Where have we got to on Theories of Change? Passing fad or paradigm shift?

Theories of change (ToCs) – will the idea stick around and shape future thinking on development, or slide back into

My favourite ToC cartoon
My favourite ToC cartoon

the bubbling morass of aid jargon, forgotten and unlamented? Last week some leading ToCistas at ODI, LSE and The Asia Foundation and a bunch of other organisations spent a whole day taking stock, and the discussion highlighted strengths, weaknesses and some looming decisions. (Summary, agenda + presentations here)

According to an excellent 2011 overview by Comic Relief, ToCs are an ‘on-going process of reflection to explore change and how it happens – and what that means for the part we play’. They locate a programme or project within a wider analysis of how change comes about, draw on external learning about development, articulate our understanding of change and acknowledge wider systems and actors that influence change.

But the concept remains very fuzzy, partly because (according to a useful survey by Isobel Vogel) ToCs originated from two very different kinds of thought: evaluation, (trying to clarify the links between inputs and outcomes) and social action, especially participatory and consciously reflexive approaches.

At the risk of gross generalization, the first group tends to treat ToCs as ‘logframes on steroids’, a useful tool to develop more complete and accurate chains of cause and effect. The second group tend to see the world in terms of complex adaptive systems, and believe the more linear approaches (if we do X then we will achieve Y) are a wild goose chase. These groups (well, actually they’re more of a spectrum) c0-exist within organisations, and even between different individuals in country offices.

One confusing consequence is that the term ‘Theory of Change’ describes both a formal document and a broader approach to thinking about development work. For some Theory of Change is a precise planning tool, most likely an extension of the ‘assumptions’ box in a logframe; for others it may be a less formal, often implicit ‘way of thinking’ about how a project is expected to work.

Craig Valters' favourite ToC cartoon
Craig Valters’ favourite ToC cartoon

And when used as a planning tool, ToCs are being implemented in addition to logframes rather than instead, and grant recipients are reporting to donors against the latter, with predictable results: ‘At end of day, we have to prioritise logframes – that’s what you’re being graded against’.

I came away feeling that this is all generating so much confusion that we actually need to find new words to describe the different understandings and approaches to ToCs.

For a start, it might help to distinguish between Theories of Action, which focus much more on whatever intervention is being discussed, and Theories of Change, which unpack how the system is changing (or might change in future) even absent our intervention. In my experience, the organizational imperative to do stuff, raise money, demonstrate impact, or just be active means that people spend far too little time studying and understanding the social, political or economic system before intervening (books like Portfolios of the Poor are a brilliant exception).

Enough ToCs are now being discussed and included in project plans and implementation for us to start seeing how they work out in practice. The leader on this is probably The Asia Foundation, with its collaboration with LSE, (discussed last year in Craig Valters’ excellent paper) but others are joining in – see the Comic Relief review or this 2012 study by CARE of the use of ToCs in 19 peacebuilding projects in Uganda, Nepal and the DRC.

Some of the more interesting benefits of ToCs identified during the seminar included:

  • Opening up a conversation with donors about more realistic programmes and aims, allowing both donors and recipients to challenge unrealistic expectations and over-claiming
  • Making explicit a lot of the implicit/tactic knowledge and analysis that underpin what aid agencies actually do (especially important given high levels of staff turnover – see below)
  • Motivation: ‘reminding everyone why they are doing this’
  • More quickly identifying things that aren’t working, so we can stop doing them (and spend the money on something else) – this is an aspiration, but Craig says he has so far seen few signs of it happening.


Some other impressions from the discussion:

1. It all comes down to people: ‘The ones who like ToCs are the development cynic types kicking against standard agendas/normative claims, who like having a language that forces people to think harder. It’s often those with more research/academic backgrounds. The others are not rejecting ToCs, but just not investing so much in them – they have lots to be getting on with.’ At the moment, all the success examples seem to require exceptional ‘development entrepreneurs’, but they are in short supply – what would a ‘where there is no Jaime Faustino’ approach look like?

2. Down with diagrams? OK, I am not a visual person (at school I was always rubbish at art), but I really hate the diagrams that

Imagine being presented with this on your first day
Imagine being presented with this on your first day

seem to accompany ToCs (see example). Others disagreed – simple clear diagrams can be a great comms tool, much more accessible than 40 page narratives. Trying to cram everything into a single diagram by just adding more and more boxes and arrows may be helpful for those participating in the exercise, but within months, they will all have left (staff turnover again) and the incomers will stare at the resulting visual spaghetti in terror and incomprehension. Maybe we should develop a special kind of ToC diagram that self destructs after 10 seconds, Mission Impossible style, and has to be endlessly redrawn by each new staff team?

3. How did everything get so top-down? think one cause lies in ToCs’ links with Political Economy Analysis: ‘a brief moment of omniscience when you can do a PEA and understand power relationships’. The ‘you’ in that sentence is always a consultant, a researcher, or an aid agency staffer – never a community or a (growl) ‘beneficiary’. Where are the bottom up, participatory ToCs? Why are IDS and Robert Chambers so absent from this discussion? ToCs talk a lot about power in other settings, but ignore power relations in our own organizations – what if they become ‘just another corporate stick to beat people with’? NGOs like Peace Direct and CAFOD are doing some bottom up work on this, but we need to be much more vocal.

4. Maybe we do need a toolkit after all: toolkits get a really bad press – restrictive, ignoring local context and tradition etc etc. But what’s the alternative? I fear it is the overpaid consultant or other expert flying in to explain all (see previous point) – ToCs could end up excluding local staff, partners and communities even more than the traditional approach. A good toolkit could include and reassure staff who don’t spend their whole lives on this stuff, but the question remains how to design it so that it doesn’t impose outside templates and preclude imagination, innovation and adaptation. Seems to me a toolkit would need to set out the kinds of questions to ask in designing a ToC, and the process for then implementing and adapting it. The Asia Foundation seems to be developing one, involving teams keeping and updating timelines of the evolving ToC, and having regular ‘time outs’ to revise it.

A close second
A close second

As for where we go from here, I think we have to consciously differentiate between ToCs as a bigger, better blueprint, all designed up front, and ToCs as a compass for helping us find our way through the fog of complex systems, discovering a path as we go along. But that will only happen if we can help people overcome their hunger for the certainty, however illusory, provided by linearity and up front design. At least donors are up for this – one DFID rep said she wanted ‘truly strategic partnerships in which not knowing is seen as a strength’.

Meanwhile, if ToCs are to endure, we need to demonstrate some quick wins from using them, and make it easier to do so. And NGOs and others need to step up and join TAF in offering themselves as guinea pigs.

[With thanks to Craig Valters for voluminous comments on my (much shorter) first draft]

Update: turns out that, as with all the most important questions of the age, South Park has it covered. Here’s the 9 second intro to ToCs [h/t Martin Clark]

[youtube height=”HEIGHT” width=”WIDTH”]https://www.youtube.com/watch?v=tO5sxLapAts[/youtube]

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.


33 Responses to “Where have we got to on Theories of Change? Passing fad or paradigm shift?”
  1. Ross Clarke

    A gripe of mine is that donors now seem to require a ToC for the proposal process (good) but provide no guidance on how this sits alongside the obligatory logframe (confusing).

    The mainstream status quo remains that logframe linear thinking trumps all (especially in the donor community), but we are now asked to add a more useful change analysis without guidance on how the two sit together. You could make a strong case that the two approaches are contradictory.

    So I echo the call for a deeper understanding on what we mean by theories of change. Well overdue. But I would add that we need greater clarity on how ToCs sit alongside logframes.

    Given the latter is unlikely to disappear any time soon, we need to determine how (if?) we can take the best from both approaches.

    We need to build a more flexible, politically grounded programming model, but if ToCs are an add-on rather than being fully integrated into program design/implementation, this runs the risk of creating confusion rather than achieving the change in how we do business that is needed. A ToC for how we fully adopt ToCs anyone??????????

    • Duncan Green

      Thanks Ross, this is clearly an issue – if ToCs are to be useful, they have to be more than a ‘nice to have’ extra. sitting alongside the ‘must have’ logframe, because, as people pointed out, that means the logframe always triumphs in the end, in terms of time, energy and £. I don’t think anyone suggested ways to overcome this, but maybe a return to the old intention of the logical framework approach, with the ToC providing some of the assumptions thinking, and an explicitly iterative, adaptive approach (e.g. time outs for reflection and adaptation every X months, with external facilitation if necessary) during implementation? Could some more practitioner types come on on this please?

  2. Catherine Dom

    Two sentences struck me in the fairly long post. First, that “it might help to distinguish between Theories of Action, which focus much more on whatever intervention is being discussed, and Theories of Change, which unpack how the system is changing (or might change in future) even absent our intervention.” and that “we have to consciously differentiate between ToCs as a bigger, better blueprint, all designed up front, and ToCs as a compass for helping us find our way through the fog of complex systems, discovering a path as we go along.”

    I agree on the distinction, but as practitioners do we not need both, ToC, and ToA rooted the ToC i.e. in a better understanding of the system and how it’s changing? And then, in addition, very regular review and revision of both ToC and ToA (in the “rapid feedback loop” mode of DDD, just to add on jargon…)?

    And, like Ross, I’d like someone to explain how logframes sit with ToC (and now ToA).

  3. Amanda Harding

    Why is it that the ToC conversation consistently separates the development wonks and the researchers? Researchers engaging in Research-for-Development have moved from talking “impact pathways” to ToC over the past six years in a genuine attempt to shift towards development outcome based research. This is transforming research design (ToC thinking) and how, with who and for whom research is undertaken (ToC-in-action). This shift (or paradigm change) is a struggle, as evidenced by the CGIAR whose research programmes perfectly reflect these dilemmas (see, the CGIAR Challenge Program on Water on Food http://waterandfood.org/our-approach-2/) but there is increasing evidence that ToC thinking is key to achieving development impact.
    Isn’t it time for these communities to come together in a joint conversation on what works, where and why? This would be true ToC thinking.

    • Duncan Green

      Good point Amanda, I am often reminded of things like the increasing use of inception phases in research projects, where funders accept that the first year is about scoping, deciding what/how to do the research, and fund it. Not yet the case in many development projects, where you are supposed to do all of that beforehand and come up with a nice neat project to finance. Some ‘purposive floundering’ needs to be included as a matter of course, surely, as well as periodic time outs for reflection and course correction?

  4. Saskia Daggett

    Fascinating stuff. I couldn’t agree more on your comment about these processes being top down. In the Africa Climate Change Resilience Alliance we went to villages and chatted to people first before coming together as a staff team and asked them about climate change and mapped decision makers together and asked them what they wanted to see. This radically altered what we would have discussed otherwise – particularly on gender. It became terrifying stark to us how much more we needed to work on issues of household conflict and gender based violence. It also gave us a new focus on consciousness. The women I talked to were blaming themselves for climate change??!!! The geopolitics of the situation were completely unknown. We are struggling to know what to do about this – but are trying (new partnerships with the Institute for Social Transformation, designing new research, discussing with colleagues to address some of these issues etc) – which we wouldn’t have been otherwise. People are nervous about wading into communities and making the best use of people’s time but surely we must be able to find better ways of being rooted, engaged and responding to the reality of people’s lives and experience?

  5. Craig Valters

    Thanks for this Duncan, it’s a great set of reflections on the workshop. A few quick thoughts based on the comments so far from Catherine and Ross:

    – I think that logframes and Theories of Change (at least in the way I understand them) are contradictory. The former emphasises linear and predictable pathways of change, and the latter does (or should) not. Their co-existence in donor demands, I believe, is a result of the contradictory ideas and forces at work within donors themselves. As we know, DFID (for example) is not a homogenous entity. If Theories of Change do become (or are) an ‘add-on’ then I doubt they will be too helpful in the long run.

    – My suggestion is that organisations try and take the opportunity presented by the relatively strong authorising environment in DFID (among some staff) to pilot new Theory of Change approaches which do take into account unpredictability, the need to adapt programmes over time and local understandings of change (no easy task I know!). If we can develop examples of useful approaches, it may go a long way to ensuring it stays relevant and helpful in the longer term.

    – Catherine, I agree on that Theories of Action are part of Theories of Change – and I think to remove ‘us’ from the Theory of Change process entirely would make this lose a lot of meaning for practitioners – although I do agree with Duncan’s point that we need to do a thorough, ongoing analysis of power/systems/context.

    As Duncan has highlighted, anyone who wants more information on this workshop can find the agenda, summary and presentations here: http://www.odi.org/events/4194-theories-change-development

    Over the coming weeks we’ll be writing up a more in-depth discussion paper for publication too.

  6. Isabel Vogel

    Nice to hear thoughts and reactions on the Asia Foundation workshop (which I couldn’t attend due to prior commitments). Good to see the outputs coming from this.

    My comments (apologies for a long post) are that there is still a huge demand from all sorts of organisations – from donors to small NGOS – for help to work with a theory of change approach. Even if prompted by top-down, external pressures, people still find it an insightful process that supports improvements in their work, stakeholder relationships and learning.

    But I totally agree about the confusion, and wanted to respond to the comments about the inherent tension in working with both ToC and log-frames. My assignments have become less about one-off ToC processes, and more about building MEL systems and organizational learning around ToC (inc. regular review and revision). This includes linking with log-frames.

    There’s too much to go into here, but as a basic principle, the practitioners that I collaborate with all concur that you have to be clear about the different purposes of ToC and log-frames: ToC is for learning about change for people – giving you a large frame to keep critical thinking going about ‘how, why and so what has really changed and for whom?’ The ‘compass’ for navigating as Catherine puts it nicely.

    Log-frames give you a summary framework, and the focus is mainly on your programme – key bits of your ToC thinking go in to contextualize the programme, but the purpose of a log-frame is to keep a focus on the ‘what’ – your strategic options, implementation and immediate results that form your contribution to the change process (informed by your Theory of Action.) However, as others have said, you need both – your ToC should help identify critical assumptions to keep front-of-mind and track through your log-frame.

    So the two frames have different purposes, but in my view, they do relate, and you can build a MEL process that encompasses both – Maureen O’Flynn (http://www.intrac.org/pages/en/maureen-oflynn.html) proposes working with ‘critical areas of enquiry’ which relate to the ‘how, why and so what’ and the assumptions that are critical within your ToC to form the basis of your strategic learning, plus an indicator-based analysis which tells you about your emerging contribution (and more testing of assumptions relating to your intervention and ToA). And applying this learning to regularly refreshing your ToC and tweaking the log-frame, as is recommended, or at least having an informed conversation about doing so with your donors.

    Amongst ‘ToC practitioners’ (if there is such a group – ToC is still only an approach), there is a lot of ongoing practical work to distil the essence of a ToC approach and critical thinking (in a toolkit-y type way), but still encouraging people to ‘hack ToC’ to find what works best for their purpose in a given point in time. The Hivos Action Learning Group on ToC is one group (resource portal at http://www.theoryofchange.nl/).

    Building on our learning from the Hivos process, Irene Guijt (https://www.linkedin.com/pub/irene-guijt/6/826/213) and I recently developed a ‘systems-informed ToC’ training module for Hivos staff, with ‘ToC essence’ plus ‘hacking options’ (there will be various guidance products coming from this in the next couple of months – a possible tool-kit?).

    As for what next, I expect it would be really helpful to link up the different strands of practical work to keep ToC useful, now that there are established portals for sharing this stuff…

  7. Albert van Zyl

    Surely we are becoming obsessed with ToCs? Isn’t the real question rather about whether we are learning, adapting and getting better and making accountability happen? You know the Zen story of the master who pointed at the moon and berated the disciple for staring at his hand…

  8. Isabel Vogel

    Yes, of course that is the real question, Albert, but I think that people like ToC as an approach because it helps them to learn, adapt and get better at making accountability happen, to have a coherent learning process to help them get closer to that moon… ToC is the learning process, not the product. And it is only one approach of many.

  9. GA

    Echo Isabel’s comments above. The fact is that many development programs *still* do not clearly articulate the purpose of their activities. They cobble together trainings with little strategic directions. Tools are sorely needed to help and encourage articulating these purposes and tying it to evidence. Theories of change and log frame are some of the better tools in that regard (if done right) right now.

  10. David Couzens

    For a while I used “Philosophy of change” to describe the more macro level hypotheses about how change happens and “Theory of Change” to describe the application of that generic philosophy to a specific context and issue – why and how do you think change will be brought about in this particular situation. The phrases are pretty horrible but I think the distinction matters.
    Really dislike the either/or discussion about logframes and ToC. I see it as a natural flow from problem/context analysis through ToC and then to the logframe. if you have done the first two then the logframe is a doddle as all the thinking has already been done.
    I also think it is important to remember that your ToC is your working hypothesis – it is almost certainly wrong in parts but having it written down gives you a more structured way to learn and adapt – ToCs should evolve throughout the life of a project. I think they help deal with complexity if seen this way.
    On diagrams I think they are essential but they should be a visual aide for the written bit and not an attempt to capture everything – if they do they rapidly become horrific, unsightly and unreadable!

  11. Sean Furey

    Good discussion. I find a ToC useful for organising a network (RWSN) because different topics are at different stages of maturity and face different challenges (e.g. handpump improvement is boring and old-hat; mobile phones are shiny and suffering from hype). Log Frames just don’t work in this context (do they work in any context?), but I have seen some ToCs that just look like Log Frames turned through 90 degrees on the page, which makes me wonder how well the concept of ToC is understood. I think ToC can bring together the right mix of struture and flexibility to give the approach lasting appeal and usefulness.

  12. Søren

    A theory of change, without a theory of being, is merely a statement of belief.

    The thing is, we all have theories of being (we’re not amoebies after all). We’re just commonly unconscious about them. Different theories of being -> very different theories of change.

    • Søren

      Accordingly, I’d like to see the funding agencies being explicit about THEIR theory of change. The exercise at the project level then becomes one of putting context-specific flesh on the bones and truthing said theory.

    • Duncan Green

      OK Soren, I want you to give me a specific example of a theory of being for a particular aid programme, and how it differs from a ToC. In words a non wonk would understand (so ontology is right out). Over to you.

      • Søren

        I’m not arguing for programme specific theories of being. That would seem schizophrenic. I’m arguing for being conscious and explicit about our perceptions of how the world works. The first step to changing something is understanding it.

        What I’m trying to say with my polemic is that there is no such thing as a theory of change without a theory of being. Whether realized or not, your type one logframes on steroids ToC is based on a theory of being where the whole is the sum of its parts – like a clockwork is the classic Cartesian example. The process-oriented ToC, on the other hand, is the natural outcome of a theory of being where the whole is an emergent property of its parts. Here a living organism is the classical example.

        Whether you abide to one or the other will affect how you approach any programme.

        A water supply project designed according to the human rights-based approach would be the work of someone with a clockwork theory of being. The assumption is that you get a law passed in parliament securing the right to water and with identification of rights holders and duty bearers. Then all that’s left is making that information available to the public. If your theory of being is on the living organism side you’d probably opt for something more like your convening and brokering approach.

        The HRBA camp will probably protest my example but I think that just supports my belief that theory of being should be centre to any theory of change. A theory of change which isn’t conscious about its theory of being is much more likely to be assumed rather than based on empirical evidence. Hence the ‘theory of change without a theory of being is a statement of belief.’

        • Jay Goulden

          Good reply, Søren – have just been doing some work on theories of How Change Happens, which might naturally be called “theories of change” if the ToC wasn’t a specific, different thing in the development industry! So, I have been referring to them as Theories of how change happens, which is a little inelegant I fear (David’s comment earlier referred to “Philosophy of change”, but that doesn’t quite feel right either).

          So, we have:
          a) The theory of being/philosophy of change/theory of how change happens, for overall theories explaining how we believe change occurs (or what Duncan called systemic meta-theories and archetypes in his blog of 21 June 2011 -http://oxfamblogs.org/fp2p/what-does-a-theory-of-change-look-like/)

          b) A Theory of Change for the specific system we are working on and the change we wish to see in it, which may include the intervention hypothesis and logic of the organization (unless that is a separate Theory of Action)

          And sometimes the LF – and it all has to be evaluable, as Rick points out…

          Anyone got a better name for a), to reduce the confusion?

  13. Jay Goulden

    Having a Theory of Change, a Theory of Action and a logframe seems a bit of an overkill, especially for a (relatively short-term) project. With the teams I’ve worked with, I’ve found Theories of Change more useful at a higher level than the project, such as for a Long-Term Programme or overall strategy, serving as broader guidance for projects that then fit within and contribute to the Theory of Change. As well as covering the desired changes in the overall system, we also included in the ToC the organization’s strategies to contribute to those changes (the theory for our intervention or action, that you refer to), thus avoiding the need for a further diagram for team members to get their heads around. Projects then develop their own planning tools – usually a logframe – fitting within the overall Theory of Change.

    If the ToC is regularly reviewed and adjusted, based on changes in the context and learning on progress seen and the validity of the hypotheses that are part of it, then it would seem to be very much compatible with learning approaches, and working in complex dynamics. That would enable them to serve as overall guidance and compass, rather than the blueprint you fear. And in that sense, I do hope they aren’t a fad, but useful processes and tools to improve the quality of our thinking, planning and reflection. Perhaps not a paradigm shift either, but a significant improvement on existing planning processes.

    A quick defence also of the all too easy straw person that is the logframe: while logframes like any tool, can be used badly or unimaginatively, they are not incompatible with adaptive programming and learning. The SAVI example in Nigeria seems a good case, with a relatively simple Theory of Change used to track overall progress, as well as a logframe that was used flexibly based on changes in the context and learning. It was apparently modified at least 12 times, and had broad outcome indicators (such as “# of demonstrable changes in policy and implementation by state government in response to public demand where there is evidence of attribution to SAVI’s approach”, around which evidence was gathered using Outcome Mapping) – see pages 2-3 of the December 2014 workshop on Monitoring and learning in politically smart and adaptive
    programmes, at https://files.ctctcdn.com/5ca16d18101/7298dacd-a030-4849-8234-5b6333ef10de.pdf. There must be plenty of other such examples beyond SAVI, surely?

  14. Jess Dart

    Thanks for the fascinating discussion. I am responding here from Australia, and I am a ToC facilitator who works both with Australian contexts and internationally. In my work I rarely get asked for a logframe — it seems to have been dropped. The Australian Aid program no longer requires logframes at all – instead they mandate program logic/Toc. Amazing as it sounds the sky did not fall in. We just use one framework (ToC/logic) and have been happily and un-confusidly using this for several years now. Many domestic programs never used logframes at all – it really is an aid thing.

  15. rick davies

    Duncan has a hangup about diagrams, it seems. If the conventions used in the diagram are clear that helps a lot. Social Network Analysis software can also help by providing simplifying filters that enable a focus on the key elements of complex network diagrams. People just need to become a bit more network- literate. You cant complain about excess linearity if you cant be bothered to learn about networks.

  16. Jake Allen

    Considering this site’s title, it seems that the thing that’s missing is the understanding of how TOCs and logframes fit within the power relationship betweeen donors and contractors/suppliers/partners.

    Quite often the decision between sideways logframe v beautiful encpasulation of dynamic change either comes from what the donor (really) wants, or importantly what the contractor thinks or knows that they want. The latter is very much a part of reinforcing the former.

    If we want TOC and LFs to work better together we really need to do more to show this, and to push for as much space and resource as possible to do this. And this leads to a second point, that both of these things are tools, and so it’s silly to expect that they will all by themselves transform how things are done (which is why TOCs have so far been so underwhelming).

    So recognise the needs of your donor, but find the balance between what you think is best in terms of design and use of the TOC. Most advisers i’m meeting at present are very much up for finding this middle ground. And remember what the purpose of a theory is – setting out a model and finding evidence to test it. How long should evidence be collected – dunno, but a year or half a programme cycle seems reasonable.

    My final bugbear on TOCs is how much they end up being this is our aspirational vision on change, totally divorced from evidence of what actually works, or how change has happened before. I realised this is part of the process of taking ideas from a Business case, say, and turning them into a reality that is manageable, but if we don’t try to do these things we can’t expect that things will much improve.

    • rick davies

      Re “My final bugbear on TOCs is how much they end up being this is our aspirational vision on change, totally divorced from evidence of what actually works, or how change has happened before”
      My sympathies…
      My modest requirement for any ToC, put forward during the recent TAF/ODI hosted meeting on ToCs, was that they should at least be evaluable. Some suggested criteria for assessing evaluability of ToC are in the presentation. See http://www.odi.org/events/4194-theories-change-development

  17. Craig Valters

    Jake and Rick – agree with lots of this. Probably my biggest bugbear with Theories of Change is also that they focus too much on working backwards from the ultimate super-goal of the programme (some form of world peace, dressed in donor-influenced language). There are, of course, world-view type assumptions embedded in organisational outlooks and programme decisions, which are important to interrogate. But if Theories of Change are to be useful – which for me means creating a space for honesty, critical reflection and shifting strategies throughout programme cycles – then they have to be based on tangible, locally-derived understandings of how change happens. Otherwise this whole process becomes too inward looking and not concerned enough with improving the lives of people as the ultimate end goal.

    More broadly, interesting to see the logframe/ToC debate. I suppose it’s not that they are inherently contradictory in purpose. But in my research with The Asia Foundation it became clear that when a Theory of Change approach is welcomed, it’s because it can be much more liberating (intellectually and from a ‘results’ perspective) than using a logframe. And where logframes were not mandated in the Philippines (which was on a DFAT programme rather than DFID), Theories of Change were used as part of a broader iterative and adaptive programming approach to good effect. Good to note that yes, it was not because of a Theory of Change approach that this iterative, adaptive programming happened – but it was a useful ally to the overall strategy. I put a few thoughts on how a Theory of Change approach can work with such programming here: http://asiafoundation.org/in-asia/2014/12/10/can-theories-of-change-help-us-do-development-differently/

  18. Kate Higgins

    Thanks to Duncan and to all those who have commented on this. A really useful piece. A few months ago, I started the process of putting together a ToC for an initiative I am now working on – the DataShift. I was struck by how little guidance was out there on how to do this well. I read nearly all the things that have been referenced here, have a research background, had a little time to do some thinking, worked with partners and colleagues, and we came up with this draft for broader comment: http://civicus.org/thedatashift/wp-content/uploads/2015/03/DataShift-Theory-of-Change-For-Consultation.pdf. A few reflections on Duncan’s piece and the comments following:
    – There needs to be some sort of relatively easy to follow guidance on ToC. Sure – there cannot be a template – but if we think that this is a useful process for smaller initiatives and organisations with limited resources to do, there needs to be some guidance. We all know what the alternative – flying in a consultant for a few days to do it – will result in.
    – I think there are many types of ToCs. For us, the ToC was about identifying how our activities connected with our objective and vision (log-framey I know) but doing this in a collaborative way that takes into account risks, assumptions, context and politics.
    – The ToC process can be really useful. For the DataShift team and our partners, it is helping us get clear understanding and buy in between multiple actors across multiple continents about what we are about and how the different moving parts – some of which can control and some of which we cannot – fit together.
    – There is a big question of scale and resources. Not all initiatives and organisations are equal. Some will want to and be able to invest loads of time and money in this process, some will not. The ToC process, and potentially the product, will likely reflect this too. I would have happily spent weeks on the DataShift ToC, consulted widely, brought together partners for workshops, commissioned expert comment and input, and had it published to look super slick, but on balance, this was not the best way to use limited resources. Sometimes all you can afford is a Toyota rather than a Rolls Royce, and that is okay.
    – On the diagrams – I also prefer text but at the end of the day, this is the part that people have found the most useful!
    Bottom line for me is that for us, this was well worth the effort and resources. But a bit of guidance that synthesizes lessons and experience from others would be really valuable, particularly for organisations and initiatives that are doing this with limited resources.

  19. Søren

    A final quip:

    I’m sceptical of the attempt to separate all these terms.

    I tend to find it useful to think of ToC as a logframe which is explicit about- and adaptable to its implied theory of being (note that theory ≠ hypothesis). It’s the different theories of beings which produce these different manifestations. You can’t bridge them by calling them something different.

    It’s my hope, of course, that forcing people to be explicit and argue for the Theory of Being implied in their Theory of change would help weed out some of the wild ‘goose chases’. And I think that’s the more constructive path compared to telling someone who’s theory of change you find invalid that it isn’t a theory of change but a theory of action or something else.

    Realising this isn’t perhaps so far from Duncan’s comment above (April 16 – 8:03 am) perhaps we just ought to call it New-Logframe?
    And then a ps in regard to the loathed consultant. I couldn’t agree more that flying in a jetlagged consultant to sit in 30 meetings and produce a ToC in a week is an offence to any thinking individual. But having a consultant, who’s able to put things in new perspectives, and challenge some of the conventional wisdoms, while facilitating the process a theory of change should be, isn’t necessarily such a bad idea.


  20. Irene Guijt

    Important and nuanced discussions here – thanks. Following up on what Isabel Vogel said, HIVOS (Dutch NGO focusing on social transformation, the ‘harder to assess’ stuff, in general) has invested in developing its home-grown critical version of ToC since 2010, based on principles rather than protocols. ‘Critical’ simply because Hivos wants to avoid formats for compliance and has therefore emphasised question-driven discussions, and because power and gender and structural analysis are central. Since then we have developed through trial and error, in-country a set of question-driven core steps with specific methods or tools simply options. Hivos has focused on it as a thought process and hammering on about assumptions at every step. Diagrams are a by-product and vary greatly.

    We have not found ToC at odds with logframes but helpful prior to submitting a LFA matrix if required. The process involves a mix of backward mapping – from the lofty goals to a more tangible change outcome, as well as forward (what is needed for that change to occur). This links the Theory of Change level discussions to those about the Theory of Action. Hivos will produce detailed guidance notes over the next couple of months.

    What I miss in these discussions is the tension between the ‘liberating’ experience that ToC discussions can offer and having to navigate against the stream of the external and internal pressures that give priority to upward accountability. It has proven pretty darn difficult to develop organisational processes that keep revisiting and updating the ToC narratives and diagrams with evidence-informed insights as the work unfolds. Problems of using ToC for regular critical reflection will continue, I expect, as long as organisational and external pressures reward a proposal culture of predictable targets and an M&E practice that focuses on checking performance.

    Albert – yes, let’s not obsess about a term or a practice. But if it helps organisations think beyond the comfort zone of their ‘off the shelf’ strategies for change – a workshop or a training event or a campaign, great.

  21. Cara Winters

    Interesting post and conversation.

    Over the last few years, the Norwegian Refugee Council developed an agency-wide M+E framework based on a contribution approach to performance measurement. A cornerstone of our system is the use of Theory of Change at multiple levels of planning and M+E. We have not found it to be in conflict with log frames, since a log frame is mostly used as a reporting tool for donors. The ToC, in contrast, is the foundation for how we understand if or how we are making a difference.

    This is embedded in our agency calendar. Every program in every country office develops a ToC as part of the program strategy process. We also used ToCs at a global level to map what information to track across the agency, which is then used to trigger evaluations and inform management decision making.

    So I guess the lesson from NRC is that ToC is relevant if tied to an holistic measurement approach and embedded within a system.


  22. Lisa Faye

    Just rereading this post as our Governance Programme Manager here at Oxfam in Tanzania will be in Oxford to (among other things) give a talk on the evolution/emerging of the Chukua Hatua theory of change. It’s interesting to see how a ToC matures in long term programming. This theory of change continues to evolve as the team goes through various learning activities and difficult discussions. Still plenty of use in ToC for us and lots of participation in developing them… Perhaps my colleagues would think differently though!?

  23. Horacio

    What a great discussion, I came a bit late though. I did a google search with exactly this: “why is ToC outdated” and this was the first item that came out. Eye opener!
    I am looking for new approaches to replace those which I think are outdated, and I believe ToCs is one of them. Why? for all the arguments against it shown elsewhere.
    I work for a NGO in Ecuador, our donor are very flexible with the approaches we use. I am starting to implement OKRs (Objectives and Key Results) in our work. I wonder whether is there any experience here on using them for rural development programs. I mean, if Google, Intel, F. Gates, Bono, and others use them, why not to start replacing some old tools we still have in our garages (logframes, ToCs, etc.), right? Any thoughts?

Leave a Reply

Your email address will not be published. Required fields are marked *