478 



-36- 



to »ocUl classes A or B, «nd to be light smokers. These trends, bovever, 

 do not resllr explsio some of the differences betveen clusters. For 

 Instance Cluster 10 eontsins one of the highest percentages of low tar 

 smokers, and yet 09 average members of this cluster are not specially 

 health conscious, not apprtciably older than average, nor upper class, 

 and they are certainly nat light smokers. The explanation of this 

 could be that one is dealing vith very small minorities of each cluster, 

 and that the profiles of these minorities in terms of health concern, 

 age and so on are rather different f^oa the overall profiles of the 



clusters to wiiich they belong. •- 



^ • 



As-'flready indicated, most of chose who have at some time smoked 

 '■'.; ■ > . 



a low tar brand regularly finish up as "former toftokers" of such s brand; 



^ O "^"^ 



in other words '^hey give it up in favour of a braa<r*ith higher delivery. 



However the proportion who give it up varies substantiM^iy beCvMo the 



various categories of <^ker shotm in Table 11. For instiate 50Z of 



social classes A snd B giv6 up ss»king low tar cigarettes as against 



60Z of class C2. Aaoog the' cni^ers as many as 90Z of Cluster 11 give 



up compared vith about SSZ of CIuYcks 9 and 10. Overall 72Z of those 



who have ever smoked a low Car brand regularly have given it up in 



favour of a higher delivery brand (168 out, of 232 smokers). This is 



not a very high success rate, and It la Lastn^(ive to examine some of 



- ' /' 

 ■ ■ ~ -^ 



the stated reasons for giving up low tar brands. Tbe three most coomonly 



-v, 

 stated reasons were: "coo mild/insufficient taste"; "Coo expensivt"; 



"just wanted a change". Table 12 shows the number of smotiers giving c 



each of these answers broken down by degree of concern for health. "^ 



o 



