Working Paper: CEPR ID: DP14260
Authors: Yann Bramoull; Habiba Djebbari; Bernard Fortin
Abstract: We survey the recent, fast-growing literature on peer effects in networks. An important recurring theme is that the causal identification of peer effects depends on the structure of the network itself. In the absence of correlated effects, the reflection problem is generally solved by network interactions even in non-linear, heterogeneous models. By contrast, microfoundations are generally not identified. We discuss and assess the various approaches developed by economists to account for correlated effects and network endogeneity in particular. We classify these approaches in four broad categories: random peers, random shocks, structural endogeneity and panel data. We review an emerging literature relaxing the assumption that the network is perfectly known. Throughout, we provide a critical reading of the existing literature and identify important gaps and directions for future research.
Keywords: networks; peer effects; identification; causal effects; randomization; measurement errors
JEL Codes: C31; C21; C90
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
Cause | Effect |
---|---|
correlated effects (C10) | causal identification of peer effects (C92) |
endogenous peer selection (C92) | correlated effects (C10) |
common shocks (E32) | correlated effects (C10) |
random peers (C92) | identification of endogenous peer effects (C92) |
random shocks (D80) | distinguish between contextual and endogenous peer effects (C92) |
structural models (E10) | account for network formation (D85) |
panel data (C23) | control for unobserved characteristics (C29) |
peers' peers characteristics (C92) | influence individual outcomes through direct peers (C92) |
network structure assumptions (D85) | identification of peer effects (C92) |
random assignment (C90) | mitigate omitted variable bias (C20) |
structural frameworks (L10) | account for correlated effects (C39) |