Adaptive programming in design, but not in practice 

Once upon a time, there was a development programme pursuing inclusive democratisation across Africa, Asia and Latin America. They worked through a network of partner organisations in programme countries based on a strong and elaborate Theory of Change (ToC) covering desired systemic and behavioural changes of key actors. With the ToC as a guiding framework, implementing partners worked out annual plans indicating the nature and number of foreseen interventions along with budget estimates. Twice a year, they reported on progress in completing their plans, including an analysis of variance between actual and planned interventions, and outputs. In most countries, this variance was substantial but generally accepted as something to be expected when working on a complex issue. A new year would start, the planning and implementation cycle would be repeated, considering contextual changes and experiences from the past year. 

So far so good, or so it seemed, as all partners accepted and went through these motions, understanding the need for regular plans and reports as a basis for fund allocation and accountability. At the same time, nagging doubts remained about the effectiveness of the programme. Questions - such as, are we making meaningful progress, are we progressing fast enough, and what are we learning to progress faster? - kept coming back without convincing answers. Besides, the annual planning became a mandatory ritual, increasingly perceived as formality disconnected from reality.  

Do you recognise anything of the above? In our work as evaluators, we encounter programme designs that are carefully thought through with Theories of Change (ToCs) with clear and interconnected actor-based pathways of change. Such ToCs do much more justice to reality than earlier log frames, which often simplified complex social change to a linear 3-step logic with some assumptions that had to hold. Thus, much progress can be seen in the quality of the design thinking of new programmes.  

However, in many programmes this progress evaporates when it comes to turning programme design into operational plans and monitoring systems. Suddenly, linear thinking is back, implying predictability of action and effects that we know do not exist in a complex reality. Consequently, M&E findings reveal substantial differences between what is planned and the reality, which is accepted as a logical consequence of unforeseen changes and progressing insights. And indeed, maybe many of these changes could not have been foreseen. However, what can be foreseen after 60 – 70 years of development cooperation is that there will be changes. What will change is uncertain, the occurrence of change is not.  

Demonstrating meaningful progress 

This is not where the story ends. A review of the programme’s performance concluded that the programme ‘failed’ if judged by its original results indicators. However, when judged by actual signs of progress towards its overall ambitions, the review found the programme to be rather successful. This caused programme management to conclude that their thorough complex-aware programming deserved to be followed up by more flexible and adaptable planning and monitoring practices. In other words, the adoption of a concept that has been around for a while but is not put into practice enough: adaptive programming. 

Putting adaptation into practice 

So, what does this mean in practice? How is the organisation moving towards adaptive programming? 

They maintain the same overall ToC as a starting point. Subsequently, partners undertake a country-specific political economy analysis to tailor the ToC to context, identifying and prioritising what adjustments are needed and in which key actors. Then, outcomes are formulated in the shape of the desired behavioural changes of targeted actors (3 to 5 actor groups per country). For each actor, a results-oriented intervention strategy is formulated. This strategy aims to include three types of outputs considered necessary and feasible in pursuit of this behavioural change (i.e. building capacity, creating opportunities and influencing motivation1). For each output, the implementing partners have a package of services to choose from, including the provision of resources, facilitation, training, advisory and brokering. Annual plans offer indicative choices in scale and types of services to be delivered, but partners have the autonomy to adapt their choices as considered necessary. For this, partners are allocated an overall budget with broad cost categories and clear parameters in terms of indirect versus direct expenses. 

This does not mean that partners can do whatever they want, as they agree to report and be accountable for the choices made and the resources used. This accountability mechanism is complemented by a learning-oriented monitoring system dedicated to capturing meaningful progress towards behavioural change along with a reflection on the effect of and learning from adaptations. 

Preconditions for success  

Is it really that simple? Yes and no. Yes, because the mechanisms of putting adaptive programming into practice are simple enough. And no, because successful adaptive programming requires preconditions that are less simple and straightforward, especially where existing cooperation systems are predominantly designed to control and minimise risks.  

In short, among such preconditions are: 
  1. Partners to embrace a mindset of ‘failing forward’. Accept that ToCs are sets of hypotheses to be tried and tested in an ever-changing context. Adaptations are not a sign of failure or changing course but essential for producing meaningful results. They are to be encouraged, explained, and learned from rather than justified. 
  2. Organisations have and demonstrate adaptive capacity. The capacity of their organisation is not measured by what they know, but by how they learn and adapt. 
  3. A sense of ‘joint’ accountability fed by a clear understanding and practice of mutual accountability for choices made and resources used.  
  4. Sufficient trust and honest info-sharing, which is something that can only be built over time, by providing security that information is used for learning and not for sanctions. 

In conclusion, adaptive programming is not a state-of-being, it requires management decisions to be put in place. It is the result of a collective agreement among partners to engage in a process that concentrates on achieving, capturing and learning from meaningful progress. All of this, under difficult and unpredictable circumstances, by remaining alert and adaptive. 

By , Managing Director and Senior Evaluator at MDF.