What we’re missing by not getting our TWP alphabet straight

TWP guru Heather Marquette does everyone a great service by explaining the important differences between all the acronyms.

I am struck by how often people say ‘TWP/PDIA/adaptive management/PEA…whatever’. Kind of like when my great-aunt calls me by various relatives’ names first before getting mine right – ‘Sheila… Mary…Lily…Heather!’ – these things may share a common genesis, and there are threads that obviously connect them, but they are actually different things.

That matters because important distinctions are getting lost, so here is my (final) attempt to unmuddle the soup.

Problem-driven iterative adaptation (PDIA)

PDIA isn’t about donor programmes. It’s about state capability. Matt Andrews and Lant Pritchett et al’s research looks at what effective states do, not what donors do directly. They describe this approach as:

  • Local solutions for local problems;
  • Pushing problem driven positive deviance;
  • Try, learn, iterate, adapt; and
  • Scale through diffusion.

Their interest in donors is limited to how donors should support states in building their own capacity to undertake reforms and, just as importantly, what not to do to undermine this.

Donors can’t ‘do’ PDIA. DFAT’s Saku Akmeemana made this point really well at the 2018 Australasian Aid Conference (see 13:19-16:06). If you’ve not watched this, stop reading this blog right now and do so! She said,

The whole idea of…the PDIA model is around learning…[PDIA] is looking at the process of national development…There’s an endogenous feedback loop from the experimentation to the adaptation. There’s a whole political and administrative system to respond, and we’re trying to mimic that somehow in a compressed project/programme cycle with something that’s externally imposed. So some of these ideas are going to be very hard to implement, because we don’t have that endogeneity in the feedback loop.

Credit: courtesy of pjevans

Referring to China’s experience with adaptive learning, she talked about the Chinese Communist Party’s incredible – and massive – system for learning, concluding that, ‘We have to have some humility and modesty in terms of what we can do through a project’.

In his (non-)retirement speech, David Booth made an important observation that PDIA isn’t just about adaptation – it’s about being problem-driven, not solutions-led. External actors, like donors, tend to come to the table with funding for particular solutions, rather than starting with asking local actors what their priority problems are.

I recently met with a programme team in Ghana, and a standout comment from a Ghanaian colleague was that he has been so proud to be able to go to country counterparts and say, ‘What problems are you grappling with? How can we help you solve these?’, not ‘here’s what I’ve got for you’. David’s work intersects with Matt and Lant’s in the sense that one of the worst things that donors can do to undermine a state’s own reform process is to insist on the solution in the beginning, or perhaps by only being interested in specific problems – which may not be the state’s own priorities. More than getting rid of log frames or developing new M&E frameworks, moving away from solutions-led aid would be an important step towards building state capability, enabling states to define their own problems and, through their own problem-driven iterative adaptation, finding their own solutions.

So PDIA = important insights into how effective country reformers work.


Adaptive management (AM)/Flexible & Adaptive (F&A) programmes

Adaptive management is, in some ways, how PDIA translates into aid practice. It is seen as good management practice in a huge range of sectors and so isn’t aid specific. At its heart, it’s about piloting, testing, learning, adapting and delivering. This is likely to look different at different scales and in different contexts and needs a good evidence base to understand how, when, where, why and who, and – just as importantly – what doesn’t work.

What is aid-specific is that adaptive programmes are being delivered in contexts that aren’t funders’ own. It is (or could be), in effect, aid’s riposte to development programming that’s solutions-led, and there is a direct line from research on PDIA to adaptive management; however, the two are not interchangeable, and that’s really important. Understanding this helps to explain why the M&E nut has been so hard to crack.What M&E (and L) needs to do in the context of externally-funded adaptive aid programmes is to try as far as possible to find ways to replicate the feedback loops that exist ‘naturally’ in local contexts, as Saku explained so well.

So AM/F&A ≠ PDIA (as David Booth pointed out).

And AM/F&A ≠ TWP.

Political Economy Analysis (PEA)

Unlike PDIA or even AM, PEA isn’t an approach. It’s a tool, an input, a sense check, a framework. Or, more accurately, it’s a range of tools, inputs and frameworks that can help with sense checking a programme, from planning through design to implementation and beyond.  It can be formal or informal. It can be national, regional, sectoral, problem-driven, ‘everyday’. It can be done once, producing a big report, or it can be integrated into a team’s daily working. It can be insightful or it can be rubbish.

It’s not the same thing as the political analysis done by diplomats, though that can make an important contribution to PEA. At it’s best, PEA helps us to understand where political will for reform may come from and, if at all possible, how to shift from political won’t to political will in order to achieve programme outcomes or – just as importantly – to avoid negative unintended consequences/do no harm. ‘Political won’t’ is not always passive, an absence of will; it can be active – from refusing to acknowledge and engage on an issue, to actively preventing an action (sometimes for good reasons), to harnessing powerful interest networks and violently sabotaging it. Good PEA can help you understand which one of these is happening and, perhaps, also help you figure out if you can do anything about it.

PEA reflects a reality where development programmes are funded, owned and implemented by outsiders. PDIA – the process by which states deliver their own reforms – shouldn’t really need the kind of ‘rules of the game’, full bells and whistles PEA analysis that outsiders need in order to understand the context. Reformers in country may pursue reforms that aren’t technically or politically feasible, of course, but the basic contextual knowledge is there. Anyone can need help thinking through how to unblock reforms or how to generate necessary political will, but only external actors need the sort of ‘full blown’ PEA that ends up in training courses for donors and INGOs. But Everyday Political Analysis, stakeholder mapping, risk analysis, horizon scanning etc can be used by national as well as external actors. The outputs might still be rubbish, and bad PEA can generate its own problems, of course. We definitely need better (any?) evidence on this.

But PEA ≠ TWP.  This is important.

Thinking & working politically (TWP)

Unlike PEA, thinking and working politically isn’t a tool, or even an approach. It’s a way of thinking and working that keeps the understanding that everything is political front and centre. It is, at its heart, how people ‘parlez-vous politics’ (see this and this). It isn’t something that only aid people do. It’s not even something that only policy makers do. My dad – a carpenter and fisherman – thinks and works politically better than anyone I’ve ever met.

Research shows how TWP has helped local reformers get from political won’t to political will. Research  looking at how a donor supported local civil society in better holding the Nigerian government and oil companies to account, to understanding how government agencies in India and China have ‘bundled’ climate change with other priorities in order to make change happen, to getting more High Net Worth Individuals paying tax in Uganda, and through this we’re learning more and more about how TWP – by local actors, sometimes supported by external actors and sometimes not – helps make progress on specific problems. The size of the impact from better evidence on this is potentially vast.

Without TWP, PDIA can build states’ capability in ways that could do harm. Capable states can do unspeakable things, after all.

Without TWP, adaptive programmes can do the wrong things flexibily and adaptively: starting with solutions and not problems, hitting the political will wall and, of course, doing harm.

Without TWP, PEA could be just another programme input that sits on a shelf and doesn’t actually do anything. And, of course, rubbish PEA, done poorly,aimed at the wrong level or coming at the wrong time could lead to poor decisions that (one more time…) do harm.

As an incredibly experienced aid practitioner and diplomat once said in a TWP Community of Practice meeting, ‘TWP should be firstly about not f***ing things up’. By getting our TWP alphabet right, and having a good evidence base, we can bring focus back onto state capability, think more clearly about how best to do adaptive programming and to use PEA, and we can stop f***ing stuff up. That seems pretty important to me.

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.


26 Responses to “What we’re missing by not getting our TWP alphabet straight”
  1. Super helpful summary – in particular the distinction between PEA and TWP. In my experience local implementation teams are always baffled with why we make them trudge through so many PEAs, while they think ‘Yes – I knew all that, I do watch TV, and speak to friends, and have lived here all my life’ – super helpful to be able to distinguish when you do a PEA for outside interveners, and what’s useful for already engaged implementers. One other clarification, the Nigeria oil sector project was much more about engaging a wide and diverse network of policy entrepreneurs than about ‘civil society’ – as your article powerfully reminds us, I do worry that inaccurate use of language can be dangerous in perpetuating entrenched methods rather than the genuine learning and change that’s needed.

    • To Sachin’s point about making locals trudge through so many PEAs, I’ll caution that just because staff are local and pay attention to local politics, doesn’t necessarily mean they understand why things are the way they are on all problems. Problem-level PEAs allow for deep dives into unpacking why these problems exists and persist. The PEA process requires triangulation of findings and encourages debate and questions assumptions. But, I agree that doing a baseline problem level PEA is one step in the arguably more essential on-going TWP learning process.

    • Heather Marquette

      That’s so true! A hugely impressive, complex programme involving a number of state and non-state actors that I’ve simplified for (hopefully!) flowing writing in a blog.

      Hopefully your comment will encourage folks to click on the link and read the paper!


    Hah. I live in a world of (mostly, these days, humanitarian) acronyms and fully appreciate their value despite the alphabet soup, but despite reading this blog regularly had to read through both the mailing list email and half the article to remind myself what TWP actually stood for. Perhaps explain the critical three letter acronym (TLA) in the heading? 😉

  3. Sina Odugbemi

    In my experience the focus should be what internal reformers need to be effective. And when top technocrats are asked to lead reform efforts in sectors you’d be amazed by how many of them disdain politics or the internal workings of their own political systems. The smart ones know how to advance their careers but getting a complex reform through is another matter. Whatever we call these approaches what is relevant is whatever helps local reformers to be effective. Which is why a needs assessment is a must before you run workshops designed to help them. Still, very helpful piece.

  4. I agree this is a very useful summary of the acronym soup. Especially enjoyed how TWP “has helped local reformers get from political won’t to political will.” TWP is the meta-concept. And, I agree that conducting a PEA does not necessarily equate to TWP; however, PEA is a tool/method/approach that can help you do get there; as described in USAID’s latest (2018) guidance on this topic “Thinking and Working Politically through Applied Political Economy Analysis,” PEA is a tool that can help development practitioners (or any one for that matter) think and work in ways that are more politically aware. Similar to PDIA, PEA is about building in learning and paying attention to how politics affects reform efforts and development investments regardless of who is behind them. Similar to Every Day Political Analysis, Collaborative, Learning and Adapting (CLA), yet another acronym courtesy of USAID’s Learning Lab, can be used as a structured approach to iterate PEA to increase the chances development practitioners are really engaging in TWP.
    We are using these acronyms in order to justify a bit more time to think before we spend. We are also asking to take more (politically informed) risks or ‘small bets’ on the local actors who are already TWPing to bring upon change.

    While we (development practitioners) all face the imperative to spend or risk having the funds redirected elsewhere, this movement/paradigm change that we are all part of is an attempt to spend more wisely. To make our investments sustainable, we know, we have always known, that we need to spend more time understanding the local context to know how to choose which local efforts to support. This acronym soup is our newly emerging vocabulary to enable us to articulate this mindset and to link with other development thinking, such as Market Systems Development, promulgated by the BEAM Exchange and Marketlinks.org to cross pollinate ideas to further improve our practice.

    • Heather Marquette

      Hey Marc, great points! I reckon that the way forward is to differentiate these ‘acronyms’ in a way that enhances their complementarities, where seeing them as totally discrete would be as bad as seeing them as being interchangeable. As the ‘alphabet soup’ expands, it would be great to think about how everything fits together better, rather than inadvertently competing for space/top acronym.

      It’s entirely possible to think & work politically (lowercase intentional) without conducting PEA, but political analysis can be essential in order to do this effectively in many situations. I was in a university meeting earlier and suggested twp to think about finding support for a possible strategic investment. Some stakeholder mapping, or everyday political analysis, would be really useful there. For some reason, PEA’s become a bit of an orphan, but it’s so important to better understand what’s happening with it.

      • Heather, I agree that twp can be done without formal PEA as you’ve noted. I think the reason PEA stands out is because it is a process or a ‘thing one does’ to make technical programming (hopefully) have greater impact. This means the standard development program model doesn’t really have to change much thereby allowing funds to be spent in accordance with budget cycles. If we were really to apply twp to programming, as some projects have, the business model for the industry may have to change, which is, of course, very political. However, as you (and Dasandi, Laws, and Robinson) note in your recent (June 2019) article, there isn’t a strong enough evidence base to prove for TWP has sufficiently improved development effectiveness. I do wonder if there is more ‘twp” (by other names and done in other sectors) initiatives than one might think?

  5. Jay

    Was wondering where DDD fits in, Heather – is that essentially the same as TWP, or does it have its own home in the adaptive acronym universe?

    • Heather Marquette

      I’ve never been part of DDD, but I don’t think it’s live any more. There isn’t a website, though the manifesto still lives on the ODI website if you look for it. I think Matt et al are working on BSC and ODI on adaptive management. But I could be totally wrong!

  6. Thanks for this Heather. I particularly appreciated the points about what PDIA, PEA and Adaptive Programmes can miss if they don’t include a TWP perspective.

    In some ways, I’d like to see that argument taken further, with a more positive spin and practical examples: so, supplementing perspective A with perspective B can be helpful in this way, and has been in this context, and supplementing C with A can be helpful in this way, and has been in this context. Or, to take that a step further, to see more analysis of the (implicit) theories of change (and therefore understandings of how change happens) that the various approaches align with/can help to put into action. I guess that point relates to your paper on “What does the evidence tell us about thinking and working politically?” https://www.cogitatiopress.com/politicsandgovernance/article/download/1904/1904

    Here’s an earlier attempt by Dave Algoso and I to get to grips with the alphabet soup by considering how central adaptive learning is to different approaches, and how the various approaches tend to deal with politics https://oxfamblogs.org/fp2p/where-have-we-got-to-on-adaptive-learning-thinking-and-working-politically-doing-development-differently-etc-getting-beyond-the-peoples-front-of-judea/

    And for those who haven’t yet seen it, I’d strongly recommend this Nigeria-focused analysis by Kate Bridges and Michael Woolcock, which is a practical example of how trying to take an adaptive approach played out in a particular context. https://openknowledge.worldbank.org/handle/10986/31907

    Finally, if any readers want access to my tagged Evernote stash of 1000+ articles on various adaptive approaches, feel free to ping me.

    • Heather Marquette

      Hi Alan! Blame Duncan, who (quite rightly…) wouldn’t let me babble on for another few hundred/thousand words. I loved your blog with Dave, and this one that Tom Parks wrote right afterwards – https://www.dlprog.org/opinions/fragmentation-of-the-thinking-and-working-politically-agenda-should-we-worry. But I’d just spent 3 days in a room full of governance specialists using the acronyms interchangeably in unhelpful ways. This is definitely my final word on this, but I’ll hold off betting the farm on it being the final word…

      Something I’m excited to see are the PhDs being written on all of this right now. It could be that someone out there will bring greater clarity than any of us trying to disentangle have been able to do. I’ve been interviewed by a few International Relations grad students, for example, who have different ways of mapping this field. It’s interesting!

        • Jamie Pett

          To respond to Alan’s request to take this further… I’m right now editing an ODI working paper looking at how these different approaches/ways of thinking can be combined with their private sector counterparts (agile, lean startup, design thinking – more language for the soup!). It’s partly an attempt to break down the adaptive – politics – innovation siloes. Hopefully out soon!

  7. Very helpful summary. I am still struggling to introduce PDIA and AM in my own administration.
    However, I tend to find that donors can and should use PDIA. Namely for their own procedures. More and more, donors are just another administration struggling to become/stay effective. Isomorphic mimicry is not only pushed on developing nations. A lot of the OECD guidelines is based on Isomorphic mimicry. Using PDIA would not only help them to become more problem driven, but it would also make them more open for the problems faced by their partners.

    • Heather Marquette

      Colleagues of mine in Birmingham run a Public Service Academy, and one component is about the ‘21st century public servant’ – https://www.birmingham.ac.uk/schools/public-service-academy/about/twentyfirst-century-ps.aspx. A lot of what they write about really resonates even though it has nothing to do with aid/donors. I think we don’t reflect enough on what it means to be a civil servant in this space, and you’re absolutely right that PDIA offers potentially important lessons, especially if we stop thinking of ‘donors’ as a discrete analytical category.

      • Heather Marquette

        Just following on from this, there’s a growing evidence base around the civil service that looks at some of the barriers to agile policy making and implementation. It’s much more specific about the ‘authorising environment’ and about domestic politics. As far as I can tell, in most adapt dev papers the authorising environment tends to be a ‘black box’ written about by people who don’t necessarily understand the civil service (which covers bilaterals and, up to a point, some multilateral). My worry in this space is that the authorising environment will change around adaptive management but in a way that’s completely apolitical, and then the window of opportunity will close. And then we risk doing bad things (or the wrong things) flexibly, eventually calling out ‘but where’s the politics???’. The field does have a history of depoliticising a range of agenda…

  8. I think the bigger problem here is the use of acronyms, period. It permits us to be sloppy because we don’t have to look at the actual words. It also is exclusionary to those who are not ‘in the know’, so does not promote equitable access to ideas.

    • Irene Guijt

      Very helpful summary, Heather – and excellent comments. Particularly important to come back to PDIA as state-focused, not donor-driven. The donor search for new ‘hows’ is important but can lead to bureaucratisation and formula creation that locks in protocols rather than mindsets and principles. I’ve been chatting with Duncan what, for example, adaptive management would look like for a civil society coalition not driven by a donor contract and governance requirements that shape the nuts and bolts. (Ditto Fred on the problem of acronyms being excluding, in-speak catchalls that allow for fuzz and buzz.

  9. Suvojit Chattopadhyay

    What would really help to take the knowledge on these subjects further is to understand how governments and private sector utilise these approaches and tools (most likely without ever using these names and acronyms). This would help aid workers engage better with the wider context, beyond their immediate counterparts.

  10. David Grocott

    I would argue that, by and large, all these approaches/tools/ways of thinking are attempts to deal with the fact that development (i.e. driving change) typically involves solving complex (hard) problems, that require individuals interested in solving them to (1) understand the prevailing situation as best they can, including the power dynamics at play [TWP & PEA]; (2) design best fit solutions to issues that people care about [PDIA]; and (3) implement ‘solutions’ flexibly and with a focus on learning [adaptive management]. Focusing on only one aspect of the alphabet soup risks missing the big picture.