Household Assets Among Australian Age Pensioners: A preliminary analysis of data from the Department of Human Services

Select | Print

Stephenson, Alec; Zhu, Zili; Toscas, Peter ORCID ID icon; Reeson, Andrew ORCID ID icon; O'Hare, Colin; Koo, Bonsoo; Cox, David




This report is the first product of a 2016 ARC Linkage Project under a non-commercial/public good research agreement between Monash University, CSIRO, Challenger Ltd and Accurium Pty Ltd. We acknowledge funding received from the ARC Linkage Grant LP160101038. The research partners identified as a priority the need for an improved empirical understanding of the circumstances of Australian retirees to inform the current public policy debate on retirement incomes policy and in particular the government's development of a framework to assist superannuation trustees to provide Comprehensive Income Products for Retirement. This report presents summaries of asset data from transactional records held by the Department of Human Services for the purpose of assessing eligibility for social security payments and concession cards. DHS removed personal identifiers before providing the data to CSIRO. Consistent with means test treatment, which is undertaken on a household basis, the data cover not just a de-identified individual's asset and income status but also their partner's income and assets. This data set, which is new to academia, offers the opportunity to gain a far more detailed and salient understanding of the financial status of older Australian households. For the purposes of informing debate on how best to provide retirement income this report focuses on the financial status of households at, and after, retirement from the workforce. Among the households represented in the data there exists a wide variation in financial circumstances. Many retirees have very limited wealth. Partnered households typically have approximately twice as much in assessable assets. There are also significant differences between the superannuation and other means testable assets of homeowner and renter households. The longitudinal nature of this data set shows continuing improvements for all groups as the superannuation system matures, however there have been significant variations in the rate of growth in superannuation assets between different household types. The data presented in this report focuses on a limited range of characteristics necessary to consider what is an appropriate product for a retiree or group of retirees. These are arranged by household type, both single and couples. They include superannuation assets, other means testable assets, total means testable assets and homeowner or renter status. This information is necessary when considering both social security entitlements and asset allocation for a retirement income stream, including the appropriate allocation to pooled longevity products. This data should be useful from a public policy perspective in better understanding the financial circumstances of the target group for CIPRs (Comprehensive Income Products for Retirement), as well as the potential to improve retirement outcomes by allowing funds to design more than one CIPR for their members. It can also be used to assist superannuation trustees when they are designing CIPRs to better understand the likely range of financial circumstances of their members that are not visible to the fund, including assets held outside the fund and on a couples basis.

CSIRO Risklab Australia

CSIRO Risklab Australia

Financial Mathematics

household_assets.pdf (pdf) (1,002KB)

This report has been placed on the CSIRO repository and may be made available to persons outside of CSIRO for non commercial purposes, in its entirety and without deletion of disclaimers and copyright information.


Technical Report (Author)


Stephenson, Alec; Zhu, Zili; Toscas, Peter; Reeson, Andrew; O'Hare, Colin; Koo, Bonsoo; Cox, David. Household Assets Among Australian Age Pensioners: A preliminary analysis of data from the Department of Human Services. CSIRO Risklab Australia: CSIRO Risklab Australia; 2017.

Loading citation data...

Citation counts
(Requires subscription to view)