Dynamic efficiency measurement: theory and application.

1987 1988 1989 Mean (%) 98.63 99.01 97.7 Median (%) 100 100 100 Std. error (%) 6.45 5.00 6.94 IQR (a) (%) 0 0 0 Maximum (%) 100 100 100 Minimum (%) 55.11 63.38 64.22 Skewness (b) -5.67 -6.29 -3.27 Excess kurtosis 33.58 41.09 10.42 Normality 0.5187 0.5128 0.4864 (p-value) (<0.01) (c) (<0.01) (c) (<0.01) (c) Percent inefficient 7% 7% 13% farmers

1990 1991 1992 Mean (%) 99.04 96.9 98.19 Median (%) 100 100 100 Std. error (%) 5.50 11.0 5.34 IQR (a) (%) 0 0 0 Maximum (%) 100 100 100 Minimum (%) 62.18 31.3 77.4 Skewness (b) -5.87 -3.35 -3.38 Excess kurtosis 34.20 12.22 10.90 Normality 0.5368 0.4478 0.4817 (p-value) (<0.01) (c) (<0.01) (c) (<0.01) (c) Percent inefficient 3% 12% 15% farmers (a) IQR is the interquartile range. (b) The skewness coefficient is the ratio of the third central moment and the cube of the standard error. (c) The 0.99 quantile is the largest quantile presented in the Lilliefors table. Thus, the p-value is some value less than 0.01. Table 5. Upper Bound on Technical Efficiency of All Inputs, by Year

1987 1988 1989 Mean (%) 88.48 90.60 88.48 Median (%) 88.81 95.81 91.09 Std. error (%) 10.11 10.73 11.85 IQR (a) (%) 18.41 18.31 20.17 Maximum (%) 100 100 100 Minimum (%) 65.61 58.81 59.01 Skewness (b) -0.432 -0.702 -0.627 Excess kurtosis -0.982 -0.586 -0.773 Normality 0.1580 0.2850 0.2142 (p-value) (<0.01) (c) (<0.01) (c) (<0.01) (c) Percent inefficient 71% 53% 64% farmers

1990 1991 1992 Mean (%) 89.05 87.69 87.76 Median (%) 87.66 87.97 87.65 Std. error (%) 9.60 11.72 11.91 IQR (a) (%) 18.63 23.83 19.96 Maximum (%) 100 100 100 Minimum (%) 70.20 63.77 62.64 Skewness (b) -0.228 -0.271 -0.509 Excess kurtosis -1.273 -1.382 -0.894 Normality 0.1740 0.2439 0.2251 (p-value) (<0.01) (c) (<0.01) (c) (<0.01) (c) Percent inefficient 72% 64% 62% farmers (a) IQR is the interquartile range. (b) The skewness coefficient is the ratio of the third central moment and the cube of the standard error. (c) The 0.99 quantilc is the largest quantile presented in the Lilliefors table. Thus, the p-value is some value less than 0.01. Table 6. Lower Bound on Technical Efficiency of All Inputs, by Year

1987 1988 1989 Mean (%) 68.67 75.62 74.29 Median (%) 63.66 72.90 72.76 Std. error (%) 14.70 13.75 14 IQR (a) (%) 18.89 19.03 17.25 Maximum (%) 100 100 100 Minimum (%) 47.1 51.08 50.29 Skewness (b) 0.75 0.435 0.498 Excess kurtosis -0.417 -0.757 -0.689 Normality 0.1574 0.1172 0.1035 (p-value) (<0.01) (c) (<0.018) (<0.096) Percent inefficient 93% 87% 89% farmers

1990 1991 1992 Mean (%) 76.02 74.86 76.88 Median (%) 70.67 73.74 72.63 Std. error (%) 12.80 14.71 13.97 IQR (a) (%) 18.58 22.19 16.2 Maximum (%) 100 100 100 Minimum (%) 56.80 49.34 48.91 Skewness (b) 0.526 0.278 0.315 Excess kurtosis -0.93 -0.944 -0.724 Normality 0.1704 0.1089 0.1314 (p-value) (<0.01) (c) (<0.088) ([approximately

equal to] 0.01) Percent inefficient 92% 92% 85% farmers (a) IQR is the interquartile range. (b) The skewness coefficient is the ratio of the third central moment and the cube of the standard error. (c) The 0.99 quantile is the largest quantile presented in the Lilliefors table. Thus. the 1-value is some value less than 0.01. Table 7. Lower Bound on Allocative Efficiency of All Inputs, [A.sub.g3], by Year

1987 1988 1989 Mean (%) 60.94 65.57 63.61 Median (%) 62.34 68.06 66.07 Std. error (%) 21.25 19.39 18.29 IQR (a) (%) 25.79 24.98 23.17 Maximum (%) 100 100 99.20 Minimum (%) 16.48 16.64 20.72 Skewness (b) -0.124 -0.39 -0.408 Excess kurtosis -0.554 -0.346 -0.302 Normality 0.0695 0.0757 0.0742 (p-value) (<0.20) (c) (<0.20) (c) (<0.20) (c) Percent inefficient 97% 97% 100% Farmers

1990 1991 1992 Mean (%) 63.92 65.82 61.27 Median (%) 63.34 67.38 69.75 Std. error (%) 19.92 20.14 25.02 IQR (a) (%) 20.89 25.68 37.95 Maximum (%) 100 100 100 Minimum (%) 10.28 15.61 5.54 Skewness (b) -0.454 -0.361 -0.613 Excess kurtosis 0.189 -0.25 -0.691 Normality 0.13 0.0605 0.1929 (p-value) (<0.25) (<0.20) (c) (<0.01) (d) Percent inefficient 98% 98% 98% Farmers (a) IQR is the interquartile range. (b) The skewness coefficient is the ratio of the third central moment and the cube of the standard error. (c) The 0.80 quantilc is the smallest quantile presented in the Lilliefors table. Thus, the p-value is some value greater than 0.20. (d) The 0.99 quantile is the largest quantilc presented in the Lilliefors table. Thus. the p-value is some value less than 0.01. Table 8. Upper Bound on Allocative Efficiency of All Inputs, Ago, by Year

1987 1988 1989 Mean (%) 86.62 99.84 87.02 Median (%) 100 100 96.86 Std. error (%) 20.85 1.28 19.12 IQR (a) (%) 16.40 0 19.47 Maximum (%) 100 100 100 Minimum (%) 19.59 90.02 25.20 Skewness (b) -1.628 -7.617 -1.595 Excess kurtosis 1.70 56.02 1.732 Normality 0.2659 0.5345 0.2486 (p-value) (<0.01) (c) (<0.01) (c) (<0.01) (c) Percent inefficient 48% 1.6% 56% farmers

1990 1991 1992 Mean (%) 82.47 83.56 81.93 Median (%) 91 97.69 90.91 Std. error (%) 19.86 21.76 21.86 IQR (a) (%) 29 28.39 30.77 Maximum (%) 100 100 100 Minimum (%) 33.46 32.06 20.89 Skewness (b) -0.922 -1.07 -1.032 Excess kurtosis -0.504 -0.229 -0.126 Normality 0.1918 0.2502 0.2043 (p-value) (<0.01) (c) (<0.01) (c) (<0.01) (c) Percent inefficient 69% 61% 64% farmers (a) IQR is the interyuartile range. (b) The skewncss coefticient is the ratio of the third central moment and the cube of the standard error. (c) The 0.99 yuantile is the largest yuantile presented in the Lilliefors table. Thus, the 1-value is some value less than 0.01.