The question is not whether we can afford a basic income. The question is whether we can afford not to have one.
Within two years, a new popular initiative for a citizens' dividend must be launched. Funding comes from an AI value creation levy. The transition occurs in three phases and replaces OASI in the long term.
OASI is based on wage contributions. When AI takes over the work of accountants, analysts and lawyers, the contributions disappear -- while the benefit entitlements remain. OASI was created in 1948 for a world in which people were employed for forty years and then drew a pension for fifteen [1].
Two developments are undermining this model:
OASI currently pays out around 50 billion francs per year. The fiscal gap from AI displacement could grow to 75 to 116 billion by 2050 [5].
The idea is not left-wing. Milton Friedman -- Nobel laureate, father of monetarism, advisor to Ronald Reagan -- proposed the negative income tax in 1962: anyone below a threshold income receives a payment, without means testing [6]. Friedrich Hayek endorsed a basic income for the same reason: it creates freedom from the state, not dependence [7].
Switzerland has a unique cooperative tradition: Migros, Coop, Raiffeisen, Mobiliar. An AI citizens' fund based on the cooperative principle would be the logical continuation:
| Phase | Period | Citizens' Dividend | OASI Status |
|---|---|---|---|
| 1 | 2028--2035 | approx. CHF 500/month as supplement | OASI remains |
| 2 | 2035--2045 | rises to CHF 2,500/month | OASI integrated into dividend |
| 3 | from 2045 | universal dividend | OASI, DI, social assistance replaced |
The BVG pension (occupational pension) continues as capital-funded supplementary income -- it is financed from assets, not from wage contributions.

Description: Three-phase transition from OASI to citizens' dividend. Phase 1 (2028–2035): ~500 CHF/month supplement. Phase 2 (2035–2045): rises to ~2'318 CHF/month. Phase 3 (from 2045): universal dividend replaces OASI. Population growth: 6.9→8.1 million adults.

| Item | Amount |
|---|---|
| Gross UBI (CHF 2,500/month x 9 million adults + CHF 1,500 x 1.5 million minors) | CHF 297 billion |
| Deduction: eliminated social transfers (OASI, DI, supplementary benefits, social assistance) | -CHF 80 billion |
| Deduction: income tax on UBI, consumption reflux | variable |
| Net funding requirement | CHF 150--180 billion |
| Current wage-related levies (which will disappear by 2050) | CHF 220 billion |
The AI value creation levy, a financial transaction tax and the reallocation of existing social budgets could cover the net requirement.
| Study | Result |
|---|---|
| Finland (2017--2018, KELA) | Recipients were happier, healthier and just as likely to be employed [10]. |
| Stockton, USA (SEED, 2019--2021) | Full-time employment rose from 28 to 40 per cent [11]. |
| Kenya (GiveDirectly, since 2016) | Long-term positive effects on entrepreneurship and health [12]. |
On 5 June 2016, Switzerland became the first country to vote on a UBI: 76.9 per cent said No [13]. Ten years later, the situation has fundamentally changed. ChatGPT did not exist. KOF ETH had not documented any measurable labour market effects. The singularity was science fiction. The world of 2016 no longer exists.
[2] FSO, Population Scenarios 2023--2055, Reference Scenario A-00-2025.
[3] McKinsey Global Institute: A Future That Works: Automation, Employment, and Productivity. 2017.
[4] KOF ETH Zurich, Labour market effects of generative AI, 2023.
[5] FSO; SNB; SECO; McKinsey; own model calculation.
[6] Friedman, Milton: Capitalism and Freedom. University of Chicago Press, 1962.
[8] Alaska Permanent Fund Corporation.
[9] Norges Bank Investment Management, Government Pension Fund Global.
[10] KELA (Finnish Social Insurance Institution): Final Report UBI Experiment, 2020.
[11] Stockton Economic Empowerment Demonstration (SEED): Final Report, 2021.
[12] GiveDirectly, Long-term study on basic income in Kenya.