How we estimate the cost of raising a child — and what the numbers do (and don't) mean.
Our offline model is a deterministic, structured estimate — it does not call a large language model and does not read live external data. It is designed to give a defensible planning range, not a quote. The model has five moving parts:
Every city has a preset 0–18 base cost in USD for one child on a moderate lifestyle with public schooling. This is the anchor of every estimate and is converted to local currency with a fixed exchange-rate table.
Rather than one global split, we assign each country a cost-structure archetype (e.g., high-housing China/Asia, welfare-state Europe, Gulf, developing markets). Each archetype defines how the base cost is divided across the 8 categories — so a Tokyo estimate spends more on housing and juku, while a Berlin estimate spends more on festivals and less on healthcare.
Choosing private/international schooling adds a region-specific premium on top of the public base — roughly 35%–75% of the base depending on country (highest in the Gulf and Singapore, lowest in welfare-state Europe). The premium is counted entirely in the Education category, not smeared across all categories.
The public portion scales with family size: 2 children ≈ 1.7× (≈15% per-child saving) and 3 children ≈ 2.3× (≈23% saving). Savings come from shared bedrooms, hand-me-downs, sibling discounts, and bulk buying. The private premium, however, is charged per child (tuition is not shared).
Undergraduate cost (age 18–22) is calculated separately and never folded into the 0–18 total, because it is optional and varies enormously by country and by local vs international student status.
All category amounts are computed in USD and sum exactly to the total. Local-currency figures use a fixed rate table (1 USD → local). CNY is shown at a fixed ≈7.2 for reference only. Treat FX as approximate.
Our published figures are our own modeled estimates, not data reproduced from any external organization. We benchmark our approach against publicly available cost-of-living research and national household-expenditure surveys — listed as reference frameworks on the Sources page. We are not affiliated with those organizations, and we do not republish their proprietary datasets.
These are structural estimates, not real quotes. Individual circumstances (neighborhood, school choice, spending habits) shift actual costs by 10–30%. We refresh the database quarterly with updated exchange rates, housing indices, and policy changes. Current figures reflect 2026-09-07 estimates.