Maximum-likelihood estimates of the stochastic frontier profit functions for loss

In this section, only the results of the profit function models incorporating weather-related variables were discussed according to the result of rejecting the hypothesis testing of no effect of weather-related variables. Maximum-likelihood estimates of the stochastic frontier profit functions of loss group and no loss group are presented in Table 4.5.

The coefficients results of the profit function in both loss group and no loss group reveal that the variables for normalized prices of human labor and land preparation are negatively and highly significant, explaining the higher the prices of these inputs, the lower the

73

Table 4. 5 Maximum-likelihood estimates of the stochastic frontier profit functions of loss group and no loss group

Variables Loss group No loss group

Coefficient Std. Err t-ratio Coefficient Std. Err. t-ratio Profit function

Constant 23.10*** 1.44 16.08 6.80*** 0.52 12.99

Rainfall -0.30*** 0.00 -115.46 -0.11*** 0.02 -5.38

Dummy for

replanting 0.69*** 0.06 11.50 - - -

PS -1.54*** 0.10 -14.72 0.07 0.07 0.98

PF -0.32*** 0.03 -10.65 0.11*** 0.02 6.42

PC -0.97*** 0.09 -10.92 0.09*** 0.03 2.91

PL -1.16*** 0.07 -17.55 -0.28** 0.11 -2.53

PLP -0.20*** 0.03 -6.56 -0.47*** 0.15 -3.21

Seed rate 0.54*** 0.13 4.07 0.34*** 0.08 4.43

Fertilizer -0.57*** 0.03 -18.45 0.10*** 0.01 10.63

Chemicals -1.01*** 0.07 -13.66 0.11*** 0.03 3.92

Human labor -1.84*** 0.08 -21.92 0.09 0.05 1.60

Variance parameters

𝜎²= 𝜎_𝑢²+ 𝜎_𝑣² 17.76 1.10 16.15 23.01 2.03 11.33 𝛾 = 𝜎_𝑢²/(𝜎_𝑢²+ 𝜎_𝑣²) 0.99*** 0.00000003 37116718 0.99*** 0.00000011 8621144 Log likelihood

function -137.77 -159.79

Profit inefficiency effect function

Constant -0.99 1.03 -0.96 -2.71** 1.00 -2.72

Gender -4.73** 1.68 -2.81 2.78** 0.99 2.81

Age 3.79*** 1.27 2.98 -10.16*** 0.92 -11.06

Experience -0.04 0.84 -0.05 8.63*** 1.13 7.61

Education -0.77 1.25 -0.62 4.17*** 0.92 4.55

Credit access -6.86*** 1.43 -4.80 -2.86*** 1.02 -2.81

Participation in farmer

organization

-1.26 1.06 -1.19 -19.72*** 1.56 -12.65

Training access -5.12*** 1.36 -3.76 3.56*** 0.99 3.60

Location -0.21 1.20 -0.17 12.07*** 1.00 12.04

Pulse area -1.79** 0.81 -2.21 -5.37*** 0.78 -6.91

Total number

of observations 73 109

Note: ***, **, * represents significance at the 1% (p<0.01), 5% (p<0.05) and 10% (p<0.10) level.

Source: Own estimates

74

profitability of the farmers. The negative significant effect of normalized wage rate of human labor is stronger in loss group than no loss group, pointing out that increase in wage rate of human labor will have more impact on the profitability of pulse farmers in loss group by reducing net return from pulse production.

In loss group, the coefficient values of normalized prices of fertilizer and chemicals, and of used quantity of fertilizer and chemicals are negatively significant at 1% level, respectively, while the results of those variables in no loss group has a positive significant impact on profitability at 1% level and are opposite to the results of loss group. The results can be interpreted that when yield loss due to rain incidence occurred, the more utilization of these inputs would hinder to achieve the maximum profitability and indicated that the efficient and effective application of these inputs is necessary for the farmers.

In loss group, the coefficient values of normalized prices of fertilizer and chemicals, and of the used quantity of fertilizer and chemicals are negatively significant at 1% level, respectively, while the results of those variables in no loss group are opposite to those of loss group and have a positive significant impact on profitability. The results can be interpreted that when yield loss due to rain incidence occurred, the more utilization of these inputs would hinder to achieve the maximum profitability and indicated that the efficient and effective application of these inputs is necessary for the farmers.

The coefficient value of normalized prices of seed is negative and significant at 1%

level and seed rate is positively significant at 1% level, denoting that the increased use of seed rate can improve profitability as the more use of seed can ensure for the profit of the farmers in the mean of replanting, while the increase in price of seed can decrease profit of the farmers for loss group. On the other hand, the coefficient of seed rate in no loss group also results in a positive significance at 1% level, and the interpretation for this result can be gone the same trend as in loss group. Moreover, in loss group, the result of human labor is also negatively

75

significant impact on profitability at 1% level, inferring that excessive use of human labor would reduce the profit from pulse production, while it is positive and not significant in no loss group, meaning that this variable are more responsive to yield loss due to rain incidence.

The results of the estimation of profit inefficiency effect function are illustrated in the lower part of Table 4.5.

In loss group, the dummy variable for gender is negatively significant on profit inefficiency at 5% level, implying that the male farmers, as expected, are more efficient in profit than female farmers. It may be because of the definitive decision of the male farmers regarding farm management under crop yield loss condition due to rain incidence. But, in no loss group, gender has a positive significant effect on profit inefficiency at 5% level, inferring that female farmers are more efficient than male farmers under no loss condition because of no necessity of any special management for production processes.

The age of the farmers is positively significant at 1% level on profit inefficiency in loss group, indicating that the older farmers are less efficient than the younger farmers due to the less motivation and enthusiasm on adopting updated modern farming technology as well as explaining that the younger farmers are more active in handling the adapted measures of encountering erratic rain incidence during the crop season. However, the coefficient result of age in no loss group is negatively significant at 1% level, indicating that older farmers are more efficient than younger farmers under no yield loss effect due to rain incidence.

In both groups, the coefficient values of credit access variable are negatively significant at 1% level on profit inefficiency, directing credit access from government can encourage farmers for more capital investment and thus enabling farmers for using farm inputs in time, consequently enhancing productivity and hence profitability for both groups.

In addition, the results of pulses area have a negative significant impact on inefficiency at 5% level in loss group and 1% level in no loss group. The results depict that large-scale

76

farmers are more efficient than small-scale farmers. The reason may be due to the vulnerability of small-scale farmers in encountering crop damage and yield loss by rain incidence is larger than large-scale farmers in term of capital and management skill. Moreover, the effect is stronger in no loss group, mentioning that the impact of yield loss by rain incidence can decrease profit efficiency for loss group compared to no loss group even the farmers are large-scale farmers.

Training access variable in loss group also has a negative significance at 1% level on inefficiency, implying the more access to training for farmers, the more efficient in profit. On the other hand, in no loss group, the result of training access variable is positively significant at 1% level, which is opposite to the loss group. The result directs that training access is more important and makes more efficient for the farmers in loss group than those in no loss group as the effectiveness training access is lower in no loss group than loss group.

In no loss group, the estimated results of experience, education and location have a positive significant effect on profit inefficiency, while those variables in loss group are negative but not significant. The results can be interpreted that the farmers with more experience, higher education level and farmers from Bago Region are less profitably efficient than those with less experience, low education level and from Yangon Region. The possible reason may be that experience is regardless for the profit efficiency improvement, farmers with higher education level may have another job other than farming making less attention on production activities, and farmers from Bago Region are more affected by crop damage due to rain incidence than those from Yangon Region.

The result of participation in farmer organization variable in no loss group has a negative and significant relationship with inefficiency while it is negative but not significant in loss group. The result reveals that farmers who participate in farmer organization can achieve

77

higher profit efficiency for no loss group, participation in farmer organization is one of a key factor for the profit efficiency improvement of the farmers.