Andrew J Leidner1, W Douglass Shaw1, Steven T Yen2. 1. Department of Agricultural Economics, Texas A&M University, College Station, TX, USA. 2. STY Health Econometrics, Knoxville, TN, USA.
Abstract
OBJECTIVE: This paper investigates the change through time in the perception of smoking-related health harm and smoking behaviour from 1949 to 1981. BACKGROUND AND CONTEXT: A variety of common behaviours can be linked to chronic disease risk-smoking, over-eating, and excessive sitting, to name a few. Changing behaviours to reduce exposure to such risks can be an effort that spans generations and decades. SETTING AND PARTICIPANTS: Respondents to Gallup Poll surveys in the United States from 1949, 1954, 1957, 1971, 1972, 1977 and 1981. METHODS: Graphical analysis and probit regression are used to investigate trends through time and statistical associations of smoking with the perception of smoking-related health risks and other socio-demographic variables. INTERVENTION AND MAIN VARIABLE STUDIED: Perceived smoking health risk. MAIN OUTCOME MEASURE: Smoking participation. RESULTS AND CONCLUSION: Our findings include the proportions of individuals who were self-reported smokers fell between 1949 and 1981, from 0.48 to 0.34. Among smokers, the proportion who believed smoking was harmful increased from 0.52 in 1949 to 0.81 in 1981. By 1981, the proportion of non-smokers who believed smoking was harmful was 0.98. A negative association between belief in smoking harm and the decision to smoke was shown in regression analysis. This association became more pronounced over the three decades under study.
OBJECTIVE: This paper investigates the change through time in the perception of smoking-related health harm and smoking behaviour from 1949 to 1981. BACKGROUND AND CONTEXT: A variety of common behaviours can be linked to chronic disease risk-smoking, over-eating, and excessive sitting, to name a few. Changing behaviours to reduce exposure to such risks can be an effort that spans generations and decades. SETTING AND PARTICIPANTS: Respondents to Gallup Poll surveys in the United States from 1949, 1954, 1957, 1971, 1972, 1977 and 1981. METHODS: Graphical analysis and probit regression are used to investigate trends through time and statistical associations of smoking with the perception of smoking-related health risks and other socio-demographic variables. INTERVENTION AND MAIN VARIABLE STUDIED: Perceived smoking health risk. MAIN OUTCOME MEASURE: Smoking participation. RESULTS AND CONCLUSION: Our findings include the proportions of individuals who were self-reported smokers fell between 1949 and 1981, from 0.48 to 0.34. Among smokers, the proportion who believed smoking was harmful increased from 0.52 in 1949 to 0.81 in 1981. By 1981, the proportion of non-smokers who believed smoking was harmful was 0.98. A negative association between belief in smoking harm and the decision to smoke was shown in regression analysis. This association became more pronounced over the three decades under study.