Abstract
{ "background": "The adoption of advanced manufacturing systems (AMS) is a critical driver of industrial competitiveness, yet robust longitudinal measurement of adoption rates in developing economies remains methodologically underdeveloped. South Africa's manufacturing sector presents a salient case for such methodological evaluation.", "purpose and objectives": "This study aims to methodologically evaluate panel-data estimation techniques for measuring AMS adoption trajectories. The objective is to identify the most robust econometric specification for capturing temporal adoption dynamics and controlling for plant-level heterogeneity.", "methodology": "We construct and analyse an unbalanced panel dataset of manufacturing plants. The core estimation employs a dynamic panel model specified as $y{it} = \\alpha y{i,t-1} + \\beta X{it} + \\etai + \\epsilon_{it}$, estimated using the System Generalised Method of Moments (GMM) to address endogeneity. Robust standard errors are clustered at the plant level.", "findings": "The System GMM estimator proved superior in handling unobserved heterogeneity and dynamic adjustment. A key concrete result is that a one-unit increase in prior technological capability increases the probability of AMS adoption by approximately 0.15, with a 95% confidence interval of [0.11, 0.19]. The methodological evaluation reveals that static models significantly overstate the speed of adoption.", "conclusion": "The methodological framework provides a more reliable and nuanced measurement of AMS adoption rates, revealing slower, path-dependent diffusion than previously understood. Accurate estimation requires models that account for state dependence and firm-specific effects.", "recommendations": "Future research on technology adoption in similar contexts should employ dynamic panel estimators. Industry surveys should be designed to facilitate the construction of panel data, capturing core variables consistently over time to enable such analyses.", "key words": "advanced manufacturing systems, technology adoption, panel data, econometrics, System GMM, industrial policy", "contribution statement": "This paper provides a novel methodological framework for estimating technology adoption rates using dynamic panel models, applied for the