Microsimulation of the tax and benefit system involves modelling the different taxes, compulsory contributions and transfers, and applying them to a representative dataset of economic agents (individuals, households, firms, etc.). Such a tool enables us not only to measure the redistributive effects of current and past legislation, but also to predict the impact of reforms that have not yet been adopted.

Three broad approaches

  • The first approach is purely static: it consists in applying a legislation particular to a given population without taking into account the behavioural responses that may be generated by the legislative changes. This approach allows us to take into account the variety of individual situations and not to rely on simple “typical cases”, but it reaches its limits when individual or company reactions to legislative change are significant.
  • The second approach takes into account behavioural reactions in order to grasp the actual impact of a policy. For example, following an increase in tobacco taxes, we may expect a reduction in cigarette consumption (as well as multiple cascade effects such as altered patterns of consumption of other goods and services, and changes to health costs). The capacity of a microsimulation model to take these induced effects into account depends, then, on our knowledge of the effects of such policies.
  • The third approach goes further by attempting to take into account the dynamic effects of public policies: in the case of pensions, the purpose is to model the impact of reforms on individuals’ behaviour throughout their life cycle. The dynamic microsimulation artificially ages a given population so as to estimate the impact of reforms over the long-term.

The IPP has developed its own microsimulation model of the French tax and benefit system: TAXIPP

This microsimulation model operates on a variety of databases and allows us to go back to 1979, and forward, annually, to 1997. Accessing TAXIPP.