Why did the distribution change?
This provides a method for causal attribution in distribution shifts, which is incremental but useful for researchers and practitioners in causal inference and fairness analysis.
The paper tackles the problem of identifying root causes of distribution changes by developing a formal approach using graphical causal models to attribute changes to specific causal mechanisms, and demonstrates its application through simulations and a real-world case study on income distribution differences between men and women.
We describe a formal approach based on graphical causal models to identify the "root causes" of the change in the probability distribution of variables. After factorizing the joint distribution into conditional distributions of each variable, given its parents (the "causal mechanisms"), we attribute the change to changes of these causal mechanisms. This attribution analysis accounts for the fact that mechanisms often change independently and sometimes only some of them change. Through simulations, we study the performance of our distribution change attribution method. We then present a real-world case study identifying the drivers of the difference in the income distribution between men and women.