Falling rates of smoking and a growing preference for smoke-free public spaces1 and homes2 have been documented. There are also numerous, less obvious indications of society’s reduced tolerance for tobacco smoke and smoking. In 1992, Sydney share accommodation advertisers named non-smoking as the most sought after attribute in prospective flatmates, even ahead of requests specifying a person’s sex.3 In the United States4 and the United Kingdom, smoking can significantly reduce the sale price of houses. Smoking smells and stains were named by 30% of London home buyers as the leading “turn off” when inspecting a home, reducing the average price by an estimated £16 000.5 Smoking status is also an important factor in other than real estate: on 6 July 2004, advertisements on the online auction site <www.eBay.com>, which auctions anything from cars to collectibles, mentioned one of the following words “non-smoking, non smoking, nonsmoking, smokefree, smoke free” in 368 323.
Here, we report an investigation into an unexplored, unobtrusive measure of the changing face of smoking in the community: the smoking status of people advertising themselves on a major dating website. We used the search function of Australia’s largest online singles websites, “RSVP” <www.rsvp.com.au>, to test two hypotheses: (i) there would be more smokers among those seeking partners or dates than in the general population;6 and (ii) the most attractive or appealing advertisers (judged by the frequency of their photos prompting people browsing the site to view their ads) would less often be smokers than all advertisers on the site.
The RSVP website permits searching for people advertising on their site by specifying or excluding various characteristics, including a compulsory self-reported smoking status field (don’t smoke/trying to quit/occasional/regular). We compared the distribution of non-smokers and smokers in the RSVP advertising population on a random day with data on Australian smoking prevalence.6
The website displays advertisers’ photos, and browsers may click on any of the photos to read about the person advertising. The “Top 100” feature of the website allows visitors to inspect the 100 men and 100 women whose advertisements are accessed most often by site visitors. This allowed us to compare the smoking status of these people with that of the other advertisers on the site. (The “Top 100” feature is now available in each age category, but, at the time we did our research, these 100 male and 100 female advertisers were drawn from the whole site and, coincidentally, were all aged 20–29 years).
Our study did not require ethical approval, and was unfunded.
Percentage of current smokers in the Australian population* compared with the percentage of self-declared smokers among advertisers on a dating website (RSVP) † ( n = 132 176), and among the “Top 100” male and female advertisers ( n = 200)
|
Smoking prevalence among “Top 100” male and female advertisers |
|||||||||||||||
On 10 February 2004, the RSVP site had 350 942 members, including 132 176 who had posted advertisements for themselves on the site. These numbers change on a daily basis as new advertisers join and others withdraw.
We compared the proportion of advertisers on the RSVP website who were smokers with the proportion of smokers in the most recent national study of adult Australian smoking.6 Our results are summarised in the . For all age groups of women and for men aged 50 years and older, a higher proportion of advertisers were smokers compared with the proportion in the general population. Among men aged 30–49 years, there was no difference in smoking prevalence between advertisers and the general population. However, for the youngest age group of men (20–29 years), advertisers were less likely to be smokers than their counterparts in the general population. For all age groups, the proportion of people describing themselves as “occasional” smokers was significantly higher among advertisers than in the general population. For example, among women aged 20–29 years, 5.5% of the general population were occasional smokers compared with 22.1% of the advertisers (χ2 = 151.40; P < 0.0001), and among men the corresponding proportions were 8.4% and 20.3% (χ2 = 163.57; P < 0.0001). (Data on smoking status by age are available from the corresponding author.)
There was a higher proportion of non-smokers in the “Top 100” advertisers (82%) compared with all RSVP advertisers aged 20–29 years (67%; χ2 = 50.73, P < 0.001): 85% of the men and 78% of the women in the “Top 100” advertisers reported being non-smokers, and 2% of both sexes were regular smokers (the others describing themselves as “occasional” or “trying to quit”).
Compared with members of the general population, women and older men advertising on the RSVP dating website were more likely to be smokers; and those aged 20–29 years whose photos were judged most attractive by visitors to the site were significantly less likely to be smokers than all advertisers in that age group. The very large numbers of advertisers on the site (and presumably a similarly large number of visitors) provide high confidence that these phenomena are real.
That smoking makes it more difficult to attract responses to dating advertisements is supported by anecdotal evidence from the largest dating service in the United Kingdom, Dateline, which reported: “Often, if one of our clients queries that they haven’t received many responses from potential partners, we can trace it back to the fact they admitted on their ‘compatibility’ form to being a heavy smoker. It’s amazing how many smokers then call us back a week later to inform us that they’ve now decided to kick the habit!”.8
Our data also indicate that smoking may be an added impediment to advertisers’ overall “presentation of self” if they hope to attract a partner. The sex differences we found suggest that women and older men who smoke may find themselves more “desperate and dateless” than younger men who smoke.
Previous research has suggested that smoking may be associated with increased facial wrinkling9,10 and hair loss.11 Further, among a group of people, about half of those who had smoked for 10 years or more were able to be identified from facial features associated with smoking.12
People “clicking on” the photographs of advertisers do not know the person’s smoking status before they view the information supplied. Thus, the “Top 100” men and women are chosen for their looks. That only 2% of these 200 people were regular smokers suggests that smoking may be associated with less attractive looks.
It is likely that our conclusions are conservative, as there was a higher proportion of self-assigned occasional smokers (18%) and a lower proportion of regular smokers among advertisers compared with national data (4.5% occasional smokers),6 indicating a possible bias in self-reported smoking status. It may be that the increasing social unacceptability of smoking caused some purposive misclassification bias in self-reporting smoking status: regular smokers may be more likely to report themselves as occasional smokers or as smokers attempting to quit than for these two groups to misreport that they are regular smokers.
Thus, as well as suggesting that female and older male smokers are more likely to use a dating website, our study provides some evidence that smokers on such a website are perceived as less attractive than those who do not smoke. Such information might provide some smokers with additional motivation to quit.
- Simon Chapman1
- Melanie A Wakefield2
- Sarah J Durkin3
- 1 School of Public Health, University of Sydney, NSW.
- 2 Centre for Behavioural Research in Cancer, The Cancer Council Victoria, Carlton, VIC.
None identified. S C is not dateless but has occasionally been described as desperate. M A W is currently dateless but not desperate. S J D is neither desperate nor dateless. S C is the guarantor.
- 1. Letcher T, Black C. Public opinion about smokefree policies in bars, nightclubs and gambling venues, 2000–2002. (CBRC Research Paper Series No. 2.) Melbourne: Centre for Behavioural Research in Cancer, The Cancer Council Victoria, July 2003. Available at: www.accv.org.au/cancer1/research/cbrc/research_areas/research_paper_series/ab_rps2_2003.htm (accessed Oct 2004).
- 2. Borland R, Mullins R, Trotter L, White V. Trends in environmental tobacco smoke restrictions in the home in Victoria, Australia. Tobacco Control 1999; 8: 266-271.
- 3. Chapman S. Shared accommodation — non-smokers wanted. Tobacco Control 1992; 1: 248.
- 4. Martin A. On tobacco road, it’s a tougher sell. New York Times 2004; 8 Feb.
- 5. McIntyre S. Smoking devalues your house. Evening Standard 2004; 4 Feb. Available at: www.thisislondon.co.uk/news/business/articles/timid73843?source (accessed Oct 2004).
- 6. Australian Institute of Health and Welfare. Statistics on drug use in Australia 2002. (Catalogue No. PHE 43.) Canberra: Australian Institute of Health and Welfare, 2003. Available at: www.aihw.gov.au/publications/phe/sdua02/sdua02-c02.pdf (accessed Oct 2004).
- 7. Australian Institute of Health and Welfare and the Commonwealth Department of Health and Aged Care. National Drug Strategy Household Survey 2001 [computer file]. Canberra: Social Science Data Archives, Australian National University, 2002.
- 8. Department of Health (UK). New statistics show smokers are losing out in love. (press release) 2000; Feb 11. Available at: www.dh.gov.uk/PublicationsAndStatistics/PressReleases/PressReleasesNotices/fs/en?CONTENT_ID=4002477&chk=ZgpFIc (accessed Oct 2004).
- 9. Leung WC, Harvey I. Is skin ageing in the elderly caused by sun exposure or smoking? Br J Dermatol 2002; 147: 1187-1191.
- 10. Yin L, Morita A, Tsuji T. Skin aging induced by ultraviolet exposure and tobacco smoking: evidence from epidemiological and molecular studies. Photodermatol Photoimmunol Photomed 2001; 17: 178-183.
- 11. Trueb RM. Association between smoking and hair loss: another opportunity for health education against smoking? Dermatology 2003; 206: 189-191.
- 12. Model D. Smoker’s face: an underrated clinical sign? BMJ 1985; 291: 1760-1762.
Abstract
Objective: To determine (i) whether people advertising themselves on a dating website were more likely to be smokers than members of the general population; and (ii) whether attractive advertisers (those whose ads were viewed most often) were less likely to smoke than all advertisers.
Design: Comparison of the number of advertisers who smoke with survey data on national smoking status.
Setting: “RSVP”, Australia’s largest web-based dating site (455 196 members on 12 October 2004).
Participants: 132 176 advertisements accessed on 10 February 2004.
Main outcome measures: Smoking status; and “votes” for advertisers’ attractiveness based on how often visitors to the site accessed individual advertisements.
Results: In every age group, there were higher proportions of women smokers among the advertisers than in the general population (P < 0.05), and this was also the case for men aged 50 years and over. There was a higher proportion of non-smokers in the “Top 100” men or women advertisers aged 20–29 years (82%) compared with total RSVP advertisers in this age group (67%) (P < 0.001). 85% of the men and 78% of the women in the two “Top 100” groups were non-smokers, with only 2% of both sexes claiming to be regular smokers.
Conclusions: Compared with the general population, a higher proportion of women and older men who advertise themselves on a dating website are smokers. Smokers may be perceived as less attractive.