Abstract
Manufacturing systems in developing economies face persistent operational and safety risks, yet methodological frameworks for their longitudinal analysis remain underdeveloped. This gap limits evidence-based interventions for industrial resilience. This study aims to evaluate methodological approaches for risk assessment and to develop a panel-data model for quantifying risk reduction in manufacturing systems. The objective is to identify robust estimators for measuring the efficacy of engineering and managerial controls over time. A balanced panel dataset was constructed from repeated surveys of manufacturing plants. The core econometric specification is a two-way fixed effects model: $Risk{it} = \beta0 + \beta1 Intervention{it} + \mui + \lambdat + \epsilon{it}$, where $\mui$ and $\lambda_t$ denote plant and time fixed effects. Inference is based on cluster-robust standard errors to account for heteroskedasticity and serial correlation. The methodological evaluation confirms the superiority of fixed effects estimators for controlling unobserved plant heterogeneity. The model estimates a statistically significant average reduction in composite risk scores of 18.2% (95% CI: 14.7% to 21.5%) associated with the implementation of integrated safety-management systems. Panel-data estimation provides a rigorous methodological foundation for measuring risk reduction in industrial settings. The findings demonstrate that systematic engineering interventions yield substantial and quantifiable improvements in system safety. Manufacturing regulators should adopt panel-data methodologies for monitoring and benchmarking plant performance. Industry practitioners should prioritise integrated system interventions, supported by longitudinal data collection to facilitate causal analysis. industrial safety, panel data, fixed effects, risk assessment, manufacturing engineering, econometric modelling This paper provides a novel application of econometric panel-data methods to engineering risk analysis, creating a replicable framework for causal inference in industrial safety studies within developing contexts.