Best Practices in Procurement, Contract Management, and Technical Supervision for Electrical Infrastructure Development: A Multi-Country Archival Study of Eight World Bank-Financed Projects

Abstract

Background: Electrical infrastructure projects combine technically interdependent equipment, long-lead supply chains, multi-party interfaces, safety-critical construction, and commissioning obligations. Procurement, contract management, and technical supervision are often managed as separate administrative functions even though delivery failure commonly propagates across their boundaries. Objective: This study identifies best practices across these three functions and tests whether their integrated project-level maturity is associated with stronger delivery outcomes. Methods: A comparative archival study was conducted across eight World Bank-financed electricity projects in Vietnam, Rwanda, Tanzania, Bangladesh, Kenya, Uganda, India, and Pakistan. Official implementation completion reports, Independent Evaluation Group assessments, procurement records, and current procurement and project-governance standards were reviewed through 20 September 2026. An original Procurement-Contract-Supervision Maturity Index (PCSMI) scored eight observable domains from 0 to 2 (total 0-16). World Bank/IEG outcome ratings were independently converted to a six-point ordinal scale. Cross-case pattern matching, Spearman rank correlation, an exact permutation test, leave-one-out analysis, and domain-drop sensitivity analysis were applied. Results: PCSMI totals ranged from 3 to 14. Higher-rated projects more consistently combined market-aware procurement, credible qualification, coherent packaging, clear contract administration, disciplined decision governance, adequately resourced supervision, and formal quality/testing controls. Lower-scoring cases showed recurring combinations of protracted procurement, weak bidder due diligence, payment or site-readiness constraints, fragmented approval structures, poor claims/change records, and late corrective supervision. PCSMI was positively associated with project outcome rating (Spearman rho=0.873; exact two-sided permutation p=0.0071). Leave-one-out correlations ranged from 0.816 to 0.952, and domain-drop correlations from 0.794 to 0.873. Conclusion: In the examined cases, procurement, contract management, and technical supervision functioned as one delivery control system rather than three independent disciplines. The findings support a stage-gated owner model that links market engagement and packaging decisions to executable contract controls, prompt commercial administration, risk-based supervision, inspection/test plans, and commissioning evidence. The small heterogeneous sample supports mechanism-oriented inference rather than causal claims.