Malawi faces critical decisions on how to allocate limited health system resources. The Thanzi La Onse (TLO) model is the first whole-population, multi-disease health system model for Malawi. By integrating economic production theory, it helps policymakers understand how shortages of health workers, supplies, and facilities impact health – and how targeted investments can deliver greater health gains.
Here, we provide a summary of two recent Thanzi papers using the TLO model, and discuss the implications for health policymaking.
1. Incorporating an Economic Approach to Production in a Health System Model (Chalkley et al., 2025)
This paper advances the Thanzi La Onse (TLO) model, Malawi’s whole-population, multi-disease health system model, by embedding principles from economics into the modelling of healthcare delivery. The central contribution is to move beyond “black box” approaches and explicitly capture the role of scarce inputs (health workers, facilities, consumables) in healthcare production.
“Our goal in incorporating the economics of production into the healthcare system module of the TLO model for Malawi is to increase its usefulness and relevance.”
Key Findings
- Resource Constraints Matter: Simulations suggest that relaxing constraints on healthcare workers and consumables could avert 5–9 million additional DALYs in Malawi over five years.
- Production Framework: The model uses production functions to capture how inputs combine. A Leontief-type function (fixed proportions of inputs) ensures treatments can only be delivered if all required inputs are available.
- Institutional Influences: Ownership and management structures (e.g., government vs. faith-based facilities) shape efficiency, quality, and service availability, with ongoing work to quantify their effects.
- Health Workforce Dynamics:
- Absenteeism in Sub-Saharan Africa averages ~35%, substantially reducing care access.
- TLO simulations incorporate healthcare worker motivation, absence, and facility-level factors.
- Time & Motion data collection in Malawi (2024) will refine input estimates.
- Consumables: Stock-outs vary by facility size, ownership, and management, creating opportunities for targeted policy interventions.
Policy Implication: Strengthening health system production (workforce expansion, improved management, stock-out reduction) could yield health gains equivalent to large-scale disease-specific interventions.
Pre-print available at arXiv: Incorporating an Economic Approach to Production in a Health System Model.
2. The Properties of a Leontief Production Technology for Health System Modeling: The Thanzi la Onse Model for Malawi (Chalkley et al., 2025)
This companion article examines the theoretical and practical use of a Leontief production function (fixed input ratios) in modelling healthcare delivery within the TLO framework. While often seen as restrictive in economics, the paper shows that in a dynamic, multi-agent, multi-disease model, Leontief technology becomes rich and realistic.
“Our conclusion is that [Health System Models] can nevertheless flourish and capture complex realities, by adopting a simple Leontief production function approach.”
Key Findings
- Feasibility in Data-Constrained Settings: Unlike more complex production functions (e.g., CES), Leontief requires fewer parameters and can be calibrated with existing data, making it well-suited for LMIC health system models.
- Emergent Complexity:
- Although Leontief is linear in theory, when applied in TLO with multiple diseases, cadres, and consumables, it generates non-linear production relationships and non-constant returns to scale.
- Adding multiple outputs (different treatments) introduces competition for resources, creating realistic trade-offs and increasing returns to scale under some conditions.
- Stochastic disease dynamics add uncertainty, further enriching production outcomes.
- Policy Usefulness: Despite its simplicity, the Leontief framework can capture the complexity of health system bottlenecks and provide credible estimates of returns on investment in workforce and supply chains.
Policy Implication: Even with limited data, robust production modelling is possible. This enables policymakers to test investment scenarios (e.g., scaling health workers vs. addressing stock-outs) and anticipate system-wide effects.
Pre-print available at arXiv: The Properties of a Leontief Production Technology for Health System Modeling: The Thanzi la Onse Model for Malawi.
Together, these studies:
- Strengthen the conceptual and empirical foundations of the TLO model.
- Demonstrate that economic production theory improves the policy relevance of health system models in Malawi.
- Provide policymakers with credible tools to explore how workforce, management, and supply chain reforms translate into health gains.
Authors:
-
Martin Chalkley, Tim Colbourn, Timothy B. Hallett, Tara D. Mangal, Margherita Molaro, Sakshi Mohan, Bingling She, Paul Revill, Wiktoria Tafesse.
-
Martin Chalkley, Sakshi Mohan, Margherita Molaro, Bingling She, Wiktoria Tafesse.
By: Kath Devlin | October 2025
