The INSEE and the DREES share the code for the Ines model

16/06/2016
logo Insee

Following the provision of the tax calculation engine OpenFisca, the Ines model of the INSEE (French National Institute of Statistics and Economic Studies) and the DREES (Research, study, evaluation and statistic management) has in turn been donated to the open source community.

On the 14th of June, this model used to produce statistical analyses on the structure of taxes and benefits in relation to household living standards was provided under the licence CeCILL v2.1.

The Ines model for the evaluation of policies

The Ines model is written in SAS language and allows the simulation of tax payments and household benefits using a sampling methodology. The model namely includes income tax, the CSG (general social contribution), collections, family benefits and the RSA (earned income supplement). Since 2014, this model has helped in the creation of balance sheets for social and tax measures for the previous year based on the household living standards as well as the provisional cost of the envisaged measures. Another use for the model is early forecasting of the trend of the poverty rate.

Populating the model with data

The Ines tool has been around for nearly 20 years. However, its versions have only been being used for a few years. The code, documentation and configuration files are available in the Adullact repository. The code using SAS language will deter developers not familiar with the world of statistics. It only covers the main branch of the tool and excludes specific ministry orders or studies.

Finally, to populate the model, access to actual tax data will depend on the presentation of the project to the committee on statistical confidentiality. There is one alternative: data made available by economists C. Landais, T. Piketty and E. Sae who have provided a website to go with their book  "Pour une révolution fiscale" (For a tax revolution) providing an ad hoc data simulator to simulate their own reform.

 

Source: http://www.insee.fr

comments powered by Disqus
top