Aims
To better understand the longstanding inequalities concerning alcohol and tobacco use, we aimed to quantify the effect of household economic security on alcohol and tobacco consumption and expenditure.

Design
Longitudinal analysis using data from the Household, Income and Labour Dynamics in Australia survey (2001–2018).

Setting
Australia

Participants
A nationally representative cohort of 24 134 adults aged 25–64 years (187 378 observations).

Measurements
Tobacco and alcohol use (Yes/No), frequency of use of each per week, household expenditure on each per week, household weekly income (Consumer Price Index [CPI]-adjusted), employment security (based on conditions of employment) and housing affordability (housing costs relative to household income).

Findings
At baseline, one-quarter of the sample used tobacco and 87% used alcohol. Annual increases in household income were associated with the increased use of both tobacco and alcohol for people in households in the lowest 40% of the national income distribution (OR = 1.13, 95% CI = 1.03–1.23 and OR = 1.12, 95% CI = 1.04–1.20, respectively) with no similar income effect observed for higher-income households. In relation to smoking, the odds of a resident's tobacco use increased when their household was unemployed (OR = 1.32, 95% CI = 1.07–1.62). In relation to alcohol, the odds of use decreased when households were insecurely employed or unemployed, or housing costs were unaffordable (OR = 0.87, 95% CI = 0.77–0.98, OR = 0.66, 95% CI = 0.55–0.80 and OR = 0.84, 95% CI = 0.75–0.93, respectively). This was also reflected in the reduced odds of risky drinking (defined in accordance with Australian guidelines) when housing became unaffordable or households became unemployed (OR = 0.90, 95% CI = 0.81–0.99; OR = 0.82, 95% CI = 0.69–0.98, respectively).

Conclusions
In Australia, smoking and drinking appear to exhibit different socio-behavioural characteristics and household unemployment appears to be a strong determinant of smoking.