Political economy, experimental or abstract and deductive science?

Pierre Cahuc and André Zylberberg raise the problem of the scientific status and the method of economics: for these authors economic science has become an experimental science searching cause and effect relationships econometrically tested. We support, on the contrary, that, logically, economics can be only deductive science; about looking for cause and effect relationships, not only it does not characterize science in general, but furthermore it is inadequate in terms of knowledge of social objects.