towards specific marginalised groups, such as adolescent girl refugees or disadvantaged ethnic minority communities.
  • Voice and accountability: CSOs hold governments to account and provide a platform for a diversity of voices and opinions. As a result, DFID believes that strengthening civil society can help to reduce corruption and make governments more accountable for their use of public funds. DFID therefore sees a healthy civil society and a conducive operating environment for CSOs as inherently important goals. DFID pursues these goals through both funding and policy influencing work.
  • Innovation: CSOs have a track record of piloting, refining, evidencing and mainstreaming development innovations and of challenging existing ways of working. Extrapolating from past innovation successes, the Civil Society Partnership Review (CSPR) says that British CSOs in particular are “crucial partners and allies” in DFID’s “efforts to modernise, update and improve the effectiveness of the global development system”. A few of DFID’s funding streams are exclusively innovation-focused, and many other streams have a dedicated innovation window or encourage innovation in other ways.

These views are expressed in DFID documents going back to at least 2010. They are widely held across the department and reflect a broad consensus on the value of CSOs among donors and practitioners in the global development and humanitarian community. We find them to be coherent, consistent and persuasive. Some respondents and documents also mentioned other valuable contributions of CSOs to the development process, such as changing social attitudes, strengthening peace efforts and strengthening livelihoods and economic empowerment.

DFID also often mentioned two reasons for supporting UK CSOs in particular:

  • the UK’s strong development CSOs are part of Britain’s soft power and global leadership on development issues
  • UK CSOs engage the UK public and have the potential to build public support for the UK aid programme (see Box 3, objective 8, and Box 4).

The CSPR mentions both reasons, and they add to the CSOs’ collective status as a “significant and enduring partner” for DFID.

DFID encourages CSOs to embrace the ‘leave no one behind’ commitment

In the CSPR’s seventh objective, DFID aims to support “interventions that will make the biggest difference to the largest number of people”. In the third objective, it wants to “meet the UK’s commitment to leave no one behind”. There is a trade-off between these objectives because, as DFID acknowledges, “reaching the most excluded people… can be expensive”.

Before our review period, CSOs that were implementing centrally managed DFID projects did not systematically prioritise the hard to reach over other groups. They pointed to evidence that centrally managed projects “that specifically include hardest to reach groups like people with disabilities can fair less well in [DFID’s value for money] assessments”. During the review period, DFID and the fund manager for UK Aid Direct and UK Aid Match have been clearer about their expectation that CSO partners contribute to the ‘leave no one behind’ commitment, and have used a range of communication channels to convey this message. Both instruments have also actively promoted the new ‘equity’ part of its value for money guidance. The inclusion of equity in the value for money metrics means that reaching hard-to-reach groups has now become a key feature of ‘value for money’. As a result, we saw a trend where centrally funded projects worked more clearly and deliberately towards the ‘leave no one behind’ commitment at the end of the review period than at the beginning.

DFID has diversified its partnerships with UK CSOs, but the stated rationale for this is not rooted in evidence

The CSPR commits DFID to diversifying its CSO partnerships. DFID’s efforts to do so are most obvious in the centrally managed instruments, which have worked to include a larger number of smaller UK CSOs from more locations across the UK. To achieve this, DFID has conducted roadshows across the UK to attract new applicants. It has also adjusted some of its eligibility criteria to be more favourable to smaller CSOs, created windows under UK Aid Direct that specifically target small CSOs – the Small Charities Challenge Fund and, later, the Jo Cox Memorial Fund – and adopted application processes and payment arrangements that are suited to smaller organisations. This effort is likely to have arrested, and possibly even begun to reverse, a decade-long trend in which institutional funding was increasingly focused on large UK CSOs (see Figure 6).

Figure 6: In the decade before the CSPR, the growth in institutional donor funding for UK CSOs had been invested in large CSOs only

One reason for DFID’s efforts to open funding opportunities for an increasing number of small organisations from across the UK is to build a wider and more diverse constituency of the UK public that is engaged in international development. The assumption is that if more of the public are exposed to aid-funded activities by organisations in their locality, they will be more favourably disposed towards the aid programme. DFID has not yet substantiated this causal link.

A second stated reason for the policy choice to support smaller CSOs is that “small and medium-sized organisations are filling an important niche, running important and innovative projects. It is often smaller, grassroots organisations that are best able to make the connections on the ground and achieve lasting impact.” DFID does not have evidence that small UK CSOs are either more innovative or more likely to achieve lasting impact than large CSOs. We did see some evidence that the opposite might be true: an assessment of DFID’s Civil Society Challenge Fund (a fund that focused on small UK CSOs and that was terminated in 2015) found that the evidence did “not support the hypothesis that public funding of small NGOs through the fund can be justified by positive empirical evidence of its impact”. A 2014 external evaluation of one of DFID’s centrally managed funding streams came to a similar finding, and also found that smaller grantees “were not especially successful at innovation” as they had a lower risk appetite than larger CSOs, smaller networks, and were not as good at capturing lessons from innovation. Moreover, DFID predicts that increasing the diversity of CSOs comes with “higher costs for managing larger numbers of applications and grants. Small CSOs may be less experienced and require more investment to manage.”

A third stated reason for DFID’s diversification efforts is that “CSOs told us that DFID should better recognise the sector’s diversity and the unique contribution a wide range of CSOs make to DFID’s work.” In this context, we note that the competitive nature of the funding rounds, in combination with DFID’s assertion that “funding processes will reward bids that use the interventions proven to be the most cost effective”, have meant that proposals were often designed to meet all of DFID’s approach preferences. These many preferences, as well as DFID’s mainstreaming and other conditions and requirements, are all individually sensible, but collectively they stifled creativity and reduced diversity and scope for CSOs to localise their approaches. As a consequence, DFID has increased the overall number of centrally managed UK CSO partners, but may have reduced the diversity of the approaches that DFID’s partners use.

We therefore conclude that the stated reasons for the policy choice to increase the diversity of DFID’s partnerships with UK CSOs, and to focus its diversification efforts on the basis of CSO size and spread across the UK, are not rooted in evidence.

DFID’s focus on UK CSOs does not further the localisation of development and humanitarian efforts

DFID has funding streams for which only UK CSOs are eligible (UK CSOs then transfer a proportion of this funding to their national partner CSOs in other countries). These are UK Aid Match, the Small Charities Challenge Fund and the Community Partnership Grants of UK Aid Direct. There are no formal restrictions in other centrally managed funding streams but, in practice, most funding is awarded to UK CSOs. For example, almost all UK Aid Connect consortia are led by UK CSOs. Within UK Aid Direct and its predecessor (the Global Poverty Action Fund), the proportion of funding that went directly to CSOs in developing countries declined from 28% (December 2011 to May 2015) to 18% (June 2015 to October 2018). This focus on UK CSOs is explicit in the foreword to the CSPR, in which the then secretary of state, Priti Patel, said that “Britain’s civil society organisations… do extraordinary good around the world” and that they are “a crucial part of Britain’s soft power and leadership around the world”.

The CSPR’s emphasis on the role of UK CSOs does not align well with the localisation agenda (see paragraph 3.9). It is also at odds with the CSPR’s observation that “CSOs emphasised the need for organisations based in developing countries to have a more prominent leadership role”, and with DFID’s recognition of the importance of national and local organisations in other DFID documents.

DFID did give us a few examples of instruments that had progressed towards its localisation goal, such as the DFID-supported DRC Humanitarian Fund, which has surpassed the Grand Bargain commitments related to the direct involvement of local and national responders to humanitarian needs in the Democratic Republic of Congo. DFID has also funded and invested in networks such as The Girl Generation, which strengthens efforts by hundreds of African organisations to reduce female genital mutilation, and has invested in efforts to improve the organisational capacity of a range of CSOs in DFID priority countries. However, within DFID’s centrally managed instruments, there has on the whole been no clear trend towards the localisation of development and humanitarian efforts. ICAI’s recent reviews of the UK’s humanitarian support in Syria and the UK’s core funding to the UN humanitarian system reported similar findings.

DFID’s approach to maximising value for money strengthens the focus on tried and tested programme interventions but does not facilitate adaptive and innovative programming

Through all the central funding instruments in our sample, DFID’s focus on value for money led to the expectation that its partners would use evidence-based approaches, provide upfront assurances about the effectiveness of their proposed programmes, and work within predefined result frameworks.

Where CSOs propose to deliver tried and tested interventions, DFID’s approach to value for money is an appropriate way of ensuring the best return on the investment. However, where the objective is to support innovative work, tight predetermined programme frameworks can be unhelpful, reducing the freedom and the incentive to experiment and learn from failure. This is a problem in particular for UK Aid Connect, which sets out to be strongly innovation-focused. Its business case states that “flexible and adaptive programming approaches will be built in from the start, building in beneficiary feedback approaches and points in the programme cycle which allow the consortium to innovate and to respond to risks and opportunities … throughout the duration of the programme”. The same document, however, also notes that proposals “will be assessed against the [value for money] of the management arrangements (economy), their impact on poverty (effectiveness and efficiency) and relative balance (are we achieving a high level of poverty reduction for every pound spent?)”.

In our interviews with recipients of centrally managed and Ethiopia-specific funding, we often heard that CSO’s desire to innovate was hampered by their fear of failing to meet predefined milestones. Grantees are required to work within detailed budgets that are not easy to change and that provide limited scope for the flexible, adaptive programming needed to support innovation. For the in-country programmes in Bangladesh, the picture was more positive. Specifically, CSOs that worked with DFID directly (rather than through intermediaries such as the United Nations Office for Project Services and the British Council) sometimes praised DFID’s willingness to repurpose funding when needed, or to provide additional funding when new programming opportunities were seen as particularly worthwhile. In Bangladesh, we also saw evidence of DFID and grantees working together to innovate, experiment and remain relevant in evolving environments.

DFID has changed its funding practices, and introduced more stringent requirements, to promote the transparency and accountability of UK aid

DFID wants the UK public to view aid spending through CSOs as legitimate, and wants CSOs to be more accountable to beneficiary communities. The CSPR therefore listed transparency and accountability, together with cost-effectiveness, as “core principles” that are “central to [DFID’s] entire approach to CSO funding”.

As part of DFID’s operationalisation of these core principles, the CSPR brought in three changes in DFID’s funding architecture:

  • DFID terminated its centrally managed unrestricted funding instrument. All centrally managed funding is now project-based.
  • DFID consolidated its “many separate funding streams for CSOs” that had “sometimes led to confusion and duplication”. DFID does not keep count of its central funding streams, so we do not know the extent of the consolidation, but we saw 17 examples of terminated instruments. The overall number of its centrally managed funding instruments is certainly lower at the end than it had been at the start of the review period.
  • For the consolidated funding streams, the CSPR stated that “DFID will revise its approach to reimbursing overhead costs.” This was the start of a process towards what DFID calls ‘cost transparency’ and ‘full cost recovery.’ It allows CSOs to include an organisation-specific percentage in project budgets that accurately covers the indirect costs of project delivery. This is in line with the 2006 UK Treasury guidance that it is not good value for money for CSOs to be “unwittingly subsidising a public service”. DFID operationalised its new cost transparency model in the last quarter of 2018.

The CSPR also expects improved practice from the CSOs that DFID funds, stating the need for “all organisations… to be open and transparent about exactly where the money goes and what it achieves”. The condition applies not just to direct funding recipients but throughout the delivery chain, and DFID “will not compromise on assurance standards” irrespective of an organisation’s type and size. DFID combines training and capacity building with formal requirements on adherence to transparency and accountability standards. DFID-funded capacity-building efforts in developing countries illustrate this focus: they are often related to transparency and accountability (towards DFID, primarily) rather than, for example, to an organisation’s ability to fulfil its mandate and achieve its objectives.

We found some evidence from our sample that CSOs had strengthened their practice in terms of accountability both towards people affected by poverty and disaster and towards DFID and the UK public. Progress of the latter type is clearer and more clearly attributable to DFID pressure and capacity building efforts. We have not seen attempts to measure the resources and costs involved for CSOs to comply with DFID’s accountability standards, nor strong attempts from DFID to keep these costs to a minimum. While transparency and accountability standards are important, the combined weight of DFID’s requirements may inadvertently contribute to a declining civic space. DFID raised this risk itself in 2015, noting that “DFID should assess its own practices to identify any cases [that] could contribute to closing [civic] space in the name of alignment, harmonisation and accountability”.

DFID’s urgent response to safeguarding scandals gained focus and balance over time

The issues of sexual exploitation, abuse and harassment did not feature prominently in DFID guidance until the 2018 scandals at Oxfam and Save the Children prompted the department to take on a leadership role on the issue. The urgent nature of DFID’s initial action was appropriate but came at the cost of delaying other DFID work, such as work in the field of civic space. Initial duplications in demands to DFID’s CSO partners were unnecessarily time-consuming for CSOs, and disrupted some of their other work. At the time of writing this report, DFID had not yet initiated a learning exercise to ensure a new crisis would not lead to similar duplication and disruption.

DFID’s approach gained focus and balance after these initial months. Towards the end of our review period, it included some CSO capacity development support, a pre-grant ‘enhanced due diligence’ process covering safeguarding as well as other issues, and a requirement for CSOs to verifiably adhere to standards that ensure protection of vulnerable populations and minimise the risk of exploitation and abuse. It is too early to assess if the sector may be on a trajectory towards verifiable compliance to safeguarding standards, or what role DFID will play in achieving this. We will therefore revisit progress in this area. DFID’s compliance-based approach can be a useful step, provided it is implemented in parallel with wider efforts to make the sector safer. This includes work to reduce the extreme imbalances of power found between aid providers and aid recipients in emergency situations, to foster a safety-first culture across the humanitarian sector, and to resolve sector-wide structural problems such as the ease with which people who have been fired for misconduct are re-employed by other organisations. DFID is aware of the complexity of the issues of sexual exploitation and abuse, and of sexual harassment.


Let there be no doubt: this international summit was not about gathering promises to tinker around the edges. This is about setting in place a fundamental rewrite, from root to branch, of the way the aid sector operates.

DFID secretary of state, Penny Mordaunt, Quote 2: Statement at the Safeguarding Summit 2018: Global aid community to take action against sexual predators

DFID’s project-based funding model adversely affects CSOs’ transaction costs, organisational adaptability and long-term strategic thinking

With the termination of DFID’s central unrestricted funding instrument, almost all funding received by DFID’s CSO partners is for the implementation of projects, with funding tightly conditioned by results frameworks. This potentially improves the transparency of DFID-funded spending, as project funding is based on detailed budgets and reporting, while unrestricted funding can be used flexibly and at the discretion of the CSO.

However, while the increased transparency of spending is an important aim, pursuing this aim through a funding model that is solely project-based can negatively affect the CSPR’s strategic aim to “maximise value for money and results from CSO programmes and engagement”. In the short term, project-based funding means higher planning and reporting costs for CSOs, lower flexibility to adapt to learning and changed circumstances, and limited financial headroom to respond to new challenges (such as safeguarding against sexual crimes and misconduct). In the long term, CSOs that get all their income fixed in the form of project funding are unable to invest in staff development, build organisational capacity, develop stronger partnerships or respond strategically to a rapidly changing global context.

Many of DFID’s current CSO funding streams include capacity-building elements, but the support is supply-driven and somewhat imposed. With the termination of unrestricted funding, DFID no longer funds experimental capacity-building approaches where results cannot be predicted and success is uncertain (see Box 5 for an example).

While the decision to move away from unrestricted funding was arguably a policy choice, DFID documents continue to note the inherent advantages of providing core funding. DFID is also aware that “the literature is clear that core funding and unrestricted funding are the best mechanisms to support innovation. Funding tied to results or specific projects discourages risk-taking. The results agenda may have inadvertently had a negative effect on experimentation.” Notwithstanding this awareness, DFID has chosen a funding approach that treats CSOs as project implementers, with funding tightly conditioned by results frameworks. This limits CSOs’ ability to manage their own capacity development and to adapt to the evolving contexts in which they operate.

The current centrally managed funding instruments have increased the volatility of some CSOs’ funding portfolios and raised the sector’s collective fundraising costs

The reduction in the number of DFID’s funding instruments has increased clarity around the available funding streams. However, it has also made it harder for some CSOs to build diverse portfolios of grants, making them more vulnerable to variations in their income. For CSOs that are largely dependent on UK funding – such as the UK members of international CSO confederations – this intensifies the phenomenon of project-based funding leading to large and sudden increases and decreases in turnover.

CSOs told us that the consolidation of instruments had compelled them to pursue any available funding, even in the case of a sub-optimal alignment between their strategy and the specific theme of a funding window. The resulting increase in competition, with more applicants responding to each funding call, made it less likely for any individual application to be successful. In line with the UK government’s guidance on competition for funding, DFID sees its highly competitive centrally managed proposal selection processes as driving value for money. The increased competition may indeed have helped to drive up standards in funding applications, but it has also driven up the collective proposal development costs for CSOs. DFID has not evaluated the impact that these costs are having on the overall pool of funding available to spend on humanitarian and development programming.

In 2018, a DFID-commissioned piece of research recommended that DFID should resist the temptation to “make more stringent data and documentary demands of potential … recipients of funding” due to, among other reasons, “the costs of doing so”. In interviews with DFID and the fund manager of UK Aid Direct and UK Aid Match, we saw an awareness of the investment levels currently required to submit proposals, and some evidence of efforts to limit application costs. The application process of the Small Charities Challenge Fund, for example, started with an automated eligibility check, so that organisations did not spend unnecessary time and effort on investing in ineligible proposals. Similarly, some funding rounds had a two-stage application process in which weak proposals were eliminated on the basis of a concept note. Further discussions on ways to reduce the level of investment needed to submit proposals are ongoing.

DFID and the wider UK government do not have clear and shared objectives for their work on the maintenance of civic space

Faced with a global decline in civic space, a 2010 review of DFID’s support for civil society noted that “there is a pressing need to focus on the operating environment”. More recent DFID documents also raised concerns about the declining civic space in many of its focus countries, and confirmed its intentions to act (see Box 6).

Given that the UK alone could not reverse the global trend of declining civic space, the 2010 review suggested that DFID “find opportunities to work closely with other donors, foundations and private donors – promoting a shared interest in the sector”. In the review period, DFID and other parts of the UK government did participate in various national and global forums and working groups, but often without clear and shared influencing objectives. One reason for this may have been the lack of a natural lead department for the issue. DFID has the largest engagement with CSOs and owns the ‘leave no one behind’ commitment, which is where declining civic space often has the most immediate effects. However, the Foreign and Commonwealth Office (FCO) leads on human rights, including civic rights. Another reason is that there is no agreement on what ‘good’ looks like in a field where anti-terrorism, security, stability and human rights agendas all play a role.

The 2010 review of DFID’s support to civil society also noted that “country programmes do not systematically focus on the regulatory environment”. In June 2015, another internal paper recommended that DFID “should publish guidance for country offices (and partners) covering available evidence on what works and methodology for analysing closing space in specific country contexts, identifying corresponding windows of opportunity and developing appropriate programmatic responses from a toolkit”. DFID has yet to publish such guidance, and the need for it was highlighted again in a survey among country offices in October 2018.

Conclusions on relevance

DFID has a clear and consistent view on the role and potential value of CSOs in the humanitarian and development sectors in general, and for the UK aid programme in particular. It sees CSOs as key partners on service delivery, holding governments to account and reaching, including, representing and giving a voice to marginalised groups. DFID recognises that civil society is often uniquely equipped to play these roles, and therefore considers a healthy civil society to be a goal in and of itself. The CSPR objectives and commitments reflect these views.

However, we are concerned that the short-term nature of DFID’s CSO funding is not conducive to the long-term health of the sector. Project-based funding combined with the stringent new requirements of DFID’s funding mechanisms have had positive effects on the accountability and transparency of its CSO partners. However, this funding model has also tied CSOs closer to DFID as its service providers, providing less scope for supporting an autonomous, strong civil society in its priority countries and in the UK. DFID has not followed up on its own recommendations to avoid unduly stringent data and other requirements that add considerable costs for CSOs and can contribute to restricting their operational space. Instead, the termination of unrestricted funding, the income volatility caused by the consolidation of funding instruments, the individually sensible but collectively overburdening conditions and requirements of these instruments, the unnecessarily high investments required to develop funding proposals, and the enhanced due diligence requirements of 2018 come together to pose challenges for DFID’s CSO partners. Furthermore, some of the assumptions underpinning DFID’s funding practices have not yet been tested, and some of the practices do not reflect the objective to localise humanitarian and development practice.

This means that DFID’s practice in this review period is not fully aligned with its commitments to foster a healthy civil society as a goal in itself, and with the evolving realities in many of its focus countries. Instead, DFID has by and large treated CSOs as implementers of short projects, allowing them limited opportunity to ensure their continued relevance, health and adaptability. This has happened at a time when global civic space continued to decline. DFID and the wider UK government do not have a clear set of objectives in relation to the global decline of civic space, and the department has not yet published a guiding framework that helps country offices analyse and respond to the decline.

In view of the above conclusions, we award DFID an amber-red score for the relevance of its approach to partnerships with CSOs.


Effectiveness: How well does DFID’s funding for CSOs and related influencing work contribute to better development results and a more effective civil society sector?

Most DFID-funded CSO projects are delivering positive results

We assessed the results of 28 DFID-funded grants and contracts, and took brief looks at many others, and found that almost all projects are delivering positive results. One of the best-achieving grants in our sample was also by far the largest one: a £223 million five-year grant to BRAC, a very large CSO headquartered in Bangladesh, with which the UK government has had strong and evolving relations since its foundation almost 50 years ago. This ‘Strategic Partnership Agreement’ is DFID’s only remaining large unrestricted grant, based on a strategic agreement ( jointly with Australia). In 2018, “the evidence from BRAC’s […] reports, DFID field visits and consultation with BRAC and partners confirms progress in achieving results that align with UK priorities, at scale, in the right places and to the right target groups”. This aligns with the findings of an independent evaluation of the previous funding period, and with our own observations when we visited BRAC projects in 2016 (described in paragraph 4.100 and Box 11).

For other funding, the overall picture is also broadly positive, and DFID-funded CSO projects have had direct life-changing or even life-saving impact for beneficiaries. However, we have also seen projects that have not delivered the anticipated results or have not had the scale or timelines to tackle the complex, interrelated problems they were seeking to address. Across DFID’s CSO portfolio, results were often affected by delays and short project cycles (we discuss this later in this section).

Within our sample, the CSOs were engaged in diverse activities. For instance:

  • Practical Action helped poor landless women in Bangladesh increase their incomes (£1 million over two years)
  • PREFA strengthened adolescent sexual and reproductive healthcare services in Uganda (£2 million over three years
  • the Start Fund disbursed quick funding for 45-day interventions in low-visibility disasters around the developing world (£27 million over four years).

Almost all the grants we looked at were closely aligned with DFID objectives, and almost all were at least to some extent gender-sensitive in their approach. Projects managed by DFID country offices (to a larger extent than the centrally managed ones) often successfully held authorities to account, from school management to national administrations, and brought about pro-poor changes in their practice. These locally funded projects were carefully implemented in constrained environments: we saw evidence of cautious timing (for example no new messages just before elections, lest they be interpreted as political manoeuvring), well-judged advocacy where CSOs were persistent without being confrontational, optics management to enable authorities to change their practice without fear of losing face, and changes in the influencing targets when one level of a public authority proved to be more receptive to evidence than another level. CSOs often ensured the acceptability of their efforts, in the eyes of public authorities, by referencing governments’ own goals, plans, policies and legislation. We saw examples of locally funded CSOs successfully using a range of approaches to strengthen the effectiveness of their interventions. These included community scorecards (a citizen-driven accountability tool for the assessment, planning, monitoring and evaluation of service delivery), strategic litigation that used hitherto unimplemented pro-poor laws, local community knowledge sharing and mobilisation around practical problems, and large-scale surveys that brought widespread problems to light. Centrally, we saw DFID supporting the representation of CSOs from developing countries in a range of global and regional development policy forums, such as in the steering committee of the Global Agriculture and Food Security Programme.

The most difficult parts of CSOs’ work often showed less consistently positive results. For instance, DFID has committed to “help CSOs put accountability at the heart of their efforts”, but accountability towards people affected by poverty and disaster does not yet have a strongly rooted tradition in the development and humanitarian sectors, and was often less strong than it should be (see Box 7). Similarly, while a number of projects within our sample showed elements of good practice on disability inclusion, more remains to be done. We saw plans to strengthen inclusiveness practice and efforts by DFID’s fund managers to help raise project staff’s awareness of disability inclusion issues, but we also found that most non-specialist CSOs do not yet have the experience required to perform consistently well in this field.

More generally, DFID’s ‘leave no one behind’ commitment remains a challenging one for DFID-funded CSOs to work towards. This is not a new finding: when assessing UK Aid Direct projects that were implemented before our review period, an external evaluation report found that these projects often had a selective and superficial focus on the ‘leave no one behind’ commitment. CSOs often identified very poor people, and designed work to benefit the groups they had identified, on the basis of assumptions rather than evidence. Moreover, the strategies used to include disadvantaged groups were often insufficient to overcome the various barriers to their participation. During our review period, from May 2015 onwards, we saw groups that were overlooked in DFID-funded programming, such as people at the bottom of the Yemeni caste system and older people in many of the projects. However, we also saw an overall improvement in projects’ attention to particularly hard-to-reach groups over the course of the review period. This improvement was achieved in part because DFID became clearer about its expectation that CSO partners contribute to the ‘leave no one behind’ commitment (see paragraph 4.6). For UK Aid Connect, which is the latest fund to become operational, DFID included indicative outcome statements as a guide for programming interventions, and these are clearly focused on leaving no one behind.

For funding managed by DFID’s country offices, adherence to the ‘leave no one behind’ commitment was more consistent throughout the review period, and CSOs were better able to reach marginalised groups. DFID’s in-country funding often targeted refugee communities, and within these communities it often focused on specific vulnerable groups (we saw work done with refugees with disabilities in Bangladesh, and with adolescent girl refugees in Ethiopia, for example). Other marginalised groups that we saw receiving attention, and whose visibility increased as a consequence, were drug users, ethnic minorities and women prisoners in Ethiopia, and women and children from the poorest rural households in Bangladesh.

Although we recognise that further improvements are possible, we conclude that the projects in our sample were achieving positive and sometimes life-changing or even life-saving results. In fragile contexts, we found that CSOs were often the only organisations with the networks, grassroots connections or humanitarian access needed to produce these results.

The limitations of short project cycles are often offset by CSO commitment to long-term engagement

The extent to which projects in our sample have achieved or will achieve durable change has not been evaluated, since DFID does not include a budget for ex-post evaluations in any of its project grants or contracts. The few such evaluations that we have seen (see Box 5 for an example) were financed by unrestricted funding, which DFID terminated in 2016. This means that we have had to base our assessment of likely sustainability of results on project design and context.

In isolation, most projects in our sample are unlikely to achieve durable change, as the project duration is generally too short to truly change systems and behaviour. An external 2016 evaluation concluded that, in the pre-CSPR period, “delivering systemic change was also not always possible within project lifetimes”, and preparatory documentation for the CSPR suggested that “longer term grants (5 years) strengthen sustainability through longer planning and relationships, and create stronger organisations”. DFID has not followed up on these observations, and project cycles might even have shortened during the course of the review period: we looked at the average project duration of one large CSO’s total development projects portfolio and found that it had reduced from 40 months in March 2016 to 30 months in September 2018.

However, most CSOs we reviewed are implementing their DFID-funded development projects as part of wider efforts that collectively do seem capable of contributing to durable change. This is because of a comparative advantage that sets CSOs apart from other key funding recipients: they often have a long-term presence in their regions of operations, where they build on pre-existing trust, and use financial sources and timeframes that go beyond DFID’s project funding. CSOs often further enhance their chances of durable results by aligning their work with national strategies, in collaboration with local authorities and through existing structures and facilities. Furthermore, DFID country offices sometimes overcome some of the limitations of short project cycles by dividing long-term development endeavours over a series of individual projects that collectively span a decade or more.

Grant making and contract award processes were almost always delayed, or even cancelled altogether

Whether arranged directly or via funding intermediaries, nearly all grants and contracts that were concluded in our review period were delayed compared to the original timelines (see Box 8 for illustrations and Box 9 for the main exception). This was the case for centrally managed funding agreements as well as funding agreements managed by country offices. Delays were the norm for individual project grants as well as for the next phases of longitudinal series of projects. Delays were common for funding of all sizes, for CSOs of all types. Sometimes calls for proposals were cancelled altogether, after CSOs had made considerable investment in developing proposals (see Box 10).

Delays continued after grant agreements had been signed. For example, for UK Aid Match projects, DFID’s own assessment was that “the inception phase and commencement of the project has, almost universally, taken longer than expected”. These delays were often not within the control of DFID, its intermediaries or the CSOs, as they were linked to issues such as political changes, government approvals, the start of the school year and recruitment bottlenecks. However, many of these obstacles could have been anticipated and incorporated in the planning.

Delays were mostly caused by weak management of discontinuity and poor process management

Delays were not limited to funding decisions. Evaluation contracts were generally delayed, as was the transfer of UK Aid Match to a fund manager. Similarly, DFID’s internal and cross-departmental conversations and policy development related to civil society (such as on civic space, the full cost recovery model and the CSPR itself) have typically progressed more slowly than originally envisioned. There appear to be two main reasons for these persistent delays: DFID’s weak management of discontinuity and change, and poor process design and management within DFID and most of its intermediaries.

There has been a lot of discontinuity in our review period. DFID has had three secretaries of state since May 2015. There have been national elections in 2015 and 2017, as well as the EU referendum and the subsequent Brexit negotiations. The safeguarding scandals abruptly added an unforeseen priority to DFID’s CSO agenda. In Ethiopia, a new prime minister took power in April 2018 and expanded civic space (enshrined in law as of February 2019), but Ethiopia also saw the highest number of newly displaced people in any country. A large number of Rohingya arrived in Bangladesh, and a terrorist attack in Dhaka turned Bangladesh into a non-family duty station in 2016, leading to immediate and large changes in staff, as well as severe travel restrictions, for DFID and all other major donors.

Given such a volatile context, we would have expected to see DFID taking flexible and pragmatic decisions to ensure continuity of the CSO operations it supported. This might have involved extending some grants and contracts on their existing terms, where DFID’s rules on bridge funding allow for this, to buy time for a more considered response. Instead, DFID opted for long delays in launching projects or moving between phases of programmes. Within ongoing programmes, elements that required considerable staff investments on the side of DFID were neglected (see paragraph 4.59 for an example).

Not all delays were caused by external events: they were also down to weak process management. Though DFID operates on the basis of Smart Rules, designed to introduce lean, effective processes, the processes we have seen were not always clear and formalised and this led, among other problems, to convoluted engagement with CSOs. We saw nine versions of a single proposal, for example, submitted in the course of six months, because the absence of a clear feedback process led different staff members from a DFID intermediary to issue different – and sometimes conflicting – requests at different times. In cases where processes had been formalised, they were sometimes unnecessarily cumbersome. In the case of one programme, minor funding decisions required eight layers of approval. These two examples are extreme versions of common problems. We saw very little evidence of DFID developing streamlined, formalised and transparent processes with a clear critical path. As a result, the department had very little ability to predict how long its own processes would take to complete.

Poor communication, overly challenging timelines and poorly written bid documents reinforce a sense of inequality in DFID-CSO partnerships

DFID’s weak process management was often exacerbated by poor communications with CSOs around its funding processes. Timelines for funding rounds were not always communicated, and when they were they often turned out to be incorrect. Information about delays was sometimes communicated months after originally announced timelines had passed, in response to repeated questions.

Conversely, DFID did not tolerate delays on the side of CSOs, though it did sometimes extend the original timelines after feedback that they were unrealistic. Even after extensions, there was usually only around five to eight weeks between a call for proposals and the submission deadline. Our evidence shows that this does not provide enough time for CSOs to prepare proposals in-house (which is more affordable than engaging consultants), to consult people affected by poverty and disaster or, for the largest proposals, to give the CSO leadership, including its trustees, enough time to fulfil their oversight role (creating tension with DFID’s emphasis on accountability). We saw examples of time allocated to prepare proposals being shorter than originally scheduled to compensate for delays on DFID’s side, and of the proposal preparation period overlapping with a holiday season, placing further pressure on the bidding CSOs.

Unclear, inconsistent and lengthy bidding documents also added to the pressure. A recent proposal round for a series of commercial framework contracts (the International Multi-Disciplinary Programme, or IMDP) had six weeks between the announcement and the submission timeline (extended from the original four and a half weeks) and came with over 300 pages of documentation, without navigation aids and with version control issues. Other requests to CSOs were often made with similarly challenging timelines.

The combination of tight timelines for CSOs, and common delays and opaque communication on the side of DFID and its intermediaries, has led to a level of disillusionment in CSOs about the nature of the partnership.

Delays, discontinuity, unreliability and time pressure during the funding process have affected partner CSOs’ overall results

The delays had practical consequences for the programmes that were funded, hampering their efficiency and effectiveness. They commonly led to reduced first year outcomes and shortened project cycles (sometimes but not always corrected in the form of project extensions), and they sometimes caused disruption to time-bound interventions (such as programming linked to the agricultural cycle or that needed to be completed before the wet season). Delays in renewing grants for long-term, multi-phase programmes caused a loss of continuity and diminished impact. There was only one instance in our sample of a new phase starting on time, at the completion of the previous phase. DFID caused discontinuity even in that case, as the approval for the latter phase was granted only weeks before the end of the former one, and by that time the organisations had planned their exits and staff had moved on.

Delays and long periods of uncertainty affected human resource and other planning, drained initial momentum and excitement, took considerable senior leadership time away from other strategic issues, and caused uncertainty and nervousness among programme staff, leadership and CSO boards. The uncertainty and delays were particularly challenging for organisations that employ people with skills that are in high demand, such as healthcare professionals and experienced humanitarians. We saw several instances where such staff moved on to organisations able to offer more certainty. Among organisations that depend on DFID funding, the delays further intensified the income volatility that characterises the development and humanitarian sectors, and that was already more pronounced because of the consolidation of funding streams (see paragraph 4.59 above).

The effects of weak process management and weak management of discontinuity during the funding process have delayed and impacted the effectiveness of DFID-funded projects. For the CSOs that are most dependent on UK funding (typically UK members of international confederations that are not permitted to fundraise outside of the UK), these delays have been a distraction throughout the organisations’ hierarchy.

Once CSOs start implementing DFID-funded projects, DFID is a supportive donor

When CSOs reflected on their relations with DFID during project implementation, they were largely positive. CSOs see DFID as an engaged donor that is at the forefront of thinking in thematic areas such as sexual and reproductive health and girls’ education. They find some of DFID’s guidance (on gender and value for money, for example) to be useful, sometimes also drawing on it for work that is not DFID-funded.

Both in our country visits and in the UK, CSOs said that the predetermined milestones of DFID’s results framework often stood in the way of responding flexibly to changing circumstances but that, depending on personalities, DFID and its intermediaries did sometimes allow for flexibility when this was clearly needed. Our CSO respondents were often similarly positive about DFID’s funding intermediaries: they were generally seen as engaged, responsive, knowledgeable, reasonable in their expectations and trustworthy, and some of them played a useful mentoring role. The multilateral agencies were the exception: they were often seen as aloof, slow, bureaucratic and unreasonable in their reporting and other requirements.

CSOs were positive about their relations with DFID country offices, once projects had reached their implementation stage. The feedback was particularly positive in Bangladesh, even though a terrorist attack in Dhaka had reduced DFID’s presence to skeleton staffing for much of the review period. They mentioned the length and strength of their relations, and the generally helpful attitude of DFID staff. They appreciated DFID’s project site visits and found that DFID provided constructive suggestions that helped overcome implementation challenges. They mentioned that DFID interventions had helped them overcome operational obstacles in Cox’s Bazar, and that DFID had successfully intervened when public authorities instructed one CSO to terminate its operations. CSOs that focus on hard-to-reach groups often said that DFID Bangladesh is a natural and engaged partner and ally. In Ethiopia, the feedback was more varied (probably due to much of DFID funding going through multi-donor channels, which limits direct contacts), but some grantees saw DFID as more pragmatic and flexible than some of the other donors.

It is not clear that the risk-averse, behind-the-scenes country office approaches to declining civic space are achieving results, but DFID does provide practical support to CSOs

Identifying what works to prevent a decline in civic space is not straightforward: lessons from the literature on the issue are limited, while national circumstances differ widely and the causal link between external interventions and changes in civic space is long and hard to identify. Neither DFID nor the FCO is currently tracking its interventions in a manner that would allow for an assessment of their effectiveness. It is therefore not known if, for example, the influencing work of the UK and the broader donor community has contributed to the recent positive developments in Ethiopia’s civic space (where the February 2019 CSO law removes many of the obstacles that CSOs have faced in the review period) and, if so, what lessons the UK could learn from this.

In the absence of central guidance, the nature of DFID’s approach to bolstering civic space differed across country offices. Our review of DFID’s governance work in Uganda found that DFID had taken an ad hoc rather than a strategic approach to declining civic space in Uganda, based on following up on problems experienced by individual CSOs. Conversely, DFID Bangladesh, in collaboration with the FCO, adopted a more deliberate, long-term, multi-layered and strategic approach, in Cox’s Bazar and elsewhere, and is monitoring the changing environment in real time so as to remain relevant.

Our analysis of country-specific DFID responses do show one commonality: the approach is risk-averse. DFID would like to see civic space expand in Ethiopia and Bangladesh, but it is not the UK’s first priority in either country. DFID will not venture into work that might risk the broader suite of UK interests in these countries. One inevitable consequence is a selective approach towards civic space influencing work: like most other donors, DFID Ethiopia will not risk public confrontations on politically sensitive issues such as LGBT+ rights, or the suppression of civil liberties in the Somali region of Ethiopia, and DFID Bangladesh’s response to extrajudicial killings and disappearances in Bangladesh has been muted. Similarly, DFID does not stand in the way of state control by, for example, providing civil society actors with technical means that lower the risk of surveillance when communicating online using social networks. Instead, DFID and the FCO lobby against further restrictions quietly, and behind the scenes, and DFID uses non-confrontational programming approaches. This includes trust-building work between CSOs and public authorities, which assumes that government action to limit civic space is caused by a lack of trust, rather than a political choice.

Within these limits, DFID has provided strong practical support to CSOs, even when their operating environments became more difficult. This was the case in Cox’s Bazar before the recent arrival of Rohingya refugees, and in Ethiopia after legislation limited the fundraising options and operational freedoms of CSOs. Within the limits of the law, DFID’s support has also helped CSOs to deal with new requirements (including by repackaging programmes to increase their acceptability) and to raise their awareness of opportunities, including practical advice, for instance pointing CSOs to government officials who were sympathetic to CSOs and able to support them.

The UK government has not been central to international efforts to stop the global decline in civic space, with the exception of the issue of CSO access to banking services

In the review period, DFID and other parts of the UK government participated in various national and global forums and working groups on the protection or expansion of civic space. However, all external observers we talked to noted that this is a field where donors in general were risk-averse and the UK’s role was modest (and some said absent or uncollaborative). These external specialists also noted that, since US President Barack Obama left office in 2017, there had been a global leadership vacuum in this field. As a result, progress has been slow and limited in most areas related to civic space. DFID’s November 2017 list of ‘achievements’ in this field consists of outputs such as a piece of research, a round table and an ‘options paper’, but not of actual successes in halting or reversing the decline in civic space.

The one exception we saw is related to the UK government’s role in the design and application of international financial regulation. Governments have often used the 2001 ‘Recommendations on Terrorist Financing’ of a global task force as a pretext to restrict civic space. Recommendation 8 asked countries to “review the adequacy of laws and regulations that relate to entities that can be abused for the financing of terrorism” and noted that “non-profit organisations are particularly vulnerable, and countries should ensure that they cannot be misused”. The UK government, led by the Charity Commission, played a significant global influencing role that eventually led to a modification of this recommendation in June 2016. The new recommendation does not consider all non-profit organisations to be “particularly vulnerable” and instead refers to “non-profit organisations which the country has identified as being vulnerable to terrorist financing abuse”. The new recommendation asks for “focused and proportionate measures”. This revision amounts to a significant influencing success, and the UK government role therein has been considerable and traceable.

The original recommendation also contributed to a process of ‘de-risking’, whereby banks withdraw their services from fragile areas to avoid inadvertently violating international rules. This continued even after the global task force modified its recommendation.

In the UK, DFID has been part of a ‘trisector working group’ on de-risking, in which different parts of the UK government, as well as CSOs and banks, are represented. This working group has enjoyed some success in addressing the issue within the UK, even if progress is slow and work is sometimes delayed – most recently because work on safeguarding took priority. The UK government has promoted this set-up in international forums, as it was unique in the world and goes some way to dealing with the potentially conflicting policy agendas on financial stability, counter-terrorism and inclusive international development.

The UK has not yet made a strong international contribution to the issue of de-risking. This may change, however, because in October 2018 DFID announced that it considered de-risking to be a “critical financial development and humanitarian issue”, a “critical issue for DFID”, and something that the UK, as a global financial capital and a leading international development actor, was “well-placed to influence”. ICAI will continue to follow the developments in this area.

Conclusions on effectiveness

DFID and most of its funding intermediaries have been supportive, if somewhat inflexible, during the implementation phase of CSO projects, and almost all CSO projects in our sample have been delivering results that improved and sometimes saved lives. CSOs delivered these results through direct service delivery as well as by holding authorities – from the leadership teams of schools to various levels of government – to account. The best results were achieved in areas where CSOs have long-standing experience and clarity on what works, such as in awareness-raising and service delivery in fields such as education and sexual and reproductive health. Results were more mixed in newer areas, such as the inclusion of people with disability and the engagement of people affected by poverty and disaster in all stages of the projects that aim to support and empower them.

CSOs maximised their chances of achieving durable change by aligning their projects with national development strategies, embedding them in plans with longer timeframes and larger budgets, and implementing them in collaboration with local authorities and through existing structures and facilities. The DFID country offices in Bangladesh and Ethiopia further enhanced the chances of sustained success by providing support to long-term development efforts with the sequencing of individual projects that collectively spanned a decade or more.

While we saw plenty of evidence that CSOs were delivering effective development and humanitarian interventions with DFID funds, we were less convinced that DFID’s funding practices had enhanced their effectiveness. DFID’s weak management of discontinuity and the poor process management of both DFID and most of its funding intermediaries have hampered the delivery of results. These weaknesses led to unreliable and unpredictable donor behaviour in the stages before project
implementation, with insufficient communication around timelines and an inability to stick to them. This, together with the unnecessarily pressurised demands on CSOs, have affected the first year results of many DFID-funded projects, and the collective results of endeavours that relied on multiple phases of DFID funding. For those CSOs that are most dependent on UK funding, DFID’s unreliable and unpredictable behaviour has distracted staff and leadership to the extent that it has affected their overall performance.

We saw evidence of the UK doing useful work in relation to civic space in both Bangladesh and Ethiopia, and CSOs found DFID’s practical support helpful at local levels. However, DFID’s overall influencing and programming work in this field has been slow, limited and unguided by a clear DFID-wide approach. Without a clearer and more rigorous approach that is actioned with a greater sense of urgency, the UK government, together with an equally risk-averse international donor community, is unlikely to reverse this global process of declining civic space.

As a result of the important and, in many cases, largely avoidable shortcomings found in this review, we have given an amber-red score for the effectiveness of DFID’s CSO partnership approach to achieve better development results and a stronger civil society sector.


Learning and innovation: How well does DFID promote learning and innovation in its partnerships with CSOs?

DFID has been slow to fill gaps in the knowledge needed to design its funding instruments

In 2015, DFID reviewed its restricted grants, unrestricted grants, commercial contracts and match funding, and found that “it was not possible to say one funding type delivers better [value for money] than another”. There were knowledge gaps within individual instruments as well: the UK Aid Direct assumption that support to small CSOs encourages innovation and provides good value for money, and the UK Aid Match assumption that match funding strengthens the UK public’s support for UK Aid, had not yet been substantiated.

At the start of our review period, DFID was planning to commission research and evaluations that would gain insight into these issues. Other knowledge gaps that DFID had identified at that time included “the added value of centrally funded CSO projects rather than funding from country offices; the optimum characteristics of competitive challenge fund type arrangements; the extent to which UK Aid Direct contributes to building capacity in funded organisations; … and the advantages and disadvantages of “open” and thematic funding rounds”.

However, DFID delayed work to fill these knowledge gaps, and no research or evaluations on these issues has yet been conducted. We did see DFID-commissioned research in a range of other fields related to its CSO work, such as on civic space, civil society in major emerging national economies, and CSOs supporting accountability in cash transfer programming We also saw research products that are not specific to civil society but do, by their nature, have a strong CSO component – such as research on what works in relation to ending female genital mutilation and in adolescent-focused interventions.

We saw evidence of DFID learning from its own experience and from other stakeholders

Not all learning requires formal research and evaluation, and we found examples of DFID incorporating lessons from its previous experience and from its engagement with CSOs and other stakeholders. Some of this learning was through informal mechanisms (for example DFID using the same technical consultants for different programmes). There was also deliberate and systematic learning, using sources and informants that were appropriate to the issue at hand. For example, DFID:

  • conducted an internal learning exercise in preparation for its recruitment of a fund manager for UK Aid Match
  • developed its cost transparency model using input from a DFID-CSO working group
  • developed its 2018-22 National Action Plan on Women, Peace and Security with active engagement with and learning from CSOs
  • sought feedback on UK Aid Connect design from the International Development Group (a group of donor countries, previously called the International Donor Group).

At the time of writing this report, DFID’s Civil Society Department has designed a ‘civil society learning journey’ to share mechanisms of working with civil society partners and challenges and advantages inherent in collaborative partnerships across DFID and between cadres and countries.

Some of DFID’s funding intermediaries actively used CSO feedback to learn and adapt. The clearest example of this was the fund manager of UK Aid Direct and UK Aid Match, which used CSO feedback to improve upon its role as a fund manager, and kept CSOs informed about these improvements through regular reports that were titled: You said, we listen.

In country offices, DFID participated in the learning events of programmes it funds, and we saw evidence of new phases of multi-phase projects using learning from previous phases. In Bangladesh, we saw a lot of evidence of DFID’s real-time learning about changes in the Bangladeshi operational context, which enabled DFID to play the supportive role mentioned in paragraph 4.64.

DFID supports good quality learning events and processes, but practical impact on the quality of programming is difficult to find

DFID’s shift to project-based funding has closed some learning avenues (such as the one described in Box 5). However, DFID and its funding intermediaries funded and facilitated many other learning systems, processes and face-to-face and online events. We saw national and international gatherings, thematic working groups and communities of practice (with and without DFID participation), cross-CSO and other monitoring visits, webinars and masterclasses, videos, case studies, recommended reading and CSO- and DFID-produced papers that capture learning and are given visibility in sharing events.

Investments in learning were generally proportionate to the nature of the funding instruments. Instruments that mostly funded projects that were grounded in established good practice required only modest learning interventions. Instruments that worked with and through small local CSOs often invested in learning events and other activities that helped the partners by identifying and sharing good practices. This was the case for the Civil Society Support Programme in Ethiopia (£14 million over three and a half years for the first phase) and for the Manusher Jonno Foundation £26 million over four years) and PROKAS (£29 million over six years) in Bangladesh (see Annex 1 for brief descriptions of these programmes). Instruments with an explicit focus on innovation were typically designed to capture real-time learning. This was the case for Start Bangladesh (£10 million over four years), an innovative DFID-funded initiative that disburses money to CSOs in just a few days in response to small-scale disasters, and that invested heavily in its own learning, through peer reviews and regular meetings. In a similar spirit, Bangladeshi CSO BRAC’s ambitious organisational change process, which includes a transition to becoming almost entirely self-financed, is accompanied by an active cross-organisational learning process that is co-financed by DFID.

UK Aid Connect, the most recent of DFID’s funding instruments, is also investing heavily in real-time learning. The instrument is designed to support learning by doing, and its projects have elaborate learning plans built into them. They are designed to generate rigorous evidence and promote wider learning around innovative development solutions. We also noted positive interactions between UK Aid Connect and DFID policy staff. These UK Aid Connect projects have only just started so it is too early to assess results.

Over the review period, UK Aid Direct has become better at capturing project data through improved reporting (which had been an early weakness of the instrument). The fund manager has used this for monitoring and reporting purposes but has not yet gone further to aggregate and analyse the data and use it to inform learning about what does and does not work.

Overall, the CSOs we interviewed generally valued the learning efforts of DFID and its intermediaries, both centrally and in-country. However, when pressed for examples, much of the learning appeared to centre around helping grantees meet DFID’s demanding terms and conditions of funding. Stakeholders were unable to offer many examples of learning around the substance of programming. We saw limited evidence of lessons on what works being absorbed across grantees or between CSO funding instruments. There are various possible explanations for this: learning events may be used to showcase project successes rather than stimulate discussions that facilitate the absorption of learning, preset plans and results frameworks may limit the application of learning, and the uncertain funding environment for CSOs forces them to focus on financial challenges rather than learning.

As a result, while there is ample evidence that DFID has promoted quality learning events and processes, concrete examples of practical learning at the programming level are difficult to identify. In part, this may be because results of learning are diffuse and long-term.

Some of DFID’s innovative funding mechanisms are potentially transformational

Most DFID funding in the review period used established project-based funding modalities, but DFID also used a few innovative funding mechanisms. For example in Bangladesh, Start Bangladesh and the BRAC-DFID-DFAT ‘Knowledge Partnership’ are new mechanisms that are potentially transformational.

The Start Fund is a largely DFID-funded initiative112 in which member CSOs can request rapid financing for sudden small- to medium-scale crises with low visibility, and for spikes in chronic humanitarian crises. The global Start Fund is innovative, and so is its first national version, which was established in Bangladesh, where seasonal flooding generates a need for regular humanitarian response and where there is a well-developed CSO sector able to operationalise such a fund. In both cases, funds are approved by the member CSOs themselves, within a few days. It is the only fund of its type in operation and its response time is unmatched in the humanitarian field. In addition to funding, DFID supports the Start Fund’s efforts to increase its visibility with potential donors, to help diversify the funding base.

The Knowledge Partnership is part of the five-year strategic partnership between BRAC, DFID and the Department of Foreign Affairs and Trade, to which DFID contributes £223 million. The partnership marks an important break in the conventional practice whereby donors attempt to transfer knowledge and expertise to their grant recipients. In this case, it is assumed that the learning will go in both directions.

So far, the principle of mutual learning between BRAC and DFID has not been put into practice. A staff reduction in DFID Bangladesh following a terrorist attack meant that there was little time to invest in learning. A planned ‘BRAC week’ in which DFID could learn from BRAC’s experience of responding in an agile manner in a volatile context has not yet materialised. This idea of mutual learning with CSOs is nonetheless an important one, with the potential to strengthen DFID’s management of discontinuity and its technical approach to areas of its programming, such as social norm change and reaching marginalised groups.

DFID funds innovation but during the review period it did not have processes in place for recognising and promoting successes

DFID defines innovation as “a new solution with the transformative ability to accelerate impact”. Developing innovative work is often more expensive than following established practice, in terms of impact per unit of spend. This is because innovative projects necessarily have a higher failure rate, and because the process of experimenting and monitoring results is itself costly. Innovation is good value for money only if successes are identified, scaled up and replicated, so as to influence wider development practice.

One effort to replicate success was supported by IDEO.org’s ‘Amplify’. This project received £10.1 million over five and a half years to assist with the scale-up of successful innovations in fields such as refugee education and disability inclusion. The initiative we saw – ‘WeLoveReading’ – came from a Jordanian organisation called Taghyeer. Taghyeer helps refugee children learn to love reading through an informal network of learning hubs. Following early success in a Jordanian refugee camp (which was not DFID-funded), Amplify designed a brand, a suite of training tools, a service model and a partnership strategy for Taghyeer to deploy the intervention quickly through partnerships with organisations working in any refugee setting.

Within our sample, there were no other examples of DFID investing in successful, replicable innovative solutions. As discussed above, DFID’s approach to value for money does not facilitate innovative programming, even though several funding instruments have an innovation window. We saw examples of innovative work within our sample projects, but most of this was local innovation, which is the use of an existing intervention in an area where it had not previously been tried. None of it was using new developments in technology, even though the then secretary of state said, in her foreword to the CSPR, that technological advances “provide huge opportunities for CSOs – and I want the UK to be in the lead in seizing them”.

DFID and its funding intermediaries are not well set up to recognise and follow up on successful innovation. They do not systematically recognise innovation and collate information about its results. CSOs told us that the bulk of effort goes towards reporting and compliance requirements, rather than conversations about what could usefully happen next, in terms of replication and scale-up. We saw examples of past innovations that remained underused, for instance when DFID did not follow up on several examples of successful innovations identified in an external evaluation of UK Aid Direct.

DFID’s relationship with BRAC in Bangladesh provides an example of the department actively engaging with a CSO to pursue a successful innovation (see Box 11). In this case, DFID provided part of the initial pilot funding and then continued to fund replication, adaptation and scale-up of the model within Bangladesh (also by other organisations), and independent assessments of the results achieved. However, DFID has not yet fully used its influence to support BRAC’s efforts to do the same outside of Bangladesh. The model is included as a ‘great buy’ in DFID’s chief economist’s overview of ‘best buys in development’.

We conclude that DFID has some track record of funding successful innovations, but that it is not yet clear that its funding during the review period has produced replicable and scalable innovative models. DFID could also do more to identify and follow up on the successes its funding has contributed to. UK Aid Connect, which is just coming on stream, looks better prepared to pursue innovation goals. We will follow its progress once it is up and running.

Conclusions on learning and innovation

We find that DFID has not been proactively filling the knowledge gaps in the design of its CSO partnerships and funding instruments, even though these gaps were acknowledged before the beginning of our review period. However, there are examples of useful research on civil society-related thematic areas. We saw evidence of active learning in a range of areas, with DFID incorporating lessons from previous experience and from its engagement with CSOs and other stakeholders.

DFID and its funding intermediaries have funded and facilitated a range of learning systems, processes and events. We found that these initiatives were generally well designed and proportionate to the nature of the projects being funded. We found less evidence that the results of learning were applied to programming. This is partly due to the organisational pressures on CSOs in a difficult funding environment. It is likely, however, that the learning efforts have contributed to gradual shifts in programming practice in fields such as disability inclusion and accountability towards people affected by crisis and poverty.

Most of DFID’s funding modalities are well established, but we also saw a few innovative funding mechanisms with the potential for transformative impact. At the programme level, DFID has created some funding windows for innovation, but its funding approach does not generally facilitate innovation. We saw a few examples of successful innovation that were identified and scaled up or replicated, but also examples that were not. We find that DFID and its funding intermediaries are not well set up to identify successful innovations, partly because of the high level of attention given to compliance processes.

Overall, we find some of DFID’s efforts to promote learning and innovation in its civil partnerships to be credible. But due to the lack of learning to fill knowledge gaps on funding types before setting up new central funding instruments, and insufficient sharing and uptake of learning and innovations, we have given DFID an amber-red score in this area.

Conclusions & recommendations

Conclusions

We have awarded DFID an amber-red score for its partnerships with civil society organisations (CSOs).

DFID has a clear and consistent view on the role and potential value of CSOs in the humanitarian and development sectors. However, some of the assumptions underpinning its practices have not yet been substantiated and there is an insufficiently robust effort to localise development practice. Moreover, within the parameters of DFID’s CSO funding instruments, CSOs largely function as implementers of short projects with standardised predetermined results frameworks. The Civil Society Partnership Review (CSPR) goal to “build a diverse, resilient and effective civil society sector and a supportive operating environment” does not sit comfortably within DFID’s funding approach, as it allows CSOs limited opportunity for self-driven capacity development and adaptation to the evolving contexts in which they operate.

The consolidation of DFID’s funding instruments, combined with individually sensible but collectively overburdening requirements and the unnecessarily high investment of time and effort required to develop funding proposals, pose additional challenges for CSOs. This risks leading to inefficiencies within the CSO system, with staff time concentrated too much on applying for DFID funding rather than delivering effective and innovative programmes for beneficiaries. These additional burdens are particularly disruptive because they have come at a time of uncertainty for UK CSOs around EU funding and the continuing shrinking of global civic space.

At country level we saw some useful work to maintain and expand civic space, but DFID’s overall influencing and programming work in this field has been slow, limited and unguided by a clear department-wide approach. Without a clearer and more rigorous approach and a greater sense of urgency, the UK government is unlikely to contribute to reverting this global process of declining civic space – particularly considering the lack of leadership and willingness to take risks on this issue within the donor community as a whole.

Once CSOs reached their project implementation phase, DFID and most of its funding intermediaries have been supportive. In some cases, this helped CSOs deliver results that improved and sometimes saved lives. CSOs maximised the value of projects by aligning them with national development strategies, embedding them in their longer timeframes and larger plans (helped by DFID country offices, which often sequenced individual projects to collectively span a decade or more), and implementing them in collaboration with local authorities and through existing structures. However, DFID’s weak handling of disruptive events, and the poor process management of both DFID and most of its funding intermediaries, led to unreliable and unpredictable donor behaviour in the stages before project implementation, with insufficient communication around timelines and an inability to stick to them. This, together with the unnecessarily pressurised demands on CSOs, has affected the results of many DFID-funded projects.

We saw a few innovative funding mechanisms with the potential for transformative impact, for which DFID could now maximise global visibility and buy-in. But at the programme level, DFID’s funding approach does not generally facilitate innovation, and we find that DFID and its funding intermediaries are not well set up to identify innovation successes. This is in part because of the high level of attention given to compliance processes.

Overall, we find DFID’s efforts to promote learning in its civil society partnerships to be credible, providing a good basis for further development. DFID learned actively and often systematically, using lessons from previous experience and from its engagement with CSOs and other stakeholders. DFID also commissioned useful research on CSO-related themes. However, DFID has not filled obvious knowledge gaps in the assumptions that underpin the design of its CSO partnerships and funding instruments. DFID and its funding intermediaries have funded and facilitated a range of learning systems, processes and events. We found that these initiatives were generally well designed and proportionate to the nature of the projects being funded. We found less evidence that the results of learning were applied to programming. This is partly due to the organisational pressures on CSOs in a difficult funding environment. It is likely, however, that the learning efforts have contributed to gradual shifts in programming practice in fields such as disability inclusion and accountability towards people affected by crisis and poverty.

Recommendations

Recommendation 1

DFID should fill gaps in the knowledge needed to optimise the design of its central funding instruments.

Problem statements

  • DFID has identified, but not done enough to fill, key knowledge gaps in its approach to funding CSOs through central funds. These include:
    • the value for money of different CSO funding approaches (such as restricted grants, unrestricted grants, commercial contracts, match funding) for particular types of CSO work
    • the added value of centrally managed funding compared with funding managed by country offices
    • the advantages and drawbacks of open versus thematic funding rounds
    • the extent to which exposure to aid-funded activities causes people in the UK to be favourably disposed towards, and engaged with, the UK’s aid programme
    • the extent to which small and medium-sized CSOs are likely to be innovative, connected to communities and able to achieve lasting impact.

Recommendation 2

Throughout DFID’s central and in-country portfolios, the process towards funding agreements should be more efficient, predictable, reliable and transparent, and should allow CSOs sufficient time to develop proposals.

Problem statements

  • In the stages before project implementation, weaknesses in process management, management of discontinuity, and communication with CSOs have led to inefficient, unreliable and unpredictable donor behaviour, regular and often lengthy delays, and unnecessary costs for CSOs.
  • Delays in DFID processes often compromise the first year results of DFID-funded projects, and the collective results of programmes relying on multiple phases of DFID funding.
  • In DFID’s funding rounds, CSOs typically have between five and eight weeks to prepare and submit their proposals. In combination with mid-process changes in bidding documentation, this leaves insufficient time for the more complex project proposals to consult people affected by poverty and disaster. It also necessitates the use of external proposal writers, and leaves insufficient time for quality assurance and oversight by CSO leadership. The result is a reduction in proposal quality, an increase in their development costs, and compromises in the CSOs’ internal accountability processes.
  • For CSOs that are most dependent on UK funding, DFID’s unreliable and unpredictable behaviour absorbs a significant amount of management attention, to the detriment of programming.

Recommendation 3

Throughout its central and in-country portfolios, DFID should have a stronger focus on the long-term results of its CSO-implemented programmes, the localisation of development and humanitarian efforts, and its CSO partners’ long-term capacity to deliver relevant results in evolving contexts.

Problem statements

  • DFID’s funding and capacity-building efforts are not sufficiently geared towards the long-term goal of locally led development and humanitarian work (the localisation agenda).
  • DFID funding is project-based and usually relatively short-term. This limits the extent to which CSOs are able to achieve sustainable results, particularly for the most marginalised groups.
  • DFID does not test the durability of project results by funding ex-post assessments.

Recommendation 4

DFID should do more to encourage CSO-led innovation, and to recognise and promote the uptake of innovation successes.

Problem statements

  • DFID’s funding approach is geared towards CSOs delivering tried and tested services and interventions in stable environments. It is insufficiently flexible, risk-taking and otherwise conducive to the creation and uptake of innovation.
  • A few DFID-funded programme innovations could potentially be scaled up, replicated and adapted to befit different contexts, but DFID and its funding intermediaries are not well set up to identify and fully use successful innovations.
  • While there are examples of successful innovation within its CSO partnerships, DFID could do more to maximise their visibility to relevant audiences.

Recommendation 5

DFID should provide a guiding framework for country offices on how to analyse and respond to closing civic space within a national context, and work with other UK government departments to agree a joint approach to addressing the decline of civic space at the international level.

Problem statements

  • Globally and in many of DFID’s priority countries, civic space continues to decline (with the notable exception of Ethiopia).
  • DFID and the wider UK government do not have a clear set of priorities, objectives or approaches in relation to the global decline of civic space.
  • DFID does not yet offer a guiding framework that helps country offices to analyse and respond to closing civic space, and country approaches range from ad hoc and responsive to proactive and strategic. There is evidence of demand for such guidance.

Annex 1 - Projects in our sample

Note: please see the PDF version of this review for full references.