Date of Award
Master of Science
David W. Brown, B. D. Raskopf
The objective of this study was to determine the extent to which secondary time series data could be used to explain the increase in beef cattle numbers in Tennessee.Two secondary objectives were to determine if the data suggest that:1. Dairy and sheep numbers are relevant variables in explaining beef cattle numbers in Tennessee.2. The concept of opportunity cost along with the cost of hired labor is a relevant variable in analyzing beef cattle numbers in Tennessee.Secondary time series data for the time period 1925-1965 was used to calculate seven regression estimating equations. The dependent variable estimated is the January 1 inventory number of beef cattle inTennessee.The independent variables consisted mainly of enterprises and their product prices that compete with the beef cattle enterprise inTennessee. The variables were used as proxy variables to evidence changes in opportunity cost associated with alternative agricultural enterprises and to determine the influence of changing competing product price on beef cattle inventory numbers.In addition, an index of farm wage rates was used to determine the influence of labor cost upon the adoption or intensification of the beef cattle enterprise in Tennessee. A 0-1 type dummy variable was used to differentiate the World War II period from the preceding and subsequent years included in the study. Two other variables used were calf prices and cow prices.Four primary models were calculated. In all the models, most data were averaged for three years with the value of the dependent variable being expressed as a function of the average of the three preceding years observations on the independent variables. The price data in three of the primary models was transformed into ratios by the use of one of three variables. They are: (1) a moving three-year average of cattle prices; (2) a moving three-year average of calf prices; and (3)an index of farm wage rates calculated by the United States Department Of Agriculture for Tennessee. The use of ratios reduced the amount of correlation between the independent variables.Using the results obtained above, three improved models were calculated after removing highly correlated variables. The improved models were calculated and selected as estimating equations on the basis of (1) the economic rationality of the variables included, (2) the amount of multicollinearity present, and (3) the size of the standard error of the estimate.Sheep and lamb numbers were entered in three of the seven models as an explanatory variable. In all three cases, the b coefficient had a negative sign attached. This indicated that an increase in the three-year average of sheep and lamb numbers had an associated decrease in beef cattle inventory numbers. The time lag was not explicitly defined,but the construction of the model required that it be a function of an average value of the past three years observations on sheep and lamb numbers, along with other variables.The seven models did not give evidence to support the hypothesis that increasing beef cattle numbers can be explained in part by the decline in dairy numbers. In all but one model, there was a positive coefficient which implied that increasing dairy cattle numbers have associated increases in beef cattle numbers. This net relationship is not acceptable on the basis of economic reasoning since the two enterprises are considered to be generally competitive. The b coefficients are significant at the .01 level. The negative b coefficient in the one model was not significant.The index of farm wage rates had a positive b coefficient in all the models in which it was used as an independent variable. This gives support to the contention that ceteris paribus increases in the cost of labor at the farm level should result in shifts from labor intensive enterprise to less time consuming ones and/or tendencies toward a greater utilization of available farm labor. In both cases, beef cattle enterprises could provide the means to accomplish the shift.The use of a 0-1 dummy variable did not give evidence to support arguments for either a larger or lower intercept value during theWorld War II years.In attempting to find an acceptable estimating model, the conclusion was made that for practical purposes, the January 1 inventory number of beef cattle can be written in terms of three explanatory variables. The model is:The model is: Yt = 2323.90 - 4883.20X5 - .5228X12 - l.52X13
(1059.60)** (.0172)** (.2072)**
Yt = The January 1 inventory number of beef cattle in Tennessee
X5[a three-year average of lagged calf prices (dollars/cwt)]/w3.
w3 = an index of farm wage rates for Tennessee.
X12 = a three-year average of lagged corn acreage harvested
X13 = a three-year average of lagged oat acreage harvested
(1000's acres)In conclusion, secondary time series data provided some relationships that may be useful in explaining beef cattle increases inTennessee. However, cross sectional analysis is needed to analyze the relationships suggested by the models in this study.
Kerley, Charles Ray, "An economic analysis of the beef cattle enterprise in Tennessee using multiple regression on secondary time series data. " Master's Thesis, University of Tennessee, 1970.