Can Evidence-based Activism still bring about change? The view from East Africa

Spent last week defrosting in Tanzania, at a fascinating conference that produced so many ideas for blogs that, twaweza conferenceeven if all the promised pieces don’t materialize, we’re going to have to have a ‘Twaweza week’ on FP2P. Here’s the first instalment.

I’m buzzing and sleep deprived after getting back from an intense two days in Dar es Salaam, reviewing the strategy of one of my favourite NGOs. Twaweza works in Tanzania, Kenya and Uganda, and is in my experience unique in its deep and simultaneous commitment to civic activism, social change at scale and rigorous research, focusing in particular on improving the lamentable quality of primary education in East Africa.

I first got involved with them back in 2013, at a point where their research had found that their hugely ambitious efforts had produced precious little in the way of results. But instead of burying the findings, they summoned people from all over the world to discuss them. Respect. That was where I first came upon my all-time favourite cartoon (see then a miracle happensleft). And I’ve been trying to understand ‘then a miracle occurs’ ever since.

We had an exchange of blog posts in 2013, and Varja Lipovsek and Rakesh Rajani, both at Twaweza at that point, summed up where they wanted to go next:

  • move beyond the ‘Magic Sauce’ of access to information as a driver of change
  • Political economy analysis as standard
  • Focus on the interstices between citizen and state
  • Understand the barriers to poor people’s agency
  • Think about Positive Deviance
  • Iterate and adapt
  • Don’t give up on scale

Fast forward to 2018, and I think this checklist is still where a lot of the more interesting work on accountability and governance is at.

But the world has changed since 2013, not least in the area of accountability and civil society participation. There has been a flowering of thinking and practice on institutional reform and governance, but there is a weird disconnect: the ‘Doing Development Differently’ and ‘Thinking and Working Politically’ types working on institutional reform seem barely on speaking terms with organizations like Twaweza and the whole ‘open government’ and transparency movement, even though they are often thinking and acting on parallel tracks. That weakens the thinking and practice of both groups: DDD people are often far too dismissive of civil society as a bottom-up driver of change; transparency and accountability people aren’t thinking hard enough about how to understand and ‘work with the grain’ of informal institutions and states.

More generally, the zeitgeist has not been kind to Twaweza’s original project, built on evidence and support for civil twaweza-logosociety. Governments and policy makers are more ready to dismiss evidence in these post-truth times, and civil society is under siege in dozens of countries (including all 3 of Twaweza’s). Which leads to two big questions:

To what extent should Twaweza focus on defending civil society space, even if it currently looks like a losing battle in many countries? Or (like the International Budget Partnership, which it resembles in some ways), should it further diversify its partners to non-CSOs, eg working more with private sector, or faith leaders, the judiciary or sympathetic politicians and officials?

I had an even bigger question on its understanding and use of ‘evidence’. Twaweza’s commitment to it is laudable, but it’s main value so far seems to have been in proving what doesn’t work – an impressively transparent process of unlearning and questioning initial assumptions, coming up with new ideas and then testing them in turn. That was particularly the case at the 2013 soul-searching. Since then, it feels like some of the engagement with academics has slightly lost its mojo – the researchers are burrowing away on questions such as what makes voters switch parties, that are of great interest to political scientists, but not obviously relevant to Twaweza. The gulf between the researchers and the activists at the seminar seemed wider this time, with the activists more sceptical about the value of some of the research, and insisting that what constitutes ‘data’ had to emerge from communities. I got very excited about updating Robert Chambers – from ‘handing over the stick’ to ‘handing over the click’ (I’m shallow like that). What’s the best thing to read on bottom up construction of data?

But I still think Twaweza should stick to its ambitions on scale (Rakesh originally got my attention by saying Twaweza didn’t want to do anything that wasn’t going to reach at least 2 million people). So if evidence and access to information aren’t enough to trigger the elusive miracle, are there any alternatives? Two emerged during the course of discussions:

Spot the outlier
Spot the outlier

Positive Deviance: Twaweza is using PD to identify high performing schools in poor areas of all 3 countries. It’s a fascinating attempt to apply one of my favourite ideas on the ground, and Sheila Wamahiu has promised to write a post for FP2P on the experience.

Social Network Analysis: I was interested in this recent LSE post about applying SNA in fragile states, and wondered if Twaweza could get to scale by identifying the nodal individuals, communities or organizations on any given issue, and targeting them – after all, as Malcolm Gladwell described years ago in Tipping Point, marketing departments have been doing it for years.

These are big challenges and big ideas, but if any organization has the ambition and the smarts to pull it off, it’s Twaweza.

Anyway, I’ve run out of space. Lots more to come. Twaweza want right of reply to this post, 8 people attending the seminar promised to write blogs on their work, and I have at least two others of my own to churn out – a very productive couple of days!

(Full disclosure: Twaweza paid for my flights and hotel)

Update: Here’s MIT GOV/LAB’s Alisa Zomer’s summary of the event

Subscribe to our Newsletter

You can unsubscribe at any time by clicking the link in the footer of our emails. For information about our privacy practices, please see our .

We use MailChimp as our marketing platform. By subscribing, you acknowledge that your information will be transferred to MailChimp for processing. Learn more about MailChimp's privacy practices here.


6 Responses to “Can Evidence-based Activism still bring about change? The view from East Africa”
  1. Duncan!! I am ex-Oxfam staff and a serious follower of the blog. I do like your sharp view and critical analysis on emerging trends in development sector. I have been on my eyes what will you write on Oxfam Hati issue. Nothing pop up. Why you are silent? Are you a positive deviant? I wish so.

  2. Cathy Shutt

    Interesting post.

    We had lots of similar debates about different kinds of evidence and its use for different types of learning during MAVC that culminated in the compilation of a learning dictionary including a range of learning definitions: experiential, evidence based, operational, contextual and evaluative, which are described here:

    I am not sure what the best thing to read on citizen data is and whether this is the kind of ‘research’ you had in mind, but one of the main findings from Itad’s macro evaluation of DFID’s empowerment and accountability programme reads “Dialogue between citizens and service providers works best when backed up by locally-generated evidence. Discussions between citizens and service providers was more focussed and people were more motivated when they had access to citizen generated data that monitored service quality and user satisfaction. For example, a social accountability programme in Bangladesh provided grants to local civil society actors which promoted public dialogue with service providers, backed by systematic user feedback gathered through tools like scorecards. As a result, safety net provisions became more transparent and were no longer captured by those who were not entitled.”

    There are several MAVC publications on citizen data that Rosie probably mentioned. Here is a link to one where citizen data has had policy level influence.

    Insights work as used in human centred design and the development of brand strategy might also be relevant?

    As no doubt you are aware, the Hewlett Foundation’s TPA learning sub-strategy is informed by the disconnect you describe and interested in supporting learning on citizen based evidence (among other things).

    • Duncan Green

      Thanks Cathy, useful links. I was wondering who’s doing work on what citizens define as data, rather than just getting citizens to collect data defined by others. Like that work to get communities to define poverty lines, when they typically come up with things like owning a bicycle or tin v thatch roofs. Has that been tried on eg education or health?

  3. Ken Smith

    Really interested to understand more about Twaweza. Is there any route for average East African citizens to get involved with them ? Who provides the funding and is it sustainable ?