Speakers

Rukmini Banerji - Director of ASER Center, Pratham, India

John Githongo - CEO of Inuka Kenya Ltd

Shanta Devarajan - Chief Economist for Middle East and North Africa. The World Bank
Marta Foresti - Head of Politics and Governance, ODI 

Chair

Kevin Watkins - Director, ODI

This event was hosted by ODI to discuss why progress in this area has remained slow and uneven, and asked the question ‘what can be done to ensure that development policy and practise do better in the future?’ It was chaired by Kevin Watkins, ODI director.

Rukmini Banerji, Director, ASER Center, Pratham (India) gave an overview of her experience in the education sector in India, reflecting on the constraints to delivering high quality and equitable education services. She highlighted one of the major challenges faced in India over the past ten years as the difficulty of convincing people that outcomes and quality need to be at the centre of debate. She explained Pratham’s innovative approach to addressing this issue, focused on problem identification and problem-solving. First, this involves a problem-identificationphase focused on collecting and producing evidence of educational quality at all levels (starting at the district level and aggregating up to the state, then national level), to get key actors to identify poor educational quality as a problem. The problem-solving element focuses on finding solutions, including those which can be scaled up. This is necessarily experimental, involving employing a variety of models and strategies at the same time in different contexts and at various levels (i.e. village communities, local government, or the state).

Rukmini identified three key elements to Pratham’s success: collecting evidence to demonstrate what works, maintaining contact with different levels simultaneously to ensure the build-up of trust, employing a flexible approach, allowing for reaction to lessons learnt.  One example she highlighted was Pratham’s work in Bihar state, where they have partnered with multiple levels of government to address the problem of children’s capacity to learn at the grade level at which they were being taught. Working with the district education teams their strategy was to teach children at their actual level rather than by their grade. This was initially implemented in some districts, and then scaled up to the whole state. This showed that a simple model with ‘easy ingredients’, and not much additional funding, was able to energise the system, and was able to be scaled-up by domestic actors.

John Githongo, CEO, Inuka, Kenya, drew on his recent experiences in looking at the impact of corruption on service delivery in Kenya to outline four key issues. Firstly, he underlined the importance of placing equity (rather than poverty) at the centre of development planning given the current context of Kenya and other African countries (i.e. the presence of economic growth and macroeconomic stability).

Secondly, he noted that corruption cannot be addressed solely from the bottom-up but that corruption also needs to be addressed in terms of the networks (top down and bottom up) through which it filters. The responsiveness of government to reform processes has to be politically beneficial and therefore it is important to have a clear mapping of engagement around dealing with corruption.

Thirdly, it is important to note that political will can be manufactured; this requires a wide variety of relationships (or networks) at the local, national and international levels to enable some of the major issues around corruption to be addressed. Networks have the capacity to capture institutions and/or processes, and therefore understanding how they operate (including in particular sectors) is key.

Finally, he argued that the answers to these questions may not reside in capital cities but rather can exist already on the ground. He argued that local people are extraordinary versatile at navigating bureaucracy and politics on the ground, and that the humility to engage with those experiences at the local level (to understand, to appreciate and to listen) should be the key starting point to dealing with these issues.

Shanta Devarajan, Chief Economist for Middle East and North Africa, World Bank, addressed how the debate around politics and service delivery has moved on since the 2004 World Development Report, Making Services Work for Poor People, of which he was a lead author. One area of evolution he highlighted was that the WDR 2004 gave equal importance to the so-called long and short routes to accountability, however growing experience suggest that the long route’ may matter more. In other words, if the underlying political system is dysfunctional, no amount of attempts to fix the organisational management of service delivery, and the relationships between service providers and users, will lead to significant change.

He argued that this means that donor agencies can no longer operate in the same way and that there has been some change in the way that agencies like the World Bank function; in this respect WDR 2004 has changed perspectives. One example is the acceptance that pressures cannot be brought to bear on a political system from the outside, and instead a focus on building up the information base so that poor people themselves can bring pressure. Shanta concluded by arguing that the way to manage political tensions within the heart of reform processes was to approach it from a sectoral rather a governance perspective; for example keeping the focus firmly fixed on the outcome, such as increasing the number of children with a quality education, rather than tackling head on the rents issues.

Marta Foresti, Head of Politics and Governance, ODI, noted that governance research has consistently demonstrated that issues of politics are undermining the effective delivery of services, but this research has been challenged to provide answers as to the ‘so what’ of these findings for policy and practice. Marta argued that there are now five reasons to be optimistic about the capacity of research to provide these answers.

Firstly, governance research has moved on from generic conclusions that context matters to specifying how that matters and what can be done differently. Secondly, diagnostic tools and practical guidance are available that can facilitate exploring the space for solutions (such as ODI work on problem driven PEA, Matt Andrews’ PDIA approach). This has the potential to build operational solutions into donor programme design. Thirdly, these more sophisticated diagnostics help to detect which elements of the problems (particularly the political blockages) can be dealt with, what the conditions are for when some of the more systemic constraints can be addressed and what external actors can do. In other words, complexity is no longer synonymous with non-tractability. There is some evidence that at least in small ways, some donors are adopting such approaches i.e. using flexible pockets of funding.  Fourthly, research has underlined the importance of going local and shows that service delivery can be successful when it starts at the point at which services are experienced and things can be done. The work of Pratham reveals that it is possible to start at the local level, allowing experimentation, and then scale up.

Finally, although research puts domestic actors at the centre of the agenda, it also shows there are effective models for intervention by external actors, including by working through intermediaries or forms of arms-length assistance.

Marta concluded by noting that the research agenda needs to grab this momentum, delving deeper into what it will take to change practice.

Following these presentations, a number of comments and questions were raised. These centred around the assumptions made by donors about how politics work, the way in which information can be used to overcome constraints to service delivery, the lessons for donor programming and the scope for optimism. The key points included:

·         Donor organisations need to avoid making generalised assumptions about how politics and political accountability work and ensure they have an in-depth knowledge of the context.  For example, assumptions about the role of elites and non-elites in blocking or supporting reform are often over-simplified. Whilst formal accountability is important, informal relationships drive these changes and often elites and non-elites have shared interests. Coalition-building is therefore essential.

·         Although information is recognised as a necessary condition, questions can still be raised as to precisely how more information leads to progressive change. Increasing access to information can play out in unexpected way. Shanta highlighted various examples including: work in Benin where a radio campaign to encourage bed net (which should be publicly funded) use did increase demand, but meant parents paid privately; in Pakistan, an information campaign about education quality meant suppliers (particular private schools) reduced their prices in anticipation of a market response instead of parents pushing for change. Two points seemed to be particularly important: Firstly, information and action need to go hand in hand; moreover in different contexts, different actors may be needed to achieve objectives and these can change over time.  Secondly, there is a need to consistently ask underlying questions about who is being influenced by information.  

·         A number of reasons to be optimistic were also discussed. Questions highlighted that  there has also been progress in questioning the assumptions which inform donor and development practice. This includes a refocus on the long route to accountability, more subtle and critical messages about the capacity of information to generate greater accountability, and reflections on the role of trust-building and coalition building, across elite and non-elite divides. 

·         Two further lessons for donor programming were highlighted.First,donors need to ensure that they have an in-depth knowledge of the local level context prior to programme design. Second, programmes need to have an inbuilt capacity for experimentation and flexibility; they should be informed by on-going action research as to what the results of the intervention are and be able to change course. 

·         Finally, there was recognition that the role of politics has hindered progress towards the MDGs and this implies that the post-2015 frameworkshould pay more attention to political dynamics. One way of doing this could be to focus more on some less ‘visible’ outcomes (such as learning and education quality) and ways of building political will for these, rather than those which are more ‘visible’ (such as school attendance and construction).