Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Agricultural Economics

Major Professor

Steven D. Mundy

Committee Members

Luther H. Keller, John R. Brooker, William E. Cole


This study focused on smallholder tea production as one strategy for Tanzanian economic development. The central research question was the economic evaluation of the potential of this strategy for increasing peasant farmer income streams in an economy comprised of mostly subsistence farming.

Two objectives were: (1) the assessment of the extent to which selected production factors explain the variability in the annual output of green tea leaf (output) and (2) the estimation of the physical resource productivities and the computation of marginal value products (MVP's) of selected resources for a sample 150 smallholder tea farms in Njombe, Tanzania to determine economic efficiency and the potential for enhanced utilization of selected resources.

A conceptualized production function was specified in a single equation model in which the annual output, the dependent variable, was a function of labor actually used, land area in tea, bush population, expenses for farm tools and expenses for fertilizer and insecticide. The model also included zero-one variables: two levels of education of the household head, no education and primary or more; and two farmer perceptions about tea processing plant capacity, adequate and inadequate. This was the main model called Model II. Three other models were specified. Two of the three were subproduction functions, referred to as Model I (output as a function of bush population categories) and Model III (output as a function of labor categories). The fourth model was a modification of Model II with discrete zero-one farm organization variables added.

The functional form for Models II and IV was specified as a modified Cobb-Douglas (except for discrete variables). The functional form for Models I and III was linear statistical analysis and was performed using ordinary Least Squares.

The results of Model II indicated that the coefficients for total labor, land area, bush population and farmer perception of inadequate processing plant capacity were statistically significant at the 5 percent level. The coefficient for household head education level of primary or more was statistically significant at 10 percent. Except for expenses on farm tools, all signs were as expected.

The results of Model II suggested constant returns to scale. Labor was the most important factor, followed by land and bush population, in explaining variation in output per farm. The potential for enhanced utilization of labor, land and bush population was evident with each respective MVP being greater than its MFC.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."