Income inequality in New Zealand: Why conventional estimates are misleading

Abstract

Considerable attention is currently being paid to establishing the extent of inequality in New Zealand and whether it has risen in recent years. This paper offers some insights into the inequality measures and interpretations that commonly feature in those debates. These typically relate to cross-sectional inequality, such as annual Gini coefficients for various income definitions, or comparisons of income growth rates across income deciles. But failure to take into account the longitudinal dimension of inequality can lead to misinterpretations of inequality data and measures. The paper shows that examining longitudinal income data for the same individuals over time strongly contradicts cross-sectional inequality evidence. For example, some recent cross-sectional inequality measures suggest that the incomes of initially low-income households grew at slower rates than those with initially higher-incomes. This has been interpreted as the poorest earners being ‘left behind’. But recent longitudinal data, at least for individuals, reveals evidence of much faster-than-average growth among initially lower, compared to higher, income earners. Thus, ‘regression to the mean’ is a dominant feature of the longitudinal data.