1.7 Trends in the prevalence of smoking by socioeconomic status

Show / hide chapter menu

1.1 A brief history of tobacco smoking in Australia

1.2 Overview of major Australian data sets

1.3 Prevalence of smoking—adults

1.4 Prevalence of smoking—young adults

1.5 Prevalence of smoking—middle-aged and older adults

1.6 Prevalence of smoking—secondary students

1.7 Trends in the prevalence of smoking by socioeconomic status

1.8 Trends in prevalence of smoking by country of birth

1.9 Prevalence of tobacco use among Aboriginal peoples and Torres Strait Islanders

1.10 Prevalence of smoking in other high risk sub-groups of the population

1.11 Prevalence of use of different types of tobacco product

1.12 Future smoking rates?

1.13 International comparisons of prevalence of smoking

Acknowledgments

References

In Australia29, 59, 60 and many other countries,61, 62 smoking behaviour is inversely related to socioeconomic status, with disadvantaged groups in the population being more likely to take up and continue smoking. The authors of a seminal British report on poverty and smoking have observed that one can "almost study social disadvantage itself through variations in smoking prevalence."63 p78

A number of sociodemographic variables are closely connected with the likelihood of smoking. Siahpush and Borland64 have examined the correlation of smoking behaviour with several factors, including education, family income, and Index of Relative Socio-economic Disadvantage (IRSD).[7] This research found that all three measures of socioeconomic status—education, income and IRSD—were independently and significantly related to the likelihood of smoking for both sexes. Of these three measures, IRSD was most strongly related to smoking status. Individuals falling within the highest IRSD category of disadvantage were about twice as likely to smoke as individuals in the lowest IRSD category, irrespective of individual levels of education and income. This finding suggests that the influence of neighbourhood is an important contributing factor to whether an individual smokes or not. Siahpush and Borland describe this as a "contextual effect", occurring because smoking is normative behaviour in a particular environment, or because there are other physical, cultural, social or economic factors in those areas that encourage or lead to smoking.64

The relationship between smoking and social disadvantage is discussed in greater detail in Chapter 9. Differences in smoking rates among socioeconomic groups result in different patterns of tobacco-caused ill health and disease. For discussion, see Chapter 3, Section 3.31.

1.7.1 Smoking and educational level

Increasing education levels are associated with decreased likelihood of smoking. Table 1.6 shows that between 1986 and 2004, people with the highest levels of education had consistently lower levels of smoking than other members of the community.

Analyses of combined data for males and females, adjusting for age and sex, show that the prevalence of smoking found in every education level in 2004 was significantly lower than that found in 1980. The decline in prevalence of smoking was most substantial among those who had graduated from university or had some university education, those who finished secondary school and for those who stayed at school up until Year 9. For people with a trade qualification and those staying in school until Years 10 or 11, most of the decline in prevalence occurred in the 1980s. For these education groups, the prevalence of smoking in 2004 was generally not significantly different from that found throughout the 1990s.

Looking at the data by year, from 1983 onwards those who completed Year 12 had significantly lower levels of smoking than those who left school after Year 9. In analyses that do not control for age and sex, the difference in smoking prevalence between those with education up to Years 9, and those with education up to Years 10 and 11 was only statistically significant in 1986, 1998, 2001 and 2004.

Population trends for higher rates of school retention65 and increased attainment of post-school qualifications66 are likely to be contributing to the overall decline in smoking in the Australian population.

1.7.2 Smoking and occupation

Smoking behaviour is clearly stratified by occupational level, increasing occupational prestige correlating with a decreased likelihood of smoking. In 2004, smoking prevalence among lower blue collar workers was about two and a half times higher than in upper white collar workers, although prevalence among all socioeconomic groups is in decline. Overall, in 2004, 14% of individuals in upper white collar employment were smokers, compared with 35% of those working in lower blue collar employment.

Table 1.7 and Figure 1.5 show trends in smoking prevalence by occupation level. Occupation level is based on the current occupation of survey respondents. Respondents who were not employed at the time of the survey (such as retirees, or people engaged in domestic duties) were not asked for a previous occupation and are classified as not being in paid work. Adjusting for age and sex, individuals in upper white collar occupations have reported significantly lower rates of smoking than all other groups for every study year except for 1980, when there was no difference in smoking levels between those not in the labour force and those in lower white collar occupations, and then in 1995 when there was no difference in smoking levels for those with upper and lower white collar occupations (see Table 1.7).

For most survey years the prevalence of smoking among lower blue collar workers has been significantly higher than for all other occupational groups, except for in 1980 and 1992, when there was no statistical significance in the difference between smoking behaviour among lower blue and upper blue collar workers.

In the 1980s, smoking among the group categorised as "not in paid work" was generally lower than for all employment groups other than upper white collar workers. From 1992 onwards, the prevalence of smoking in the "not in paid work" group has been closer to those in lower white collar occupations. While the "not in paid work" category is likely to include a proportion of socioeconomically disadvantaged people, who as a demographic group demonstrate a higher prevalence of smoking, as indicated above this category also includes retired people (older people being less likely to smoke than younger people—see Table 1.4), those engaged in domestic duties (more likely to be female than male, and hence to have a lower smoking prevalence—Table 1.2), and students in post-secondary education (who are less likely to be smokers than those with a lower level of education level—Table 1.6).

Comparing smoking behaviour within individual occupational groups over time and adjusting for age and sex, the prevalence of smoking declined during the 1980s for each occupational group, before levelling off in the 1990s. The decline in prevalence of smoking was significant between 1980 and 2004 for all occupational groups. However for lower white collar and upper blue collar workers there has been little change in smoking prevalence between 1995 and 2004. Among lower blue collar workers, smoking prevalence in 2004 was significantly lower than estimates found in all other years except 1992 and 2001.

The decline in smoking between 1980 and 2004 has been proportionately greater among white collar workers than blue collar workers. Overall, prevalence of smoking among upper white collar workers halved over this study period, while among lower blue collar workers, prevalence dropped by about 30%, suggesting that the differential in prevalence between the highest and the lowest occupational groups is widening over time, a pattern which has been observed in the United Kingdom as well.67 However the lower prevalence of smoking among those with higher occupational status is a result of fewer individuals in these groups having taken up smoking in the first place, rather than being more successful at quitting smoking.28