Adaptive development: great progress and some niggles..
Pete Vowles 🇰🇪🇬🇧

Great blog. These ‘niggles’ are all indeed real problems. This sentence is spot-on: We want programmes that actively seek to use learning and experience to dictate change not just reduce forecasts as they hit reality.

But the biggest problem in log-frames is how they are created and how they incentivise or dis-incentivise. I say that as someone who is often needing to create advise or negotiate log-frames as part of my job.

I try to get project teams to develop their theory of change first, which allows for more complexity and vision. Results that are not sure (high risk) but are exciting (high return) get discussed. But then when we start boxing things into log-frames, the ambition simply disappears. The exciting part of a programme may be its potential to influence take-up by others, or to crack a new code. But no programme manager will want to put that most challenging bit into the log-frame, because they cannot be confident to deliver it. So we end up proposing a dry log-frame with achievable indicators, and this incentivises decent delivery not ambition.

I echo Mike Johnson: log-frames disincentivise risk-taking and ambition.

There is no easy solution to this for an organisation that wants accountability. It’s like the dilemma with teachers. The best teachers are inspiring, and boxing them in with rules, plans, and requirements just hampers them or pushes them out. The evidence of impact is in the letters and cards they get from kids whose minds they transformed, year after year. But if you leave teachers alone to empower them, you don’t catch or improve the poor performers. Some kids end up with a bad deal. So with projects. If you set up a system that encourages risk and aspiration and allows failure, there should be more stella performers. But flexibility and tolerance of risk might also allow some poor operations to go unchecked. How do you know when an under-performing programme just needs to adapt and ‘pivot’ and when it is simply not a good use of resources?

The suggestions below for log-frames that encourage adaptive learning are good. That will encourage the ‘pivot’. So does payment-by-results which gives the implementor maximum flexibility to adapt operations. But this still doesn’t address the tolerance of failure from innovation. And a focus on outcomes instead of outputs only works if standard project time frames are extended.

I completely understand why civil servants opt for log-frames with SMART indicators and boring milestones rather than risk-taking, given pressures for accountability and delivery. And I understand why DFID Partners (as you kindly call them) opt for boring achievable outputs and indicators that might score A. Understandable, but not ideal.