Comparability — Austria [top]

For Austria 1992, there is no income variable, but there is a monthly housing costs variable, which MTUS uses in place of income as this is the only alternative. Nonetheless, the data is missing for 68.4% of households.

Comparability — Netherlands [top]

For 2000, we have selected the cut-off point for the first quartile to be 2,400 euros, and the cut-off point for the third quartile to be 4,700 euros. We coded income as '1' (lowest 25%) for cases that fell below the cut-off point (i.e. excluding the cut-off point), and we coded income as '3' (highest 25%) for cases that were equal or fell above the cut-off point (i.e. including the cut-off point).

For 2005, the selected cut-off points for the first quartile are 501 euros-1,000 euros, and the cut-off points for the third quartile are 1,501 euros-2,000 euros. We coded income as '1' (lowest 25%) for cases that fell below the cut-off point (i.e. excluding the cut-off point), and we coded income as '3' (highest 25%) for cases that were equal or fell above the cut-off point (i.e. including the cut-off point).

Comparability — South Africa [top]

In 2000 and 2010, since the original income variable (INCORIG) was categorical, the proportions of household income actually represented by each of the values of INCOME differ slightly from what the value labels specify.

Comparability — South Korea [top]

In 2004 and 2009, no household income variable is available in the original data. The data for this variable was estimated from the original monthly individual income, and as the original income was coded as income groups, the percentiles do not match exactly to the source data.

Comparability — Spain [top]

For Spain 2002, data on income was collected and/or coded in income groups rather than interval values. For this reason, we have selected the cut-off point for the first quartile as part of code '2' (500e to 999.99e), and the cut-off point for the third quartile as part of code '5' (1500e to 1999.99e). We coded as 1 (lowest 25%) cases that fell below the cut-off point (i.e. excluding the cut-off point), and we coded as 3 (highest 25%) cases that were equal or fell above the cut-off point (i.e. including the cut-off point). As a result, the category identified as belonging to the lowest 25% does not represent exactly 25% of cases and the category identified as belonging to the highest 25% does not represent exactly 25% of cases.

Comparability — United Kingdom [top]

In 1974, there are no income variables, though there are two related variables. FOS, family occupation status (A, B, C1, C2, D, E + unemployed + pension, students), and SOCLASS (social grade of household, A, B, C). The combination of these creates a 3 category distribution which will be related to income level of the household. Nonetheless, this is not the same as income. Users should consider carefully whether to use this variable for this study.

Comparability — United States [top]

In the 2003-2012 data, the income categories are specific to each year of the ATUS data. The categories are created based on approximately the lowest 25%, middle 50%, and highest 25% of the distributions of INCORIG in each year of data.