Latent Causal Socioeconomic Health Index

TL;DR

This paper introduces a Bayesian hierarchical model that combines latent health assessment, spatial analysis, and causal inference to evaluate national socioeconomic health, demonstrated through global case studies and validated by simulations.

Contribution

It develops the first integrated framework combining latent health modeling, spatial analysis, and causal inference for national socioeconomic health assessment.

Findings

The LACSH model effectively captures spatially correlated latent health.

The framework successfully identifies causal effects of policies on societal health.

Visualization techniques aid in covariate balance assessment.

Abstract

This research develops a model-based LAtent Causal Socioeconomic Health (LACSH) index at the national level. Motivated by the need for a holistic national well-being index, we build upon the latent health factor index (LHFI) approach that has been used to assess the unobservable ecological/ecosystem health. LHFI integratively models the relationship between metrics, latent health, and covariates that drive the notion of health. In this paper, the LHFI structure is integrated with spatial modeling and statistical causal modeling. Our efforts are focused on developing the integrated framework to facilitate the understanding of how an observational continuous variable might have causally affected a latent trait that exhibits spatial correlation. A novel visualization technique to evaluate covariate balance is also introduced for the case of a continuous policy (treatment) variable. Our resulting LACSH framework and visualization tool are illustrated through two global case studies on national socioeconomic health (latent trait), each with various metrics and covariates pertaining to different aspects of societal health, and the treatment variable being mandatory maternity leave days and government expenditure on healthcare, respectively. We validate our model by two simulation studies. All approaches are structured in a Bayesian hierarchical framework and results are obtained by Markov chain Monte Carlo techniques.