A reform proposal — and why doing nothing is the expensive option
Austria runs more than 30 distinct personal identifiers and over 50 separate logins across its administration. A single person moving house, changing employment or starting a business crosses a dozen systems that do not know about one another. This is not a usability problem. It is an architecture problem with a price tag, and it is measurable.
Fragmentation is structural, not accidental. Identifiers were created per domain, by different statutes, at different times: the central register number, the source identification number, sector-specific personal identifiers, the social insurance number, company and association numbers. Each is legally sound in isolation. Together they produce a system nobody designed.
Federal structure multiplies it. Nine provinces run their own portal ecosystems — five in a shared association, four independently. Moving between provinces adds six to ten further identifiers and logins.
Electronic identity does not resolve it. Austria's ID Austria reached 4.8 million users by December 2025 and works well for what it does. But it is a login method, not an identifier consolidation: it covers roughly a fifth of the procedures a person encounters, and leaves the underlying identifier landscape untouched.
The cost falls on those least able to bear it. Non-take-up of means-tested benefits runs at roughly 50 % across Europe. Complexity is a filter, and it filters by capacity to navigate bureaucracy — not by need.
| Item | Figure | Basis |
|---|---|---|
| Investment and migration, six years | EUR 1.9–3.1 bn | three-scenario cost model |
| Annual savings from year 7 | EUR 0.5–1.3 bn | procedure-level bottom-up |
| Break-even | 7–10 years | including pessimistic scenario |
| Cost of inaction over 15 years | EUR 8–20 bn | legacy maintenance, EU penalties, lost productivity |
| EU penalty framework 2026 | EUR 3,560/day × 0.87 | Eur-Lex C/2026/1653, Austrian n-factor |
| Pilot municipality, first year | approx. EUR 680,000 | EU-fundable, no national budget line |
All assumptions are documented and adjustable; a sensitivity analysis across twelve variables is part of the corpus. The proposal does not depend on the optimistic scenario holding.
Every reform of this kind produces the same failure mode, documented in the Netherlands (SyRI, childcare benefits), Australia (Robodebt) and elsewhere: automated decisions cause harm at scale, and nobody is answerable. The people with disposition over the system bear no risk; the people bearing the risk have no disposition.
Five of these are already anchorable in existing law. Two require the legislature. All eight are written to be usable as procurement and authorisation criteria, not as principles.
A corpus of 41 documents under CC BY-SA 4.0: the main proposal, 26 annexes covering discretion, architecture, costs, liability, constitutional review, sensitivity analysis, international comparison (Germany's OZG, Switzerland's two e-ID referendums, Estonia, Finland, Denmark, France, UK), human rights case law, political economy, civil society, security operations and implementation; a citable source apparatus; interactive cost and simulation tools; and a peer-reviewable working paper on the attribution problem.
Written without commission, without funding, without fee, by one person. Everything is open for reuse, adaptation and contradiction.
Printing: A4, one page. Use the browser print dialogue — background graphics are not required.
German version: one-pager.html