The way you formulate your problem at least weakly determines your solution space. Let’s take a simple example. In Sweden, we have set a goal to have zero deaths in traffic every year. This is proudly referred to as a zero-vision, and is the overarching narrative around how we think about traffic safety.
Sweden has a low rate of traffic deaths compared with many other countries, and the common assumption is that the zero-vision is responsible for this — but the zero vision was put in place when the number of deaths were already declining fast, in 1997 (the graph is of the number of people dead in traffic:
The current state of affairs is good – Sweden finds itself at the bottom of the list of deaths in traffic by any number of measures. But we have quite obviously also plateaued. And this where it gets interesting.
Are we likely to get anywhere if we continue to iterate the zero-vision here? If we do, the kinds of solutions we are looking for tend to be absolute – we will ask the question “how do we get to zero?” – but those solutions, that solution space, seems limited.
What happens if, instead, we start thinking about relative goals? Decrease the number of dead 5% every year. If we do, we will make progress toward the goal because there will always be something we can do, some small measure that helps.
This is where incrementalism beats big vision – because in incrementalism you deal with solution spaces that are orders of magnitude bigger than in “vision”-spaces.
Now, this is not just relevant for traffic issues. It is even more relevant for the big vision political pieces we see coming out with frightening regularity. The EU has turned this into an artform. Every decade or five years it comes out with a paper arguing that in a certain amount of time it will become “leading” at this or that. The Lisbon Agenda was a great example — the EU set the goal to become the world’s leading R&D region with government spending on science the highest in the world.
This strategy fails, over and over again, and it is not because the EU is not trying — it has summits and writes papers and sets up projects. But the projects are all within the solution space of the question – they deal with how to become leading.
Would it not be much better to look at relative, incremental goals instead? To increase the spending on R&D 15% a year would be a start. To increase investment in R&D – whether private or public – 15% a year even better. The solution space here is much larger and much more interesting.
This post almost ends up being a post in praise of incrementalism, you may say. Should we not have visions? Should we not set Big Hair Goals? Of course we should, but we need paths there – and where the vision is the end of the process, the solution space remains small. We need to get much better at mechanism design, and realizing that within the visions and missions we should try to maximize our solution space by asking the relative, incrementalist questions, questions like:
- Can you make this grow by 10% YoY?
- Can you increase the pace of this by 5% a year?
- Are there ways to chip away at the corners here and get us 3% more revenue?
- How would you go about decreasing complaints just 5%?
These questions are more interesting than you think, because they compound. And that is where the visionary change hides, in compounding increments.
(H/T to Hosuk Lee-Maikyama for the pointer on yet another EU document looking at becoming leading – and as a European I applaud the ambition, heck, it is necessary – but the targets are in absolutes which limits us in growing in two ways – in limiting the solution space as suggest above, and in locking us in to targets that may well be overtaken by the often fast and unpredictable changes in the world. If it is absolute numbers (20 million this, 10000 that) or “cutting edge of quantum computing”, we still end up with the wrong kinds of targets).