Why are some foresight tools more equal than others?

My work has recently brought me into contact with the world of academic foresight. The group I am working with includes not only futurists, but also people from other disciplines. The focus of the work is on foresight, so those from other disciplines are asking the not unreasonable question of foresight is about? One answer placed me in contact with Popper's Foresight Diamond (see picture), which, I have to admit, I had never encountered prior to this work. I'll leave why that is as a question for later. My first reaction to the analysis was one of deep unease.

This was followed by reference to a piece that suggested that the point of foresight was to predict change (don't agree with that) and how foresight tools are useful for thinking about the future (agree with that). In this particular piece, the author then goes on to reference the Three Horizons Model and Causal Layered Analysis. This left me deeply dissatisfied and extremely uncomfortable. It is worth tracking the source of this discomfort.

The second piece gave me a better clue to the source of the discomfort than the first, but let's start a bit further back. What is the point of foresight? For the insights into a range of emergent futures that they can provide. How do we unlock those insights? By using a range of foresight tools developed for that purpose. The key point is the range of tools because that suggests an intention worth uncovering. Tools are used because they are useful. A tool that has little utility is a fairly poor tool. It follows that if the purpose of foresight is to be useful, then it needs to embrace tools that have a high degree of utility and to discard tools that have little utility. This is right to the point.

The second piece was binary. Only the 3H Model and CLA were mentioned. The two are not equivalent in practice. The 3H Model is encountered more frequently in practice nowadays, but, outside of the public sector and a few voluntary agencies, CLA is hardly encountered at all. In the private sector it is extremely rare to encounter CLA. Why is that? I think that it can be ascribed to three factors. First, there is the relative complexity of the competing models. CLA intentionally delves into complex layers of meaning and intention. This is intuitively difficult to grasp and needs a great deal of explanation just to arrive at the starting point of a project. The 3H Model is simple and intuitive as a descriptor of change. There is the present (Horizon 1), the future (Horizon 3), and the transition between now and then (Horizon 2). Nothing more complex is needed to understand the model and it can start straight away.

The second problem area relates to what the models are looking at, their strategic intent. CLA aims to examine deeper layers of meaning that most commercial organisations are uninterested in exploring. They take the view that Jungian archetypes and bedrock stories might be interesting conversations, but they aren't exactly on message for the future running of the organisation. The 3H Model, by way of contrast, has the strategic problem as the centre of the exercise. It focuses on the problems that are worrying those who commissioned the investigation. That speaks to the third problem area - the cost of project. Because of the developmental time involved and the breadth of the staff whose input is needed on a project, CLA is far more expensive to deliver a project than the 3H Model. If resources are constrained, then the uncertainty over whether or not a tool will deliver a useful result can be minimised by using the less expensive model. The cost of writing off a project that delivered no appreciably valid or impactful results is lessened.

This is why the foresight diamond makes me feel so uneasy. It identifies a large number of foresight tools, but then ascribes to them a degree of equivalence that I consider false. I appreciate that an academic work has to include all possible outcomes for the sake of completeness. I imagine the peer reviewers looking carefully at what had been excluded. However, the final result is misleading because not all tools have an equivalence. Perhaps that's why I hadn't encountered the foresight diamond before? As a practitioner, it has very little of use for me. 

If the foresight diamond were to be reworked as a word cloud, with the size of the entry determined by the frequency that it is encountered in practice, then I wouldn't mind betting that virtually all of the diamond would be occupied by the 2x2 Matrix. The 2x2 Matrix is by far the most commonly encountered foresight tool. It is easy to understand, it is quick to deploy, and useful results can be derived with fairly minimal cost. Whereas the 3H Model might yield useful results over an afternoon, the 2x2 Matrix can yield useful results over a cup of tea. If a 2x2 Matrix doesn't yield useful results, you will have lost a tea break. If a CLA project doesn't yield useful results, you will have lost a much greater sum of resources, be they time or money. This is why I am rebelling against the foresight diamond. It creates a false equivalence between foresight tools.

Coming back to the original question, why are some foresight tools more equal than others? One reason is that some foresight tools have a much lower cost to generate more useful results than others. These are the ones more frequently encountered in practice. This helps managing the uncertainty around the ability of different tools to generate useful results because the cost of a write off of results that aren't useful is much lower. I'm afraid to say that, at the end of the day, money talks.

Stephen Aguilar-Millan

© The European Futures Observatory 2021