Impact of Digitalization on the Employment in the Manufacturing Industries of Bangladesh

Abstract

The rapid and continuous development of digital technologies is transforming the global manufacturing landscape, restructuring employment patterns, organizational structures, and strategic decision-making process by implementing and adopting Artificial Intelligence, Internet of Things (IoT), and Data-Driven technologies. While digitalization enhances operational efficiency, productivity, and competitiveness, its effect on employment remains widely debated, insufficiently examined, inadequately addressed and empirically underexplored in the existing literature, particularly in developing countries. The main aim of this research is to examine the impact of digitalization on employment dynamics in the manufacturing industry in Bangladesh. The study investigates and analyzes workforce transformation, changing skill requirements, organizational strategies, Research and Development (R&D) investment, and policy implications associated with digitalization. This study carries out a bibliometric analysis, which comprehends the global research trends over the past three decades, with a systematic literature review, which summarizes the major theories, facts and contextual background. The conceptual model was constructed through integrating existing theoretical models, such as Creative Destruction, Labor Market Segmentation, Human Capital Theory, Diffusion of Innovation, Skill-Biased Technological Change Model and Technology Acceptance Model, the study develops a theoretically grounded “Moderated plus Direct Effect Model Framework” integrating multiple perspectives on technological change and Labour Market Dynamics. In this study, a mixed-methods research approach is used. In the qualitative analysis, the purposive non-probability sampling method is employed, generating data from 53 semi-structured interviews and four focus groups across diverse sub-sectors of the manufacturing industries. In the quantitative phase, data were collected from 213 respondents representing different manufacturing industries, providing a broad and representative foundation for statistical analysis. Across both phases, participants included employees, industry leaders, policymakers, technical and operational professionals, and academic experts. Responses were examined and analyzed through SPSS Statistics 26.0, and SmartPLS v.4. Partial Least Squares Structural Equation Modeling (PLS-SEM) was used P a g e iii to test Measurement Reliability, Construct validation, and Structural Relationships among digitalization, employment outcomes, organizational strategies, R&D, and firm level moderating characteristics. A pilot study is done before the full-scale study to optimize the research design and evaluate the measurement model. Qualitative analysis shows that workplace structures are rapidly transforming due to automation and AI. There are concerns about job displacement and growing demand for digital skills, and the need for investment, collaboration, and structured upskilling for sustainable employment transitions. In Quantitative analysis, measurement model validates a high level of internal consistency, with Cronbach Alpha and Composite Reliability rating higher than the usual level of 0.70 and AVE higher than 0.50, and this establishes Convergent and Discriminant Validity. The results of Structural Equation Modeling (SEM) were found to support path coefficients (β), t-statistics, and p-values, stating that digitalization has heterogeneous effects on employment in the manufacturing industry with significant explanatory power (R2 = 0.812; Adjusted R2 = 0.796), significant predictive relevance (Q2 > 0.35), and acceptable model fit. The effect sizes estimates (f2 = 0.05-0.23) show moderate to strong effects of IT investment, R&D, Strategy and Policy and Digital Skills that are emerging as the key determinants on Labour Market Dynamics. The moderation analysis also indicates that Industry Size and Type have a significant impact on the association between digitalization and employment adaptation. The multicollinearity test (VIF < 3.3) and Bootstrapping procedures prove the stability and strength of the model. This research contributes to the digitalization and employment of literature by integrating multiple theoretical perspectives into a unified empirical framework applicable to developing industrial contexts as well as offering practical recommendations to the policymakers, industrial players, and institutions. The research also offers a replicable analytical model for examining technology-driven labor transformation in emerging industrial economies.