Finding Failure


I have been pondering (again) about how well international donors cope with failure. What continually strikes me is that we rarely conclude that an approach to a problem isn’t working as we thought it would despite our best efforts and intentions. In conversations with my counterparts in bilateral and multilateral organisations, it’s a similar story: few – if any – programme approaches appear to fail.

A sign of high quality programming, right?

There is no doubt that international development organisations have achieved a huge amount over the years. But, considering what we are trying to achieve, it just can't be right that everything we do will work as intended.

So, how would we know if we are failing?

One way of looking at project performance is to use the 3 dimensions of time, cost and quality.

In the IT world (arguably simpler than the complex world of international development?) Bent Flyvjberg estimates that one in six projects have an average cost overrun of 200% and a time overrun of 70%. If this is the case for IT projects, it would be plausible that projects designed to achieve change in complex political and risky delivery environments would have equivalent cost and time over-runs.

The ‘quality’ metric is harder still. Does the project intervention actually achieve the intended outcome? Given the challenges of international development projects it would be deeply surprising if they all do. Many are high risk, operate in unstable environments, are dependent on a range of external factors and are based on assumptions of how change happens.

So why is it hard to find failure?

There are lots of potential reasons, including:

  • There are lots of ‘everyday failures’ happening all the time where we and our delivery partners are – as you’d expect – taking routine corrective action to get them back on course. These may not be covered in formal programme documentation.
  • We are constantly adapting and revising performance metrics of time, cost and quality so programmes change but continue to achieve against our (new) expectations. So, by constantly adjusting we never appear to fail.
  • We don’t have sufficient data and feedback loops to actually know how well programmes are really performing (or our partners aren’t telling us).
  • We are measuring activities and not actual change (which is harder to attribute and identify). Perhaps the real change happens a long time after a programme has finished, beyond the reporting period.
  • We aren’t taking sufficient risks, opting for easier tried and tested approaches. Perhaps we extend programmes that work and discontinue those that don’t, but never explicitly describe the discontinued ones as failed projects.
  • We just don’t like talking about failure for fear of undermining the case for international development or — more parochially — our own reputations.

I suspect the answer is often a bit of all of these (and probably other factors?) Whatever the reason, we have a responsibility to be more explicit about what isn’t working (or how we are adjusting and adapting) so we can learn for next time and improve the impact of the work for the communities we serve. In a world of increasing public scrutiny of aid this can be difficult but essential.

How to find failure

So how do we actually identify where things aren’t working? A few thoughts:

Make it safe to talk about failure

First of all, we need to create an environment where it is safe to fail, safe to get things wrong and safe to not know the answer. Arguably, that is the role of leadership teams, but relies on all of us to be sufficiently curious to want to know what isn’t working. We need to do the same for our partners, building their confidence that failure doesn’t necessarily mean the end of funding (in my view, the opposite — I’d far rather we funded organisations who could talk openly about what they are getting wrong and learning than one who reports 100% success all the time).

Think like an ‘aid skeptic’

We need to learn to think like ‘aid skeptics’ from time to time, take a critical and reflective look at what we are doing and be prepared to openly discuss it. Our strong values, dedication and commitment can sometimes make this hard, especially if we have put a huge amount of emotional energy into an idea or an approach (after all, who wants an internal audit or, worse, journalist uncovering our failings, undermining our work?) One way to help do this would be to actively seek external challenge by asking people to deconstruct our programmes, looking for the negatives and then being open to discuss them seriously. Behavioral economists and project specialists talk about the importance of the ‘outside view’.

Find ways to listen to local views

We need to think about what ordinary people would say about our work, continually putting ourselves in their shoes. What would people in a bus say about a project? What would the gossip at the local church, borehole, school tell us? I have been inspired recently by the work being done by DFID and on the Amplify Project (more on a later blog) that attempts to solve problems through experimentation and iteration around people’s needs.

So, what do we do when we find failure?

At this point it’s worth stating that we should have a very low appetite for failure when we – or our partners – aren’t doing the job properly. So, let’s assume we are…

When things go wrong, it is likely that the problem we are trying to resolve will continue to exist and that we continue to be committed to finding a solution. This can make it hard to make objective decisions on what to do next. While a private investor might decide to walk away (for example, when boss of ‘The New Day’ shut down the paper after 9 weeks), failure doesn’t necessarily mean the end in the public sector.

Let’s be open about what is not working, be quick to respond and then work hard to share the experience and lessons. Of course there will be times when we need to re-design or even cut our losses and start again. But there is nothing wrong with that. After all, the task of poverty reduction is one of the most challenging there is.

What do you think we need to do to move the ‘failure’ debate on in international development?