Why rethinking how we work on market systems and the private sector is really hard

May 17, 2017

     By Duncan Green     

Whatever your ideological biases about ‘the private sector’ (often weirdly conflated with transnational corporations in NGO-land), markets really matter to poor people (feeding families, earning a living, that kind of thing).  But ‘making markets work for the poor’ turns out to be really difficult and, just as with attempts to tackle corruption or improve institutions, there is a rethink going on in the aid business. Critics of conventional approaches (of which I am one) argue that systems thinking and complexity both explain why a lot of previous approaches haven’t worked that well, and suggest some new ways to tackle the problem.

To catch up on some new research on all this, I spent a fascinating afternoon at DFID last week. The ‘knowledge hub’ BEAM Exchange (they don’t like to be called a thinktank) Beam exchangepresented a discussion paper and technical paper on ‘rethinking systemic change’, along with a warts-and-all case study from Palladium on the difficulty of trying to put this into practice in a large market development programme in Uganda. Some highlights.

The discussion and technical papers reviewed the lessons from the 3 elements of New Economic Thinking: evolutionary economics, new institutional economics and complex adaptive systems. Eric Beinhocker’s influence much in evidence. The papers draw some useful and pretty challenging conclusions:

‘The aim of development must be to enhance the evolutionary process in an economy and create access to this process for all levels of the society, both politically and economically.’

‘Institutional change cannot be put in place by solving problems and scaling them up. Institutional change can only occur if the local actors who are part of the system become aware of their role in the system and have options to purposefully influence institutional evolution. A group that is targeted by an intervention is likely to become more marginalised, not less – a boundary is drawn around the people, and the group is given an often highly simplistic and idiosyncratic label, such as ‘the poor’. Instead of labelling groups of people, the aim of market systems development should be to create opportunities for all people to engage in a functioning market economy.’

market-systemThe authors, Shawn Cunningham and Marcus Jenal, find that a lot of people say they are thinking in systems, but they really aren’t:

‘Too often in our work we see programmes claiming to work towards achieving systemic change while they are busy fixing narrowly defined problems guided by very specific objectives in very short time frames, while ignoring the broader systemic context. They then expect to be able to ‘scale-up’ the solution they have found and through that make it ‘systemic’.’ Ouch.

As is often the way, the critique is more powerful than the proposed solution – a set of 7 pretty generic and unobjectionable ‘general principles’ for those working on market systems (e.g. ‘create and maintain situational awareness’).

But what really made it real for me was the case study on Uganda. The paper, by Palladium’s Andrew Koleros, discusses the challenges faced during the inception phase of a £15m, 6 year programme: “Northern Uganda: Transforming the Economy through Climate Smart Agribusiness – Market Development (NUTEC-MD)”. It compares the project’s experience during its first year with the insights provided by the BEAM think pieces.

What came graphically across in the case study and seminar discussion is the tension between good intentions in design and the subsequent pressures to deliver short term, predictable results in tight timeframes. Systems thinking informed NUTEC’s initial analyses and approach. That initial desire to understand and ‘dance with’ the system however, was swiftly undermined by the pressures of running a multi-million pound project and in particular, the demand to both predict in advance, and then achieve, pre-agreed results. In practice this seems to have generated a lot of anxiety, pressure and an emphasis on ‘us’: what do we do? How do we hit the target of 75,000 families in 6 years? How do we find some quick wins to keep DFID happy? This is in conflict with and rapidly squeezes out efforts to walk the walk in terms of dancing with the system, bottom up solutions, engaging with positive deviance and understanding and supporting endogenous change processes.

This tension also led to a shift to safe bets – a 9 month inception phase was meant to produce proposals to reach a quarter of the 75k households. So the team ended up importing and adapting successful interventions from elsewhere in Uganda to ensure it met its donor targets, rather than trying to identify and support solutions emerging locally.

NUTEC’s inception phase ended a little over a year ago. Since then, the team says it has had more breathing space. It has been able to step back and take a more holistic, participatory approach to engagement with local market actors. But tensions remain.

Hats off both to Palladium for being candid about some of the problems it faced and to DFID, for creating the space for critical reflection.

Some wider thoughts from the ensuing discussion:

There is no guru: local business people often have fixed ideas about what is wrong, and they may not be the best ones. Outsiders have partial knowledge at context intervention 2x2best. So surely a good approach is to find ways to programme in uncertainty – eg multiple parallel experiments (see my current favourite 2×2 – there seems to be a huge resistance from aid agencies to going below the x axis).

The relative power of outsiders and endogenous drivers of change seems to matter. It may be harder to ‘dance with the system’ when the aid business is a 300kg gorilla, and the system a relatively flimsy stripling, as in Uganda. Harvey Koh presented FSG’s Rockefeller-funded study of the Indian dairy cooperative, Amul – did that do better because of the size and self confidence (especially post independence) of Indian reformers, and the secondary role of outsiders? (I’ll blog on the Amul story when FSG finally publish its paper).

How could a portfolio approach help: as long as each project is under pressure to hit its individual targets, you get plain vanilla and safe bets. Can you take a portfolio approach within a project, e.g. by establishing a minimum acceptable target for results, and then having innovation on top? Could Payment by Results have a role? Or a portfolio approach across a region, country or donor, that looks for a blend of safe bets and high risk/high innovation projects?

Finally, it still feels like there is a real need for market systems people to catch up with the work on institutional reform by the governance people – Doing Development Differently, Building State Capability and so on. Governance people seem to have made a lot more headway in understanding what goes on in the black box of institutions, and how to influence it. Some of those insights might well be relevant to market institutions too.

And we really need some more iconic positive market systems stories, like SAVI on governance – hurry up on Amul, please!