Insee (Institut national de la statistique), the French public agency for statistics, and Drees (Direction des études du Ministère des Affaires sociales et de la santé), which is in charge of surveys at the Ministry of Social Affairs and Health, has published the source code of the microsimulation algorithmic model called Ines.
Developed since 1996 by the two public organisations, Ines simulates the impact of social and fiscal regulations in France. “It assesses the budgetary and redistributive impact of many social tax and benefit reforms: family allowances, housing allowances, welfare benefits, social security contributions, income tax”, the project page states.
Ines uses data collected for social and fiscal revenue surveys in France. These surveys provide information (more than 1,000 pieces of information on each individual, according to Drees) on a sample composed of 50,000 French households.
Ines is used for annual surveys (like France Portrait Social) or deep research on economical, fiscal and social impacts.
Ines is also used to simulate social contributions, taxes and various social support instruments, including the RSA (Revenu de solidarité active – an in-work welfare benefit).
Insee and Drees said that opening the source code of Ines “allows research teams to access data on which the model is based, to carry out sharp and detailed work, especially for the evaluation of public policies”.
Last April, France also opened the source code of its tax calculator in order to improve fiscal transparency in the country. The release of Ines is another step in this direction.
Freely accessible on Adullact
Now open source, the source code is written in the SAS language and is composed of several programmes, Insee said. The Adullact association, which promotes and centralises free software for the French administration, hosts the source code on its platform. Detailed documentation is also provided through a dedicated wiki.
Experts are now able to register on the Adullact website to download and use Ines. Users can also contribute to the source code to improve Ines, the two public organisations said. A dedicated forum was also developed to collect contributions. Insee and Dress will review the suggestions received.