ICT adoption and firm-level productivity gaps in Sub-Saharan Africa

This study examines the relationship between information and communication technology (ICT) adoption and firm-level labour productivity in Sub-Saharan Africa using the 2025 World Bank Enterprise Survey, covering 5315 firms across 14 countries. ICT adoption is measured using a composite binary indicator capturing whether a firm owns a website or uses the internet for commercial activities. The analysis applies a five-stage empirical framework comprising a Probit model of ICT adoption determinants, an endogenous switching regression to address self-selection bias through the inverse Mills ratio, estimation of average treatment effects (ATT, ATU and ATE), Oaxaca–Blinder threefold decomposition and Machado–Mata quantile decomposition to assess distributional heterogeneity. Findings show that ICT adoption significantly improves labour productivity, increasing productivity by about 33.0% for adopters and by 16.0% for non-adopters had they adopted ICT. The Oaxaca–Blinder decomposition indicates that 69.5% of the productivity gap is explained by differences in firm characteristics, while 30.5% is attributed to differences in returns to those characteristics. Quantile decomposition results further reveal widening productivity gaps across the distribution. Robustness checks confirm the consistency and reliability of the findings.