Pakistan Rural Household Panel Survey (PRHPS) 2014

Data Paper: A User’s Guide to Data from Round 3 of the Rural Household Panel Survey

July, 2016

International Food Policy Research Institute (IFPRI) Innovative Development Strategies (IDS)

Preferred citation: IFPRI/IDS (International Food Policy Research Institute/Innovative Development Strategies). 2016. Pakistan Rural Household Panel Survey (PRHPS) 2014, Round 3. Washington, D.C./, Pakistan: IFPRI/IDS.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE The International Food Policy Research Institute (IFPRI), established in 1975, provides evidence-based policy solutions to sustainably end hunger and malnutrition and reduce poverty. The Institute conducts research, communicates results, optimizes partnerships, and builds capacity to ensure sustainable food production, promote healthy food systems, improve markets and trade, transform agriculture, build resilience, and strengthen institutions and governance. Gender is considered in all of the Institute’s work. IFPRI collaborates with partners around the world, including development implementers, public institutions, the private sector, and farmers’ organizations, to ensure that local, national, regional, and global food policies are based on evidence. IFPRI is a member of the CGIAR Consortium.

INNOVATIVE DEVELOPMENT STRATEGIES Established in 2002, Innovative Development Strategies (IDS) is a private sector consulting organization which carries out diagnostic and evaluative exercises. Its research and outreach is devoted to the identification and analysis of economic, social, cultural, political, business, and institutional problems connected most particularly to economic development and poverty alleviation. IDS services reflect the spectrum of support in all areas of economic development, poverty reduction, social sector to public sector reforms and governance, and from agriculture and rural development to industrial development and trade. In all these areas, IDS offers support in conducting research, sector capacity building, and development of Monitoring and Evaluation capability at all levels. IDS has vast experience in developing and delivering large and complex projects with a proven track record of success. IDS supports donors, NGOs, and government bodies working towards economic development and poverty reduction in Pakistan.

Copyright 2014 International Food Policy Research Institute. All rights reserved. Sections of this material may be reproduced for personal and not-for-profit use without the express written permission of but with acknowledgment to IFPRI. To reproduce the material contained herein for profit or commercial use requires express written permission. To obtain permission, contact the Communications Division at [email protected].

1. Introduction

Responding to a request from the , the Pakistan Strategy Support Program (PSSP) was launched in July 2011. This program is a flexible country-led and country-wide policy analysis and capacity strengthening program which provides analytical support on a range of economic policies affecting agricultural growth and food security in the country. The core purpose of the program is to contribute to pro-poor economic growth and enhanced food security through strengthened national capacity for designing and implementing evidence-based policy reforms.

PSSP’s four primary research priorities are as follows: a. Agricultural production and productivity b. Water management and irrigation c. Macroeconomics, markets, and trade d. Poverty reduction (Income dynamics) and job creation (social safety nets)

The Pakistan Rural Household Panel Survey (PRHPS) 2014is the third round of the PRHPS; Rounds 1 and 2 were conducted in 2012 and 2013 respectively. This survey aims to provide a quantitative basis to identify and address urgent economic development and policy priorities in Pakistan. Many modules and questions in Round 3 are consistent with the prior rounds. PRHPS Round 3 was able to collect complete data from 1,876 households in the rural areas of three provinces namely: (i) Punjab; (ii) ; and (iii) (KPK). The sample is representative of the rural areas of Punjab and Sindh provinces and of the rural areas in 11 of the districts in KPK province. The survey collected information on many topics including sources of income, nature of employment, consumption patterns, time use, assets and savings, loans and credit, education, migration, women’s decision making, economic shocks, transfers in and out, health and nutrition, and participation in social safety nets. Four survey instruments were developed to collect this information. These included two household questionnaires (designed to collect individual- and household-level information from a main male and a main female respondent who were interviewed separately), a community questionnaire, and a price questionnaire.

2. Sampling Methodology

The third round of the Pakistan Rural Household Panel Survey (PRHPS) was conducted in 76 villages in Punjab, Sindh, and Khyber-Pakhtunkhwa (KPK) in April-May 2014. The sampling frame was based on the . Household and population data were available for 1998 at the national, provincial, district, tehsil, union council, and mouza1 (revenue village) level. The population and number of households were projected to 2012 for each of these levels using district and tehsil-level population growth rates.

All enumeration blocks classified as urban in the 1998 Census were removed from the sampling frame, as this is a rural household survey. All enumeration blocks with a projected population greater than 25,000 in 2012 were also considered urban and removed from the sampling frame. The sample excludes rural areas in Balochistan and the Federally Administered Tribal Areas, because they were considered unsafe for

1 A mouza is an administrative unit based on land revenue records and may correspond to a specific land with one or more settlements.

enumeration. Additionally, 13 districts of KPK were excluded from the sampling frame due to safety concerns. The remaining 11 districts of KPK were part of the sampling frame.

A multistage, stratified sampling technique was used to select the sample. We first used the proportion of rural households in each province to determine the number of districts that would be chosen from that province. A total of 19 districts were selected from within the three provinces; 12 from Punjab, 5 from Sindh, and 2 from KPK. We then used Probability Proportionate to Size (PPS) to select districts from each province. PPS ensures that, within a province, districts with more rural households have a greater probability of being selected in the sample.

For each province, the total number of households was calculated, and then the districts were arranged in a random order. The sampling interval was calculated by dividing the total number of households in the province by the number of districts that were chosen from the province. For example, for Punjab the sampling interval is 10,143,181/ 12 = 845,265.

The next step was to generate a random start ‘r’, which was a random number between zero and the sampling interval. The cumulative number of households was calculated in each province across districts. The district that contained ‘r’ within the range of its cumulative number of households was selected. The district that contained the sampling interval plus ‘r’ within the range of its cumulative number of households was selected as the second district. The nth district selected contained the sampling interval plus n-1 multiplied by ‘r’. The process was repeated until the required number of districts were chosen from within the province.

Within each district, 4 mouzas were chosen using an equal probability, systematic selection. In other words, mouzas with smaller populations had the same probability of being selected as highly populated ones. Using PPS at this stage would have meant that each household had the same probability of being in the sample. However, that would bias our sample towards more populous mouzas and possibly ignore the smallest mouzas. Since our survey aims to understand the dynamics of different kinds of villages in rural Pakistan, it is imperative to include mouzas of different sizes.

The mouzas were arranged in a random order and were assigned a serial number. The sampling interval was calculated by dividing the total mouzas in the district by 4 (the number of mouzas chosen from each district). A random start ‘r’ was generated, which was a random number between zero and the sampling interval. The mouza with the serial number ‘r’ was selected as the first mouza. The mouza with serial number sampling interval plus ‘r’ was selected as the second mouza. The third mouza had serial number sampling interval plus two ‘r’, while the fourth mouza had serial number sampling interval plus three ‘r’.

For each mouza in the sample, the enumeration team conducted reconnaissance and created a map. All mouzas were divided into enumeration blocks. Each enumeration block was the same size containing 200 or fewer households. If there were fewer than 200 households in a mouza, the entire mouza was considered a single enumeration block. In each mouza, one enumeration block was randomly selected for enumeration.

A complete household listing was conducted of the enumeration block that was selected. Twenty-eight households were then randomly selected from this list using an equal probability, systematic selection. The listing form gave every household a serial number. The sampling interval was calculated by dividing the total number of households in the enumeration block by 28 (the number of households chosen in each mouza). A random start ‘r’ was generated, which was a random number between zero and the sampling interval. The household with serial number ‘r’ was the first household selected for enumeration. The household with the serial number sampling interval plus ‘r’ was the next household selected for enumeration. The nth household selected for enumeration had a serial number of the sampling interval plus

n-1 multiplied by ‘r’. The process was continued until 28 households were chosen in each mouza. There was no replacement for households that refused to participate in the survey.

Table 1: Pakistan Rural Household Panel Survey Sample Province Number of Number of Mouza Number of Total Number of Districts per District Households per Households in Province Mouza Punjab 12 4 28 1309 Sindh 5 4 28 557 KPK 2 4 28 224 Total 19 76 2090 2090

The twelve districts surveyed in Punjab were Kasur, Bhakkar, Khanewal, Attock, Vehari, Jhang, Dera Ghazi Khan, Bahawalnagar, Rahim Yaar Khan, , , and Sargodha. The five districts surveyed in Sindh were Thatta, Dadu, Sanghar, Jaccobabad, and Hyderabad, while the two districts surveyed in KPK were Mansehra and Nowshera.

The 3rd round of PRHPS covers 1,876 households out of the 2,090 households in Round 1. Information on these households was collected from males or females or both. Information on 214 households could not be collected in Round 3. Of these, 71 households in 3 mouzas2 could not be surveyed due to restrictions by local district administrations,3 while the remaining 143 households could not be surveyed due to refusal to respond, household migration, or non-availability of respondent.

2 Two mouzas (Chit Sarkani, Jan Pur) of Dera Ghazi Khan and one mouza (Andheji-Kasi) of Dadu District. 3 This led to an adjustment in the calculation of household weights in the relevant districts; households in mouzas that could be covered (within the same district) were assigned a larger weight.

3. Overview of Analysis

The following is an overview of the analysis completed using the PRHPS Round 3 data. The household survey collected information on many topics, and below are some of the key takeaways.

Pakistan’s population is relatively young. The data indicates that the rural population of Pakistan is relatively young. The average age of the rural population is 25. Over one quarter of the population is 9 or younger, and 60 percent is under the age of 25. The current demographics, if combined with a sustained decline in fertility rates, present a substantial window of opportunity for reaping demographic dividends in the coming decades. However, this will require investments in quality schooling and human capital. Job creation will also be necessary to ensure that the young rural population has economic opportunities.

The majority of Pakistan’s population is illiterate. Fifty percent of population, 10 years and older, can read and write, and the literacy rate4 is 64 percent for males and 35 percent for females. However, there is considerable geographic variation in literacy rates. While the literacy rate in KPK is about 61 percent, it is 53 percent in Punjab and only 35 percent in Sindh. The data also shows that literacy is considerably higher in younger people. Only 49 percent of individuals have ever attended school, and only about 29 percent of individuals had completed primary education. Just close to 10 percent of individuals completed matric (10th grade). The proportion of individuals who have completed more than 10 years of education, including professional education, is only 4 percent.

School enrollment rates significantly differ across gender and provinces. The data shows that the enrollment rate among school going children aged 5 to 14 years is 68 percent. However, this aggregate number masks significant variation across gender and province. While school enrollment is 74 percent for boys, it is only 61 percent for girls. The aggregate also varies from 93 percent in KPK to 77 percent in Punjab and only 42 percent in Sindh. We also observe that girls have a higher dropout rate than boys. While poverty is cited as the main reason for non-enrollment, many respondents expressed security concerns about sending their daughters to school. Pakistan’s young rural population, which makes up a significant proportion of the total population, presents opportunities as well as challenges for policymakers. In the coming years, it will be crucial for policymakers to invest in education and provide people with the necessary skills and appropriate employment opportunities.

Over 45 percent of the rural households are engaged in agriculture. The average amount of land operated is 5.7 acres, while the average cultivated land is 4.5 acres. Nearly 58 percent of farm households are owner- operators, 14 percent are owner-cum-tenants, and 28 percent are tenants. Most of the land in Punjab and KPK is operated by owners and in Sindh by tenants. Most of the farmers are growing more than one crop during the two agricultural seasons. Almost one-third of the area is devoted to wheat irrespective of farm size. Among other crops, another one-third of the area is allocated to pulses and cotton by farmers with farms larger than 12.5 acres.

The ownership of livestock is not limited to farm households. Nearly 88 percent of farm households, and 71 percent of nonfarm households, own livestock. On average, nonfarm households have less animals than the farm households. The number of small animals (goats, sheep) is higher than the large animals (cows, buffaloes, bullock, cattle) for nonfarm households compared to farm households.

Rural households in Pakistan earn their livelihood from a wide range of activities. Nearly half of the labour force is working as an unpaid family worker. The labour force of Punjab and Sindh is dominated by agricultural wage workers (26 percent and 21 percent), while about 11 percent of Sindh’s labour force is involved in non-agricultural business. This proportion is lower for Punjab (8 percent) and KPK (7 percent).

4 Literacy rate is defined here using the standard definition, i.e. a person can read and write

The data shows few households have only one source of income (18%). A majority, 57%, of farm households derive their livelihood, not only from farming and livestock, but also through nonfarm economic activities.

The majority of the rural households use firewood as a primary source of energy for cooking and heating. Electricity is consumed by most households (90 percent), even though electricity outages remain high at an average of 12.5 hours a day in the summer and 9 hours per day in the winter. Over 92 percent of households with electricity use it for lighting purposes. Firewood is the main source for cooking and heating, which is used by almost two-thirds of households. Animal and plant residue is another source for cooking and heating, and 18 percent of households use this source for cooking, while 22 percent use residues for heating. A smaller percentage (15 percent) of households uses natural gas for cooking and heating. Most households (32 percent) use emergency lights and (10 percent) candles as an alternative source during an electricity outage. Households who use gas, face gas outages; on average, 2.30 hours per day in winter and 1.64 hours per day in summer. In the case of gas outages, over one-third of households use firewood, and 38 percent use other sources (e.g., petrol, diesel, or coal) as an alternative source.

A considerable proportion of households do not have adequate provision of piped water facilities. Access to safe drinking water is crucial for a healthy life. The data shows that boreholes or tube-wells are the most common source of drinking water across the three provinces. Piped water is only available to about 19 percent households. However, when asked how much households would be willing to pay for safe drinking water, almost 62 percent stated that they would not be willing to pay anything, mostly because they were satisfied with the current water arrangements (71 percent). This, coupled with the fact that 41 percent of households are not aware of water borne diseases, provides a worrisome picture.

Households have extremely low access to sanitation facilities. About 28 percent of households do not have access to any kind of toilet, while only 61 percent of households use a toilet facility with plumbing. The remaining 12 percent use dry pit latrines. Province wise, we find flush toilets to be more common in Punjab and KPK, and dry pit latrines more common in Sindh. We also find that 33 percent of all houses do not have a drainage system, while only 13 percent of houses have a covered drainage system in place.

An increase in food prices is the most common negative shock for rural households. Seventy-one percent of households reported experiencing some sort of an economic shock (such as an unusual increase in food and/or other prices, a loss of or reduction in income or remittance, repayment of debt, fall in the price of business products (including agricultural products), and loss of assistance from government, etc.) in the past year and, as a result, also suffered from monetary loss. The highest monetary loss of about 58 percent occurred due to economic shocks, followed by natural/agriculture related calamities at 25 percent.

Households used various coping strategies to deal with negative shocks, such as reducing consumption, substituting to cheaper goods, and working more. However, 22 percent of households who faced shocks during the last year did not have any coping strategies to deal with one or more negative shocks they experienced. We also find that reduction in food consumption and/or substitution of cheaper food is the most common strategy to deal with shocks due to price hikes. Migration was found to be the most common coping strategy in case of natural disasters; about 61 percent of the households who were affected by floods migrated to other locations.

Most loans were taken from friends and relatives. Rural household borrow from diverse sources, such as commercial banks, public sector financial institutions, as well as from informal sources, such as family members and friends. Over 64 percent of the rural households who took loans in the last year borrowed from friends and relatives, while only about 16 percent borrowed from financial institutions. Furthermore, we find that formal credit is even less common in KPK and Sindh as compared to Punjab. Ease of access to credit was stated as the most common reason for borrowing from friends and relatives, followed by the

no collateral requirement. The most common reason for obtaining loans was to meet consumption requirements (25 percent), followed by to cover medical expenses (22 percent). For the households who paid an interest rate on loans, an average rate of 19 percent was paid to banks, a rate of 22 percent was paid to informal lenders, such as shopkeepers and aarthi, 13 percent was paid to relatives and friends, and the highest rate was charged by money lenders at 24 percent.

4. Contributors and Acknowledgments

The survey was designed and supervised by the International Food Policy Research Institute (IFPRI) and was administered by the collaborative partner in Pakistan, Innovative Development Strategies (IDS), Islamabad, Pakistan. The funding for the survey came from the U.S. Agency for International Development (USAID). IDS served as the data collector and handled all of the survey logistics, from enumerator training to the processing of the completed questionnaires. The units of analysis varied by module and included Individual, Household, and Community.

Gratitude is owed to the 150 enumerators who worked diligently in extremely difficult circumstances. Despite several security issues, they put out exemplary efforts to collect the data. Additionally, the 1,876 households, 73 key informants, and 592 shopkeepers at district, union council, and mouza level who participated in this survey and provided their valuable time and useful information also deserve commendation.

This work would not have been possible without the guidance and support provided by Dr. Sohail Jehangir Malik, Chairman of IDS, Dr. Paul Dorosh, Director of the Development Strategy and Governance Division at IFPRI, and Dr. Stephen Davies, Program Leader of Pakistan Strategy Support Program and Senior Research Fellow at IFPRI. Their invaluable input and advice at each stage of this process, from survey design to the production of these discussion papers, is gratefully acknowledged.

Additionally, the following individuals contributed to the production of the dataset: Dr. Katrina Kosec, Dr. Valerie Mueller, Fatima Zaidi, Dr. David Spielman, Dr. Nuzhat Ahmad, Brian Holtemeyer, Huma Khan, and Edward Whitney of IFPRI and Nishat Malik, Arshad Khurshid, Mubashir Ijaz, Munazza Akhtar, Uzooba Hureem, Ali Asghar Jilani, Abdual Hameed Lehgari, Adeel Khalid, Muhammad Hamza, Umar Afzal, Muhammad Imran, Munib ur Rehman, Zahid Masood, Amjad Iqbal Nasir Hameed Khan, Kamran Hakim, and Munazza Saboohi of IDS. The support provided by the PSSP team members, Dr. Hina Nazli, Asjad Tariq, Hassan Shafiq, Saqib Shahzad, Amina Mehmood, Asma Shahzad, Sara Rafi, Sheheryar Rashid, Wajiha Saeed, Saad Moeen, Hira Channa, Omer Majeed, Amna Ejaz, Saira Malik, Faryal Ahmed, and Tahir Ahmad—is gratefully acknowledged. Further gratitude is owed to Col. Imran Afzal Malik and

his team, Haji Afsar Khan, Afzaal Ahmed, and Asjad Iqbal, for providing logistic and administrative support.

Finally, the United States Agency for International Development (USAID) is again recognized for their generous funding, without which this survey would not have been possible.

1. Disclaimer

The International Food Policy Research Institute (IFPRI) requests that users of the data acknowledge the source of the Pakistan Rural Household Panel Survey 2014 dataset in all publications, conference papers, and manuscripts, as described under preferred citation.

IFPRI adheres to the principle of unrestricted public access to its own final research outputs and will make such outputs freely available. The Institute encourages the use of the Pakistan Rural Household Panel Survey 2014 dataset; for detailed information on its use, please refer to IFPRI’s Intellectual Property Policy. The data files in this dataset are unit record or ‘raw’ data files. Information that would allow survey respondents to be identified has been deleted from the files, but all other information remains. IFPRI’s decision not to alter the contents of the data files means that the user of these files will need to take care in handling missing observations, outliers, and violations of logical consistency.

The data are provided ‘as is’ and in no event shall IFPRI be liable for any damages resulting from use of the data. While great effort was taken to obtain high-quality data, the accuracy or reliability of the data is not guaranteed or warranted in any way.

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE www.ifpri.org

IFPRI HEADQUARTERS

2033 K Street, NW

Washington, DC 20006-1002 USA

Tel.: +1-202-862-5600

Fax: +1-202-467-4439

Email: [email protected]