Raising the pension age for women has reduced Australia's birth rate by delaying grandmothers' retirement and limiting their availability for childcare, new research by the e61 Institute has found.
Using data from the Household, Income and Labour Dynamics in Australia (HILDA) Survey from 2001 to 2022, the study finds that women whose mothers were not yet old enough for the pension were less likely to have children, with the likelihood decreasing from 73.5% to 69%.
Those women also had fewer children on average at 1.47 compared to 1.56 for women whose mothers had qualified for the pension.
Between 1995 and 2023, the pension age for women gradually increased from 60 to 67, delaying grandmothers' retirement. The employment rate among grandmothers eligible for the pension is 25% compared to 36% for those who are not, and eligible grandmothers work 3.9 fewer hours per week, the research found.
"Raising the pension age has kept women contributing in the workforce for longer, but the trade-off looks like a reduction in their daughters' fertility rates," said e61 Research Manager Pelin Akyol.
"This may be because women are more likely to have children when their mothers are available to help them with childcare.
"We find that grandmothers' pension eligibility has a large impact on fertility, similar to the impact of the introduction of paid parental leave, which increased the average number of births by five per cent.
"The effects were even greater among women with lower wealth and education levels, suggesting they rely more on grandparental childcare.
"With Australia's fertility rate at a historic low of 1.5 births per woman, policymakers should carefully consider impacts on fertility when designing policies that encourage labour force participation among the older population."
The study isolated the impact of pension eligibility on fertility by comparing similar women whose mothers just qualified for the pension based on their age to those whose mothers had to keep working due to the gradual pension age increase. The study controlled for observed and unobserved time-invariant factors, such as cohort-specific differences and year-fixed effects, as well as potentially time-varying characteristics such as education, health and region.
