114 CIRCULAR 249, U. S. DEPARTMENT OF AGRICULTURE 
age of ingredients other than tobacco present in chewing tobacco, 
especially plug, comparisons of per capita consumption of individual 
prea according to their manufactured weight are somewhat dis- 
torted. 
There is no actual measure of the ratio of smokers to total popula- 
tion, but it is common knowledge that changes in the ratio have 
occurred. The most important factor in this was the great increase 
of smoking among women beginning about 1915. An additional 
factor which may not be so well understood lies in the changing ratio 
to the total population of the group of individuals of smoking age. 
The advent of widespread smoking by women meant only that within 
a certain age group, say 20 years or older, many potential smokers 
became actual smokers. 
This discussion is about the factors that affect the ratio to 
total population of the age group which includes both actual and 
potential smokers. For various uses population statistics are broken 
down into several age groups, including some comprised of children 
obviously not of smoking age. For present purposes it may be 
assumed that individuals 20 years of age or over are either actual 
or potential smokers, and, as a rather strained corollary, that persons 
under 20 do not use tobacco. In computing per capita consumption 
of tobacco, however, the total population is used as the divisor. 
If the consumption as thus computed is to be strictly comparable one 
year with another, it becomes important to consider whether the 
relative importance of the respective age groups has remained reason- 
ably constant. Factors affecting the age distribution in population 
statistics which bear directly upon this question of comparability are 
the decreasing death and birth rates. 3 
The effect of a progressive decrease in the death rate is to leave larger 
numbers, year by year, remaining in the ranks of actual or potential 
smokers. It is a change that has been progressing slowly over a long 
period of years. The decreasing birth rate, on the contrary, has the 
effect of diminishing the ranks of those too young to smoke. This 
change is of much more recent development and supplements the 
decreasing death rate in its effect on per capita consumption computa- 
tions. Their combined effect on the age-group comparisons has been 
especially significant since 1930. For example, in 1930, 61.2 percent 
of the total population was 20 years of age or older; by 1940 this group 
had increased to 65.5 percent. The effect of changes of this kind on to- 
bacco consumption may be illustrated as follows: Assume that for 
2 years under comparison the total population remained unchanged, 
and that the average rate of consumption within the group of 20 
years or over likewise remained unchanged. Then any change in 
total per capita consumption would be due to a change in the relative 
importance of that age group. For example: 
Year X Year Y 
Population group Million persons Million persons 
(1) Nonsmoking age—19 years or under__-_---- -- 40 30 
(2) Smoking age—20 years or over____-_------- 90 100 
Total populations 222-222 ae 130 130 
Pounds of tobacco Pounds of tobacco 
Per capita consumption within group 2_______-_-- 5 5 
Total consumption. =. oe eee 450, 000, 000 500, 000, 000 
Per capita consumption based on total population— 3. 46 3. 85 
