National Bureau of Statistics

TANZANIA NATIONAL PANEL SURVEY (NPS)

Progress report April 8th 2010

Summary

The Tanzania National Panel Survey (NPS) is a multi-year initiative led by the Tanzania National Bureau of Statistics (NBS), with the main objective to provide high-quality household-level data to the Tanzanian government and other stakeholders to monitor and understand poverty dynamics, track the progress of the Mkukuta poverty reduction strategy, and evaluate the impact of other major, national-level government policy initiatives. The NPS is part of a system of surveys undertaken by the NBS. It complements other survey efforts with its two unique features. First, it is an integrated survey covering a wide range of socioeconomic factors. By comparison, other prominent NBS national household surveys focus on a single topic or a narrow range of sectors, such as the DHS on health, the ILFS on labour markets, the HBS on expenditure, and the NSCA on agriculture. Second, the NPS is a panel household survey where households are revisited over time, allowing for the study of poverty and welfare transitions and the determinants of living standard changes, rather than only static statistics.

The NPS instruments include a Household and an Agricultural Questionnaire, as well as a Community Questionnaire. Information is collected on multiple levels: individual level information (such as labour, education, health), household level information (such as food consumption expenditures, asset ownership, housing amenities), and various farm-level information for plots, crops, and other activities.

The Household Questionnaire is detailed and multi-topic, and in each survey round new topics of policy relevance may be considered. In the 2008/09 wave, the Household Questionnaire included modules on basic demographics, education, health, labour, remittances, governance, violence against women, consumption, assets, social safety nets, credit, crime and justice, welfare shocks, deaths of household members and anthropometric measurements. The Agricultural Questionnaire records information on each household plot, farm inputs, crop harvest and sales information, processed agricultural products, livestock, fisheries, farm implements and machinery, and farm extension services. The Community Questionnaire collected information on physical and economic infrastructure and events in surveyed communities.

The NPS Technical Committee, chaired by NBS, includes representatives from various line ministries, government agencies and several development partners, including the Ministry of Agriculture, Food Security and Cooperatives (MAFC), Ministry of Finance, Millennium Challenge Authority Tanzania, World Bank, DFID, UNICEF, UNFPA, and JICA. The Technical Committee advises the NBS on technical issues related to survey design and implementation.

The first wave of the NPS was supported by several donors, including the World Bank, the Bill and Melinda Gates Foundation, UNICEF, UNFPA, and the Royal Danish Embassy, as well as the Government of Tanzania through the pooled Mkukuta funding. Each wave of the panel survey is fielded over a period of 12 months. Financial support for the next 3 waves of the NPS has been secured from the EU, the World Bank, and the Government of Tanzania.

Year 1 Activities

Sample: The 2008/2009 NPS sample size was 3,280 households in 410 Enumeration Areas (2,064 rural household and 1,216 urban households). It was constructed based on the National Master Sample frame, and includes a partial subsample of households interviewed during the 2006/2007 Household Budget Survey. The NPS sample is representative at the urban/rural level and at the level of the major agro climatic zones of Tanzania. The sample was split into two cycles, each nationally representative, to reduce account for seasonality and to facilitate the monitoring of the fieldwork.

Preparation of questionnaires and data entry system: Questionnaires were developed in collaboration with line ministries and donor partners. The NBS solicited feedback from various stakeholders in terms of survey content and design. The questionnaires were piloted in the Morogoro region in June 2008. After piloting, the questionnaires were further revised and finalized. Questionnaire manuals were developed to outline detailed instructions for field staff for use during training and as a reference over the course of the field work. Data from the NPS were entered using a Computer Assisted Field Entry (CAFE) system developed in CSPro by NBS with support from the World Bank.

Field staff and training: The NBS recruited 7 supervisors, 21 enumerators, and 7 data entry technicians to comprise 7 mobile field teams. Training for the data collection staff and alternates was conducted in Korogwe in September 2008. A follow-up training was held in Tanga in April 2009, after which the second cycle of fieldwork began.

Data collection: The first cycle (months 1 to 6) of field work began immediately following the training in October 2009 and continued until March 2009. During this cycle, data was collected from 1,638 households in 207 enumeration areas. Data entry was done concurrently with data collection by a dedicated data entry technician that traveled with each mobile field team. During the second cycle (April – October 2009), information was collected from another 1,645 households in 203 enumeration areas. Data collection was completed in October 2009.

Post-data collection data cleaning and documentation: Following the completion of fieldwork and data checks during data entry, an extensive review of data files was conducted, including verification against the paper questionnaire of outliers and missing values.

Data dissemination and access: An interim NPS report as been prepared and is being finalized. The content of the report included a description of the survey, analysis on a wide-range of socioeconomic indicators, and poverty projections from the Household Budget Survey (2006/07) to the NPS households. A formal review of the report by independent academics is under way. Upon receipt of these reviews, the NPS report will be finalized, using the final (cleaned) version of the data. The final NPS report is expected to be released in April 2010. The complete dataset and full documentation is being prepared. Complete release of 2008/2009 NPS (data files with documentation) will follow soon after the release of the NPS report.

Year 2 Activities

Initial preparations for the second wave of the NPS have begun, and a detailed work plan is being finalized for endorsement by the Technical Committee.

Sample: The sample for the NPS Year 2 is the 3,280 households interviewed in the 2008/2009 NPS. Households that have moved intact to a new location within Tanzania will be interviewed at their new location. Additionally, a subsample of individuals who left households that were part of the 2008/2009 NPS will also be interviewed, along with the members of their current (new) household.

Preparation of questionnaires and data entry system: Preparation of the revised questionnaire and training materials has started. Requests for additional questions / sections are being collected by the NBS and will be presented to the Technical Committee for consideration.  Following pilot testing and the final approval of the revised questionnaire from the Technical Committee, a CAFE data entry system will be developed in CSPro by the NBS with support from the World Bank. 

Field staff and training: Field work for the second year of the NPS is scheduled to begin in September 2010 and continue for 12 months.  The composition of the field teams and fieldwork calendar will remain largely unchanged from the year 1 structures.  Training for year 2 is planned to begin in August 2010. 

Data collection: Fieldwork for year 2 will again be conducted in two nationally representative cycles, with all households being revisited during the same month as their year 1 interview.  Though tracking protocols are still being finalized, an additional number of members who have left their original households will be interviewed in their new households.  Data entry will again be done concurrently in the field by a dedicated data entry technician. 

Post-data collection data cleaning and documentation: In the second year of the NPS, the data will undergo both an extensive system of checks in the field and be double entered by a headquarters-based data entry team.  Comparisons of the data entered in the field with that entered at headquarters will greatly reduce the time between the completion of field work and the finalizing of the dataset.

Data dissemination and access: While a detailed plan for year 2 data has not yet been developed, many additional areas of analysis will be available with two completed rounds of this longitudinal survey.  Following the completion of data cleaning and the year 2 NPS report, the 2010/2011 NPS data files and accompanying documentation will be made publicly available following the procedures and conditions stated under section 20 of the Statistics Act No. 1, 2002.