Rent increases and mental health among low-income renters in Australia: a longitudinal analysis

Regulation of within-tenancy rent increases is weak in Australia. While links between unaffordable and insecure housing and poor mental health have been established, the mental health burden of landlord-imposed rent increases – which represent a specific type of housing shock - has not been examined empirically. We used fixed-effects regression analyses of 9,808 observations (from 3,604 individuals) to investigate the relationship between within-tenancy rent increases and mental health among 20-64-year-old low-income private renters in the Household Income and Labour Dynamics in Australia (HILDA) Study (2002-2019). We tested whether the relationship was the same for younger (20-44 years old) and middle-aged (45-64 years) renters and explored effect modification by pre-existing affordability stress and markers of potential rental market disadvantage. In sensitivity analyses we assessed the likely impact of potential selection bias due to the necessary exclusion of movers from our analytic sample. Compared to years when rent remained stable, people’s mental health scores were worse following a rent increase of at least 10%, with effects concentrated in younger renters. Sensitivity analysis revealed that structural limitations precluding identification of renters who move in response to a rent increase make it likely that mental health effects of rent increases are underestimated. There was evidence of effect measure modification by pre-existing affordability stress. Policy debate should consider the potential protective impact of rent caps on the mental health of younger renters. • Regulation of within-tenancy rent increases is weak in Australia • We investigated relationship between within-tenancy rent increases and mental health • People’s mental health scores were worse following a rent increase of 10% or more. • Effects were concentrated in younger renters and those already in affordability stress • Sensitivity analysis showed structural selection bias likely underestimated effect sizes