How the model works

ScenarioPulse runs one modeling pipeline in four stages. Each stage is its own kind of scenario, and later stages build on top of earlier ones.

The four stages

1. Demography — projects the population forward by age and sex, given fertility, mortality and migration assumptions. See how to model demography.
2. Coverage — turns a demographic projection into headcounts: how many people contribute to the pension system (by regime) and how many draw a benefit (by benefit type), each year.
3. Macro — GDP growth, inflation, and the wage/pension indexation rates that carry incomes and benefit amounts forward year over year.
4. Finance — combines Coverage’s headcounts with Macro’s rates to compute contribution revenue, benefit expenditure, the deficit or surplus, the intergenerational reserve, and the present value of obligations to current beneficiaries.

See how to model pensions for the screen-by-screen version of stages 2–4.

Scenarios chain by pinning

A Coverage scenario is built on top of a specific Demography scenario; a Finance scenario is built on top of a specific Coverage scenario and a specific Macro scenario. You choose which one to build on by pinning it (📌 next to a scenario in its list) — a pinned scenario becomes available as a “base” the next stage can pick from.

Nothing stops you from pinning several Demography scenarios and building several Coverage scenarios on different ones, or building several Finance scenarios that mix different Coverage/Macro combinations — that’s exactly how you isolate the effect of one assumption while holding everything else fixed. In Pensions/Outputs, a small chip strip under the country selector always shows which Demography, Coverage, Macro and Finance scenario the chart you’re currently looking at is built from — click a chip to jump to that layer.

One workflow for every scenario type

Whichever stage you’re working in, the workflow is the same:

• Change one assumption, keep the rest fixed, and click Add scenario — each run adds a line to that layer’s chart and an entry to its scenario list, so you accumulate a fan of forecasts.
• Show/hide scenarios with the checkbox next to each one, and use the Outputs tab’s charts to compare them.
• Export any chart as CSV — no sign-in required.
• Save a scenario to your account (sign-in with Google or LinkedIn) and Share a public, read-only link — this works identically for Demography, Macro, Coverage and Finance scenarios. See saving and sharing.

How accurate is this?

Each country’s baseline scenario — the one every other scenario starts from before you change an assumption — is checked against real, published official statistics at its own base year (2026 for Spain, 2024 for the USA): contributor and beneficiary headcounts by type reproduce the official totals almost exactly, and computed aggregates like total contribution revenue and total benefit expenditure land within a documented tolerance (typically within 5–10%, and for the USA, the Trust Fund’s real projected depletion year — 2034, per the Social Security Trustees’ own report — reproduces exactly). These checks run automatically with every change to the project and are re-verified whenever a country’s underlying data is updated.

That check covers the base year — and, for the USA’s reserve, roughly a decade out. It does not extend to the rest of a projection. A typical scenario runs 25 years or more past the base year, and nothing that far out can be checked against reality, because it hasn’t happened yet. Past the base year, a chart shows what follows from the assumptions you picked — demographic, macroeconomic, and policy — not a verified prediction. The further out you look, the more the result reflects those choices rather than any fact that can be checked today. That is the entire premise of comparing several scenarios side by side rather than presenting one number as “the future”: every projection is an argument about which assumptions are reasonable, and the tool’s job is to make that argument’s consequences visible, not to settle it for you.

Present Value, briefly

The Present Value view (Pensions/Outputs → Finance → Present Value) answers a different question than the Deficit/Reserve charts: instead of one year at a time, it sums all future payments to today’s beneficiaries, discounted back to today, and groups the total by birth year. The discount rate defaults to the picked Macro scenario’s own nominal GDP growth — the standard choice for an unfunded, pay-as-you-go pension system, since it has no funded portfolio return to discount against — or you can set a flat override rate instead.