The super power of modelling
In this time of change, I have been part of many discussions where big, complex problems are discussed. Our culture prides itself on information and intellect: can't we put that scientific way of working to use to solve our problems?
Well, yes and no! But what we certainly can do, is harness the super-power of engineering to understand problems better and develop actual insight.
What super-power I am referring to? It's what engineers always do when they encounter complex problems: create a model that represents all potential interactions of a system.
The bay rights use case
Let's try demystifying the idea of "a model". I often find that the best way to understand a concept is through an example. The problem described in the previous post is a great fit. In blog #67 a real life problem of many southern European resorts was described: rich yacht owners are visiting en masse, enjoying the landscapes and the tourist services offered to them but negatively impacts the local community (through waste disposal or noise, among others) while the local authorities are trying to best manage the situation. A complex, actual societal problem with many stakeholders is an excellent example to try our game theory, model based approach on!
Formalizing the problem description
Let's start by describing our problem with a more formalized, math-based language. To start, we need to have a metric that will track down the net benefits that each of the 4 stakeholders enjoy. For this example we will stick with one, generic metric to capture all potential benefits (money, status, enjoyment, etc.). We will call this "social score". We could go on and create more metrics, but let's keep it simple for our example.
What we need to do now is define all interactions that we think are important to capture:
- The yacht owners enjoy the services that are offered to them. The more services that exist, they happier they are. So the yacht owner social score grows proportionally to the tourism business score. For example, we can say that for each ten social score point of the Tourism business stakeholders, the yacht owners gain one (Y = Y + 0.1 * T)
- Similarly, the tourism business sector grows proportionally to the yacht owners. We can formalize this relationship as T = T + 0.2 * Y
- The citizens also enjoy the benefits of a booming tourism sector because the local economy grows. Let's add C = C + 0.1 * T to our list of relationships.
- But the Citizens also have a worse quality of life as a consequence of the Yacht tourism. This creates a negative relationship that we can formalize as C = C - 0.2 * Y
- The authorities have the toughest role, since they want to balance all interests.
- They depend on the Citizens for votes, so their social score rises when the Citizens are happy (A = A + 0.1 * C)
- However, the Citizens are also complaining to the Authorities proportionally to how many yachts enter the bay. This is another negative relationship (A = A - 0.2 * Y)
- The tax revenue of the tourism business is very important to the Authorities, so we can say that A = A + 0.2 * T
- Lastly, the Authorities want to impose restrictions to the yacht owners, proportional to how many yachts enter the bay (Y = Y - 0.1*Y)
The following graph visually shows all the codified relationships.
Putting the model to work
Having codified the relationships to mathematical rules, we can now "run" our model by calculating the social scores after each round. We can now say that running this model simulates what happens in the real world.
To run the simulation, we would need to determine the initial social scores of the stakeholders (also known as starting conditions). It is important to note that the starting conditions and the parameters of the relationships have a big impact to the model and normally should be very carefully assigned, documented and verified, but for the sake of our example, let's assume uniform conditions and start all stakeholders with a social score of 1. The results of this model after 100 rounds is presented in the following graph:
The trends of this system become clear: everybody wins (especially the Authorities and the tourism businesses) but the citizens do not share the collective gains. This insight can be useful to draft policy: somehow the voice of the citizens should be taken into account.
Also important to note, that our simulation suggests ever growing scores, which is something that is rarely observed in nature. This doesn't necessarily mean that our model is wrong, but rather that our description likely only describes initial development and after some time other mechanisms would likely come into play. Perhaps an idea to explore another time!
If you are curious for the model itself, just reach out and I will send you the spreadsheet I modelled this on. Happy thinking!
Comments
Post a Comment