As we have demonstrated in a farm-level anal-
ysis, a farm-level demand elasticity that is “too”
elastic will inflate the types of costs evaluated in
Johnson and McManus’ article. Given the -.05
farm-level elasticity assumed in this comment,
“public costs” would be reduced $676.8 million
from the $716 million reported in their article.
Proportional decreases in their other values were
found, with “reduction in public costs” declining
from $325 to $17.8 million; “producer-consumer
surplus loss” declining from $74 to $4.1 million;
and “net reduction in social costs” declining
from $251 to $13.8 million.
If their article were intended only to present a
theoretical framework, perhaps any criticism of
their assumed elasticity values would be unmer-
ited. However, while elasticity values have no
impact on the mathematical operations per-
formed, the quantitative results are extremely
sensitive to the values assumed for elasticities,
and, thus, the results of their application have no
meaning. Researchers who wish to employ John-
son and McManus’ technique should recognize
this fact and exercise extreme care in selecting
the elasticity values to be used.
While supply elasticity does not appear to be a
problem in this analysis, one should also recog-
nize that its accuracy is as critical as the demand
elasticity. A supply elasticity that is “too” elas-
tic will deflate the types of costs evaluated by
Johnson and McManus; conversely, if “too” in-
elastic, costs will be inflated.
In summary, Johnson and McManus have
presented a theoretical framework that can prove
useful in analyzing a wide range of policy issues
involving social costs. However, anyone wishing
to employ their technique should recognize that
the results of the analysis will be very much de-
pendent upon the elasticities assumed, and that
the validity of the findings will necessarily be lim-
ited by the accuracy of their elasticities.
Finally, we would suggest that any evaluation
of social costs should also consider the tax bur-
den on the commodity or product involved.6 In
cases where punitive taxation is involved, as is
the case with tobacco, the social costs still exist,
but they may be indemnified by taxes on the
product—in some cases there may even be a net
gain to non-consumers of the product. Thus,
questions of equity should be examined along
with social cost considerations.
REFERENCES
Johnson, Ruth C. and B. R. McManus. “A Theoretical Framework for Analyzing Social Costs of The
Tobacco Program.” S. J. Agr. Econ., 11(1979):103-06.
Maier, Frank H. “Consumer Demand for Cigarettes Estimated from State Data.” J. Farm Econ.,
37(1955):690-704.
Sackrin, S. M. “Factors Affecting the Demand for Cigarettes.” Agr. Econ. Res., 14(1962):81-88.
Sutton, Russell W. “An Econometric Analysis of the Structure of the U.S. Tobacco Industry.” Ph.D.
dissertation, University of Kentucky, 1974.
Tobacco Tax Council, Inc. The Tax Burden on Tobacco. 14(1979):5, 6.
U.S. Department of Agriculture. Agricultural Statistics. Washington, D.C., 1977, p. 98.
6 According to the Tobacco Tax Council, in 1979, the total tax directly imposed on one pound of tobacco totaled approximately $5.00 at the retail level (Tobacco Tax
Council, pp. 5, 6).
158