Structural Regularization
It addresses the challenge of incorporating theoretical insights into statistical modeling to improve predictions and causal inference, particularly in economics and other scientific domains where models are often misspecified.
The paper tackles the problem of modeling data by using structural models from economic theory as regularizers for statistical models, showing that this approach can outperform both misspecified structural models and unregularized statistical models, with simulation experiments demonstrating its potential in settings like first-price auctions and demand estimation.
We propose a novel method for modeling data by using structural models based on economic theory as regularizers for statistical models. We show that even if a structural model is misspecified, as long as it is informative about the data-generating mechanism, our method can outperform both the (misspecified) structural model and un-structural-regularized statistical models. Our method permits a Bayesian interpretation of theory as prior knowledge and can be used both for statistical prediction and causal inference. It contributes to transfer learning by showing how incorporating theory into statistical modeling can significantly improve out-of-domain predictions and offers a way to synthesize reduced-form and structural approaches for causal effect estimation. Simulation experiments demonstrate the potential of our method in various settings, including first-price auctions, dynamic models of entry and exit, and demand estimation with instrumental variables. Our method has potential applications not only in economics, but in other scientific disciplines whose theoretical models offer important insight but are subject to significant misspecification concerns.