Computational simulation models are a family of methods aiming at simulating (complex) systems and allowing to analyse ex-ante possible outcomes under different scenarios and hypotheses. These models are particularly well-suited for complex problems involving different actors and complex interactions. The family of computations simulations models includes, for example, agent-based models, microsimulation and system dynamics.

Today many problems are complex because they involve different actors with not necessarily aligned goals. Using linear models to solve complex issues is doomed to fail because they ignore the system effects. In contrast, computational simulation models are built for the analysis of systems and can yield highly non-linear analyses. The simulation will give us an idea of how the system would look like under different policy measures. Many times these simulations shed light on unexpected consequences of measures and can, therefore, avoid testing bad policies in the real context - saving time and money.

The range of possible applications is extremely wide, let us focus on two examples. The first example is the health system of a country, which is characterised by a large number of different actors with different goals and many interactions. The political discussion is dominated by interest groups trying to get the best solution for their party. Rarely the discussion is based on a thorough analysis of the system-wide effects and side-effects of policy measures.

The second example of such models is the topic of gender equality in the labour force of private companies. We often hear about unequal salaries, but this is only an indicator or outcome of a complex human resource system within a company. Simulations can help us to virtually recreate this system and then analyse different policy measures. They will also allow us to estimate the necessary time frame to see actual changes in the key outcome variables.

The quality of such models crucially depends on the knowledge the modellers have about the system they want to analyse. Data and the scientific literature are great ingredients to achieve a good model, but the knowledge among the actors of the real system is even more important, especially when the goal is to obtain a realistic model.

At CORESO we believe that computational simulation models should be an important element for the discussion of policy measures to tackle complex issues. At the same time, we strongly believe that the collective intelligence of the stakeholders of an issue is a key ingredient to build a meaningful model. We use our collective intelligence techniques with stakeholders to help modellers make more appropriate models and then use the model to provide stakeholders with insights for their decision-making process.