points C or D would be expected to approximate
the optimal level of beef cow production. Rates of
transformation included in Table 2 indicate that
increasing cows above 78 at point D results in a
reduction of profits of $62.49 per cow, while in-
creasing cows up to 71 at point C has a rate of
transformation of 20.59 per cow. Thus, C or D
would indicate maximum utility for a considerable
range in the rate of substitution of profits for cows
in the utility function of farm owners. Of course,
preference functions could exist which would re-
suit in the optimum being between D and E. How-
ever, such a high rate of transformation between
cows and profits would not likely be consistent
with behavior strongly influenced by status con-
siderations. Profits, and particularly uses of profits
for consumption or investment, are also associated
with status, so that preferences which can be
characterized by Veblen’s concepts would be ex-
pected to value both cows and profits. There, a
farmer with preferences cognizant to social status
and resources of the representative farm would
be expected to have a beef cow herd between 71
and 78 cows.
Empirical Relevance of the Model
Survey data collected by Allison [1], [2] on
beef cattle production in Georgia provide a source
of information to test the utility maximization
model. In his report on owners’ conceptions of
the beef enterprise, Allison does support the con-
spicuous production hypothesis when he states: “A
sizeable portion of those farmers who have just
increased herd sizes gave psychological factors
(father was cattleman or enjoyed raising beef) as
the main reason for including the beef operation
in their organization.” [1, p. 24]. More import-
antly, actual data on beef herd sizes and land
utilization can be contrasted with the beef profit
frontier for the representative analytical farm.
Allison reports that average acres of open land
for herds of 50 to 99 cows in the Piedmont was
297 acres which was the closest to the represen-
tative farm of any averages for other herd sizes
[l,p. 16].
Additional evidence was obtained from tabu-
lation of survey data on the 16 sample farms in
the Piedmont with 150 to 350 acres open land:
four had 20 to 49 beef cows, seven had 50 to 99
cows, and six had no cows (2). The most striking
feature of this tabulation is the absence of any
herds of less than 20 cows. This is in the range
of the profit maximizing herd size for the rep-
resentative analytical farm. In addition, nearly half
the cases had more than 50 cows, which is within
the utility maximization range of the theoretical
and empirical analysis. The existence of six farms
with no beef cows does confound the evaluation
of the model, in that the same utility maximization
model does not apply to every farm unit in the
sample. However, farms with no cows may have
a different land base than the representative analy-
tical farm. Allison reports that 99 percent of the
open land was cropland on farms with 0-9 cows
and 100-199 acres of open land, and 97 percent
cropland on farms with 0-9 cows and 200 or more
acres of open land. For farms with 50-99 cows,
the percentage of cropland was 80, which is simi-
lar to the 77 percent on the analytical farm [1,
p. 16]. Thus, crops may have been more com-
petitive with beef on the nonbeef farms because
of the higher percentage of cropland than on the
analytical farm.
Survey results on land utilization provide
further support for the utility maximization model.
Allison reports 16 percent idle openland on farms
with 50-99 cows in the Piedmont [1, p. 16]. These
observed acres more closely correspond to acre-
ages predicted by profit maximization. With profit
maximization, 120 of the 243 acres on the repre-
sentative farm were idle. The increased numbers
of cows predicted by utility maximization are
associated with more complete land utilization.
With 53 cows, 42 acres of idle land is present;
however, with 68 or more cows, no idle land
exists. Since Allison’s data indicate that some
idle land is characteristic of the Piedmont, an
optimal herd would likely be no larger than 71
cows (point C).
CONCLUSIONS AND IMPLICATIONS
This paper evaluates the possibility that a
utility maximization model could explain the level
of beef cow production in the Georgia Piedmont
more accurately than a profit maximization model.
The utility model incorporated the hypothesis that
beef cows are a form of conspicuous production,
resulting from their historical association with
agricultural indicators of social status. Based on
an analysis of a representative farm situation, the
optimal organization was in the range of 71 beef
cows and profits of $9,332 compared to 16 cows
and profits of $10,041 at profit maximization. The
rate of substitution between profits and beef cows
under utility maximization is approximately $20
per cow. Survey data on beef production in the
Georgia Piedmont collaborated the utility maxi-
93