Why don’t people in power do the right thing – supply, demand or collective action problem? And what do we do about it?

My last few days have been dominated by conversations around ‘convening and brokering’, including an exchange between assorted ODI wonks and a bunch of NGOs on the findings of the Africa Power and Politics Programme, and a ‘webinar’ (ugh), with our Latin American staff on the nature of ‘leverage’ (a closely associated development fuzzword). Yesterday I set out the best example of this approach that I’ve found to date, the Tajikistan water and sanitation network. Today it’s some overall conclusions from the various discussions.

David Booth from ODI described the question he is trying to answer as ‘why don’t people in power do the right thing?’ He thinks aid agencies (both official and NGOs) have moved from thinking that the answer is building capacity in government (supply side) to strengthening the voice of citizens to demand better services (demand side), but argues that both approaches are wrong.

The mistake, he argues is seeing power as a zero sum game, whereas often the barrier to progress is better seen as a collective action problem: ‘doing the right thing involves cooperating with others and people aren’t prepared to take risks and bear the costs of working with others, unless they believe that everyone else will do so too.’

That requires a different approach, getting everyone into a room to build trust and find joint solutions to a common problem.

ODI argues that on the ground, a lot of aid agencies realize this, and are doing it already. But the official line (often driven by donors’ funding decisions) is that they are exclusively building demand-side accountability, so their reports and narrative airbrush out all that ‘collaborationist’ activity with local government officials, politicians etc. That’s a problem because it inhibits their ability to share experiences and learn how to do things better.

As evidence, ODI cited an evaluation it did for Plan of a ‘Community Scorecards’ project in Malawi that was proving remarkably successful. The programme design was classic demand-side: entitlements, rights, holding duty bearers to account etc. But when ODI investigated, they found that reality involved brokering local-level reform processes and working with local officials to help them raise concerns with central government. Solutions included communities agreeing to help with school construction. In agriculture, problems included fertilizer subsidies being traded on secondary markets, sometimes for sex. The project brokered contacts with police and the courts to help sort it out. Little of this appeared in the official project narrative.

All well and good, but Oxfam’s Jo Rowlands argued that the NGOs’ approach is different to ODI’s in one important aspect. While ODI argues for ‘going with the grain’ of existing institutions and traditions, the NGOs are more normative – going with the grain but at the same time seeking to change it, through a kind of ‘affirmative action convening and brokering’ that ensures the voices of previously excluded groups are at the table. So for example, our work with protection committees in the DRC involves helping them build trust with local government and ‘armed actors’, but also ensuring the committees have an even gender balance, which has transformed the confidence and self-perception of many women participants.

would they get more results from a meeting?

This kind of transformative approach usually involves something additional to just convening and brokering (Tajikistan is a bit of an exception). In livelihoods it involves investing in technical assistance and building organizational capacity so that smallholder assocations can benefit from value chains. In women’s rights it involves building ‘power within’ as well as brokering the kinds of discussions the protection committees have in the DRC. Elsewhere it may involve running pilot programmes to demonstrate new solutions around which the discussions can take place.

Which leads me to a key dividing line between two kinds of convening and brokering. The more innovative kind involves acknowledging that there is a problem, but admitting that we don’t have a solution, and want to get everyone in the room to try and find one. That’s the Tajikistan model, but is still something of a rarity (NGOs often think they know the answer, even when they don’t….). That is very different from merely trying to build an alliance around a predetermined policy demand (a much more common approach).

Which all left some important questions and dilemmas hanging in the air. They include:

• Given that social change often takes place through a cycle of cooperation and conflict (see diag), when and where is the ‘problem-solving approach’ the best answer? Just during the kiss and make up phase, or more generally?
• Is this approach easier in some sectors (children, water) than others (taxation, livelihoods)? Or is it easier in service delivery work (more pragmatic) than influencing (more normative)?
• ODI argues that the trick is to pick the moments when the stars are aligned for some kind of collective action breakthrough, but how do you recognize such moments, apart from in hindsight (not a lot of use for practitioners)?
• What kinds of people are good at this, and do they work for aid agencies? In my experience, lovers of ambiguity, policy entrepreneurs willing to take risks, and networkers happy to talk to people they disagree with or even dislike are in pretty short supply in the aid world
• David Booth argues that ‘meetings are of the essence’, but what distinguishes useful convening-type meetings from pointless NGO gabfests. (JAM – Just Another Meeting)?
• Which brings us to the role of donors. To what extent can they cope with the uncertainty over attribution and the long timescales involved in this kind of work? How do we take them with us?

Finally, we agreed to ask for your help. David Booth reckons we need a good snappy name for this new approach – open-minded on solutions, trust-building, convening and brokering, problem-solving etc. Any ideas?

And since ODI is funky and digital these days, here’s my 3 minute download, which they filmed straight after the meeting

What have we learned from 5 years of research on African power and politics?

The Africa Power and Politics Programme (APPP) is winding down as its five year funding from DFID comes to an end, and I’ve been wading through the 120 page synthesis report as well as the strictly-for-wimps Policy Brief. Both are entitled ‘Development as a collective action problem: Addressing the real challenges of African governance’.

Like previous APPP work, the papers are intriguing and frustrating in equal measure. David Booth from the ODI, the principal author, appears torn: his comfort zone is the abstruse conceptual landscape and language of political science. But his paymasters are practical men and women who insist on their ‘so whats’. ‘Researchers have a duty to provide more than negative messages and evidence of complexity. There needs to be a meeting point between researchers’ recognition of complexity and practitioners’ hunger for guidance.’ He does his best, and promises much, but it doesn’t come easy, with conclusions that often stop just as they get interesting (at least to prosaic practitioner types like me).

First comes the standard take-down: a comprehensive and persuasive rubbishing of mistaken approaches. Yes, the development world may have moved on from ‘magic bullet’ approaches, accepting the APPP’s core argument in favour of adopting ‘best fit’ approaches, ‘going with the grain’ of existing histories and institutions in any given place. But in practice ‘even the most reflective country activists and the best governance advisers have trouble imagining what to do differently.’

On current aid practices, the synthesis report is far more damning than the Policy Brief (perhaps in deference to DFID’s sensitivities). Booth lambasts the ‘per diem culture’ (or as my colleague Ben Phillips puts it, ‘carpe per diem’) that undermines genuine attempts to resolve local collective action problems, as well as

• ‘the distortions caused by the availability of donor money and organisational templates, now delivered to the remotest rural areas by local governments and NGOs, and
• mechanical application of donor-inspired policy guidelines by sector ministries in ways that not only contribute to policy incoherence but prevent local actors from coming together to provide their own solutions.

He’s particularly critical of direct funding to grassroots organizations, arguing that ‘in Pakistan and elsewhere, civic groups that get funding from development assistance end up with no members.’ But he then acknowledges that in Malawi ‘in fact, some of the promising experiences do involve NGOs as actors, and some involve the use of project funds’ although frustratingly, there are no further details.

APPP reckons a more profound conceptual shift is required, ditching the ‘straitjacket of principal-agent thinking’ on governance. It pours equal scorn on supply side governance (governments are keen to run the place better, they just need training) and the more recent switch to demand side (just help citizens’ groups who are dying to hold governments to account, and that will lead to development). ‘The conventional idea of supporting a pent-up ‘demand for good governance’ must be put aside.’ ‘Citizen pressure will normally lead to more effective clientelism, not better public policies.’ [ouch]

Instead ‘governance challenges are not fundamentally about one set of people getting another set of people to behave better. They are about both sets of people finding ways of being able to act collectively in their own best interests. They are about collective problem-solving in fragmented societies hampered by low levels of trust.’

How? First engage with states, but not all of them: learn how to spot more developmentalist bits, and where you find them, reject the ‘dominant view for the last 25 years that African governments cannot be trusted with interventionist policies.’

I’ve written before about David’s apparent love affair with Rwanda’s Paul Kagame (left), but here he accepts that just wishing all African leaders were benevolent autocrats is not really good enough, not least because such ‘developmental patrimonialist’ regimes tend to emerge from major wars and national liberation struggles and/or major threats to national survival like the Rwandan genocide, which are unlikely to be repeated. Instead, he acknowledges ‘the most urgent policy questions relate to options for the modal type of contemporary African regime, where clientelism is competitive and operating under a formally democratic political constitution.’ Still, there is a lingering fondness for what the synthesis report terms ‘strong, visionary leaders’, combined with a rather old school notion of development as economic transformation first, and we’ll worry about all that fuzzy human rights, well-being and agency stuff later.

What of more specific so whats? There are tantalising glimpses here and there, never fleshed out fully. These include:

• New forms of power-sharing to deal with ethnic conflict
• Ring-fencing long term development priorities (eg infrastructure, smallholder agriculture) from party politics
• Pursue what APPP dubs ‘practical hybrids’, the result of ‘conscious efforts by elements of the modern state to adapt to local preferences and ways of doing things.’
• Learn from successful governance turnarounds in Latin America (Brazil, Bogotá), which ‘worked less by changing the formal rules of the political game, and more by bringing informal social norms and moral sentiments into line with the high ideals articulated in national constitutions, making creative use of mass media and the power of example.’
• ‘Official agencies should do more things ‘at arm’s length’, delegating assistance to organisations that have demonstrated an ability to work in the ways that are required to make a positive difference.’ Would that include ODI by any chance?

Stepping back, the underlying challenge identified by APPP seems to be how both governments and citizens can move to a less short-termist mindset and agree on the kind of institutional development that underpins long term development, finding ways to overcome the paralysis of collective action problems: ‘Where positive outcomes are achieved, the reasons are almost always that circumstances have permitted a collective action log-jam to be overcome, usually at several levels simultaneously and interactively.’

I think this analysis fits with some thinking we’re doing in Oxfam around the topic of ‘convening and brokering’. In certain circumstances, the best role for an outside player like us is not to build stuff, or dispense large amounts of cash, but to get disparate local players into a room and encourage them to find their own solutions. In Oxfam the iconic programme story is in Tajikistan, where we convened a bunch of ministries, private companies and civil society organizations to discuss water and sanitation. We don’t lobby for a particular agenda or institutional template, we just keep them talking – an afternoon every two months. Already the process has yielded an inter-ministerial coordinating committee on water, a new water law, and specific projects are now starting to emerge. The secret to success in this is often the human skills of the facilitator (in this case a rather charismatic water engineer who is now the Tajikistan country director) and acceptance by all parties of the credibility and independence of the convenor.

It also reminds me of Dani Rodrik’s work on growth diagnostics and bottlenecks: ‘development progress is about overcoming institutional blockages, usually underpinned by collective action problems. It is not, for the most part, about resource shortages or funding gaps.’

This seems to be heading towards some kind of ‘participatory institutional appraisal’ approach, where development actors specialize in convening discussions of local players to get over these logjams in ways that reflect and adapt local traditions and values. This runs up against the way aid agencies currently work: high staff turnover, massive pressure to dole out funds in large amounts, demands to show ‘value for money’ via an increasingly demanding and imposed system of governance, monitoring, evaluation etc etc

A suggestion: APPP should present this work to a group of practitioners (bilateral, NGOs etc), then brainstorm on examples where they are successfully pursuing this kind of approach. They should then write them up in plain English and use them to illustrate their arguments – I think I can guarantee a significant improvement in research take up and impact. Any takers?

Harnessing religion to improve education in Africa

The Africa Power and Politics Programme is thought-provoking, innovative and infuriating in equal measure. ‘Religion and education reform in Africa: harnessing religious values to developmental ends’, a fascinating new APPP paper by Leonardo A. Villalón and Mahaman Tidjani-Alou, examines recent educational reforms in Mali, Niger and Senegal, three overwhelmingly Muslim, francophone countries in West Africa. All three have tried to ‘harness the strength of popular religiosity’ in different ways ‘to make  schools more attractive to parents by incorporating  elements into schooling that reflect Muslim values  and expectations and ensure training for future  employment.’

Some background: The education system inherited from the French colonial period is deeply secular: ‘Under these systems, only a tiny percentage of the population ever completed secondary education. Despite calls for reforms early in the post-colonial period, few changes were made.  As a result, most parents have seen official state schools as unattractive options at best, and often resist efforts to enrol their children.

Across the Sahel, another response to the bad fit between the provision of public education and social expectations has been the development of a vast parallel system of informal and religiously-based education functioning outside the official state system.’

Things got worse in the 1980s as structural adjustment programmes cut into both state jobs (previously a motive to put kids through state schools) and the schools themselves. Reforms since then have largely ignored (or even tried to suppress) the flourishing parallel world of religious schools. Then the three governments decided instead to ‘go with the grain’ (a recurrent theme of the APPP), ‘both bringing unofficial  schools more squarely into the formal state system  and reforming the formal system by borrowing  characteristics from the informal, such as introducing  religious education in state schools.

In each of the countries, states embarked on reform projects inspired and justified by what one key actor called ‘giving parents the educational options they want for their children’. Across the region, parents interviewed by APPP said that they want schools that incorporate religious values, but also schools that provide some hope of access to employment and practical life skills. Attempting to balance these dual demands, the reform projects have tried to recognise the parallel educational systems while imposing some degree of formalisation, or have tried to reform the official system by borrowing elements – such as religious instruction – from the informal. The result has been the creation of what are in effect ‘hybrid’ systems.’

In Niger, this involved expanding the existing network of ‘Franco-Arabic’ schools; in Mali, the government created incentives for previously unrecognized madrassas to adopt the official state curriculum without abandoning their religious mission. In Senegal, state schools have included religious instruction since 2002 in an effort to compete with the informal system.

Some findings:

‘Hybrid schools in their various formats have been extremely popular with parents, and the major challenge to the State is how to meet the high and growing demand they have created.

Despite the fears of some observers, the reforms have not exacerbated gender imbalances. At primary school level, for example, the emphasis on religion has proven particularly attractive to parents of girls. In many hybrid schools, girls outnumber boys, sometimes significantly. Finally,  preliminary indications suggest that the success  rates of the hybrid schools, as measured by  the number of students passing state exams,  is as good as or better than that of the classic  francophone schools.’

Conclusion?

‘The educational reform experiments in the Sahel provide strong evidence to support one of APPP’s core hypotheses. In the Sahelian educational context, building institutions that work with or tap into prevailing moral orders and cultural values shows real promise as a means to address some deeply entrenched obstacles to better development outcomes. Strikingly, however, while the cases suggest the importance of local values, they do not suggest a rejection of the state as a primary actor in development. Significant popular demand for education in the Sahel takes the state model as its point of departure, but asks that it be adjusted to local values. We find that the ‘grain’ of popular demand  in contemporary Africa is not a desire for ‘traditional‘  institutions, but rather for modern state structures  that have been adapted to, or infused with,  contemporary local values.’

The authors do, however, acknowledge that there may be trade-offs involved – for example on what the kids are actually being taught about gender rights, but think they should be acknowledged and debated. What I particularly like about this conclusion is that it doesn’t give up on state provision, but tries to make it speak to local culture and values, rather than those of some long dead, Voltaire-quoting colonial master.

[Postscript: make sure you read the comments - really good debate]

Africa Power and Politics – David Booth responds

ODI’s David Booth responds to my post on the ODI’s Africa Power and Politics Programme

“The APPP could hardly have hoped for a more encouraging reception for its first policy brief than the one provided by Duncan’s blog of 15 April. Encouraging and suitably challenging!

The point of a policy brief is to be, well, brief, and focused on implications. So it’s not surprising if Duncan finds some of our formulations a bit too pithy and in need of substantiation.

Duncan quotes us generously, but there is much more where that came from, freely downloadable from the APPP website. It’s also worth saying that the evidence base we are drawing on is not just APPP research but a large body of other work, including stuff to which Duncan refers us.

Let me pick up just a couple of points of apparent disagreement.

Citizen pressure and public goods

Obviously, one of them has to be our proposition that citizen pressure and bottom-up demand for accountability is a weak factor in improving the governance of public services. This poses a rather direct challenge to the way most international NGOs view their mission, so Duncan’s ‘alarm’ is understandable.

We are aware of walking into a lion’s den in posing this issue so bluntly. But we are doing so with good reason. There is a real conflict here between the conventional ‘progressive’ viewpoint and what virtually all documented experience shows.

In elaborating, we said: ‘… client “voice” is a weak source of results-based accountability unless accompanied by strong top-down pressures of some kind’’. The qualification is important. It is meant to capture two things.

One is the almost universal finding that public service improvement comes when there is successful action to improve provider motivations. And this doesn’t come mainly from the bottom-up (even in the UK, where service users are much more empowered, providers respond poorly to having ‘demands’ placed on them). The other is the fact that the documented successes in so-called ‘social accountability’ almost always involve political forces (new parties, new leaders or something of the kind); rarely are they just movements of ‘citizens’.

This is actually what the IDS research and related ODI and IDS evaluations tell us. None of them provide a blanket endorsement of civil society action on accountability. If this seems surprising, it’s because there has been a huge amount of over-selling of ‘demand side’ interventions on the basis of partial reading of the evidence.

So I hope Duncan’s alarm will soon give way to agreement. The challenge to conventional thinking about ‘good governance’ applies to all of us, not just to the soft target provided by official donor governance work.

The politics of aid

There are many other points I might pick up, but I will leave democracy, Amartya Sen and universal rights for another occasion. And on the important question of how to avoid sliding back into ‘decent chap-ism’, I will defer to my APPP colleague Tim Kelsall.

For now, let me just comment on Duncan’ scepticism about uptake of APPP findings by official aid. Of course, yes, the obstacles are pretty daunting. Backing off and concentrating on the ‘do no harm’ agenda at home, as suggested by Mick Moore and Sue Unsworth, has much to be said for it. But it is too easy to take the existing aid set-up as a given, and see the function of research as limited to feeding its appetite for ‘take-aways’.

As we said in the brief, aid needs to fit the needs of development, not the other way round. For sure, this is not a simple matter. It is about public attitudes and the long-term shaping of opinion in the North, and not just about the civil servants or the ministers who are currently in charge of things – which is why, as Duncan notes, drivers of change and political economy analysis don’t change behaviour all that much. But public attitudes to development have been substantially remoulded in recent years by NGO campaigns, not always for the good but significantly nonetheless – and with Oxfam often in the lead.

Given this success, it seems a bit soon to give up on the possibility of shaping a more mature general consensus on what matters and what doesn’t in governance for development. I think we should talk about how we can work together on this.”

Duncan: Thanks David, and I really hope you’re right about getting aid donors to think more about politics and context. But using blogger’s prerogative to have the last word, I’d just like to say that there’s at least one straw man in here – the idea that NGOs like Oxfam see change coming about purely by bottom-up civil society organization. In fact, both our theory of change, and the practice of our advocacy usually involves cross sectoral alliances with business, allies within government, churches as well as community organizations, peasant associations and the like. I’d go further, often, NGOs’ most important role is to act as catalysts and convenors, bringing groups together who would not normally talk to each other. So where I differ with you is not that change comes about from a combination of top down and bottom up, but in the APPP Policy Brief’s far too dismissive line that ‘citizen pressure is at best a weak factor and at worst a distraction’. I think events in North Africa are just the latest reason (on top of all that work by the IDS Citizenship team) to say that that statement is simply wrong. US civil rights movement, anyone? The struggle against apartheid? I notice you don’t repeat it in your post – perhaps it wasn’t quite what you meant to say?