Methodology

This is the complete version, and it is the authoritative one: the arithmetic with real numbers, the boundary of each of our three data sources, how confidence scores are assigned, the rules we apply to reader-submitted prices, and the exact origin and retrieval date of the underlying dataset. If you are citing this site in research or journalism, this is the page to read.

1. What the Big Mac Index measures

The Big Mac Index was introduced by The Economist in 1986 as an accessible illustration of purchasing-power parity (PPP). The theory says that in the long run, exchange rates should move towards the rate at which an identical basket of goods costs the same in any two countries. The index replaces the basket with a single item — a Big Mac — on the argument that it is a near-identical product, assembled locally from local inputs, sold in more than 100 countries.

The output is not a price comparison. It is a statement about a currency: if a Big Mac costs more in dollar terms in country X than in the United States, then by this measure country X's currency is overvalued against the dollar, and the percentage gap is the size of that overvaluation.

2. The calculation, with a worked example

Two steps, no adjustments:

Here it is on the euro area entry from the 1 July 2026 survey, which is the reference point for the euro-area countries we research individually:

• Euro area local price: €6.19
• Market exchange rate that day: 0.87439 EUR per USD
• Dollar price: €6.19 ÷ 0.87439 = $7.08
• United States dollar price the same day: $6.22
• Valuation: ($7.08 − $6.22) ÷ $6.22 = +13.8% — the euro is overvalued against the dollar on this measure

The implied-PPP route gives the same answer from the other direction. Dividing the two local prices, €6.19 ÷ $6.22 = 0.9952 EUR per USD, is the exchange rate at which a Big Mac would cost the same in both places. The market rate that day was 0.87439, so the euro is trading roughly 13.8% stronger than burger parity implies. Both routes are arithmetic on the same two numbers, which is the whole appeal of the index: nothing is modelled, weighted, or smoothed.

Check this example in the calculator — the link opens it already set to United States → Euro area on $100, so it lands on the same +13.8% figure derived above, next to the euros that $100 actually buys at the market rate and the euros it would buy at burger parity. Changing either country re-runs the identical two steps on that pair.

The dataset we publish also carries The Economist's GDP-adjusted series — the adj_price and *_adjusted columns described on the data download page — which regresses dollar prices against GDP per person and reports the gap from that trend line. That version answers a different question, namely whether a burger is expensive for a country at this income level, and it is the more defensible comparison between rich and poor economies. The headline figures on this site are the raw index, not the adjusted one.

Which exchange rate is used

By default, every dollar price on this site uses the exchange rate recorded in the survey release itself — the rate on the survey date, the same one used for every other country in that release. That is what keeps the ranking internally consistent, and it is what you see with JavaScript disabled or on first load.

The Live Rates toggle in the header switches this: with it on, prices are recalculated in your browser against current exchange rates pulled from a third-party FX API and cached locally for an hour. That answers "what is this burger worth in dollars today", which is a fair question, but it is no longer the published index — local prices stay frozen at the survey date while the currencies move, so gaps drift for reasons that have nothing to do with burger prices. The toggle is off unless you turn it on.

3. Three data source tiers

Not every country on this site comes from the same place, and pretending otherwise would be the single most misleading thing we could do. Every country record carries a source_tier field, and the tier determines what the number means, how much to trust it, and what we are willing to claim about it.

Economist tier — the synchronised survey

The Economist publishes the index twice a year, in January and July, covering 54 entries in the current release. These are the numbers we treat as canonical, and the reason is synchronisation rather than accuracy: every country in that release is priced on the same date, using the same definition of the product, converted at the same day's exchange rates. That is what makes a cross-country ranking meaningful at all.

It also explains a rule that surprises people: we do not refresh a single Economist country with today's menu price, even when we know today's menu price. Swapping one country to a later date while the other 53 stay on the survey date means that country is no longer being compared like for like — in a high-inflation market it would appear more expensive than reality, purely because of the date mismatch. Where we have a newer observation, it is published as an annotation next to the official figure instead.

Editorial tier — countries the survey does not break out

The Economist's release contains a single line for the entire euro area. For a reader in France or Germany that is close to useless: the euro area is not a price zone, VAT rates differ, franchise structures differ, and the national averages diverge by more than the index's own headline numbers. We therefore research individual countries ourselves and publish them as a separate tier. Currently: France, Germany, Italy, Netherlands, Spain.

Every editorial record carries the same fields, and all of them are shown on the country page:

• Price and exchange rate — a national average, not a single restaurant's menu price
• Confidence, 0 to 1 — scored by source quality and recency, using the rules set out below
• Sources — every source we used, each with a URL, a description of what it contributed, and the date we accessed it
• Last reviewed — the date a human last checked the entry, not the date of the last site build
• Notes — the caveats, including anything that argues against our own figure
• Cities — where we have defensible city-level prices, they are listed separately with their own confidence and sources

As a concrete example, the Germany entry publishes €6.30 ($7.31) with a confidence of 0.82, last reviewed on 18 May 2026, backed by 5 listed sources ranging from the manufacturer's recommended price to newspaper-published customer receipts. That figure sits between the recommended price and observed till prices, and the notes say so.

Editorial data never overrides Economist data. If a country appears in the survey CSV, the CSV wins and the editorial record is skipped at build time. The tier exists to fill gaps, not to second-guess a synchronised survey.

How confidence is scored

Economist-tier records are fixed at 1.0 by definition. For editorial and community-sourced data, the confidence score between 0 and 1 is assembled from the following adjustments:

• McDonald's official menu page: +0.4
• Multiple independent sources agreeing: +0.3
• Local news reporting with date and city: +0.15
• Reddit thread with multiple corroborating commenters: +0.2
• Source older than 12 months: −0.3
• Single source only: −0.2
• Ambiguous price range (e.g., "₹200–250"): −0.15

The score floor is 0.3 and the cap is 1.0. The resulting value drives the source badge shown next to the price on every country page: high confidence, researched, or tentative. A tentative badge is not a disclaimer we hide behind — it is the signal to go and check the country page's source list before quoting the number.

Closed tier — countries with no McDonald's

17 countries in our dataset have no McDonald's restaurants, so they have no Big Mac price and no honest way to fake one. They fall into two groups: markets the company left, such as Russia and Belarus (2022), Kazakhstan (2023), North Macedonia (2013), Iceland (2009) and Bolivia (2002); and markets it never entered, such as North Korea, Iran, Bhutan, Cuba and Yemen. Each record stores which of the two it is, and the exit year where one applies. These records carry no price, are excluded from the price ranking, and are drawn in a distinct colour on the map rather than being given a fabricated position on the price scale. Each links to a written explanation of why the country has no data, which is the honest answer to "why is this one grey".

4. Reader price reports

Readers email us prices, usually with a photograph of a till receipt or a link to a local McDonald's delivery site. We accept a submission if it meets three conditions:

1. It is the standalone Big Mac, not a meal, not a promotion, not an app-exclusive price
2. It has a checkable source — an official delivery or menu page is best, a receipt or menu photograph next best
3. It has an explicit observation date, and for a single-store price, an explicit location

A qualifying report is published as an annotation on the country page. It does not replace the published figure, and the reason differs by tier. Against an Economist figure, replacement would break the same-date comparison described above. Against an editorial figure, the reason is statistical: our number is a national average, and one receipt is a single point inside that distribution rather than a new average — in France the gap between the cheapest and most expensive outlets has been measured at 55%, so a single till slip is nowhere near enough to move a national mean.

There is one case where we do change the number: if the published figure is simply wrong — a typo, a misaligned row, a unit error. That is a bug, and bugs get fixed rather than annotated.

We publish reports that agree with us as well as reports that contradict us. In August 2026 a reader sent a receipt from a McDonald's in Saint-Martin-Boulogne, Pas-de-Calais, showing a dine-in Big Mac at €6.40 against our published French average of €6.35 — a difference of 0.79%. Nothing about the national figure changed, but the receipt became a city-level data point, France's confidence rose to 0.88, and the observation is credited on the country page. A physical receipt confirming a desk-research estimate is the strongest validation available to us, and discarding it because it was not surprising would be both bad practice and a good way to never receive a second submission.

On attribution and privacy: contributors are credited by first name only by default. Sending an email is not consent to have a full legal name indexed by search engines, so we ask each contributor how they would like to be credited — full name, first name and initial, first name only, anonymous, or with a link — and we stay on the most conservative option until they answer. Receipt images are not published without explicit permission, and where one is published, the payment block, card issuer, terminal number, exact transaction time and staff identifiers are redacted first. A receipt is the sender's transaction record, not our stock photography.

5. Update frequency

An editorial entry is re-reviewed when something makes it stale rather than on a fixed calendar: a VAT or minimum-wage change in the market, a reader report, a source passing twelve months old, or a new Economist release that changes the US benchmark it is compared against.

The "last updated" date in the site footer is derived from the data — the later of the current survey date and the most recent editorial review — and deliberately not from the build timestamp. A build date would tick over every deploy and imply a freshness the data does not have.

6. Known limitations

The index is a teaching device that happens to be quantitative. Its limits are real, and most of them cannot be engineered away:

• One product is not an economy. A Big Mac is heavily weighted towards local rent, local wages and local beef. It says little about tradable goods, rent, healthcare or education, which is most of what a household actually spends money on.
• The euro area is an aggregate, not a country. The survey's single euro-area line averages across economies with different tax rates, wage floors and franchise structures. This is the gap our editorial tier exists to fill, and it is also why our euro-area country figures will not match the headline euro-area number.
• Prices vary enormously inside a country. The only national French survey we can point to found a 55% spread between the cheapest and most expensive outlets; German franchise prices vary by roughly 16%. Any single national price is a summary of a wide distribution, and a traveller's experience can legitimately differ by a euro or more.
• Channel and format change the price. Dine-in, takeaway, delivery and in-app promotions are frequently different prices for the same burger, sometimes with different VAT treatment. We track the standalone dine-in or counter price where we can identify it.
• Exchange rates move; survey prices do not. Between two Economist releases, the local price is frozen while currencies float, so valuation gaps drift for reasons that have nothing to do with burgers.
• It is not investment advice. Currency misalignment on this measure has historically taken years to close, when it closes at all. Nothing here should be used to trade.

7. Source data and attribution

The Big Mac price series originates with The Economist, which has published the index since 1986 and distributes the underlying data as a public CSV. The exact file behind the Economist tier on this site is:

• File: big-mac-full-index-jul-26.csv
• URL: https://cdn.economistdatateam.com/big-mac/data/big-mac-full-index-jul-26.csv
• Published: 29 July 2026
• Retrieved by us: 16 August 2026
• Coverage: 2,056 country-date records, 1 April 2000 to 1 July 2026
• Upstream repository: TheEconomist/big-mac-data on GitHub

The Economist retains ownership of that series and of the original methodology. BigMacIndex.app is an independent project with no affiliation to The Economist or to McDonald's Corporation, and we make no claim to any rights in the underlying data — if you intend to redistribute the original series, take your licensing question to The Economist rather than to us. "Big Mac" is a trademark of McDonald's Corporation and is used here descriptively.

The editorial and closed-market research on this site is our own work, and every country page lists the sources it rests on. The processed dataset we publish, including our editorial figures, is available on the data download page in CSV and JSON.

8. Corrections

If a price on this site does not match what you paid, we want to hear about it — including when the answer turns out to be that our figure was right. The contact page lists what to include so that a report can actually be verified, and every correction that changes a published number is credited on the country page it affects.

9. What the data is useful for

Everything above is one division and one subtraction. The rest of this site is that same arithmetic arranged for four questions readers actually turn up with. Each link below opens on a worked case rather than an empty form.

Working out whether somewhere is actually cheap

Advice about a country being cheap or expensive is usually a claim about the exchange rate on the day it was written, and it goes stale without announcing it. The index gives the same claim with a date and a calculation attached. On the 1 July 2026 release the yen sits 50.4% below burger parity: $100 changed into yen buys about 2.0× as many Big Macs as the same $100 spent in the United States. That multiple is roughly what a dollar earner gains on anything priced out of local rent and local wages, and roughly nothing on imported goods, which is the useful half of a travel budget. Open that United States → Japan comparison and swap either side for the pair you are actually deciding between.

Setting regional prices for an app or a subscription

One worldwide dollar price is a different product in every market it reaches: the same monthly figure is a rounding error in one country and a real share of a week's grocery budget in another, and store price tiers make it easy never to notice. The App Store Pricing Advisor takes one US base price and gives you, for each of the 59 markets with a burger price, the dollar amount that costs a local buyer the same number of burgers — the whole table at once, with a CSV export to paste against a store's price matrix. It is the same ratio of two dollar burger prices used in the worked example above, applied to your price instead of to a burger. Read the output as a floor to argue with, not a recommendation: it accounts for purchasing power and for nothing else — not local tax, not payment fees, not what competitors already charge.

Reading one country all the way down

A national average hides a distribution, and where the figure is our own editorial research rather than the Economist's, it also rests on sources you are entitled to inspect before quoting it. Every country page carries the lot: the price, the confidence score and the sources behind it with access dates, and, where we have defensible city figures, the spread inside the country. France is the fullest worked example — 4 city entries and 8 listed sources, with a reader's till receipt from a town in Pas-de-Calais sitting alongside press-survey and aggregator readings for Paris, Lyon and Marseille, each carrying its own confidence. It is the source-tier rules and the reader-report rules above, applied end to end to one country.

Doing your own arithmetic on the numbers

If the question is how a currency has moved across releases, or what these prices look like joined to your own data, this site's views are in the way rather than helpful. The data download page has the full Economist series as CSV and JSON with every column documented, including the GDP-adjusted variant noted earlier. If you would rather read it live than keep a copy, the API page documents those same files as read-only HTTPS endpoints with open CORS — no key and no rate limit, so a fetch() from a notebook or a browser console works without a proxy in front of it. Neither file carries the editorial or closed tiers; those live on the country pages only, for the reason set out in the source-tier section above.