Survey ID Number
PAK_2009_BISP_v01_M
Title
Rapid Assessment Survey of the Benazir Income Support Program 2009
Sampling Procedure
The sampling Survey for the Rapid Assessment of Parliamentarian selected beneficiaries was conducted in the two stages: a test phase stages and the main survey.
1.Test Phase.A test of the planned survey was conducted by IDS in May 2009. The test was based on a random sample of 101 beneficiaries in two districts Nowshera (NWFP) and Rajanpur (Punjab). The 101 beneficiaries came from 101 households composed of 613 household members. The results and lessons from the test-phase were used to refine the final survey questionnaires.
2.Main Survey. The main survey was conducted in 15 randomly selected districts of Pakistan, during the months of August and September before Ramadan.
The sample beneficiaries were selected from the list of the Parliamentarian identified beneficiaries provided by the National Database and Registration Authority's (NADRA). The sample was drawn from the NADRA computer records in April, 2009. There were 1,660,584 existing beneficiaries in that record, composed of 46.23 percent from Punjab, 26.04 from Sindh, 18.35 percent from NWFP, 6.08 percent from Baluchistan, 2.38 percent from FATA, 0.73 percent from Islamabad and 0.14 percent from Northern areas.
The study used the probability to Proportion to Size method to obtain a sample as representative as possible with the sample size of 2,500 households. For this purpose a randomly selected sample of 5680 beneficiaries' names, addresses, NIC numbers and names of spouses/fathers was provided by NADRA for the 15 selected Tehsils in the 13 districts of the four Provinces, and Islamabad in the Federal Capital and Muzafarabad in the Azad Jammu and Kashmir.
The total number of Districts/Tehsils was purposively decided on the basis of an eventually representative sample size to capture the large diversity in levels of development across districts and the resources available. The 13 districts from the four Provinces were then allocated to roughly represent the relative share of the beneficiaries from each Province in the NADRA records on that day. As such 7 districts on Punjab, 3 in Sindh, 2 in NWFP and 1 in Baluchistan were selected.
The literature on Poverty in Pakistan divides provinces into distinct poverty or deprivation bands. For example, Punjab, is looked upon as three regional disparity bands i.e. Northern Punjab (best off), central Punjab (moderately well off) and southern Punjab (worst off). The seven districts of Punjab represent this classification with the number of districts selected in each poverty band being selected on the basis of the share of each region in the population of Punjab. As such, Attack was chosen from the North Punjab, Rajanpur was chosen from the South, and the remaining was chosen from the Center to represent the region as moving from North to South i.e from Sargodha to Faisalabad to Vehari to Multan to Bahawalpur. The same rationale applies to other provinces. For example, in Sindh province, the selected districts ranged from Karachi (the most developed) to Larkana (deprived but politically privileged) to Sanghar (most deprived). In NWFP province, Nowshera is amongst the least deprived in NWFP and Karak is one of the most deprived. Similarly, Muzafarabad was chosen to represent AJK and Islamabad to represent the Federal Area.
In Baluchistan, the unrest and law and order situation made survey almost impossible except in Quetta where a small sample was selected.
The 13 districts in the four provinces for which this list of beneficiaries was selected were chosen to represent 1) the total number of districts in each province, 2) the regional level of deprivation in each province (based on existing literature, districts in each province in Pakistan are grouped by levels of poverty - for example Punjab has three groups namely North (low poverty), central (higher poverty) and Southern (highest poverty) and 3) the share of beneficiaries in each district. One district was selected in Azad Jammu and Kashmir and Islamabad Capital territory was selected as a stand alone.The selected districts represent 70 Members of the National Assembly and 143 members of the Provincial Assembly. There are 2 designated Senate seats for Islamabad while the other Senate seats are allocated on no regional criteria. Having selected the districts, one tehsil in each district was drawn randomly. However, the list of beneficiaries provided by NADRA did not include any coding of addresses by Union Council. The Unions Council names were therefore manually coded into the information on the beneficiaries. From this list sample UCs were designated in each tehsil in such a way that at least a minimum of 10 beneficiaries would be covered in a particular sample UC.
Sample UCs were selected using Probability proportionate to Size (PPS) method of selection. Beneficiaries in each UC will be treated as measure of size. Due to nature of the Survey and to avoid unnecessary field problems as already stated UCs with 10+ beneficiaries were selected in the sample. However, in case if in any Tehsil where the number of UCs/Beneficiaries is small, the UCs with less than 10 beneficiaries were also selected.
Sample beneficiaries to be covered from each sample UC in a Tehsil will be in proportion to the size of the sample UCs.
Beneficiaries in a sample UC were selected by Simple Random Sample method of selection. In case if a sample beneficiary is not traceable then subsequent beneficiary in the list of beneficiaries of that UC will be selected as replacement.
The issue of replacement HHs in case a HH could not be traced due to incorrect address required clarification. After exhaustive deliberations with the World Bank team, it was agreed that the sample size would be increased to 2595. Out of the sample of 2,595 households (HHs), a total of 2,540 HHs were initially successfully enumerated before the Eid break. The enumeration teams could not reach 55 HHs. These 55 HHs were later verified / enumerated after the break due to Ramazan and Eid holidays. Post Eid all 2595 HHs were successfully enumerated. Out of the 2,595 HHs enumerated, the enumeration teams, in consultation with their field supervisors and the designated IDS supervisors, had to drop 347 HHs. The reasons for dropping these HHS are discussed in detail subsequently. In addition to these 347 dropped households (for whom both female and male questionnaires were not filled), there are 78 households for whom only female questionnaires were filled as either there were no male member in the HH or male members, even after three visits to the HH (refer to table 7 of the BISP Rapid Assessment Report for sample summary).