Data on annual median equivalised household disposable income came from the income distribution project (OECD, 2008). Disposable income was gross household income after deduction of direct taxes and payment of social security contributions. It excluded in-kind services provided to households by governments and private entities, consumption taxes, and imputed income flows due to home ownership. People were attributed the income of their household. Household income was adjusted for household size by assuming a common equivalence scale of 0.5.
In previous editions, net national income (NNI) per capita was used as the income measure. Following recommendations of the Stiglitz-Sen-Fitoussi commission, the income measure in Society at a Glance changes in 2011 to a household-based one. Median household income is conceptually stronger for social purposes. It more closely relates to family income than NNI per capita. In addition, median household income creates a link with the poverty data (see EQ1 and EQ2), which uses the median household income data in its calculation. Data was provided to the OECD by national consultants and was based on common methods and definitions applied to national micro data. While this approach improves cross-country comparability, national data sets still differ from one another in ways not readily standardised. In some countries, median income come from different data sources over time, and
this adds further data error. It is likely that household income measures, while conceptually stronger for social purposes, have a lower degree of international comparability than the national income aggregates. To reflect this imprecision, household income figures were rounded to the nearest USD 1 000.
For cross-country comparison, national currency measures of income were converted into the United States dollars (USD) using purchasing power parity exchange rates (PPPs). These PPPs reflect the amount of a national currency required in each country to buy the same basket of goods and services as a dollar does in the United States. Both income and PPP estimates are affected by statistical errors, so differences between countries of 5% or less are not considered significant.
After subtracting taxes and adding welfare benefits household income provides an indication of the goods and services families can purchase on the market. It is thus an absolute objective indication of material quality of life. Household income is adjusted for family size and the adjusted measure is attributed to every person in that household. Half of all people have higher and half lower income than the median.
In 2007 half the people in Turkey and Mexico had household incomes less than USD 5 000. Half the people in Luxembourg had incomes about seven times higher (Figure GE1.1). Low household income countries included those in Southern Europe and the Mediterranean and much of Eastern Europe, as well as the two Latin American countries - Chile and Mexico. Higher household income countries included Norway and the United States. Country income rankings using equivalised household income were similar to those calculated using per capita net national income (NNI). However, using a household rather than an NNI measure changed country income rankings considerably for some countries. Sweden fell seven places and New Zealand and Korea rose by five places.
Household income growth between the mid-1980s and 2007 averaged about 1.5% across the OECD (Figure GE1.2). By way of comparison, per capita gross domestic product (GDP) growth was more than half a per cent higher, and NNI growth was higher by a similar amount. Such differences can cumulate considerably over a generation. OECD average growth also hides huge country variations. Mexico, Portugal and Spain were countries where household income growth was equal to or higher than conventional national aggregates. Household income growth was especially low relative to national aggregates in Belgium, Chile, Luxembourg, Hungary, and Japan. Reasons for differences between household and aggregate production growth measures could be due to the household focus,
rather than of the nation as a whole, the focus on medians, rather than averages, the different methods of adjusting for numbers of people, or measurement errors in the statistics.
OECD (2008), Growing Unequal? Income Distribution and Poverty in OECD Countries, OECD Publishing, Paris.
Figure GE1.1, Panel B: Median income changes over a 10-20-year period are not available for Estonia, Iceland, Korea, Poland, the Slovak Republic, Slovenia and Switzerland. Changes are available from mid-1990s for Australia, Chile, Israel and Portugal.