Case Overview: Belgian Government Procurement System

Case Metadata

• Case ID: case-005-belgian-procurement-system
• Status: Active Investigation
• Date Opened: 2026-01-28
• Last Updated: 2026-01-28
• Priority: High
• Lead Investigator: DOGE Europe Research Team

Executive Summary

Belgian government IT procurement is systematically broken. Across multiple high-profile failures — Persona (€16M), iPolice (€28M), Belgian Justice digitalization (€51.8M) — the same patterns emerge: multi-year tender cycles, scope creep, vendor lock-in, and projects that deliver a fraction of promised functionality at multiples of the original budget.

This investigation examines the structural failures in Belgian public procurement, focusing on:

• The framework contract system (raamcontracten) that concentrates IT work among a small group of vendors
• The procurement regulations that prioritize process compliance over outcomes
• The lack of accountability mechanisms for failed projects
• The expertise gap between government and its IT contractors

Key Finding: Belgium's procurement system is designed to minimize legal risk, not to deliver working software. The result is billions in wasted taxpayer money and critical government systems that don't work.

Key Figures

The Problem

1. Multi-Year Procurement Cycles

Before a single line of code is written, Belgian government IT projects must navigate:

• Needs assessment and approval (6-12 months)
• Tender preparation and legal review (6-12 months)
• Publication and bidding period (3-6 months)
• Evaluation and award (3-6 months)
• Contract negotiation (3-6 months)
• Mobilization and onboarding (3-6 months)

Total: 2-4 years from problem identification to development start

By the time development begins, requirements are outdated, stakeholders have changed, and the original problem may have evolved.

2. Framework Contract Oligopoly

Both federal and regional governments use framework contracts (raamcontracten) that pre-select a small group of "preferred vendors" for all IT work. This creates:

• Vendor concentration: The same 4-6 companies (Cronos, Cegeka, DXC, Atos, NRB, Proximus) win virtually all major contracts
• Reduced competition: Mini-competitions within framework contracts lack true competitive pressure
• Knowledge asymmetry: Vendors understand the system better than their government clients
• Accountability gaps: When projects fail, the framework contract continues

3. Process Over Outcomes

Belgian procurement law (implementing EU directives) focuses on:

• Procedural compliance
• Equal treatment of bidders
• Formal evaluation criteria

What it doesn't measure:

• Actual delivery of working software
• Value for taxpayer money
• Vendor track record on previous projects

A vendor can fail spectacularly on one project and immediately win the next.

4. The Expertise Gap

Government IT departments are chronically understaffed and underpaid compared to the private sector. This creates:

• Dependence on external consultants for basic project management
• Inability to evaluate vendor proposals technically
• No capacity to challenge vendor estimates or timelines
• Asymmetric negotiating positions in contract disputes

Case Studies

Persona Project (Flemish Government)

• Budget: €16M spent, €10M more requested
• Delivered: 8-10% of functionality
• Duration: 5 years
• Status: Halted
• Vendors: Cronos, DXC-Cegeka, Atos, Proximus
• See: Case 004

iPolice (Federal Police)

• Budget: €28M+ (original estimate: €10M)
• Delivered: Partial functionality with critical bugs
• Duration: 7+ years
• Status: Troubled
• See: Case 003

Belgian Justice Digitalization

• Budget: €51.8M+ across multiple projects
• Delivered: Fragmented systems, no integration
• Duration: Ongoing failures over 15+ years
• Status: Perpetual remediation
• See: Case 002

Key Questions

Primary Investigation Questions

1. Systemic Failure: Why does the same procurement process produce the same failures across different agencies, governments, and vendors?

2. Accountability Void: What mechanisms exist to hold vendors accountable for failed projects? Why don't they work?

3. Framework Contracts: Do framework contracts serve the public interest, or do they create protected markets for incumbent vendors?

4. Cost of Compliance: How much of project budgets goes to procurement compliance vs. actual software development?

5. International Comparison: How does Belgium compare to other EU countries? Are there models that work better?

Secondary Questions

1. What is the total annual cost of failed government IT projects in Belgium?

2. How many public servants work in IT procurement vs. actual IT development?

3. What role do consulting firms play in perpetuating the current system?

4. Are there successful government IT projects? What made them different?

5. What would it take to reform the system?

Key Entities

Regulatory Framework

• Wet Overheidsopdrachten (2016) — Belgian Public Procurement Act
• EU Directive 2014/24/EU — EU Public Procurement Directive
• KB Plaatsing (2017) — Royal Decree on placement procedures

Federal Procurement

• FOD BOSA — Federal Public Service Policy and Support (manages federal IT procurement)
• e-Procurement — Federal electronic procurement platform
• Rekenhof / Cour des comptes — Court of Audit

Flemish Procurement

• Digitaal Vlaanderen — Flemish digital agency (manages ICT framework contract)
• Audit Vlaanderen — Flemish internal audit
• Vlaamse ICT Raamcontract (2022-2029) — ~€1B framework contract

Major IT Vendors (Framework Contract Participants)

• Cronos Group — Project management, development
• Cegeka — Development, infrastructure
• DXC Technology — Development, outsourcing
• Atos — Development, infrastructure
• NRB — Development (Wallonia focus)
• Proximus — Telecommunications, development

Oversight Bodies

• Rekenhof / Cour des comptes — Federal/regional Court of Audit
• Federal Parliament — Oversight committees
• Regional Parliaments — Regional oversight
• Ombudsman — Citizen complaints

Investigation Scope

In Scope

• Framework contract structures and outcomes
• Procurement timelines and their impact
• Vendor performance tracking (or lack thereof)
• Cost overruns and delivery failures
• Legal and regulatory barriers to reform
• International comparisons and best practices
• Recommendations for structural reform

Out of Scope

• Individual contract disputes (unless systemic)
• Classified/defense procurement
• Non-IT procurement
• Municipal procurement (too fragmented)

Deliverables

1. Procurement Process Analysis: Mapping the full procurement lifecycle with time and cost analysis

2. Framework Contract Deep Dive: Analysis of the Flemish ICT framework contract as a case study

3. Vendor Performance Database: Track record of major vendors across government projects

4. International Benchmark: Comparison with UK (GDS), Estonia (e-Estonia), Netherlands (DICTU)

5. Reform Proposals: Concrete recommendations for legislative and administrative changes

6. Open Source Alternatives: Identify opportunities for government adoption of open-source solutions

Initial Hypotheses

1. The system works as designed — Procurement law is designed to prevent corruption and ensure fairness, not to deliver working software. Reform requires changing the objectives.

2. Vendor incentives are misaligned — Fixed-price contracts incentivize scope reduction and change requests; time-and-materials contracts incentivize delays. Neither incentivizes delivery.

3. No institutional memory — Each failed project is treated as unique, preventing pattern recognition and systemic learning.

4. Political economy of reform — The current system benefits incumbents (large vendors, procurement lawyers, consultants). Reform faces organized opposition.

Research Methods

• Parliamentary document analysis (questions, committee reports, audit reports)
• Freedom of Information requests (openbaarheid van bestuur)
• Public contract database analysis
• Expert interviews (former civil servants, vendors, academics)
• International comparative research
• Quantitative analysis of procurement data

Related Investigations

• Case 002: Belgian Justice Digitalization
• Case 003: iPolice
• Case 004: Persona Project