UBC Theses and Dissertations
Production function analysis of paddy farming in Sri Lanka Abeysekara, W. A. Terrence
The primary purpose of this thesis was to analyse underlying input-output relationships in paddy farming in Sri Lanka. Cross-section study data involved 107 paddy farms from five major paddy districts. The' period under study was the 1972-73 Maha paddy season. Data collection was based on farm record keeping. Production function analysis was applied in the study. Factor shares and least squares regression methods were used to estimate production functions. Results .from the factor shares method were not completely satisfactory in the context of the study. Accordingly the least squares method assumed most importance in the analysis. Both multi-linear and Cobb-Douglas functions were fitted to the data and the latter gave the best statistical fit. Functional analysis was also used at the regional level. The dummy variable technique and discriminant analysis identified two regions within the main sample. Productivity index comparisons were made among districts composing the two regions. The analysis with respect to the over-all sample indicated the presence of resource mis-allocation on paddy farms. Typical paddy farmers were found to be employing land and draft services efficiently when all other resources remained at geometric mean levels of use. Nevertheless, under similar-conditions of geometric mean level resource use, fertilizer and labour were not used intensively enough. In particular fertilizer was seriously under-utilized. Calculation of expansion path resource combinations and various productivity estimates confirmed these results. Therefore, the analysis showed that paddy output can be increased by more intensive application of fertilizer and labour. The latter calls for intensive practices such as transplanting and manual weeding. But study findings also suggested the existence of labour shortages during peak periods of paddy farming. Marginal productivities of fertilizer and labour in both low and high response designated regions, showed once again that at geometric mean levels of resource application, they were substantially higher than their prices. In the regional analysis the draft service input (including animal and tractor services) was found to be typically over-utilized in the low response region and under-utilized in the high response region. The latter points to a shortage of draft services in the high response region which can act as a constraint to increased paddy production. Expansion path resource combinations were also calculated for each region to act as guide-lines for efficient resource app1i cat ion. Analysis at the district level for ascertaining the productivity of all inputs other than land in paddy farming showed that Polonnaruwa district was twice as productive as Kurunegala district. In the same context Hambantota came close to Polonnaruwa, whereas, Kandy and Colombo closely matched Kurunegala.
Item Citations and Data