The adoption of advanced manufacturing systems is critical for industrial development, yet longitudinal analyses of their uptake in emerging economies are scarce. This gap impedes evidence-based policy and engineering investment. This study aims to methodologically evaluate and compare the adoption trajectories of different manufacturing systems within the nation's industrial sector. It seeks to quantify adoption rates and identify key determinants influencing technological diffusion. A comparative panel-data analysis was conducted using a unique longitudinal dataset of manufacturing plants. The core estimation employs a fixed-effects model: Adoption₈ₓ = ᵢ + ₁X₈ₓ + ₜ + ₈ₓ, where ᵢ denotes plant-specific effects and ₜ time effects. Robust standard errors are used for inference. The analysis reveals a pronounced divergence in adoption rates between discrete and process manufacturing sectors. Specifically, the adoption of computer-integrated manufacturing systems in discrete sectors increased at approximately three times the rate observed in process industries. This disparity is statistically significant at the 1% level. Adoption patterns are heterogeneous and strongly influenced by sectoral characteristics, not merely firm size or age. This indicates that a one-size-fits-all policy for promoting advanced manufacturing is unlikely to be effective. Industrial policy should be sector-specific, with targeted support for process industries lagging in adoption. Engineering curricula should be adapted to address the distinct skill requirements of different manufacturing paradigms. panel data, manufacturing systems, technology adoption, industrial policy, fixed-effects model, Kenya This paper provides the first longitudinal, plant-level econometric analysis of manufacturing systems adoption in the region, introducing a novel application of panel-data methods to compare technological diffusion across engineering sectors.
Mwangi et al. (Fri,) studied this question.