Formal Harmonisation, Informal Divergence: Administrative Culture in European Public Procurement

Abstract

Purpose: This study investigates whether three decades of EU procurement harmonisation have created unified administrative practices or whether document-observable patterns persist, demonstrating proof-of-concept for transforming procurement documents into strategic intelligence addressing cross-border B2B information asymmetry. Design/Methodology/Approach: This study analyses 2,370 software packages and information systems public procurement notices (CPV 48000000) from 17 European countries (2023-2024) using an automated Cultural Intelligence Extraction System (CIES), achieving 95.4% expert validation accuracy. One hundred one parameters across twelve dimensions are extracted. ANOVA with effect-size prioritisation (η²) and hierarchical multi-level clustering are employed. Findings: Document-observable patterns demonstrate extreme variation despite harmonisation: 8.8× verbosity range, 10.1× technical specification range, 5.6× scope range, with country explaining 50% of transparency variance (η² = 0.500). Multi-level clustering identifies five procurement strategies: Extreme Verbosity (Hungary), Multi-Domain Integration (Romania), Efficient Minimalism (a cross-regional cluster of Czech, Finland, and Switzerland), Western/Nordic Mainstream (comprising nine countries), and Eastern/Southern Moderate Complexity (Bulgaria, Greece, and Poland). Economic analysis reveals a multi-lot value paradox: 20% of tenders capture 72% of contract value, with a 294-fold variation in value concentration. Research Implications: The results validate the institutional theory's formal/informal distinction: EU directives harmonise legal procedures, while administrative practices diverge. Findings enable a document-to-intelligence transformation, providing suppliers with market selection guidance, authorities with efficiency benchmarking, and policymakers with frameworks to distinguish harmonizable dimensions from legitimate diversity. Originality/Value: Large-scale automated administrative pattern extraction demonstrating systematic document-to-intelligence transformation for B2B strategic positioning. A hybrid LLM-rule-based system with anti-hallucination validation establishes a methodological blueprint for extracting procurement intelligence at scale.