"background": "The modernisation of manufacturing systems is critical for industrial development, yet policy formulation in this sector is often hindered by a lack of robust, predictive tools for assessing technology adoption. In Nigeria, this gap has led to inefficient resource allocation and suboptimal policy interventions aimed at enhancing industrial productivity. ", "purpose and objectives": "This policy analysis develops and evaluates a novel Bayesian hierarchical model to estimate and forecast the adoption rates of advanced manufacturing systems. The objective is to provide a methodological framework that offers policymakers quantifiable insights into adoption dynamics, enabling more targeted and effective industrial policy. ", "methodology": "A Bayesian hierarchical model is constructed, formalised as y{it \ (\, (1-) \) with () = + \ Xit, where \ \ N (\\, \\²). This structure accounts for plant-level heterogeneity and sectoral clustering. Inference is based on posterior distributions derived from Markov chain Monte Carlo sampling, with credible intervals used to quantify uncertainty. ", "findings": "The model indicates a persistent, positive relationship between targeted fiscal incentives and adoption probability, with a posterior mean odds ratio of 2. 1 (95% credible interval: 1. 7, 2. 6). However, adoption rates are forecast to remain below 40% for most system types without significant policy intervention, highlighting a substantial implementation gap. ", "conclusion": "The Bayesian hierarchical approach provides a superior framework for analysing technology adoption, capturing inherent uncertainties and heterogeneous effects often missed by conventional methods. This allows for more nuanced and reliable policy analysis. ", "recommendations": "Policy should prioritise direct fiscal incentives for early adopters, as modelled evidence shows their strong efficacy. Furthermore, government agencies should institutionalise the use of such probabilistic models for long-term industrial strategy planning and monitoring. ", "
Adeyemi et al. (Tue,) studied this question.