Abstract/Summary

The existence of “yield gap” highlight that the objective of food security in Pakistan could be achieved by increasing efficiency of the farming systems. The agriculture sector is dominated by small farms, which are often resource constrained and mostly not benefiting from the government’s policies. The improvement in efficiency seems to be most feasible alternative. Therefore, the technical, scale, allocative, cost and bias-corrected efficiency of the small farms in the mixed farming system of Pakistan is investigated at crop and farm level to see the possibility of production enhancement within the available resources. In this thesis, Data envelopment analysis (DEA) is preferred because none of the studies in Pakistan have used bootstrapping in estimation, to draw a reliable conclusion within the models explaining efficiency scores, while simultaneously producing standard errors and confidence intervals. For this analysis, the data were collected from two purposively selected Tehsils of the mixed farming system during the year 2012-13. The small farms were found 66, 71, 83, and 58 percent technically efficient for wheat, cotton, maize, and sugarcane crops, implying that the farms have wasted 44, 29, 17, and 42 percent of their resources at crop level for attaining the current level of output, respectively. These farms can still achieve the same output level by substantially reducing their inputs. The impact of scale was found more prominent in per acre analysis than in per farm analysis. The analysis was further extended to estimate the cost and allocative efficiency by using the reliable price information collected during the survey and the findings revealed a considerable room for the wheat, cotton, maize, and sugarcane farmers to reduce their input cost up to 46, 40, 33, and 49 percent, respectively. The mean allocative efficiency was estimated at 81, 86, .81, and .89, suggesting that the wheat, cotton, maize, and sugarcane farmers are 19, 14, 19, and 11 percent inefficient in the allocation of resources, respectively. At an aggregated level, the small farms were found 71, 96, 60, 85 percent technical, scale, cost and allocative efficient, respectively. The results generated by solving bootstrap DEA model for 2000 iterations revealed a significant difference of .17, .19, .08, and .19 points at crop and .13 at farm level between the original and biased corrected efficiency. This difference appears because bootstrap DEA model incorporates noise component in the model and accounts for the inefficiency caused by exogenous factors. This substantial difference further implies that results achieved through the application of standard DEA models can be misleading and must need to be overlooked again. The efficiency scores estimated in the first stage are correlated with the environmental variables used in the second stage, therefore, a double bootstrap truncated regression instead of Tobit and OLS is used to find the possible determinants of technical efficiency. The results of truncated regression analysis revealed that the contact with extension, participation in training, household size, practicing according to extension recommendations, tractor and tubewell ownership, soil quality were the significant determinants, positively influencing the TE at both crop and farm level. A strong inverse relationship between farm size and TE was observed at both levels of estimation. In conclusion, the provision of tractors to the small farms on subsidized rates and installation of tractor driven tubewell could be beneficial for improvement in efficiency. A positive impact of extension also stress that the participation of a large number of small farmers in various extension activities is indispensible in order to operate on efficient frontier. There is need to establish a strong linkage between farmers and extension that ensure the provision of timely and relevant information to the farmers through personal contact and training. More efforts are also required from government to ensure the involvement of different stakeholders in the innovation process.