Authors: Alfred Mwatika Mghanga, Eric Mathuva, Egondi Patrick
Abstract: This study examines the effect of e-procurement systems on cost performance in Water Service Providers (WSPs) in Kenya. The study is motivated by persistent inefficiencies in public procurement despite the adoption of digital procurement reforms such as IFMIS. A descriptive cross-sectional survey design employing a mixed methods approach was used. Data were collected from 221 respondents across 81 WSPs in Kenya. Spearman Rank Correlation and binary logistic regression models were applied due to non-normality and dichotomisation of the dependent variable (cost performance: 0 = low, 1 = high). Results indicate a strong positive and statistically significant relationship between e-procurement and cost performance (ρ = 0.821, p < .01). Binary logistic regression results further confirm that e-procurement significantly predicts high-cost performance in both univariate (B = 4.556, p < .001, Exp(B) = 95.230) and multivariate models (B = 3.383, p < .001, Exp(B) = 29.456). The multivariate model explained 96.4% of variance (Nagelkerke R² = .964), indicating strong predictive power. However, implementation gaps were observed, particularly low utilisation of IFMIS systems (M = 2.09). The study concludes that e-procurement is the most influential determinant of cost performance in WSPs, reinforcing Resource-Based View and Technological Determinism theories. Policy recommendations include full digital integration, ICT capacity strengthening, and enforcement of e-procurement compliance frameworks.
