pterCHAPTER-I INTRODUCTION
Household Income and Expenditure Survey (HIES) is one of the core activities of the BBS; it contains a wide range of socio-economic information at the household level that has strong influence in the decision making process for the government. It is widely used across the world, particularly in the low income developing countries, for assessing poverty level and the living standard of the people at large. Considering its importance, the national governments and the international agencies have been striving for the improvement of survey methodology and to establish of HIES technical standard.

This survey provides valuable data on household income, expenditure, consumption, savings, housing condition, education, employment, health and sanitation, water supply and electricity, etc. The survey data can also be used for compilation of national accounts of the household sectors, analysis of poverty situation and other information on household related characteristics. It also provides the weights for computation of Current Price Index (CPI). It becomes the main data source for preparation of the Poverty Reduction Strategy (PRS) and Five Year Plan (FYP). It is also used for monitoring the progress of poverty reduction and the Millennium Development Goals (MDGs) indicators. 1.1 Historical Background As statistical tool, the Household Expenditure Survey is practiced over hundred years. It can be traced back to 1857, when Ernst Engel first collected data on 153 Belgian family budgets from a group of homogeneous families in respect of taste and prices of commodities they used and that encouraged him to formulated a law that, the percentage of expenditure on food is on the average follows a decreasing function of income.

In 1904 another inquiry was made by the British Board of Trade on 2000 families of wage earners in urban areas in England. In 1920s and 30s such family budget surveys were conducted in several industrial areas in India to provide weights for construction of cost of living index numbers. The first family budget survey was conducted in Japan in 1925 covering 4785 households. Thus, during the early part of the 20th century, this sort of survey spreaded over many parts of the world covering various sections of population. 1

Household Expenditure Survey (HES) was first conducted in our part of the world, now comprising Bangladesh, during the mid fifties. The geographical coverage of that survey was only limited to four selected cities of the country. In an attempt to provide the national estimates, coverage of the survey, thereafter, was extended to rural areas.

After independence, Household Expenditure Survey was first carried out in 1973-74 and the result was published in two volumes. HES data collected for the years 1974-75 and 1975-76 were not published. Some selected tables of the surveys 1976-77, 1977-78 and 197879 were published in the Statistical Yearbooks of 1980, 1982 and 1983-84 respectively. Detailed reports could not be published due to delay in data processing. In HES 1981-82 provision was made to collect data on several socio-demographic characteristics to correlate consumption and expenditure pattern with different segments of population. Since 1973-74 up to 1981-82 data were collected using recall method.

A combination of both recall and diary methods were introduced during HES 1983-84. For this purpose, two types of schedules were introduced. One was called “Diary” to collect data on food and beverage consumed by the household on daily basis for one month by locally recruited person designated as “Diary Keeper”. The other was called “Schedule” to collect non-food expenditure with varying reference period by the BBS field staff at the end of the month. Almost similar methodology was followed in the subsequent surveys held during 198589, 1988-86, 1991-92 and 1995-96.

In the latest Household Income and Expenditure Survey (HIES) conducted in 2010, many of changes have been introduced in the methodology in order to improve the quality of data. In addition, 4 additional modules were added. These are; crises and coping measures, micro-credit, migration and remittance and disability. This is the third survey in which almost instant data entry was done using laptop in each sample area. In the HIES-2010 a team based approach was taken in the data collection and data entry. As many as 34 teams engaged across the country for field operation. Each team consisted of two enumerators cum data entry operators, one supervisor and 2 female facilitators. Female facilitators were recruited from the resident of the area where enumeration was taken place. Her task was to ease the work of the enumerators. Engagement of female enumerators yield very effective results as access to the household became easier for collection of data; particularly of food consumption data from the 2

housewife of the sample household. Each team was provided with a laptop to make data entry at the field level. Therefore, data collection and data entry were done almost simultaneously at the field. All the survey teams were supplied with an internet modem to send data to the headquaters. Thus, data of the previous month were sent to Dhaka Head Office through Internet within the 7th of the following month. This was a significant improvement over the other previous HIES surveys by making the best use of ICT technology for quick data transfer. Since 2000 HES, emphasis was given for collect information on income in addition to expenditure and consumption. Besides, in 2010 HIES, the questionnaire included more comprehensive coverage of different income sources of households compared to earlier rounds. Therefore, 2005 HIES can truely be termed as proper Household Income and Expenditure Survey (HIES-2005). This nomenculture is also followed in 2010. In HIES-2010, a separate module on Social Safety Nets Programmes (SSNP) was used as was done in HIES 2005. In 1995-96 survey, a separate module was also used to collect data on the community in the rural PSU. Data Entry Program had the provision of detecting validity and consistency errors in the data. In case of the 2010 survey, there was scope of data correction at field level by revisiting PSUs by field supervisors, if any case of there detection of error by the data entry program. 1.2 Objectives of the Survey

The main objectives of HIES 2010 were to: • • • • • • • • • • • Obtain detailed data on household income, expenditure and consumption. Determine poverty profile with urban and rural breakdown. Provide information about standard of living and nutritional status of the population. Provide data to determine the weights of Consumer Price Indices (CPI). Provide household level consumption data for compiling national accounts estimates. Provide detailed information on health status and educational level of the population. Determine poverty estimates by administrative divisions and detailed socio-economic characteristics of the population and household. Provide benchmark data for formulation of appropriate policy for poverty reduction, improvement in standard of living and nutritional status of the population. Provide relevant data for monitoring of the Poverty Reduction Strategy (PRS), Five year plan and the Millennium Development Goals (MDGs). Provide data on type, volume and distribution of resources under different Social Safety Nets Programmes (SSNP). Provide data on disability, migration, remittances, microcredit and disasters management.

3

1.3

Sample Design A two stage stratified random sampling technique was followed in drawing sample of

HIES 2010 under the framework of Integrated Multipurpose Sample (IMPS) design developed on the basis of the sampling frame based on the Population and Housing Census 2001. The IMPS design consisted of 1000 Primary Sampling Units (PSUs) throughout the country. There were 640 rural and 360 urban PSUs in the sample. The PSU was defined as contiguous two of more enumeration areas (EA) used in Population and Housing Census 2001. Each PSU comprised of around 200 households. In the first stage about one half, 612 is in exact out of total 1000 IMPS PSUs, were drawn. These PSUs were selected from 16 different strata. There were 6 rural, 6 urban and 4 SMA strata. In the second stage, 20 households were selected from each of the rural PSUs and also PSUs located in the municipal areas and SMAs. Thus, the HIES is a sub-set of IMPS. According to the sample design, the survey has completed in one complete year (1st February, 2010 to 31st January, 2011). This has been done in order to capture the seasonal variations in a cycle of one year in income, expenditure and consumption pattern. This entire period of one year has been divided into 18 terms. In each term a total of 34 PSUs are covered to collect data from a total of 680 sample households. In HIES-2010, 12240 households were selected where 7840 from rural area and 4400 from urban area. The number of sample PSUs, Households and Population covered in HIES-2010 next table.

4

Number of Sample PSUs, Households and Population covered in HIES-2010 by Residence Division
1

Number of Sample PSUs, Households and Population 2010 National Rural Urban
2 3 4

Sample PSUs Total Barisal Chittagong Dhaka Khulna Rajshahi (Former) -Rajshahi (New) -Rangpur Sylhet Total Barisal Chittagong Dhaka Khulna Rajshahi (Former) -Rajshahi (New) -Rangpur Sylhet Total Barisal Chittagong Dhaka Khulna Rajshahi (Former) -Rajshahi (New) -Rangpur Sylhet 612 392 49 34 110 71 177 105 90 55 143 94 79 49 64 45 43 33 Sample Households 12240 7840 980 680 2200 1420 3540 2100 1800 1100 1880 2860 980 1580 900 1280 860 660 Sample Population 55580 35894 4464 3107 10943 7201 15658 9383 7673 4666 7867 12138 4038 6607 3829 5531 4704 3670 220 15 39 72 35 49 30 19 10 4400 300 780 1440 700 980 600 380 200 19686 1357 3742 6275 3007 4271 2569 1702 1034

It may be mentioned that Rangpur Division has been created by the Government after the sample selection was finalized for HIES 2010. Originally this Division was a part of Rajshahi Division and as it existed before creation of Rangpur Division. In order to fulfill the requirement of planning and administration of this newly created Division, it has been decided to tabulate the data separately for both Rajshahi and Rangpur Divisions as exist at present and at the same time also tabulate the data for the former Rajshahi Division as per original sample design to meet the requirement of comparability with the previous HIES.

5

1.4

Training and Field Operation Enumerators, data entry operators and the supervisors were given detail training on the

survey before the field operation. There were 36 (including two reserve) enumeration teams for the survey. Each enumeration team was comprised of a supervisor, 2 enumerator cum data entry operators and two female facilitators. This team of five members was assigned to 1 PSU to collect data for a continuous period of 20 days.

During this period, for collecting information on food consumption, the households were divided into two groups each consisting of 10 households. Each enumerator collected information on food consumption of the households for 14 days by paying 7 visits. Iinformation on food consumption of previous two days were collected during each visit.

This time refresher training was organised for enumerators cum data entry operators, supervising officers and divisional co-ordinators after completion of the first round of data collection. The problems identified by the enumerators cum data entry operators during data collection and data entry, were discussed in the refreshers training and the possible solutions were given. This resulted in the improvement of the data quality in the subsequent rounds of the survey.

1.5

Survey Operation

Detail description of survey management, survey questionnaire, training, field test, staffing and organization of field work are annexed (Appendix-A). 1.6 Concepts, Definitions and Limitations Concepts and definitions used in Household Income and Expenditure Survey-2010; problems and limitations encountered in the survey have been described and annexed (Appendix-B).

6

1.7

Improvements in Household Income and Expenditure survey-2010 The following measures were taken in order to ensure quality and increase coverage of

Household Income and Expenditure Survey-2010 data. • •

Data capturing in the sample area using laptop with the help of an improved data entry software. Data entry program was designed to detect inconsistencies and errors in data at data entry stage and made way to correct data by revisiting the relevant sample households at data collection stage.

• • • • •

Refreshers training were conducted after 1st round of data collection. Correction of data through instant field visit by supervisors during data entry period. Data were transferred to HQ through Internet. Sample size has been significantly increased in HIES-2010 for roboustness of the findings as well as to enhance reliability. The aforementioned measures and adoption of technologies helped BBS to publish HIES 2010 report in much shorter time span compared to previous surveys

1.8

Supervision and Quality Control Strong supervision and quality control measures were adopted in HIES 2010. As

mentioned earlier, there were 34 teams, each team comprising of 2 enumerators-cum data entry operators and two female facilitators and one supervisor. In order to improve the quality of data collection, 34 supervisors were appointed to supervise the work of 34 teams. The Regional Statistical Officers from 23 Regional Statistical Offices and 11 Statistical Officers/Assistant Statistical Officer from the HQ were engaged as supervisors. In addition, 2 Supervisors and 4 enumerators were kept reserve for addressing emergencies Thus, the number of enumerators and supervisors were in fact 70 and 36 respectively. There were also eight Divisional coordinators to supervise and coordinate the activities of the enumerators and supervisors in their respective divisions. Thus, each team was composed of 5 members; 2 enumerators/cum data entry operators, 2 female facilitators and 1 supervisors. In HIES 2010, the tasks of supervisors were determined before commencement of the field work. The supervisors were given the authority to examine all questionnaires completed by the field staff and verify that each interview has been carried out correctly and the questionnaires properly completed. Aside from the supervision by the field supervisors, the Divisional Coordinators from the 7

headquarters frequently visited the sample areas and observed one or more interviews carried out by enumerators in every term and provide solution to any problem faced by the enumerators. Provisions were made for holding term-wise discussions of the supervisors with all of his team members in order to access the progress of the interviewer. The enumerators and female facilitator used this opportunity to inform the supervisors of any problem they faced during the period and the supervisors, in their turn, helped the enumerators in solving their problems. The control and supervision measures as discussed above vastly enhanced the quality of enumeration in HIES 2010.

1.9

Data Entry, Processing and Validation The interviewers regularly entered all the information collected during the interview

into the laptop at the end of the day. If they have noticed any inconsistency in the data, they went back to the relevant households and made required changes to remove the discrepancy. Once they has completed and checked the information, they must also “validate” the data entered through data entry programme that has the mechanism to check the information for accuracy. Thus, the data were cleaned and validated at the field level. The entered data were sent to the headquarters through Internet by the first week of the following month. Thus, data entry, cleaning and validation were completed in the HIES headquarters located in Dhaka, simultaneously along the survey work. After receiving all the data from the field, they were merged together for tabulation and thoroughly edited for any missing data or inconsistency in the data. Finally, tables were generated from the cleaned data using data analysis software like STATA, FoxPro and SPSS.

8