Five ways of thinking about failure in sanitation

I’m inspired back to blogging by a thought-provoking workshop yesterday on Learning from Failure in Sanitation, organised by the UK Sanitation Community of Practice (SanCoP). Huge thanks to all who contributed. We will share notes and ideas going forward very soon. If you are interested in joining the network, email us and join our Linkedin group. For now, these are my thoughts on one of the key questions of the day: what do we mean when we talk about failure? There are at least five approaches to thinking about failure that I picked out from the discussions:

Acknowledge past failure – in a ‘safe space’ – in order to learn and adapt. For example, EWB-Canada’s Admitting Failure initiative aims to build community and create ‘safe spaces’ within and between organisations so that people can discuss and learn from previous failures.

Avoid unnecessary failures that we can predict in advance. Although being more open about discussing failures may be a good thing, participants emphasised that this relies on the quality of the conversations we hold. There is a danger of superficial ‘admitting failure’ for the sake of supposed transparency without deeper reflection and learning on whether the failure should have been avoided and how it could be in future.

Plan for ‘safe’ future failures through deliberate experimentation and innovation. This is Tim Harford’s argument in Adapt. The Triple-S project suggests how experimenting can help rural water services. EWB-Canada conceptualise their approach in a similar way, and argue that the aid sector’s approach to risk, failure and innovation should be more like venture capitalists. Note that ‘experiments’ in this sense will rarely mean RCTs, for the reasons of appropriate paradigms highlighted below.

Recognise failures of paradigm, myths or lock-ins. Robert Chambers argued that many failures – in sanitation and elsewhere – arise because of thinking in the wrong paradigm. Neo-Newtonian practice (things, order, linear predictability) leads to failure in contexts where an adaptive pluralism paradigm (people, complexity, non-linear unpredictability) applies. Yet ‘myths and collective fantasies’ lead to us being locked into the wrong paradigms.

Admit failures of political economy. The failure of India’s Total Sanitation Campaign was the big example presented. Politics, power, vested interests and bureaucratic inertia all play their part. At this level perhaps ‘admitting tension’ between competing priorities and interests makes more sense than trying to ‘admit failure’. Approaches to political economy analysis may help identify certain drivers or blockers but also highlight the limited influence of external actors. Perhaps we should place greater emphasis on the role of NGOs as convenors of discussion and supporting ‘participatory institutional appraisal’ where local actors can explore these issues themselves?

A final thought: there is a danger of focusing on the smaller failures that I started with, while paying less attention to the bigger ones or working in the wrong paradigm. Yet perhaps we can address this danger by not emphasising the words ‘admitting failure’. These words still have connotations of blame and binary outcomes (even though the ‘admitting failure’ initiative knows this and stresses the need to go beyond this). But a show of hands at the end of the day suggested that a majority preferred us to get this discussion going by starting from another phrase instead. Perhaps adaptive action learning, rapid realism, course correction, or embracing complexity?

I’m not able to fully credit every participant whose ideas I’ve drawn on here, but for more work from some of the key presenters see Admitting Failure and thoughts from Ashley Good, Robert Chambers’ book Provocations for Development, and work by Andrés Hueso Gonzaléz on rural sanitation in India.


5 Comments on “Five ways of thinking about failure in sanitation”

  1. nicnaknoe says:

    Thanks Steve, some nice points pulled from yesterday – happy to see you back blogging! Personally I’m still quite keen on using ‘failure’ (at least until this movement gets going more) as I feel the other suggestions e.g embracing complexity, fade into the background too easily; that said I totally agreed with Robert Chambers that a failure shouldn’t be something to be ‘admitted’ as that makes the whole thing very negative.

    I think any honest ‘reflection’ or ‘lessons learned’ section even from a highly successful project would always include some pitfalls. Perhaps we just need top be encouraging better reflection and sharing of lessons learned post projects – be they a success…or a complete f******!

  2. mtega says:

    Admitting failure, however you term it, is not easy. I speak from experience (see

    For what it’s worth, I feel that avoiding the word “failure” would be a mistake, an attempt to brush things under the carpet. Instead, we need to change how we see failure. It should be seen as acceptable, even encouraged, especially for innovative projects, provided that a decent amount of learning goes on around the failure. I like the idea of dropping the word “admitting”, perhaps it could be replaced with “embracing”, or something of that nature.

    This is a problem for donors, especially those who put a premium on narrow judgement of value for money that focusses on outputs at the expense of learning. But by running away from the word “failure”, we aren’t going to change how it is perceived. The private sector talk about “failing fast, failing often” as a good thing. The development industry could do with a bit of that.

    • Thanks Nic and Mtega for the comments, and thanks for reminding me of the Daraja example, which is a very thoughtful discussion of what lies behind a ‘failure’. I think I agree with you both that keeping the word failure is important in some contexts – probably the first and third categories I give above. These are the two that EWB-Canada and the admitting failure initative are most concerned with. I think the danger comes when a failure that should have been predicted and avoided happens, which then gets post-rationalised into one of the two other types.

  3. […] and pitfalls of learning from failure in sanitation (read the blogposts by Aliki Zeri and by Stephen Jones about the key issues of debate, conclusions and dilemmas). This community of practice and makes a […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s