Real-time small area estimation of food security in Zimbabwe: integrating mobile-phone and face-to-face surveys using joint multilevel regression and poststratification

2025-05-06

AI Summary

In resource-constrained settings like Zimbabwe, timely and representative food security monitoring remains challenging. To address this, we propose jMRP—a joint Multilevel Regression and Post-stratification framework integrating high-frequency mobile phone surveys (broad coverage) with annual in-person household surveys (high representativeness). jMRP jointly corrects for both sampling and modality biases and quantifies uncertainty via a fully Bayesian approach. To our knowledge, it is the first method to produce monthly, subnational (e.g., district-level) estimates of food insecurity dynamics. Empirical evaluation demonstrates substantial improvements in estimation accuracy over single-source approaches and more comprehensive uncertainty characterization. By enabling scalable, real-time, and spatially granular monitoring, jMRP provides a robust methodological foundation for evidence-based policy interventions toward achieving Zero Hunger.

Technology Category

• Intelligent Robots: State Estimation
• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Security and Privacy: Large-scale security measurements
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Real-time, fine-grained monitoring of food security is essential for enabling timely and targeted interventions, thereby supporting the global goal of achieving zero hunger - a key objective of the 2030 Agenda for Sustainable Development. Mobile phone surveys provide a scalable and temporally rich data source that can be tailored to different administrative levels. However, due to cost and operational constraints, maintaining high-frequency data collection while ensuring representativeness at lower administrative levels is often infeasible. We propose a joint multilevel regression and poststratification (jMRP) approach that combines high-frequency and up-to-date mobile phone survey data, designed for higher administrative levels, with an annual face-to-face survey representative at lower levels to produce reliable food security estimates at spatially and temporally finer scales than those originally targeted by the surveys. This methodology accounts for systematic differences in survey responses due to modality and socio-economic characteristics, reducing both sampling and modality bias. We implement the approach in a fully Bayesian manner to quantify uncertainty. We demonstrate the effectiveness of our method using data from Zimbabwe, thus offering a cost-effective solution for real-time monitoring and strengthening decision-making in resource-constrained settings.

Problem

• Real-time food security estimation in Zimbabwe using mobile and face-to-face surveys
• Combining high-frequency mobile data with annual surveys for accurate local estimates
• Reducing sampling and modality bias in food security monitoring

Innovation

• Joint multilevel regression and poststratification (jMRP) approach
• Integrates mobile-phone and face-to-face survey data
• Fully Bayesian implementation for uncertainty quantification