In this appendix we present our regression analysis of the determinants of output for the three components rabi (non-sugar), kharif (non-sugar), and sugar. The rabi results for the linear model are given in Table 6.A.1 and for the three logarithmic cases in 6.A.2, 6.A.3, and 6.A.4. We have separate logarithmic results since, as explained in § 6.5, the number of observations decreases with the number of included explanatory variables: in Table 6.A.2 we have as logged explanatory variables, land cultivated in rabi and F3; for Table 6.A.3 we add VI, and for Table 6.A.4rabi fertilizer. Table 6.A.5 gives khariflinear results and 6.A.6 and 6.A.7 the logarithmic cases. There are just two such cases since we do not include fertilizer (we have data only for rabi fertilizer). Table 6.A.8 contains the linear sugar results and 6.A.9 and 6.A.10 the logarithmic.
Our criterion for the inclusion of an observation in both rabi linear and kharif linear regressions was that the household concerned be involved in cultivation in the sense of strictly positive GAO. For the case of the single crop sugar we included in the linear regression only those households with strictly positive sugar output. In the logarithmic regressions those households with zero outputs are automatically excluded.
The main features of the results have been summarized at the end of § 6.5 and certain other aspects such as the role of fertilizer in the rabi season, in the main body of the chapter. We do not therefore present any further discussion here.
|Multiple R||0.93||Number of cases||93|
|R Squared||0.86||Number of variables||1|
|Standard error||764.89||F-Value for equation||575.79|
|Variables in the equation|
|Variable||B||Standard error of B||F|
|Variables not in the equation|