Reflective Governance: The MELD Model

The MELD Model is a blend, or ‘melding’, of the COBIT5 EDM Governance Model and the Triple-Loop Learning Model, which has its origins in the work of Profs Chris Argyris and Donald Schon.

MELD stands for Measure, Evaluate, Learn, and Direct. See the header image above.

This contrivance is simply an elaboration of the Evaluate, Direct, and Monitor (EDM) Model. It aims to focus attention on the reflective practice (represented by the ‘Learn’ sphere) which can occur in relation to each of the components. The ‘Measure’ sphere is substituted for Monitor, partly to employ a more active verb (avoiding the risk that monitoring is seen as a remote and hands-off ‘activity’), but also to provide a context for future comment about methods used to measure performance and conformance.

The model also draws upon a relatively recent evolution of the monitoring and evaluation (M&E) frameworks popular with development and aid agencies. Many traditional M&E Frameworks are now being converted into Monitoring Evaluation and Learning (MEL) Frameworks (see, for example, SEforALL). MEL Frameworks promote reflective practice in project management and seek to enhance the formative aspects of evaluation, alongside the usual (and continuing) focus on summative accountability evaluations.

Readers of my recent posts on monitoring and evaluation (refer to links below) will have noted my intention to offer material on the development of your M&E Framework. As preparation for that post was underway, I first recognised that a MEL Framework was preferable. On further reflection, however, I came to the view that isolating that framework from the other key element of governance (the ‘Direct’ sphere), would undermine recognition that the four spheres belong together in a dynamic relationship.

In my view, the inclusion of the learning element in the MEL framework is really only an elaboration of an existing aspect of the evaluation element. For many organisations, reflection on their systems and mental models (double-loop learning) has long been part of their governance work. With the ever-increasing volatility, uncertainty, complexity, and ambiguity (VUCA) being experienced by every non-profit organisation, however, the importance of adaptability and agility has become more evident.

These ‘operating environment’ challenges require each of us in director or management roles to consider more fundamental questions regarding our organisation’s purpose and core principles. Questions of what practices and procedures we should use or improve (single-loop learning), or even what system dynamics we need to change (double-loop learning), often only involve modest changes to “the way we do things”. When we face questions of how we need to be different (triple-loop learning), then more substantial adaptation becomes possible. Sometimes even ‘transformative’ change occurs.

The MELD cycle and phases

Future posts will unpack each of the major elements or phases of the MELD cycle (illustrated in the table above) in more detail, but some initial observations can be made by way of introduction.

While the sequence of the phases illustrated suggests that you start with measuring (in the 12 o’clock position), it is necessary to determine purposes, goals, and targets before decisions can be made as to what needs to be measured and how. Once a start has been made at any point in the cycle though, the other phases follow in a ‘logical’ order. Each phase is dependent on the previous one. This is why the outer arrows point only in one direction (clockwise, for those still able to relate to an analogue clock face) around the circle. The logic runs as follows:

  • You can’t measure effectively before you have set directions, goals and targets.
  • You can’t evaluate effectively before you have collected and analysed relevant high-quality data.
  • You can’t learn from your experience, or draw conclusions and record insights, until you have completed your evidence-based evaluation
  • You can’t determine your next best strategic goals until you have mined your evaluation findings for insights and updated your needs analysis.

This is not to say that the process is always drawn out over weeks or months – like an annual cycle for instance. Adaptive and reflective governance is continuous. Sometimes the entire cycle of phases needs to be conducted within tight time frames. In a crisis or emergency, it may need to be done in moments. Both rapid and longer-term (planning cycle) deliberations are required to effectively and efficiently govern your organisation’s affairs.

Reflective layers

For each of the four governance spheres defined by the model, three layers of reflection are proposed. The model invites us to ask ‘What have we learned from (and about) measuring, evaluating, and directing, along with our learning processes and systems?’

In the model chart, three sets of arrows originating in the ‘Learn’ sphere each point to the other spheres, symbolising the three levels of reflection on each phase of the cycle. Those reflections focus on the improvement, development, and renewal layers as follows:

  • What do we need to do differently? (improvement)
  • How do we need to think differently? (development)
  • How do we need to be different? (renewal or transformation)

The single arrow pointing back at the ‘Learn’ sphere (learning to learn) implies questions such as:

  • ‘What have we learned about our reflective practices?
  • Do we need to become more effective in our governance approach?
  • Does the model need further refinement or revision?
  • Do we need to reassess our beliefs and attitudes?’

Model selection and adaptation

No one model is ever likely to meet the needs of all non-profit entities – especially given the diversity of purposes and structures in the sector. Nevertheless, I trust that the MELD Model may offer ‘food for thought’ about ways you could enhance your governance.

Future posts will explore the model’s four spheres, and put some ‘flesh on the bones’ of this brief introduction.

See also:

One thought on “Reflective Governance: The MELD Model

Leave a Reply