"background": "The South African manufacturing sector faces persistent systemic risks, including infrastructure instability and supply chain disruptions. Existing engineering risk assessment methods often lack robust counterfactual analysis, making causal attribution of intervention efficacy difficult. ", "purpose and objectives": "This article presents a methodological framework for the causal evaluation of engineering interventions aimed at reducing systemic risk in manufacturing plants. The objective is to provide a rigorous, quasi-experimental design suitable for longitudinal plant-level data. ", "methodology": "We detail a difference-in-differences (DiD) model for panel data. The core specification is Y{it = \ + \ (Treati \ Postt) + \ + \ +, where Y₈ₓ is a composite risk index. The parallel trends assumption is tested using event-study plots, and inference relies on cluster-robust standard errors at the plant level. ", "findings": "The framework, applied to a simulated dataset, demonstrates that a hypothetical intervention targeting energy resilience is associated with a 15% reduction in the composite risk index. The key methodological finding is the critical importance of pre-trend validation; violations of the parallel trends assumption can lead to a 95% confidence interval for the treatment effect that erroneously includes zero. ", "conclusion": "The DiD framework provides a statistically rigorous methodology for isolating the causal effect of engineering interventions on manufacturing systems risk, moving beyond descriptive correlation. ", "recommendations": "Practitioners should adopt this DiD model for retrospective programme evaluation. Future work should integrate high-frequency sensor data into the risk index and explore synthetic control methods for single-unit case studies. ", "key words": "difference-in-differences, causal inference, risk assessment, manufacturing systems, quasi-experimental design, industrial engineering", "contribution statement": "This paper provides the first formalised difference-in-differences econometric framework tailored for evaluating engineering risk reduction programmes in an industrial context, explicitly linking intervention timelines to plant-level panel data. "
Botha et al. (Wed,) studied this question.