The adoption of advanced manufacturing systems in developing economies is critical for industrial growth, yet empirical analysis of adoption drivers remains limited. Existing studies often lack robust methodological frameworks to account for the hierarchical structure of plant-level data. This study aims to methodologically evaluate factors influencing the adoption of manufacturing systems and to quantify their effects on adoption rates within the industrial sector. A multilevel regression model was employed, nesting manufacturing plants within industrial sectors and geographical regions. Data were collected via a stratified survey of plant managers and operational leads. The core statistical model is specified as y₈₉ = ₀ + ₁X₈₉ + u₉ + e₈₉, where u₉ represents random intercepts for sector j. Inference was based on robust standard errors. Plant size and technical workforce skill level were the strongest predictors of adoption. A one-standard-deviation increase in skilled labour proportion was associated with a 17. 2% increase in the adoption likelihood (95% CI: 12. 5% to 21. 9%). Sector-level variability accounted for approximately 31% of the total variance in adoption rates. The adoption of manufacturing systems is predominantly driven by firm-level resource capabilities, with significant variation explained by sectoral context. This underscores the need for targeted, sector-specific intervention strategies. Policy should focus on enhancing technical skills development within the workforce. Industry associations should facilitate knowledge transfer on system benefits, particularly for small and medium-sized enterprises. manufacturing systems, adoption, multilevel modelling, regression analysis, industrial policy This paper provides a novel application of multilevel modelling to manufacturing adoption data, generating the first sector-level variance estimates for Ghana and demonstrating the significant clustering effect of industrial sector.
Agyeman-Badu et al. (Thu,) studied this question.