The salary comparison you keep seeing online is wrong

You have seen the post. Average salary in country A, converted to dollars. Average salary in country B, converted to dollars. A ratio, and a conclusion about which country is doing better.

The arithmetic is fine. The comparison is close to meaningless, and understanding why is the single most useful thing on this page.

What an exchange rate actually prices

A market exchange rate is set by the people trading currencies. Importers, exporters, investors moving capital, central banks, speculators. What they are collectively pricing is the demand to hold one currency against another, which is dominated by things that cross borders: goods in containers, bonds, shares, direct investment.

Now think about what a salary is spent on. Rent, or a mortgage on a building that cannot be moved. Food, most of it prepared or retailed locally. Transport. Childcare. A haircut. Electricity. A phone contract. The great majority of ordinary spending goes on things that never cross a border and whose prices are therefore never touched by the mechanism that sets the exchange rate.

Converting a salary at the market rate implicitly asks: what would this money buy if it were spent on internationally traded goods? That is a real question. It is just not the question anyone means when they compare wages.

How a PPP is built

The serious version is the World Bank's International Comparison Program, which is one of the largest statistical exercises on earth and rather underappreciated for it.

Statistical agencies in participating economies price a common list of hundreds of tightly specified items: not "a shirt" but a shirt of a stated fibre composition, weight and construction, so that two countries are genuinely pricing the same thing. Those prices are aggregated into a conversion factor that says how many units of local currency buy what one dollar buys in the United States.

The result is the number behind every "GDP at PPP" figure you have ever read, and the difference between it and the market rate is frequently enormous. In lower-income economies, a PPP conversion routinely values local income at two to four times what the market rate implies, precisely because so much of the basket is non-tradable and therefore cheap.

The Big Mac index is a PPP with one item in the basket

The Economist's Big Mac index, published since 1986, is the joke that turned out to be useful. Take one product that is close to identical worldwide, compare its local price with its American price, and you get an implied exchange rate. Compare that with the actual exchange rate and you get a claim about whether a currency is over- or under-valued.

It works better than it has any right to. A Big Mac is a small, standardised bundle of exactly the ingredients that matter: beef and bread, which trade globally; plus labour, rent, electricity and local tax, which do not. It is a basket of one, but it is a well-chosen one.

It also fails in exactly the way the theory predicts, and the failure is the interesting half. Burgers look systematically cheap in poor countries even after adjusting, because so much of the burger is local labour. The Economist publishes a second version adjusted for income per head for this reason, and the adjusted index is the one worth reading.

Four traps in PPP figures

PPP is the right tool and it is still routinely misused. These are the ones that bite.

1. 1.The basket is not your basket. A PPP is built from a representative national basket. If your spending is unusual, heavy on imported goods, or concentrated in a capital city, the national PPP does not describe you.
2. 2.Quality is hard to hold constant. Two countries can price a "comparable" item where one version is materially better. Statisticians work hard at this and it remains the largest source of argument in the field.
3. 3.Capital cities distort everything. Most people in a country do not live in its most expensive city, but most price data, and nearly all the anecdotes on the internet, come from there.
4. 4.PPP is wrong for anything international. If you are paying off a dollar debt, buying imported equipment, or planning to move abroad, the market rate is the correct one and PPP is the misleading one. The right conversion depends entirely on where the money is going to be spent.

A worked way of thinking about it

Suppose you are weighing a job in a high-cost country against one at home paying a third as much in dollar terms.

The market-rate comparison says the foreign job pays three times more. The PPP comparison might say it pays perhaps a third more once local prices are accounted for, because the higher salary is buying groceries and rent at higher prices.

Both numbers are correct and they answer different questions. Everything you spend locally should be judged at PPP. Everything you send home, save in a foreign currency, or spend on travel and imported goods should be judged at the market rate. Almost nobody is entirely in one category, so the honest answer to "which job pays more" is usually a split.

The reason this game shows both a dollar price and a work-time figure is the same reason. The dollar column is the market-rate view, useful for comparing countries. The hours column is the PPP-flavoured view, useful for understanding what a price means to the person paying it. They frequently rank countries in opposite orders, and that disagreement is the most informative thing on the page.