Methodology

The Economist has published its Big Mac Index since 1986 as an accessible test of purchasing-power parity. It collects local-currency prices from McDonald's corporate and franchisee sources across more than 50 economies and publishes the raw data openly. Pricele uses the January 2026 edition.

The index has one well-known limitation that applies directly here: the burger is not identical everywhere. India's entry is a Maharaja Mac, because McDonald's does not sell beef there. The Economist also publishes a single euro-area price rather than one per member state, so Germany, France, Spain, Italy, Ireland, the Netherlands and Portugal all show the same figure on this site, labelled as the euro-area price. Real Big Mac prices do differ between those countries.

The dataset is available at github.com/TheEconomist/big-mac-data.

Cappuccino, milk, eggs, apples and petrol prices come from Numbeo, which aggregates prices contributed by users around the world and publishes national averages. Figures here were retrieved in August 2026 and are national averages in US dollars.

Crowd-sourced data has real weaknesses. Contributions skew towards large cities and towards people who use the internet in English, so a national average can lean urban and middle-class. Sample sizes vary a lot between countries: a figure for Switzerland rests on far more submissions than one for Egypt. Numbeo filters outliers, but a country with few contributors will always be noisier than one with many.

Diesel, residential electricity and residential natural gas come from GlobalPetrolPrices.com, which assembles them from government sources, regulators, fuel retailers and national media. Road fuels update weekly; the two household utilities are collected quarterly and include the cost of the energy, the network and every tax and fee on the bill.

The two utilities are published per kilowatt-hour, which lands between two and forty cents and rounds to something unguessable. They are shown here per 100 kWh instead, which is the same figure multiplied by a hundred and roughly a small flat's monthly draw. Nothing else in the game is rescaled.

Coverage is uneven, and deliberately so. Natural gas appears only where households are on a gas grid. An absent row means the country does not have the thing, not that we could not find a number for it.

Cigarettes, vape e-liquid, beer and spirits come from the WHO Global Health Observatory, specifically the 2024 rounds of its global tobacco and alcohol tax surveys. These are the most rigorously comparable numbers on the site: WHO collects the price of a named most-sold brand from supermarkets in each capital, standardises it to a fixed container — a pack of 20, 330ml of beer, a 750ml bottle of spirits, 1ml of e-liquid — and publishes it in local currency and in dollars at the official rate.

Two things to know when reading them. The surveys run every two years, so a 2024 figure is the current one even in 2026, and in a high-inflation country it will lag badly. And “most sold brand” is not one product: the most sold spirit is soju in South Korea, vodka in Poland and whisky in India, which is a large part of why the spirits column has the widest spread in the game.

The price of 1GB comes from Cable.co.uk's worldwide mobile data pricing league table, which averages more than 5,600 prepaid and postpaid plans across 237 countries.

This is the oldest data on the site. Cable's own dataset download and historical series are both dated 2023, and although plenty of other sites republish these exact figures under the current year, nothing on the publisher's page supports that. So the rows say 2023, and they are shown per the year they were collected in. Mobile data has been getting cheaper by roughly a fifth a year, so read these as a ceiling.

It is also the only item published in dollars and nothing else. Those rows show no local-currency price rather than a converted one.

The one item here that nobody actually buys. The Food Prices for Nutrition database, published by the World Bank with the FAO, prices the cheapest combination of locally available foods that would meet national dietary guidelines, per person per day. It is a floor rather than an average, and it is the figure used to count how many people cannot afford to eat well.

It is published in local currency only, so the dollar figure here is converted at the World Bank's own official exchange rate for the same year. Both halves are published; the multiplication is ours.

Most sources publish both figures themselves, and where they do, both are printed as published and neither is derived. That covers the Big Mac, all four energy items and all four WHO items: the dollar price is the publisher's own conversion at the rate on its collection date.

Two groups work the other way. Numbeo publishes in US dollars, so the local-currency figure for cappuccino, milk, eggs, apples and petrol is converted at the rate on the collection day. The cost of a healthy diet is published in local currency only, so its dollar figure is the converted one.

Either way, a converted figure is not an observed one. For most currencies that distinction is invisible. For fast-moving currencies like the Argentine peso, the Turkish lira and the Egyptian pound, it matters, and a figure can drift noticeably from what a shopper actually pays within weeks. Argentina is the extreme case, because official and parallel exchange rates have historically differed enough that any single conversion is contestable.

Mobile data is published in dollars with no local figure at all. Rather than invent one, those rows leave the local price blank. A price nobody published would look exactly as authoritative on the page as one that was.

Every price page shows roughly how long someone earning the average local wage works to buy one unit. That figure divides the dollar price by an estimated average hourly wage for the country.

Wage figures are the least rigorous numbers on the site, and they do not all come from the same place. The 33 countries the game launched with carry our own approximations. The 16 added since carry the International Labour Organization's published average hourly earnings of employees, converted to dollars by the ILO itself. The second kind is plainly better, and it exists because a new country had no estimate to inherit and we would rather use an official series than invent one. The mix is listed below rather than averaged over.

Replacing the remaining estimates would move every work-time figure on the site at once, including on archive pages for puzzles already played, so it is a decision on its own rather than a side effect of adding countries. It also would not finish the job: the ILO publishes no hourly figure for six of them, so some estimates survive whatever happens.

Either way, an hourly wage compresses an enormous amount of variation: formal versus informal employment, regional differences, and the gap between mean and median earnings, which is very wide in some of these countries. Treat work-time figures as illustrative of the general shape of affordability, not as a measurement. We round them hard for exactly that reason.

Each country page states that country's standard VAT, GST or consumption tax rate, and how ordinary groceries are treated under it. Consumption tax is the largest single wedge between what two countries charge for a physically identical item, and it is the one component of a price that is set deliberately rather than emerging from a market, so it is worth reading separately from the price itself.

Standard rates come from PwC's Worldwide Tax Summaries VAT quick chart. That is one source for all 49 countries, used deliberately: rates assembled from 49 different national pages would be internally inconsistent in exactly the way that makes cross-country comparison meaningless. Reduced, super-reduced and zero rates for European countries are cross-checked against the Tax Foundation's annual VAT rates in Europe table.

The grocery treatment matters more here than the headline rate does, because nine of the 16 items are food or drink. A country can charge 23% on most things and nothing at all on bread, as Ireland does, so quoting only its standard rate would actively mislead you about its food prices.

Two honest limits. First, where a country's food treatment could not be confirmed from either source, the field is left empty rather than filled in from memory; a missing line means we could not verify it, not that no relief exists. Second, three countries genuinely cannot be reduced to one number: Brazil overlaps federal, state and municipal levies, the United States has no VAT and leaves sales tax to states and municipalities, and India and China both band their rates by category. Those are described in words instead, and their rows should be read as national approximations.

A rate is not a full explanation of a price. Switzerland charges 2.6% on food and is still the most expensive country in this table; the United States has no VAT at all and sits mid-range. Tax tells you what the state adds. Wages, distance, tariffs, subsidies and what a country grows for itself do the rest, which is what the note at the top of each country page is for.

Every figure here is one number standing in for a whole country. That is a large simplification. Prices differ between a capital city and a rural town, between a supermarket and a corner shop, between seasons, and between brands. Eggs in particular swing hard with avian influenza outbreaks; fuel changes weekly; fresh produce is seasonal.

If a price here doesn't match what you paid this morning, the most likely explanation is that both numbers are correct and they are measuring different things. Pricele is published for general interest and as the basis for a game. It is not shopping advice, financial advice, or a substitute for an official statistical series.

The price tables here are a compilation of third-party data, so the underlying numbers are not ours to place under an open licence. The Big Mac figures remain The Economist's and are governed by the terms on its published dataset; the cappuccino, milk, egg, apple and petrol figures remain Numbeo's and are governed by Numbeo's terms of use.

The rest carry their publishers' own terms, and two of them are more restrictive than this site: GlobalPetrolPrices.com states a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 licence at the foot of every page, and WHO data is issued under CC BY-NC-SA 3.0 IGO. Both carry a non-commercial condition, and this site carries advertising. The energy and WHO figures are reproduced here with attribution and a link to the source on every price, and we will take either set down on request from the publisher. World Bank and FAO data is CC BY 4.0, which has no such condition.

What is ours is the compilation itself: the country and item selection, the local-currency and work-time calculations, the wage estimates and the writing. You may quote individual figures or a small extract of a table for editorial, educational or non-commercial use, with attribution to Pricele and a link to the page you took it from. Republishing a table wholesale, or any commercial use, needs permission from the original source rather than from us.

For anything this doesn't cover, email flaviuscojocaru19@gmail.com and ask.

If you live somewhere in the 49 countries covered here and a price looks clearly wrong, we would rather hear about it than not. Email flaviuscojocaru19@gmail.com with the item, the country and what you actually see locally. Corrections are applied to the dataset and take effect across the game and every reference page at once.

Data on this site is refreshed periodically rather than live. The source and collection month is printed next to every price, so you can always see how old a figure is.