[jltsiren]

The median income on that page is the median disposable household income, divided by the square root of household size. Because American households are large for developed countries, measures like that overestimate the income of the median American relative to the median European.

There is an easy sanity check: Swiss GDP per capita is higher than in the US, both in absolute terms and in PPP terms. Their Gini coefficient is lower, meaning that the income distribution is more equal than in the US. If a comparison shows that the average/median disposable income is substantially higher in the US than in Switzerland, it is measuring something weird.

[pembrook]

I think you may have that backwards, if trying to disentangle the individual from the household, a larger household would make the median income per person lower not larger.

Also, I’d imagine Swiss housing costs could easily account for the difference in disposable income.

I’m totally open to the idea the economists at the OECD are dumb and put out bad numbers, but I’d argue GDP per capita (GDP being a flawed measure to begin with due to only including consumption incl. government spending) is a far worse measure of on the ground reality than something that accounts for actual household income and expenses.

[jltsiren]

Housing costs are paid from disposable income under the OECD definitions.

Larger households lead to larger reported incomes, because the total household income is divided by the square root of household size. If income is $50k/person and household size is 2, the reported income is $71k/person. If household size increases to 3, reported income increases to $87k/person.

On the average, Americans move out of their parents' home in their mid-20s. Swiss people typically do it a couple of years earlier. When people in their 20s live with their parents, it often indicates the lack of affordable housing. But if you combine this with normalizing household incomes by dividing by the square root of household size, higher housing costs lead to higher reported disposable incomes.

International comparisons are difficult, because pretty much every way to do them is wrong in some sense. But in general, it's better to collect data that is consistently wrong in the same way than to try to correct the issues you encounter. If your data is consistently wrong, you can at least make reasonable comparisons between a country today and the same country 10 or 20 years ago.

Subsidized services such as education and healthcare are tricky. You could try to calculate the subsidies based on the actual costs of providing the services, or as the difference between the nominal price and the subsidized price. These can lead to very different amounts, and it's not always possible to use the same approach with every service.

If the degree of subsidies varies between countries, you could try to normalize the situation by either adding the subsidies to disposable incomes or subtracting the costs the individual has to pay. In a country where the services are more expensive to produce (and maybe also less efficient), the former leads to higher and the latter lower disposable incomes relative to other countries.

Retirement contributions are another tricky question. Voluntary savings are often included in disposable income, while mandatory pension contributions often count as taxes. But you can't always tell the difference between voluntary retirement savings and other savings. But if you then count retirement income based on voluntary contributions as disposable income, you have to be careful to avoid counting the same income twice.

[klipt]

> larger household would make the median income per person lower not larger.

Depends whether the household is larger due to children, or due to more adults, no?

That is, if a bunch of single adults pair up into households, income divided by household size will go up.