You hear it a lot—the pension system can't be relied upon: due to demographic trends, currency devaluation, and an expanding state, one shouldn't count on pensions being paid out.
In the case of the Swiss pension system, however, I argue that the system is not in danger of collapsing, even if benefits (for certain groups) will become smaller.
In a "bank run," depositors withdraw their funds when they fear the bank lacks sufficient substance. This worsens the bank's position, which in turn spurs other depositors to withdraw their money.
What would a "bank run" on the pension system look like? People would try to pull their money out of the system. How is this possible? The following options exist, among others:
- when moving away from Switzerland
- when purchasing real estate
- when starting a business
- upon retirement
The fact that you can withdraw capital from the pension system actually makes it attractive: an additional incentive to pay into the system. But in a run, when many people do this at once, it becomes a problem. How can the state counteract this? With taxes (pun intended)—they're already levied today, varying by canton, but generally much lower than income tax. It would then suffice to equalize the withdrawal tax with income tax, with cantonal adjustments, to seal the leak.
Furthermore, the rules can be tightened:
- when moving away from Switzerland: only outside the EU
- when purchasing real estate: only the mandatory portion can be withdrawn
- when starting a business: not necessary—few people do this anyway
- upon retirement: only a maximum of 50% as a lump sum
So today the state has sufficient means to prevent a run.
However, it doesn't have to come to a run to destabilize the system. The system can also "leak" substance over time until it eventually stands at the edge of a cliff. If the system leaks long enough, a run also becomes more likely. What actually is a "leak"? It's actually part of the system design that the money gets paid back at some point. One can only speak of a "leak" when more is paid out than paid in. This isn't easy to determine definitively—it also depends on the assumptions. But what generally leads to a leak: - life expectancy decoupled from the minimum conversion rate - minimum conversion rate decoupled from the nominal interest rate - retirement age decoupled from life expectancy
The current minimum conversion rate of 6.8% was decided (*) at a time (2003-2005) when life expectancy was about 3 years lower than today. The original conversion rate in 1985 was 7.2% with a life expectancy that was 4-5 years shorter than in 2003. In 2024, voters rejected at the ballot box a proposal to lower the rate to 6%.
The relevant figures over the last 40 years as a table:
|------|------------------------|---------------------------|-------------------|-----------------|------| | Year | Life expectancy (M/F) | Remaining life exp. (M/F) | Retirement age (M/F) | Conversion rate | Interest | |------|------------------------|---------------------------|-------------------|-----------------|------| | 1985 | 76.9 (73.5/80.2) | 16.5 (14.5/18.5) | 65 / 62 | 7.2% | 4% | |------|------------------------|---------------------------|-------------------|-----------------|------| | 2005 | 81.3 (78.7/83.9) | 20 (18.1/21.8) | 65 / 65 | 7.1% | 1% | |------|------------------------|---------------------------|-------------------|-----------------|------| | 2025 | 84.5 (82.7/86) | 22.1 (20.6/23.5) | 65 / 64.5 | 6.8% | 0% | |------|------------------------|---------------------------|-------------------|-----------------|------|
(*) The 2003 decision was to gradually reduce the rate from 7.2% to 6.8% over 20 years (2005-2025).
From a systems perspective, the conversion rate was already reduced too little in 2003 (technically wrong, but that doesn't mean it was "fair" or "socially" wrong). The fact that interest rates in recent years have been near zero exacerbates the problem (the system cannot be stabilized with extra returns).
One can see that political resistance is too great to defuse the situation, so how does the system remain stable? By paying out less from Pillar 2b ("extra-mandatory"). Theoretically, one could set the minimum conversion rate for the extra-mandatory portion to 1% (if not 0%...).
This means there's essentially redistribution between sub-pillars 2a and 2b.
While the contradiction can be resolved within the current framework, it could still become a problem in the long term. Here are additional payments into and out of the system that are built in:
- voluntary purchases into Pillar 2 (2b)
- increased contributions to Pillar 2 (2b)
- 80% work schedule
In the long term, more and more employees are forgoing voluntary purchases (or making smaller ones than before), paying only the minimum into Pillar 2, and reducing their work schedule (going from 100% to 80% is relatively easy). So currently, sub-pillar 2b is losing attractiveness.
At the same time, one shouldn't lose sight of the rest of Europe. Even though the Swiss pension system has certain problems, they're even more serious in the rest of Western Europe. The relatively better position strengthens overall attractiveness.
What are the possible options to defuse the problem? Raising the retirement age and lowering the minimum conversion rate meets political resistance that is nearly impossible to overcome in the current political landscape. Proposals keep coming that further throw things out of balance: for couples, for women, etc. This may be fair and desirable, but from a systems perspective it must be balanced: through taxes/levies, currency devaluation, redistribution, etc. At the same time, a purely technical view is absolutely wrong—otherwise one might come up with the idea of shortening life expectancy, which from a purely technical standpoint would bring the system back into balance.
The easiest way out is naturally qualitative growth (i.e., through productivity gains).
At the same time, however, new challenges arise. The pension system was focused on lifelong employment, with a different family structure and role for women. Due to many factors, it's harder today to find work after 50, while current labor law additionally diminishes the attractiveness of this age category: you have to pay much higher contributions than for younger employees, and termination protection is stronger. This is socially fair and worked well in the world of lifelong employment—but no longer today.
Technological and geopolitical challenges complete the overall picture. If AI takes over more and more work while pension contributions continue to come from employment income, who is supposed to pay for future pensions? With the current system and continuation of trends, for example, 80% of retirees would end up in Pillar 1, which would then be cross-financed (e.g., from value-added tax, like the newly approved 13th AHV payment).
The Swiss pension is secure, but on a smaller scale, and for lower incomes.