Pension systems across the developed world are heading toward a dependency ratio that roughly doubles by 2050: from about 28 elderly people per 100 workers today to something closer to 53. Italy’s pension spending is on a path from roughly 16% of GDP to 28% by 2036. Japan is projected to reach 81 elderly per 100 workers by mid-century. These aren’t projections from my own research, they’re publicly available demographic data, and I cite them here only to set the stakes, not to claim I discovered them.
The usual response to numbers like these is political: raise the retirement age, cut benefits, argue about who deserves what. I don’t think that’s the right frame. A demographic shift is not an opinion, and the size of the gap it creates is not something a one or two year bump in retirement age can close. Most proposed fixes are scaled for a crisis a fraction the size of the one actually coming.
The mechanism I’m working on instead is simple to state, even if the modeling underneath it isn’t: instead of paying for pensions out of current contributions forever (the pay-as-you-go model every major system uses today), build a fund large enough that its investment returns alone cover the annual cost, permanently.
Fund × Annual Return Rate = Annual Solution Cost
Once a fund reaches that size, the system stops depending on the ratio of workers to retirees at all. It runs in three phases: a building phase, funded by a fixed contribution rate, growing the fund through compounding returns; a transition phase as the fund approaches its target size; and a self-sustaining phase, where returns alone cover the cost every year, indefinitely. The dependency ratio can keep getting worse and the fund doesn’t care, because it was never depending on next year’s workers to pay for this year’s retirees in the first place.
I want to be honest about where this actually stands. There’s a full book outline (aimed at a general reader, no equations, the plain-language case for why this matters and how it would work) and an academic paper structure aimed at peer-reviewed publication. But most of the paper itself is still a placeholder. The idea is real and the framework is built; the empirical validation, the multi-country modeling, the actual manuscript, that work is in progress, not finished. This page will update as it moves.