From Price to Performance: Implementing the Best Value Approach in Czech Public Procurement

Public procurement in many European Union member states remains strongly price-oriented, often at the expense of delivery performance, innovation, and effective risk management. This study examines how the Best Value Approach (BVA) operates within a post-transition, legality-focused administrative environment through a document-based embedded case study of a major public construction contract in the Czech Republic. By analysing artefacts from the Selection, Clarification, and Execution phases, the study traces how BVA’s core governance mechanisms—expert signalling, vendor-led risk ownership, and information-centric oversight—functioned under locally constraining conditions. The findings show that BVA improved capability sorting, surfaced risks earlier, and enhanced transparency through structured reporting instruments such as Weekly Risk Reports (WRRs), Directors’ Reports (DRs), and Key Performance Indicators (KPI)s. However, the performance effects were partial. Three boundary conditions attenuated BVA’s mechanisms: a 40% price weighting that constrained qualitative differentiation, the omission of a formal Value-Added (VA) pathway for supplier-initiated optimisation, and the absence of continuous expert facilitation to support methodological fidelity. A documented execution-phase cost variance of approximately five percent further indicates residual volatility where key BVA complements are incomplete. The study integrates Principal–Agent theory, New Public Governance, and institutional isomorphism to explain why BVA’s governance architecture activated only in attenuated form and identifies the institutional conditions that moderate its effectiveness. While limited to a single revelatory case, the findings support analytical generalisation to similarly price-dominant, audit-driven procurement regimes in post-transition EU member states and offer practical guidance for evaluation design, innovation pathways, and facilitation models.