Working Paper: NBER ID: w31945
Authors: James J. Heckman; Rodrigo Pinto
Abstract: This paper examines the econometric causal model and the interpretation of empirical evidence based on thought experiments that was developed by Ragnar Frisch and Trygve Haavelmo. We compare the econometric causal model with two currently popular causal frameworks: the Neyman-Rubin causal model and the Do-Calculus. The Neyman-Rubin causal model is based on the language of potential outcomes and was largely developed by statisticians. Instead of being based on thought experiments, it takes statistical experiments as its foundation. The Do-Calculus, developed by Judea Pearl and co-authors, relies on Directed Acyclic Graphs (DAGs) and is a popular causal framework in computer science and applied mathematics. We make the case that economists who uncritically use these frameworks often discard the substantial benefits of the econometric causal model to the detriment of more informative analyses. We illustrate the versatility and capabilities of the econometric framework using causal models developed in economics.
Keywords: No keywords provided
JEL Codes: C10; C18
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| econometric causal model (C20) | understanding causality (C22) |
| econometric model (C51) | quantify policy impacts (C54) |
| econometric approach (C51) | capture complexities of social interactions (Z13) |
| Neyman-Rubin model (C52) | incomplete understanding of causal relationships (D80) |
| do-calculus (C65) | inadequate causal inference (C20) |
| thought experiments (C93) | establish causal claims (C99) |