4.5. Checking for Common Support, Balancing We first checked the common support region. It is the region where the propensity score has a positive density for both participants (treated) and non-participants (non-treated) units. Figure-3 gives the frequency distribution of the propensity scores calculated above by using Probit model for the participants and non-participants of the PDS programme. [Figure-3] The propensity score ranges from 0.047 to 1.000 for both the groups. It can be seen from the figure that the common support condition is satisfied. There is substantial overlap in the distribution of the propensity scores of both treated and control group and a severe common support problem does not exist between them. Given that substantial …show more content…
4.6. Estimation of treatment effect checking of Sensitivity between the treated and the control group The analysis of the treatment effect is conducted by focusing on two parameters, namely- the average treatment effect on the treated (ATT), and average treatment effect (ATE). The results are presented in table-10. [Table-10] Table-10 depicts that the ATT due to PDS participation for the household on their food security index is higher than the non-participants by 2.30 and the ATE, in this case, is 1.33 which is not statistically significant. Though the ATT of PDS rice and wheat on the calorie gains of the households show negative figures indicating their level of caloric deficiency, yet after participation in the PDS, the ATT of PDS rice on the Calorie gains has increased by 33819.76 calories which is significant at 1 percent level and the calorie gain from PDS rice for any household drawn from the population, as depicted by the ATE is 36899.11 calories. Similarly, calorie gain from the PDS wheat as given by the ATT shows an increment by 2491.44 calories, though it is not statistically significant. But the income gains out of PDS participation from rice and wheat witness mark of significantly …show more content…
The output shows that we get a significant positive treatment effect on the treated of 2.30, 33819.76, 2491.44, 407.11 and 10.03 for FSI, calorie gain from rice and wheat and income gain from rice and wheat at the household level respectively. That is, the rate of food security, measured in terms of FSI, calorie gain from rice and wheat and income gain from rice and wheat of participants are 2.30, 33819.76, 2491.44, 407.11 and 10.03 percent higher than that of matched control group members. Under the assumption of no hidden bias ( ), the Mantel-Haenszel (1959) bounds, test statistic gives a similar result, indicating a significant treatment effect (table-A.6 in