Cost research methodology

Content and linked guidance reviewed on 14 October 2026. Each city source retains its own data period and checked date.

Our purpose is to show what a source establishes, what it leaves open and how to turn that into a personal budget. Listing a city does not mean that a complete local cost survey exists.

Two kinds of evidence

Fellowship-location evidence identifies a destination from a programme or sister directory. Living-cost evidence supports a particular amount or allowance. The two are recorded separately. Housing and service links are practical starting points unless the cost table explicitly cites them as evidence.

Six categories, with visible gaps

We organize recurring costs into housing, food and daily needs, transport, utilities and phone, personal and leisure, and health insurance. Some source baskets combine expenses differently; notes record those differences rather than silently making their totals comparable.

• Range: the source supports both bounds, or a documented calculation produces them.
• Point estimate: one published amount. It remains one amount.
• Minimum-only: a stated starting amount without a supported upper bound.
• Partial: the available figure covers only part of the category.
• Research pending: no suitable priced evidence is recorded. It is not zero.

Scope and dates matter

City-specific guidance relates to the named location; regional and national benchmarks have broader scope. A national student allowance is not a market-price survey of every city. An editorial scenario is explicitly identified as a planning assumption, not attributed to an official source.

The checked date records when a source was reviewed. The data period records what time the figures describe. An undated page remains undated even when checked today. Forecasts and older surveys keep their own periods. No update date implies every provider price is current.

Source selection and practical limitations

We prefer host institutions, universities, governments, public transport operators and providers. We keep currency, units, eligibility and conversion assumptions. A source may describe enrolled students rather than visiting staff, subsidized rooms rather than private rentals, or an insurance minimum rather than complete cover.

Lower, midpoint and upper controls apply supported bounded figures. Missing or minimum-only entries require your own input; they are not completed with zero. A source’s total may contain tuition or omit a planner category, so it is shown separately from the total of your own inputs.

What the calculator calculates

The monthly balance compares entered stipend with entered recurring costs. The stay funding gap uses the chosen duration. Arrival cash estimates deposits, a payment-delay period and one-off costs; it is a separate liquidity view. Do not add it indiscriminately to the full-stay living total because some spending appears in both.

Exchange rates come from Frankfurter ↗, a bundled fallback or your manual entry. Dates are displayed by currency. Bank spreads and transfer fees are not included. Saved budgets are personal inputs, not new research evidence.

See category coverage, editorial standards and the stipend comparison guide.