As empirical work in identifying social effects becomes more prevalent, researchers are beginning to struggle with identifying the composition of social interactions within any given reference group. In this paper, we present a simple econometric methodology for the separate identification of multiple social interactions. The setting under which we achieve separation is special, but is likely to be appropriate in many applications.
A subsequent version of this paper is forthcoming in Economic Inquiry.
JEL code: Z13
Keywords: social interactions, conflation problem