Abstract
Purpose - The tobacco industry consistently opposes measures that would limit their marketing, but provides little empirical evidence to support its position. This paper aims to test claims that pictorial health warnings on tobacco products would be no more effective than text-only warnings.Design methodology approach - Three studies used face-to-face interviews with smokers and non-smokers to compare pictorial and text-only warnings. Two studies used semantic differential scales to estimate cognitive and affective responses to pictorial and text-only warnings, and the Juster Scale to provide behavioural estimates. The final study used best worst scaling to compare paired pictorial and text-only warnings.Findings - Images featuring medical and social images elicited stronger affective, cognitive and behavioural responses than a control, text-only message. Comparisons of refreshed text and pictorial warnings found the latter elicited stronger reactions while the former produced similar results to the control. Updating text warnings did not render these more effective; however, adding an image to an existing text warning made this more effective than the control.Research limitations implications - Arguments advanced by the tobacco industry need empirical analysis to assess their validity.Social implications - This study provides evidence that pictorial health warnings are more effective than text warnings and suggests that refreshing the text used in warning labels, the alternative promoted by the tobacco industry, would be less effective than introducing pictorial warnings.Originality value - This is the first comparison of pictorial and refreshed text warnings; the findings challenge the tobacco industry's position on tobacco warning labels and contradict arguments used to oppose the introduction of pictorial warning labels.