Sharp bounds for population and principal effects under latent sequential ignorability
Abstract
Treatment decisions are often sequential and depend on information absent from the recorded history. Latent sequential ignorability models such assignment by assuming that baseline covariates and principal strata, defined by joint potential intermediate responses, account for confounding of later treatment. For two treatment periods with a binary intermediate response, we derive sharp population and principal effect bounds under a specified association between the potential responses and show that the two targets can require different optimizations to obtain sharp bounds. For population effects, optimization over latent treatment probabilities reduces to their feasible endpoints. Stronger selection between response types can be offset by making their outcome distributions more similar, preserving both the observed distribution and population marginal outcome distributions. This replacement can change effects within principal strata, whose sharp bounds may instead require interior treatment probabilities. Restrictions on differences in conditional effects can tighten the bounds without equating outcome levels. For binary outcomes, we give an exact finite algorithm that includes these interior assignments. We also develop inference under smoothness and estimation conditions. We illustrate the bounds in a study of gym attendance after an incentive program.
Related
- Causal Inference for Sequential Settings under Interference and Latent Confounding
- Posterior Regimes and Latent Deception: Variational Bayesian Inference in Hidden Markov Models for Sequential Fraud Detection in Financial Transactions
- Sequential reduction for discrete latent variables in ecological and evolutionary models using RTMB
- Explicit Asymptotic Bounds for Sequential Calibration Beyond T2/3
- Wasserstein Bounds for Ambient and Latent Smoothing on Data Manifolds
BibTeX
@misc{ohnishi2026sharp,
title = {{Sharp bounds for population and principal effects under latent sequential ignorability}},
author = {Yuki Ohnishi},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.14330},
eprint = {2610.14330},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.14330},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Sharp bounds for population and principal effects under latent sequential ignorability
AU - Yuki Ohnishi
PY - 2026
DA - 2026/10/09
PB - arXiv
DO - 10.48550/arxiv.2610.14330
UR - https://arxiv.org/abs/2610.14330
AB - Treatment decisions are often sequential and depend on information absent from the recorded history. Latent sequential ignorability models such assignment by assuming that baseline covariates and principal strata, defined by joint potential intermediate responses, account for confounding of later treatment. For two treatment periods with a binary intermediate response, we derive sharp population and principal effect bounds under a specified association between the potential responses and show that the two targets can require different optimizations to obtain sharp bounds. For population effects, optimization over latent treatment probabilities reduces to their feasible endpoints. Stronger selection between response types can be offset by making their outcome distributions more similar, preserving both the observed distribution and population marginal outcome distributions. This replacement can change effects within principal strata, whose sharp bounds may instead require interior treatment probabilities. Restrictions on differences in conditional effects can tighten the bounds without equating outcome levels. For binary outcomes, we give an exact finite algorithm that includes these interior assignments. We also develop inference under smoothness and estimation conditions. We illustrate the bounds in a study of gym attendance after an incentive program.
ER -