RESULTS AND DISCUSSIONS
5.5 Summary
The present chapter discussed (i) the estimation of exploitation of the small tea growers by the buyers of tea leaves and (ii) the estimation of factors determining the production function and possibility of presence of technical inefficiency. The major findings of the chapter are summarised as follows:
1. Supply function was estimated considering the factors price of the output, price of urea, price of MOP, price of SSP, price of cow-dung, price of vitamin, price of herbicides, wage rate of male workers, wage rate of female workers, farm-size and dummy for linguistic community.
The four instrumental variables for price of tea leaves were: number of sales option of a farmer, distance from the farm to the point of sale, dummy for districts with high concentration of STGs and dummy for self sale of tea leaves. Some explanatory variables are found to be statistically significant. Price of urea (PU), price of SSP (PS), wage rate of male labour (wa_m), wage rate of female labour (wa_f) and the linguistic community of the grower belongs to are statistically significant.
2. From the supply function we have calculated the price elasticity of supply. This was used to estimate the monopsonistic exploitation of the small tea growers by the buyers of small tea leaves. The degree or the index of exploitation is found to be 0.19.
3. The data has been chategorized into four groups according to certain criteria to check intergroup variations of the coefficients of price elasticity of supply and in turn compare their degree of exploitation. Exploitation was found to be higher in the districts where concentration of STGs are less (compared to districts where the concentration of STGs are more), who belong to the Assamese community (compared to non-Assamese), who have a single sales options (compared to those with multiple sales option).
4. We used two alternative models to estimate production function of tea leaves with the
function of agricultural goods. In Model 2 along with farm specific regular economic variables some social and non-farm specific variables are also included. The explanatory variables considered in the Model 1 were labour, land, fertilizer, pesticides and capital cost, whereas along with these five explanatory variables Model 2 includes two non-farm specific factors viz.
dummy for linguistic community and dummy for districts with high concentration of STGs. Our preferred specification is Model 2 since it is more comprehensive. We are reporting results of Model 2 here, although the results of Model 1 are not very different.
5. The Cobb-Douglas production function was used. Out of all the explanatory variables, land, fertilizer, capital cost and dummy for districts with high concentration of STGs are found to be statistically significant. The coefficient of elasticity of output with respect to land is 0.71, showing one percent change in land directly changes the production of tea leaves by 0.71 percent. The coefficient of elasticity of output with respect to fertilizer is 0.07, one percent change in fertilizer changes the production of output by 0.07 percent. In the same way, the coefficient of elasticity of output with respect to capital cost is 0.09, one percent change in capital cost changes the production of output by 0.09 percent. Lastly, the dummy for districts with high concentration of small tea growers are found to be statistically significant. Compared to districts with low concentration of STGs, in the districts with high concentration of STGs the output is significantly higher. This is indicated by 0.64.
6. It is observed that between Model 1 and Model 2 there is only minor change in the results of the coefficients of the explanatory variables or their significance levels. So we can say that results we have obtained are statistically robust.
7. Land is always significant. In both the models the elasticity of output with respect to land is found to be 0.71 which is higher than the elasticity of output with respect to any other explanatory variable. Fertilizer is also important although the effect on output is lower. Capital cost is also found to be significant in case of Model 2 (it was not significant in case of Model 1).
8. In reality farms may not use the available resources efficiently, this results in technical inefficiency and a consequent loss in output. Hence in the next step the stochastic frontier (SF) model is used to estimate the production function and to find the presence of technical inefficiency. The results are robust as the same set of explanatory variables are found to be statistically significant in the OLS estimation and in SF estimation. The relative importance of the explanatory variables has also remained the same. In either case of Model 1 and Model 2
high concentration of STGs was found to be statistically significant, and this result is not very different from what we obtained in the OLS case. This consistency leads us to conclude with a degree of confidence that these variables are vital as far as production of tea leaves is concerned.
It is observed that age of the farm, age of the farmer, availability of own transportation facility of the farmers and TB registration of the farmers have dampening effects on inefficiency.
The mean technical inefficiency is found to be 4.82.