Electricity Reliability, Self-Generation Costs and Manufacturing Productivity in Nigeria: Evidence from 1990–2024

Abstract

Nigeria’s manufacturers operate in an electricity system characterised by limited access, network losses, recurrent supply disturbances and widespread recourse to private generation. This study examines how electricity reliability and self-generation fuel-cost exposure are associated with manufacturing productivity over 1990–2024. A reproducible annual dataset is assembled from the World Development Indicators, International Labour Organization modelled estimates, International Energy Agency electricity-loss data disseminated through the World Bank, U.S. Energy Information Administration Brent prices, International Monetary Fund price and exchange-rate series, and Nigerian Electricity Regulatory Commission reports. Manufacturing productivity is measured primarily as real manufacturing value added per labour-force participant and, for robustness, per estimated industrial worker. Because a consistent annual outage-duration series does not span the full period, an Effective Grid Reliability Index is constructed as electricity access multiplied by network delivery efficiency. Self-generation cost exposure is proxied by a real import-parity fuel-cost index based on Brent prices, the naira–dollar exchange rate and consumer prices, normalised to 2015=100. Augmented Dickey–Fuller tests identify a mixture of I(0) and I(1) variables. A bounds-test screening does not establish a stable level relationship; therefore, the preferred model explains annual productivity growth using first differences and Newey–West heteroskedasticity-and-autocorrelation-consistent standard errors. Grid-reliability growth is statistically insignificant, whereas the contemporaneous fuel-cost index is positive and significant. This positive coefficient is interpreted as procyclicality and simultaneity, not as evidence that expensive self-generation raises productivity because fuel demand and import-parity costs increase during stronger production years. Real GDP growth is strongly positive, while inflation is weakly negative. Robustness checks preserve the absence of a stable positive reliability effect and the procyclical cost association. The results show that annual macro series are informative about system-wide co-movement but insufficient for causal inference on firm-level outage losses. Policy should prioritise measurable service continuity, industrial-feeder reliability, competitive embedded generation, fuel-efficiency improvements and public reporting of outage-duration indicators