Belief Convergence under Misspecified Learning: A Martingale Approach

We present an approach to analyze learning outcomes in a broad class of misspeci?ed environments, spanning both single-agent and social learning. We introduce a novel “prediction accuracy” order over subjective models, and observe that this makes it possible to partially restore standard martingale convergence arguments that apply under correctly speci?ed learning. Based on this, we derive general conditions to determine when beliefs in a given environment converge to some long-run belief either locally or globally (i.e., from some or all initial beliefs). We show that these conditions can be applied, ?rst, to unify and generalize various convergence results in previously studied settings. Second, they enable us to analyze environments where learning is “slow,” such as costly information acquisition and sequential social learning. In such environments, we illustrate that even if agents learn the truth when they are correctly speci?ed, vanishingly small amounts of misspeci?cation can generate extreme failures of learning.