Article Text

Download PDFPDF

Blocking access to online tobacco sales sites
  1. K A Reagan,
  2. T Hong,
  3. E L Cohen,
  4. M J Cody
  1. Annenberg School for Communication, 3502 Watt Way, ASC 101C Annenberg Building, University of Southern California, Los Angeles, CA 90089-0281
  1. Correspondence to:
 Michael J Cody;
 ctheory{at}usc.edu

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Recent research expresses concern about adolescents attempting to buy cigarettes on the internet.1 Since the Master Settlement Agreement restrictions on the tobacco industry do not apply to the internet, the internet is an open channel for pro-tobacco images and promotions. According to Forrester Research, sales will approach $5 billion by 2003, potentially causing states to loose $1.3 billion dollars in tax revenues.2 Frequent exposure to icons and symbols increases liking, and can make unhealthy activities appear “normative”. Craving or possessing tobacco promotional materials is related to positive attitudes toward tobacco and to susceptibility.3, 4 The images of pro-tobacco sites can make tobacco use appear glamorous, as tobacco websites portray smokers as young, thin, and attractive, and often feed into young girls' insecurities.5–7

A number of “sting” operations highlight the fact that underaged individuals have ready access to buying tobacco online.8, 9 Given the inevitable use of filters for schools, libraries, and some places of employment, we believe that filtering programs should be effective in limiting access to sites they monitor, and that “stealth” blocking be avoided. The Center for Media Education (CME) tested the ability of programs to block access to 45 tobacco and alcohol sites, and concluded: “ . . .filters do not effectively screen promotional alcohol and tobacco content.”10 To test this concern and to evaluate internet monitoring access products, we reviewed 28 programs available for blocking access and selected four that included tobacco as a category for blocking: Bess/N2H2, Cyber Patrol, CYBERsitter, and iWay Patrol. Each program was tested separately. Testing occurred during the last two weeks of March of 2001, and re-tests were done 11–13 April 2001 for pages that did not load in March.

Random samples were drawn for a content analysis project that ultimately included 316 pro-tobacco websites.5 Of these, 154 sold tobacco products and were used in the present analysis. Most sold cigarettes (67 sites), or cigars (49 sites), while some sites sold multiple products as well as “other” products (pipe tobacco, snuff, chew) (38 sites).

Table 1 presents a summary of the results concerning blocking access. The only program to block more than half of the websites was Bess/N2H2, which blocked 65%. It is also alarming that the programs tested disagreed on what to block. At best, Bess/N2H2 and Cyber Patrol blocked the same 33 sites, which is a small amount of agreement.

To ameliorate this problem, we believe that subscribers should be empowered to add to the “not” lists. Lists of blocked sites should be transparent so subscribers know what is or is not accurately blocked by the filtering programs they use. Additionally, tobacco control advocates can become actively involved in this process. First, most filtering programs welcome input and allow individuals to submit websites at a location on their home page. At a higher level of involvement, tobacco control advocates can create a rating system like RACi for coding content based on different levels. Importantly, one category could be created for tobacco (and alcohol) sponsorships that can be activated by those parents, school teachers or library officials who prefer not to have children view or download tobacco related materials (for example, highlighting tobacco sponsored NASCAR races) from the web.

Table 1

Filter performance of four software programs

Acknowledgments

This research is made possible with funding from Proposition 99 to University of Southern California contract 99-85316 by CDHS/TCS.

References

Footnotes

  • * Level 1 could include sex and smoking, underaged smoking, erotic posturing, smoking, and bondage. Level 2 could include erroneous or harmful information. Level 3 could include sites that simply sell cigarettes, cigars, and so on.