The association between unpaid labour and mental health in working-age adults in Australia from 2002 to 2020: a longitudinal population-based cohort study.

BACKGROUND

Unpaid labour is a daily part of most people's lives, none more so than for women. Yet, in comparison to paid work, the effect of unpaid labour on mental health is an under-researched area. This study aims to address key gaps in the extant literature, examining how unpaid labour is associated with mental health in working-age men and women, and whether gender differences exist.

METHODS

In this longitudinal population-based cohort study, 19 waves of the Household, Income, and Labour Dynamics in Australia (HILDA) survey were used to employ a fixed effects regression analysis to examine the associations between unpaid labour and mental health in working-age (aged 25-64 years) Australian adults. Mental health was assessed using the MHI-5 scale. Both the individual and the combined effects of four different domains of unpaid labour (household work, childcare, care for adults, and outdoor tasks) were interrogated, as were the gender differences.

FINDINGS

Of the 37 352 participants (297 036 observations) in waves 2002-20 of the HILDA Survey, 22 832 people (190 207 observations) were aged 25-64 years, and after excluding participants with missing data, 21 014 participants (150 163 observations) were included in the analysis. Increasing time in household work was negatively associated with mental health in both men (β coefficient=-0·026 [95% CI -0·04 to -0·01]) and women (β coefficient=-0·009 [-0·02 to 0·001]), as was care for adults (disabled or older people) in women (β coefficient=-0·027 [-0·04 to -0·01]). Conversely, increasing time in childcare for women (β coefficient=0·016 [0·01 to 0·02]) and outdoor tasks for men, was positively associated with mental health (β coefficient=0·067 [0·04 to 0·09]). A null finding for the overall cumulative total unpaid labour exposure for both men and women was probably attributable to the opposing direction of effects between the individual domains that constituted the total load.

INTERPRETATION

This study reveals considerable variance and nuance in how different domains of unpaid labour affect mental health, as well as continued inequity in the division of unpaid labour in households, with women doing considerably more unpaid labour than men. This study also exposes important challenges associated with measuring and understanding total (combined) unpaid labour as a determinant of health.

FUNDING

University of Melbourne Research Training Scholarship, Australian Research Council Discovery Early Career Award, Australian Research Council Linkage Project.