568 



.'-.azaris ind stresses. 



Ocr 1S70 putiic cpi.-.mn survey shcwevj -.-.a- 4 .T4::r-.:v 

 (52%1 ttli.e'-'ed eh«t ci^aretta* are onW 3ra_st '-^e -a.-.v ca.sji 

 of 5moV«r» h*vir<3 mere illnoases. It aisc lhcw«d t>.4t ha 1.5 of 

 th« pecple who telievad th*c 3mo<«rs hav« mora illaaaa eh4ft 

 nen-smo'Kers agcepced tn« consticucicnil hypothtais as th« 

 «jcplan*cion. 



Thu», th«r« ir« millions Af p«epl« who would tot rectptiva 

 to « r.«w m«9sa<3«. stating t 



Ci9ar«tt« tmekin<j iTi«y not b« th« health 

 hacard that th« antl'smokin^ p«dpl« say 

 it is b«caus« ethT alternativts af «t 

 I»ast as probable 



Tht Roper Proposal would b« « persuasive (i^ r.ot strictly 

 leientific) mediua for this message , which we h4ve done little 

 to develop in a systematic or «©<oprehenslve way. 



Pellowin<a in my outline ot the ftepa required to start 



a shift in public oplflion xl the Roper Proposal i« accepted. 



A SCSHARIO pR ACTIQK 



1) Select a panel of experts to consult on the design 

 of the study. Ideally they would be prestige :igur»J who would 

 xAitially have a solid contribution to tuKe and who would aUo 

 be Willing to endorse the study publicly »t a later staqe. 



2) Conduct the pilot study. 



3) If ftvortblB, present the results to carefully selected 



members of the followiftg Vey f roups t 



Senate 

 House 



CAbintt 



White House 



State Governors 



Medical School and University Fresid<»nu» 



Selantififi bodies 



T 



OOIJIOOI 



