Introduction
Countries that have adopted best practice tobacco control measures have experienced significant reductions in smoking rates [
1]. However, inequalities in smoking prevalence across socioeconomic position (SEP) remain [
2‐
4]. In 2019 the percentage of daily smokers in Australia was 34% in the lowest socioeconomic group, compared to 9% in the highest [
5]. There are also disparities in the number of cigarettes smoked daily by SEP, with the most disadvantaged group in Australia smoking on average about 40 more cigarettes per week [
6]. Spending on tobacco products, particularly by disadvantaged individuals and households, can negatively affect expenditure on other goods and services.
Several international studies report on the likelihood of financial stress and smoking-induced deprivation (spending money on cigarettes instead of on household essentials) among smokers compared with non-smokers [
7‐
9]. Financial stress is measured using survey items such as being unable to pay utility bills on time or going without meals. In Australia, the probability of experiencing smoking-induced deprivation is greater for those from low-income groups [
7,
10]. Late bill payments, going without meals and having insufficient money for petrol, clothing and family leisure activities are some of the ways smoking displaces other spending [
11]. Experiencing financial stress is more prevalent among smokers than ex-smokers or never smokers [
12,
13].
Smoking cessation may lead to positive changes in household spending; however, few studies have compared the spending of smokers and ex-smokers. The most current research from Australia is based on data from 2006 showing that ex-smokers from low SEP groups, who had made changes to spending, spent more on food and clothing [
14]. Ex-smokers in 2001–2005 had a 25–42% reduction in the odds of financial stress [
15]. More recently in the US, when ex-smokers were tracked over a one-year period, the lowest SEP groups had significantly lower spending on alcohol, food at home, transportation and entertainment [
16]. Compared to Australia, the US has at least 30% lower excise tax rates on cigarettes as well as less smoke-free restrictions in hospitality venues.
Changes in the discretionary income of adults who quit may be just one factor influencing purchases. Tobacco use has been strongly linked with other health-related behaviours, including higher alcohol consumption and less expenditure on food. Smoking cessation is associated with changed eating patterns and increases in food intake [
17]. There are neurobiological mechanisms that make co-administration of nicotine and alcohol intake more pleasurable [
18]. In addition, smokers can consume more alcohol due to nicotine being able to directly offset its sedative properties. Smoking also counteracts the cognitive deficits associated with alcohol intoxication in the short term [
19]. Alcohol consumption also has the effect of reducing the usual brakes on smoking whereby people try to minimise their consumption. This may therefore result in ex-smokers reducing or avoiding alcohol [
20,
21].
To our knowledge, the expenditure patterns of smokers compared to ex-smokers by SEP group in Australia has not been previously investigated using quantitative data. We explored several household expenses such as groceries, utilities, alcohol, meals eaten out and healthcare. Our study aimed to compare the household expenditure of smokers and ex-smokers across socioeconomic groups in Australia using data over an 7-year period. This analysis will provide insights into whether smoking cessation may help to reallocate household expenditure in a way that benefits economic and health outcomes.
Discussion
This study aimed to compare household expenditure of smokers and ex-smokers, particularly among the lowest SEP groups. We found that in the lowest SEP group (quintile 1), smokers spent on average $4335 per annum on cigarettes and tobacco products and ex-smokers spent more on meals eaten out, education, motor vehicle fuel and insurance than smokers. Ex-smokers in SEIFA quintile 2 had the most differences in spending categories, with more spending in the same categories as quintile 1, along with groceries, medicine and health practitioners and less on alcohol. The apparent expenditure shifts from tobacco to spending in other areas indicate there are societal benefits of smoking cessation, beyond the direct health effects in the form of increased expenditure in the local community, such as to grocery stores and restaurants and cafes. For the Australian economy, spending on other goods and services is beneficial and would increase the income and prosperity for local businesses given that tobacco companies are transnational. Our results indicate that the interaction effects of marital status, having children, remoteness index and gender on cigarettes, alcohol, meals eaten out and medicine expenditure were not significant.
Less expenditure on alcohol by ex-smokers indicates a possible joint health effect improvement: smoking cessation and reduced alcohol intake. Our results highlight previous research that nicotine increases alcohol reinforcement and cravings in smokers and its absence results in a decrease in consumption [
31,
32]. Greater alcohol expenditure in smokers may also reflect previous findings that they are able to consume more due to nicotine directly offsetting the cognitive deficits associated with alcohol intoxication [
33]. Alcohol consumption is also related to relapses in smoking cessation attempts [
20]. Our findings of a reduction in alcohol expenditure support prior research from a longitudinal study that examined recent adults who quit over a 12-month period from the U.S. [
16] Rogers et al also found that quitting reduced households expenditure on other items that facilitate or complement smoking cessation such as food, entertainment, and transport [
16]. However, our findings indicate an increase in eating out for adults who quit. Rogers et al considered restaurants to be a smoking trigger; households with a smoker who had relapsed tended to have higher spending on food away from home. Differences in our findings may be due to smoke-free restaurants and cafes being more regulated in Australia. Expenditure on food at home was not collected as part of the HILDA survey so the extent to which overall food expenditure has changed is not known. Our results could indicate that adults who quit switched from consuming alcohol to eating out.
For adults who quit, higher health expenditure on health practitioners and medicines was found overall. Smoking cessation often occurs during a costly healthcare episode which prompts them to quit [
34]. This has been found by several previous studies, that followed smokers and ex-smokers over several years [
34‐
37]. Ex-smokers showed an increase in healthcare use and costs that began just before cessation and further increase after cessation, often over a period of 1 year, with rates of healthcare utilisation declining in subsequent years. Within 4-7 years after cessation, adults who quit returned to their baseline levels of healthcare use or lower [
34,
35]. Another possible explanation could be that lower socioeconomic quintiles prioritise purchasing tobacco products over medications or that they forego seeing a health practitioner due to the expense. However, for health practitioner expenditure there was a significant interaction between smoking status and remoteness index, indicating that the influence of adults who quit on increased health practitioner expenditure is reduced for those living in outer regional areas. This could be because those living in rural areas generally experience poorer health outcomes due to multiple factors, such as lifestyle and access to healthcare [
37].
Several interactions between education expenditure and smoking status were found. Overall, adults who quit had increased spending on education, and this was strengthened by not having children, and being male. As age increased, smokers spent more on education than ex-smokers but these differences in spending were very low. Previous research indicates that time preferences are a key component of the theory of rational addition, whereby present-oriented people are more prone to addiction and ex-smokers were less present-oriented and less impulsive than smokers [
38]. Ex-smokers may be less present-oriented and more able to envisage medium- to longer-term consequences than smokers who may be more impulsive [
38] and this may diminish with age. A similar explanation could be delayed reward discounting (DRD), a concept from behavioural economics that describes a specific type of impulsive decision-making reflecting how quickly a reward loses its value based on its delay in time [
39]. For example, substance dependence manifests behaviourally as an individual’s preference for smaller immediate rewards at the expense of considerable benefits in the future from not using the drug. A review of DRD studies indicates strong evidence of greater DRD in individuals exhibiting addictive behaviour [
39]. The lower expenditure on insurance by smokers compared to ex-smokers is consistent with a previous cross-sectional study from Australia that found similar patterns of insurance expenditure to our study between smokers and non-smokers [
40]. We found that the influence of smoking on reduced insurance expenditure was strengthened by living in remote areas and being married.
Our study has several strengths. We were able to analyse a large nation-wide representative longitudinal sample of smokers and ex-smokers over a seven-year period by socioeconomic quintiles. The only previous study that has examined these groups’ expenditure longitudinally is from the US using a follow up period of 12 months [
16]. Previous studies on expenditure of smokers in Australia by SEP have been based on a national cross-sectional surveys and qualitative interviews that compared them with adults who do not smoke [
11,
40]. Our GEE model selection approach based on
p-value< 0.05 is justified given that a post-hoc analysis showed that based on an average sample size of 2500 per year with compound symmetry correlation structure (corr = 0.5) across annual measure, the models can detect sufficiently small effects (i.e. an standardised effect size of 0.08, equivalent to 8% of SD) with 80% power. Our study could have been strengthened using only smokers pre- and post-cessation, however this was not possible due to insufficient numbers in the sample. We use an area-level socioeconomic measure of socioeconomic position, which may omit substantial proportions of individual variation in education and income [
41]. However, there is evidence that area-based measures capture the complex relationship between various economic and social phenomena that cannot be picked up by individual-based measures [
42]. Nonetheless, we have provided results by measure of income and occupation (Table
3) with very similar findings.
Because tobacco and alcohol use are highly prevalent in several other high-income countries, our findings of a reduction in alcohol expenditure associated with cessation are also of relevance outside of Australia. In the UK, for example, alcohol expenditure as a proportion of income is highest amongst the most disadvantaged [
43]. Understanding how households reallocate spending when consumption of tobacco and alcohol are reduced may alleviate financial strain amongst disadvantaged groups as well as improve health. While prevalence of smoking remains high among people with mental illness [
44,
45], future research could explore whether mental health impacts household expenditure of smokers and ex-smokers especially amongst disadvantaged groups.
Conclusions
Smoking cessation not only results in direct health benefits, but also appears to have societal economic benefits. In SEIFA quintile 2, ex-smokers, had the most significant differences in spending categories. SEIFA quintiles 1 and 2 had higher spending on meals eaten out, education, motor vehicle fuel and insurance compared to smokers daily. The reduction in alcohol spending by ex-smokers overall indicates a joint health benefit that could be used to encourage policymakers, funders, primary healthcare and the alcohol and other drug sectors, to address smoking cessation more actively. Amongst low-SEP households and across the sample, spending by ex-smokers indicates positive impacts on households and increased spending in their local communities on non-tobacco products.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.