Sixty or Seventy

In 2025, the Romanian state paid 35.44 billion lei — 1.85% of GDP — to cover the gap between what the public pension system collects in contributions and what it pays out to pensioners. Special pensions for military personnel, police, and intelligence officers, paid separately from ministry budgets, cost a further 0.77% of GDP in full. EU treaties set the deficit ceiling for a member state's entire budget at 3% of GDP. The two figures above, added together, reach 2.62% of GDP: thirty-eight hundredths of a percentage point short of that benchmark.

You can test any of the four levers below — retirement age, special-pension growth, net migration, the fertility rate — and the engine recalculates, year by year through 2060, the population pyramid, the contributor-to-pensioner ratio, and the point where the burden crosses the line.

Three systems, one word

Where the 3% threshold comes from — and why it matters

Three percent of GDP is the deficit target Romania must gradually bring its entire state budget down to, under the EU's excessive-deficit procedure: the ceiling covers everything the state spends beyond what it collects — salaries, roads, healthcare, defense, all of it. Romania is currently under that procedure precisely because its overall deficit still exceeds it.

The comparison above shows how little room that would leave once Romania gets there. If pensions alone already cost, net of contributions, nearly as much as the entire final target, reaching that target without pension reform would mean the rest of the state — salaries, roads, healthcare, defense — running on essentially no deficit at all.

What the engine moves

The engine starts from Romania's actual population pyramid (Eurostat, single-year-of-age, 2025) and ages it forward year by year using a real mortality table (Eurostat, 2023) and four levers you control:

• Retirement age — Law 360/2023 sets 65 for men (already reached) and a phased rise for women, from 63 today to 65 by 2035. The engine uses one unified retirement age as a declared simplification: move the slider to test a symmetric reform, not the exact birth-month schedule in the law's annex. It's the slider that changes the answer the most, because it instantly shifts the line between who pays in and who draws out.
• Special-pension growth — between 2023 and 2024, the cost of military pensions rose 14.9%, police pensions 12.8%, intelligence-service pensions 16.7% — far outpacing ordinary pension indexation. The slider controls how far this premium runs above standard indexation.
• Net migration — after a decade of mass emigration, Romania has logged three consecutive years of positive net migration: +97,114 (2022), +66,065 (2023), +36,200 (2024) — a positive but narrowing trend. The slider also lets you test a return to net emigration.
• Fertility rate — Romania's rate fell from 1.54 (2023) to an estimated 1.39 children per woman (2024), the sharpest year-on-year drop in the EU. This slider can never change the number of pensioners in this window — everyone who will retire by 2060 is already born — but it does change the number of contributors from the 2040s onward, as today's children reach working age: across the slider's full range, that difference reaches nearly a million contributors by 2060. The effect is real, but it arrives much later and builds much more slowly than retirement age, which reclassifies millions of people in a single move. A child born in 2027 can affect contributions only around 2047, at age 20. The slider shows the demographic effect of a given fertility rate; the cost of a policy designed to change that rate is a separate calculation.

What the engine does not simulate

The model is a declared simplification, not an official forecast.

At every selected retirement age, the engine moves the boundary between the working-age and retired populations immediately. For ages above 65, it assumes that people aged 65–69 join the contributing population from the first projected year at the same average contribution rate as everyone else, and no longer counts them as pensioners. There is no effective date, transition period, or age-specific employment rate for people close to retirement; in reality, employment falls at older ages, so the effect of a higher retirement age may be overstated, especially in the short term.

Pillar II and Pillar III don't enter this calculation: they're individually accumulated accounts that never pass through the state budget and don't, by themselves, reduce the subsidy the state pays today's Pillar I pensioners. They'll matter for future retirees' own income, but they don't change the fiscal burden calculated here — the two are answers to different questions. Mortality stays fixed at the 2023 table — it doesn't assume the medical progress of coming decades, unlike the European Commission's own projections. The "formalization rate" (how much of the working-age population actually contributes) and the "coverage rate" (how much of the older population actually draws a pension) are calibrated once, on 2025 data, and held constant — the engine doesn't model shifts in informality, labor migration, or eligibility reform. Average wages, average pensions, and GDP all grow along one assumed macroeconomic path (inflation converging toward 2.5%, real growth converging toward 1.5% a year), not a separate slider — that path doesn't change when you move migration or the retirement age, because the engine doesn't assume more active workers grow GDP, or that emigration shrinks it. The result is a coherent exploration of four real demographic and fiscal levers — not a point prediction of the exact year.

What the official data already show

The European Commission's 2024 Ageing Report — which uses the same demographic assumptions as its baseline for every EU country — projects Romania's old-age dependency ratio (population 65+ over population 20-64) rising from 33.5% in 2022 to a peak of 60% in 2056. In the same projection, the public pension system's balance, currently -2.6% of GDP, deepens to -5.4% of GDP by 2047 before easing later in the century, as already-legislated reforms — the retirement age, the indexation formula — take effect. The engine above recalculates this kind of trajectory starting from your own assumptions, not the Commission's single scenario.