2.3  Previous New Zealand studies

Studies of final household incomes, similar to this one, are regularly undertaken in some other countries including the United Kingdom (Lakin 2004), Australia (Australian Bureau of Statistics 2001) and the United States (DeNavas-Walt, Cleveland and Webster 2003). In contrast, there have been relatively few studies of final household incomes in New Zealand. The last were those undertaken by the Department of Statistics (1990), by the Income Distribution Group (1988) and by Brashares (1990), each studying the 1987/88 year. These three studies used a wide definition of final income, apportioning all government revenue and expenditure to households (see Section 3.5). Earlier in the 1980s, Snively (1986, 1988) pioneered the study of final household incomes in New Zealand.

No studies of final household incomes in New Zealand have been undertaken over the late 1980s and throughout the 1990s. Household income studies over this period have used one or more of the other definitions of income (see Section 2.1). The most extensive household-based studies have been conducted by Statistics New Zealand (1999), covering the period 1981/82 to 1995/96, and by Mowbray (2001), covering the period 1981/82 to 1997/98. Both studies use data from the HES and report changes in households’ market income, gross income, disposable income and equivalent disposable income. They find that the average household income, however defined, fell from the early 1980s to reach a low point in the early 1990s, and then recovered throughout the rest of the 1990s (Figure 3). Whichever definition is used, Mowbray finds that the average household income rose between 1987/88 and 1997/98—the period which the current study covers.

Figure 3 – Average household income, by different definitions of income, 1981/82 to 1997/98

Source: Mowbray (2001), Tables A5 and A10

The trends shown in Figure 3, however, disguise considerable changes in the distribution of income. Statistics New Zealand (1999) and Mowbray (2001) find that incomes, however defined, fell in most deciles between 1982 and the late 1990s. The notable exception was in decile 10, where incomes increased markedly (Figure 4). Mowbray finds that the average income in deciles 1, 8, 9 and 10 was higher in 1997/98 than it was in 1987/88.

Figure 4 – Average equivalent disposable income, by income decile, 1981/82 to 1997/98

Source: Mowbray (2001), Table A10.

Statistics New Zealand (1999) measures changes in the relative income levels of different deciles using Gini coefficients.[4] The Gini coefficient for equivalent disposable income and for market income increased markedly between 1985/86 and 1990/91, but not during other time periods. The late 1980s and early 1990s therefore appears to have been a time of increasing income inequality in New Zealand.

Two recent New Zealand studies have attempted to explain changes in the distribution of household income over the 1980s and 1990s, using data from the HES. Podder and Chatterjee (2002) look at equivalised gross household income between 1983/84 and 1995/96 by decomposing changes in the Gini coefficient. They show that a change in the distribution of earned income was the main contributor to increased income inequality over this period. Podder and Chatterjee consider that high unemployment, nominal interest rate rises, the introduction of GST and cuts in welfare benefits may have contributed to rising income inequality. Hyslop and Maré (2001) look at changes in gross household income between 1982/83 and 1997/98. They find that changes in the proportions of different household types account for between 10% and one-third of the increase in income inequality over this period, depending on the specific measure of inequality used. Changing socio-demographic attributes of households, such as changes in the age-mix and educational qualifications of household members, account for a similar fraction of the observed changes. Hyslop and Maré find that job losses had only modest effects on overall income inequality.