Household Income - International Comparison

International Comparison

Internationally comparable data on household income are difficult to find. Definitions differ frequently, as does the treatment of taxes (i.e., gross versus net income). Fortunately, the Luxembourg Income Study (LIS) has recently added a publicly available database with comparable statistics on household incomes for several countries, as has the OECD. These are the sources used.

Below are presented the mean and median disposable household incomes, adjusted for differences in household size. Thus, the figures presented are per person (equivalized) and after all income taxes and mandatory social contributions are paid. All figures were converted using respective year purchasing power parities (PPP) for private consumption, which is recommended when comparing incomes internationally. The PPP conversion rates are taken directly from the OECD database. All incomes are in the prices when income was earned, and refer to year 2004, except for Australia (2003), UK (2004–2005), and Sweden (2005). The exact definition of income can be seen in the LIS website (variable DPI). Generally, it includes all cash income (e.g., earnings, pensions, interests, dividends, rental income, social transfers) and excludes most non-cash income (e.g., employer contributions to social insurances, governmental health care, education). Note that capital gains are excluded from the income definition.

Caution should be made when comparing countries based on a strict ranking, since not all datasets are capturing income equally. For instance, income spent on private health insurance, which in 2004 was about $1300 per household in the US, is not subtracted. However, because PPP rates are used, the differences in prices for all other expenditres, include health and education are already taken into account by definition. When compared to national accounts data (adjusted for differences in definition), the datasets capture anywhere between 75 and 95% of the true income. More specifically, countries where surveys are used range from around 70 to 85%, while register countries (e.g., the Netherlands, and the Scandinavian nations) are capturing on average more than 90%. The U.S. dataset captures only 75% of the disposable income aggregate as of 2004. Thus, a true ranking would see a different ranking, as countries with low coverage would move above those with high coverage.