{"id":"arxiv/2604.15288","url":"https://arxiv.org/abs/2604.15288","tags":["math.ST"],"score":0,"title":"Generalization of Pearl's Front-Door Criterion","domain":"math.ST","source":"arxiv","authors":["Carol Wu","Elina Robeva"],"reviews":0,"subject":"math.ST","summary":"Pearl's front-door criterion provides a set of sufficient conditions for estimating the total causal effect from observational data in the presence of latent confounding, using the functional P(y | do(x := x*)) = \\sum_z P(z | x*) \\sum_x P(y | x, z) P(x). An open question is whether these conditions can be generalized to be both necessary and sufficient for the validity of this functional, similar to the generalization achieved for the back-door adjustment criterion by Shpitser. In this paper, we present a new, weakened set of graph-based conditions sufficient for the front-door formula to…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-04-16","reactions":{}}