Combination of EUROMOD and LIAM2 tools for the development of dynamic microsimulation models, Feasibility, example and conditions for sustainability of the linkage

LISER has implemented, within Work Package 8 of the EU-’HORIZON 2020’-funded ‘InGRID-2’project (http://www.inclusivegrowth.eu/), a combination of tools for the development of dynamic microsimulation models which objective is ‘to establish a linkage between two important microsimulation platforms: EUROMOD and LIAM2’.EUROMOD is an efficient and well-known platform for the development of static microsimulation models. It is also a model in itself. LIAM2 is a free, open source and user-friendly modelling and simulation framework, mainly implemented for the development of dynamic discrete-time microsimulation models with cross-sectional population ageing. LIAM2 has strong similarities in its built-in logic and grammar to EUROMOD. It is not a model but has been adopted by modellers in at least nine countries (AR, BE, CN, FR, IT, HU, LU, PT and SK) and has led to the creation of several dynamic microsimulation models with a similar structure (based on MiDAS_BE, Federal Planning Bureau, Brussels).In examining the possible avenues for implementing a link between the two platforms, one track was favoured, known as ‘Dynamic in Static/DiS’. It aims at dynamising to a certain extent a static model such as EUROMOD in order to carry out short-term projections, as is already done in the well-known developments of ‘nowcasting’. But our aim is to derive a device that is both as transparent as possible, and therefore more easily reproducible and improvable, and that links EUROMOD to another innovative development platform targeting the development and simulation of dynamic models. This could open up avenues for a more sophisticated implementation of the dynamic side of the combination, which might be valuable in a particularly volatile context. This implementation may also generate positive by-products of interest for the development of dynamic microsimulation models in a broader sense. This includes a ‘Static in Dynamic/SiD’ approach in which EUROMOD could become the tax-benefit module of a larger dynamic model operating over the long term. While proceeding, we emphasise the ‘infrastructural’ nature of the InGRID-2 project, hence the need to draw synergies from existing experiences to innovate at the level of the development and simulation tools mobilised in the background (EUROMOD-LIAM2 linkage), more than to rewrite from scratch or on a large scale, or even to deepen, the short-term projection procedures. What is at stake here is a method, considered as an intermediate good, rather than applications, considered as final goods, yet keeping in mind that the latter remain of utmost importance with regard to the sustainability of the path proposed here. The methodology of combination is described, implemented and first outcomes reported, including possible drawbacks.