In the realm of research, several theories coexist to explain a given mechanism. Each one stresses on certain type of driver. For instance human behaviors can be analyzed via learning process [education studies], risk perception [occupational health studies], rational allocation of resources [standard economics], institutional embeddedness [institutional economics, sociology of habitus], mechanical constraints [ergonomics], individual biological characteristics [medicine])... Each of these theories is based on a research program (Lakatos 1070) that simplifies the reality in order to provide evidence of a certain mechanism and build observation frameworks accordingly. Therefore, in the realm of research theories may compete to impose what is the major driver of behavior change (e.g. learning process versus habitus) and draw on hard core that defend contradictory hypotheses (e.g. homo oeconomicus versus behaviors are channeled by social structures internalized into mental structures).

From research to practice, a double flip is needed in order to be properly use scientific knowledge to inform decision making. As behaviors are always driven by complex causalities and have consequences on complex phenomena, no single approach can pretend to provide relevant knowledge to properly inform the practice. The actual crises (e.g. environmental crisis) call for in depth collaboration between a plurality of theories that coexist in social science, but it also call for a wide interdisciplinarity that combine social sciences with natural sciences and technical approaches.

 The aim of the proposed communication is to analyze the consequences of this epistemic situation for the research in economics and for policies orientations.

It will be based on a set of research works dealing with biodiversity conservation on the one hand and with occupational health and pesticide use in agriculture on the other hand (Laurent et al. 2014, 2015, 2016). It will show the consequences on development models of the different patterns of knowledge integration put forward in various institutional environments (government prevention schemes, agencies, trade unions...).