Three Major Issues Concerning Randomised Social Experimentation in France

In the 1960s, tools were developed in France to test laws before they were adopted. This form of assessment eventually acquired a constitutional basis with the Act of 28 March 2003 authorising normative experimentation, both nationally and locally. This innovation was the key to a specific form of experimentation, namely the randomised experiment. It borrows its methodology from social sciences, and its perimeter from international policies on poverty. This form of experimentation raises three issues, which are discussed in three successive sections. The first one is a matter of legal science: what role does this form of experimentation play within the scope of normative experiments? The second question is a scientific one: what lessons does social experimentation bring to experimental science, applied to human behaviour? The third issue is socio-political: what is its contribution to the evaluation of public policies? The conclusion recalls the ethical and scientific requirements that are necessary in the conduct and evaluation of experiments on human behaviour.