This paper contributes to the existing literature on income mobility by
developing and applying a two‑stage panel regression model and assessing
the effects of using different levels of occupational (dis)aggregation and
different earnings measures on the magnitude of father–son earnings
elasticities in Australia. We find that the overall intergenerational earnings
elasticity in Australia between 2001 and 2013 ranges from 0.11 to 0.30. Our
preferred estimates lie between 0.24 and 0.28. Elasticity estimates vary
depending on the level of occupational (dis)aggregation and earnings
measure used: they are highest when two‑digit level occupations and
hourly earnings are used, and lowest when four‑digit level occupations and
annual earnings are used. We read these findings as indicating that elasticity
estimates are sensitive to the use of different data and methods, and
researchers should be careful when undertaking cross‑study, cross‑temporal
or cross‑national comparisons.