Author: nzz.ch

Executive Summary

Digitalization projects in the public sector fail systematically because politicians create unrealistic expectations and authorities pursue overly complex solutions. Prominent examples such as the social welfare software Citysoftnet in Zurich and Bern demonstrate: massive cost overruns, delays of years, and demoralizing employee experiences are the consequence. The problem is self-inflicted – through lack of complexity control, opaque project management, and ignoring of warnings.

People

Topics

  • IT project management in public administration
  • Digitalization initiatives and their implementation chaos
  • Cost overruns in government projects
  • Software rollout and organizational change
  • Governance and project controlling

Clarus Lead

The public sector regularly fails at major IT projects – not due to technical impossibilities, but because of strategic errors: politicians announce tailor-made universal solutions, authorities overload projects with special requests, and executives ignore warning signs. The Citysoftnet example illustrates the consequences: ten years of development, deficient software, faulty billing, and in Bern already eleven million francs in supplementary credits. For decision-makers this means: without consistent complexity reduction and close monitoring, the next disasters are already programmed.

Detailed Summary

The central problem lies in the failure to focus on what is achievable rather than what is desirable. Government officials without IT expertise create wish lists that grandiose politicians tout. The result is major projects that are supposed to link multiple platforms, replace outdated systems, and simultaneously offer new functions – overwhelming complexity.

Citysoftnet exemplifies this pattern. Announced in 2014 by Basel, Bern, and Zurich as a "Swiss standard," the software was supposed to enable paperless processes, better data protection standards, and simplified workflows. After more than a decade, employees of Switzerland's largest social welfare authority report using software that "cannot handle social welfare." Faulty billing, unreliability in elementary work steps – the problems in Bern should have warned Zurich. There, the system led to "sheer chaos" in 2023, with two offices on the brink of collapse. Today, reports speak of at least eleven million francs in supplementary credits.

Further examples reinforce the pattern: care homes incorrectly declared hundreds of seniors in need of care because the software could not represent other categories. At unemployment insurance offices, unemployment benefit payments were significantly delayed because a system migration brought technical problems. In the military, IT failures multiplied in digitalization platforms, airspace monitoring, and drone projects.

Key Statements

  • Too complex, too large, too poorly controlled: Projects fail through overwhelming requirements, lack of milestone checks, and insufficient transparency
  • Custom development instead of standards: Tailor-made solutions are more expensive and error-prone than proven standard products
  • Long project durations exacerbate the problem: Nine years to implementation means technology and organizational structures become outdated along the way
  • Warnings are ignored: Experts point to risks, management downplays them as "teething troubles"

Critical Questions

  1. Evidence: What objective success criteria (timeline, budget, error rates) were defined for Citysoftnet, and how were deviations documented? Are the reported problems based on systematic data collection or predominantly on anecdotal complaints?

  2. Conflicts of Interest: What economic incentives do software providers have to complete projects quickly versus continuously expanding them? Are there contract structures with cost guarantees or penalties for delays?

  3. Causality: Are the software problems primarily technical in nature, or is the main cause insufficient organizational preparation and employee training? Why have other Swiss cantons had better experiences with similar systems – or have they not?

  4. Feasibility of Solutions: How realistic is it for government officials to restrict themselves to standard solutions when their political principals sell universal solutions with specific citizen promises? Who bears responsibility for failure: principal, supplier, or project management?

  5. Control Mechanisms: Why have six-monthly reports to oversight bodies not led to earlier cancellation decisions so far? Are there consequences for failure?

  6. Data Security vs. Functionality: To what extent are increasing security requirements themselves a complexity driver that slows down projects, and how can these be reconciled with pressure for faster implementation?

Further Reports

  • Justice IT Scandal Zurich: Justice authority data ended up with a fringe figure – evidence of government nonchalance regarding data protection (affair not further dated)
  • UBS Project "Rigi": Private sector is no better – the major bank stopped a monster project after million-franc losses
  • Military Failures: Years of delays and cost overruns in drones, logistics, and airspace monitoring

Source List

Primary Source:

Baumgartner, Fabian: "There are no miracle solutions: When politicians tout IT projects grandiosely, it too often ends in disaster" – Neue Zürcher Zeitung, 26.02.2026 https://www.nzz.ch/meinung/it-chaos-bei-behoerden-angebliche-wunder-loesungen-enden-zu-oft-im-debakel-ld.1925390

Verification Status: ✓ 26.02.2026

This text was created with the support of an AI model.

Editorial Responsibility: clarus.news | Fact-Check: 26.02.2026