| Project Detail |
The core objective of this work is to survey households in 2016/17 for the Fourth Integrated Household Survey (IHS4) and in 2019/20 for the Fifth Integrated Household Survey (IHS5). Both the IHS4 and the IHS5 samples will be comprised of a cross-sectional sample and a panel sample. The cross-sectional sample will be composed of 12,480 households in each round and will be the source of official poverty and all other statistics for which the IHS is designated by the NSO as the primary source. The cross-sectional interviews will be spread over a 12-month period, in line with the historical practice, in order to take into account seasonality in food and non-food consumption. The panel households will come from the sample that were initially interviewed in 2010 as part of IHS3 and were re-surveyed in 2013 during the Integrated Household Panel Survey (IHPS). The IHPS attempted to track all tracking-eligible individuals associated with 3,246 households that were previously surveyed by the IHS3. At the end, by tracking split-off individuals and bringing into the sample the new households that they had formed between the IHS3 and the IHPS, the IHPS sample grew to 4,000 households. The panel components for the IHS4 and the IHS5 will survey a sub-sample of the original 3,246 IHS3 households selected for the IHPS as well as the descendants of this sub-sample that were interviewed by the IHPS in 2013. More specifically, the original households and their descendants tied to 102 out of 204 IHPS enumeration areas (EAs) will be targeted starting with the IHS4, translating into an initial panel sample of approximately 2,000 households. This sample can still expand in IHS5 by tracking split-off individuals and bringing into the sample the households that they will have formed between the IHS4 and IHS5. The IHS4 and IHS5 panel components will build on the impressive tracking success achieved during the IHPS, and will work towards building a longterm panel of households and individuals. The panel data will (i) ultimately span a period of 9 years between the IHS3 and the IHS5, (ii) include data from 4 points in time and (iii) be instrumental in understanding poverty and nutrition dynamics, livelihood transitions, structural changes in the rural economy, internal migration patterns, household formation and dissolution dynamics, idiosyncratic and covariate shocks, and inter-linkages among these changes. The panel components of the IHS4 and the IHS5 will be smaller in size than the past panel rounds and the reduced sample size is in line with the approach to building long-term panels in other countries supported by the LSMS-ISA (including Tanzania and Uganda) and provides the minimum sample size for analysis at the national level, and urban/rural levels. Given the nature of panel surveys, it is expected that the attrition rate will increase (due to increased opportunities for household dissolution, household formation and individual migration) as the panel ages. With the reduced sample size greater attention can be also given towards improving quality control over the individual tracking process and limiting attrition. The continued panel data platform will also help address impact evaluation and targeting assessment needs of large-scale development programs in the country, such as the Farm Input Subsidy Program. The IHS4 and the IHS5 will be implemented using a computer-assisted personal interviewing (CAPI platform designed using the World Bank Survey Solutions CAPI software. Methodological survey experiments may be built into the IHS4 and the IHS5 to validate key survey methods and measures to improve the quality of survey data. The methodological experimentation that may be conducted under the IHS4 and IHS5 umbrella will be in line with the priorities set forth by the Research Component of the Global Strategy to Improve Agricultural and Rural Statistics, and will be vetted through the IHS4/5 Technical Working Group. |
