Daniel Homn, Ph.D.
THE SURVEY on which I am reporting was done for the Public Health Service from late December 1967 through early February 1968 by the Chilton Research Services of Philadelphia. The figures are based on telephone interviews conducted as a representative sample of the 85 percent of the U.S. households which have telephone service. These figures are being augmented by personal interviews in households without telephones. Households without telephones tend to be either in rural areas or impoverished sections of the central city. I believe levels of smoking are lower than average in the country and higher than average in the central city. It is, therefore, unlikely that adding the personal interview sample, which is now being carried out, will alter the figures by more than 1 or 2 percentage points. The telephone sample consisted of 4,414 interviews conducted among approximately 315 boys and 315 girls at each single year of age from 12 through 18 (see table). A random selection of teenagers was made from a computer that was fed information on area codes, exchanges, and banks of numbers in use in households throughout the United States. Since any possible number could be selected, even unlisted numbers fall into the sample with their appropriate frequency. Observations The proportion of smokers among teenagers appears to have declined appreciably from levels which have been reported in numerous studies over the past 10 years. Defining "regular" smoking among teenagers as smoking regularly either daily or weekly, one boy in seven and one girl in 12 is so characterized for the entire group between the ages of 12 and 458
18. This frequency varies from only 1.3 percent of the 12-year-old boys and 0.3 percent of the 12year-old girls to 35.5 percent of the 18-yearold boys and 21.3 percent of the 18-year-old girls. In 1957, 34.7 percent of the 17-year-old boys studied in Portland, Oreg. (1), were smoking at this level compared with 25.6 percent in our present sample; 25.5 percent of the 17-yearold girls compared with 15.7 percent in the present sample-a drop of about 9 or 10 percentage points in each group. In the national sample studied by Eugene Gilbert and Company and described in an unpublished report for the American Cancer Society in 1959, all teenage groups had a consistently higher proportion of smokers than their counterparts in the Oregon study. Variations in definitions of what constitutes a smoker make it difficult to compare different studies, but even the study conducted during 1967 in San Diego and another 3-year study completed a year and a half ago by the University of Illinois show significantly higher smoking than has been found in this more recent study. Expectations about smoking in the future are remarkably low. Salber's studies have shown that statements by children in junior and senior high school as to whether they expect to become smokers are accurate predictors of whether they actually do so (2). Only 2.6 percent of the entire sample say they definitely expect to be smokers 5 years from now, but 45.1 percent say Dr. Horn is director, National Clearinghouse for Smoking and Health, National Center for Chronic Disease Control, Public Health Service. This paper u'as presented at the American Cancer Society's Science Writers' Seminar, San Diego, Calif., March 25,1968. Public Health Reports
they definitely do not expect to be cigarette smokers in 5 years. Another 12.3 percent say they probably will be smokers then, and 34.2 percent say they will probably not become smokers. Even including the 5.9 percent who do not know what they expect with those who say they defnitely or probably will be smokers in 5 years, the expectation is 20.8 percent will be smokers (23.1 percent for boys, 18.4 percent for girls) when this group is aged 17-23, in contrast to rates about twice as high for comparable age groups in recent years. If these expectations are indeed...