Description:
The IPDM questionnaire survey data is part of the IPDM/STEPSA project's spatial planning exercise, and was aimed at supplying detailed migration history data for 8 provinces (Gauteng, Limpopo, North West, Mpumalanga, KZN, Eastern Cape, Western Cape and Free State), together with respondent data on household structure and economy, transport access and needs, and housing access and needs.
The data set contains 3114 variables and 5916 cases.
Abstract:
The IPDM survey data is part of the IPDM / STEPSA project's spatial planning exercise, and was aimed at supplying detailed migration history data for 8 provinces (Gauteng, Limpopo, North West, Mpumalanga, KZN, Eastern Cape, Western Cape and Free State), together with respondent data on household structure and economy, transport access and needs, and housing access and needs.
The data has been collected using a standard stratified equal-probability sampling procedure, using sample points derived from 2001 Census data.
The main purpose is to set individual migration into the economic and spatial context in which it takes place, so as to provide policy-relevant migration information in reasonable depth, with particular reference to where housing and infrastructure delivery should best be situated for youth, women and couples with children. The key question which the survey work addressed was how best to use spatial planning of housing and infrastructure to promote better access to the labour market for socially excluded groups, with special reference to youth and women.
The survey data itself was collected in two tranches, with the first survey of 2865 cases in 2008-9 and the second survey of 3051 cases in 2010. The two component surveys are aimed at providing in-depth data that can then potentially be used in conjunction with national Census data for wider coverage. These two survey data sets have been combined into one composite migration data set, which caters for the differences in the questionnaires between the two surveys and allows for the lapse of time; however, the individual 2008 and 2010 surveys can also be accessed separately in the data set, and contain some information that the combined data set does not include due to the divergence of the detailed questionnaire content in the two surveys. The 2008 Phase 1a data set contains more detailed migration and housing information, while 2010 Phase 1b data contains proportionately more transport information and devotes less space to migration.
Major variables include standard respondent household information linked to migration history, housing particulars, and transport activity and costs. Migration data is provided in some depth with reference to spatial location and is broken down by distance zone relative to metro city centres.
Face-to-face interview
Households in predominantly poor (more than 50% below R9 600 per annum in total in 2001) enumerator areas (EAs) of the country within all the major migration corridors, with an expected predominance of Africans (more than 80% of the estimated EA population) in 2001.
The two surveys (Phases 1a and 1b) used similar stratified, clustered, random sampling frames with 2001 census EAs as primary sampling units (PSUs), which were randomly selected from the list of all EAs from within the universe.
The strata used were (a) weighted distance from the nearest large central business district (CBD), taking into account the distances from other 'competing' major CBDs, (b) age of the settlement (ensuring sufficient representation of both older and younger settlements), and (c) city type (metro, secondary city or non-city). Within every PSU a sample of six eligible households -- the ultimate sampling units -- was drawn by means of systematic sampling with a random starting point.
Within households an informed adult member (preferably the head or spouse) being present at the time of the survey provided some information for the household in general and on behalf of all its members and also provided some information for himself/herself specifically).
