LCC Working Paper Series: 2015-14


Yangtao Huang, Francisco Perales and Mark Western

Non-technical Summary

Strengthening and maintaining the land of a “fair go” has been a policy aim in Australia for decades. A fair society may feature a certain amount of economic inequality, provided that there is equality of opportunity. Equal opportunity means that people’s life chances rely more on individual effort and hard work than on parental socioeconomic status and family background. Intergenerational income mobility is an important measure of linkages between parental socioeconomic status and adult children’s outcomes. It is a measure of the extent to which parental income determines grown-up children’s economic performance. It is quantified by intergenerational earnings elasticity, also known as intergenerational earnings persistence, a measure of how offsprings’ earnings are related to parents’ earnings. A larger earnings elasticity indicates less income mobility.

While an extensive body of literature has estimated intergenerational earnings elasticities in developed and developing countries, and cross-national comparative studies have flourished in recent years, there is surprisingly little research on earnings elasticity in Australia. Existing evidence is based predominantly on cross-sectional data, which obscures underlying dynamics. Trend analyses rely on different datasets for different time points, which faces issues of inconsistent income measures. Therefore, new and more robust estimates of earnings elasticity for Australia using recent longitudinal data and panel regression models are warranted.

We use the Household, Income and Labour Dynamics in Australia Survey and the Longitudinal Labour Force Survey to examine the patterns and dynamics of father-son earnings elasticity in contemporary Australia. We contribute to the literature in the following ways. Methodologically, we implement two stage regression models that take advantage of longitudinal and other characteristics of the data. Substantive improvements are made by comparing different approaches to see how sensitive elasticity estimates are to analytic choices. First, detailed occupation categories enable us to estimate earnings elasticity based on the level at which occupations are disaggregated. Second, we examine changing patterns in earnings elasticity by considering linear and curvilinear trends over time. Third, comprehensive earnings measures allow us to assess the effect of earnings volatility on elasticity estimates.

We find that intergenerational earnings persistence in contemporary Australia lies between 11% and 30%, situating Australia internationally as a country with moderately high income mobility. The overall trend of earnings elasticity since 2001 is upward, although there is a declining tendency after 2011. More data are required before we can conclude that the decline after 2011 signals a new trend. Elasticity estimates are larger using two-digit level occupations and two-sample approach than the estimates using one-digit occupations and one-sample approach, respectively, and using different earnings measures results in substantially different elasticity estimates. We read these findings as indicating that elasticities are sensitive to the use of methods and data.


July 6, 2015