Measuring Living Standards within Cities, Durban 2015
Living Standards Measurement Study [hh/lsms]
The World Bank, in collaboration with Oxford Policy Management (OPM), developed a living standards measurement survey (LSMS) at the household level to investigate the relationship between household living standards and location risk in Durban, South Africa. This project is a sister project of those measuring living standards in cities conducted by OPM in Dar-es-Salaam, Tanzania (2014-2015) and in Antananarivo, Madagascar (2016).
The Measuring Living Standards in Cities (MLSC) survey is a new instrument designed to enhance understanding of cities in Africa and support evidence based policy design. The instrument was developed under the World Bank's Spatial Development of African Cities Program, and was piloted in Dar es Salaam (Tanzania) and Durban (South Africa) over the course of 2014/15.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
The 2015 Measuring Living Standards within Cities (Durban) study covered the following topics:
• Hosehold roster
• Household characteristics
• Ownership of property
• Residential history
• Time use and commuting
• Consumption expenditure
The survey covered households in Durban, South Africa.
Producers and sponsors
Oxford Policy Management
Collaborated in the implementation of the survey
Funded the survey
The total nominal sample of 2400 households in Durban was, selected in four stages rather than two. These were: (i) selection of 200 EAs with probability proportional to size; (ii) large EAs were segmented into area units of roughly the same size (using GIS data), and one segments was selected randomly with equal probability; (iii) following listing of buildings, 15 were selected using systematic equal probability sampling; (iv) households in the 15 selected buildings were listed so that 12 households could then be selected per EA by systematic equal probability sampling. This approach reduced the need to enter as many buildings as would otherwise have been necessary, without reducing the representativeness of the sample.
For further details on sampling strategy, see Survey Methodology section of World Measuring Living Standards within Cities report.
The refusal and non-completion rate was 41 percent.
Sample weights were designed to deliver unbiased estimates from the sample. The 'raw sampling weight' is a raising factor applied to each household that is equal to the inverse of its selection probability. Any given household's selection probability is the product of the probability of selection at each stage of the sampling. The weights reflect the probability of selecting the EA in the first stage of the sampling, and the conditional probability of selecting the household in the second stage. The approach in Durban is the same, but the weights account for the additional steps after selecting the EA of: selecting that segment of the EA, if it was subdivided; selecting the building within the EA; and selecting the household within the building.
The raw sampling weights are then further adjusted to account for non-response rates. To account for non-response rates, the number of 'usable households' in each EA is calculated. Non-response rates were so high that EA groupings had to be made before a weighting could be calculated. The weights were further adjusted through benchmarking to allow for post-stratification by ethnicity, considering high levels of non-response among white households.
For further details, please see Annex 1 of World Measuring Living Standards within Cities report.
Dates of Data Collection
Data Collection Mode
Computer Assisted Personal Interview [capi]
Data Collection Notes
The Durban data collection team initially consisted of 8 enumerators and 1 supervisor, but was later reduced to 4 enumerators and 1 supervisors. Although enumerators initially worked individually - to avoid attracting unnecessary attention - the approach was modified in favor of groups of two or more following an armed attack on the team.
The Durban survey fieldwork was done between February and July 2015. The extended duration of the fieldwork in Durban was largely the result of: (i) constraints in finding a greater number of interviewers with the aptitude to implement the relatively complex questionnaire; and (ii) security concerns in the field, which caused delays in accessing specific EAs.
Cases were assigned to interviewers using Survey Solutions. Interviewers were provided with both an electronic and hardcopy map, as well as a printed completion form, and could contact the listing manager through email, WhatsApp, or google hangouts if they were unable to find the assigned house.
Completing the survey often required repeat visits. This is because the survey required input from up to three separate respondents: the main respondent, who could be any present household member, and answered questions on household composition, basic information on members, assets, remittances, grants, housing, properties and consumption; the household head, who answered questions on residential history, satisfaction, employment, time use and commuting; and a random respondent, who was randomly selected from household members over the age of 12 (not including the head), who responded questions on satisfaction, employment, time use and commuting. Enumerators visited each house at least twice before a component could be marked as unavailable - in many cases, however, more than two visits were conducted.
Quality assurance procedures included: (i) In-interview feedback from CAPI, which provided a check that modules or questions were not missing, and alerted interviewers to mistakes and inconsistencies in given answers, so that these could be addressed while the interviewer was still with the respondent; (ii) Aggregate checks conducted using the Survey Solutions Supervisor application, which allows supervisors to identify common mistakes (applied to all initial interviews, and then through spot checks); interviewer performance and completion monitoring conducted by the implementing firm, through interviewer and EA level summaries of response rates, interview completion, and progress; (iii) weekly summaries of key indictors provided by the World Bank team (following each data delivery); (iv) direct observation of fieldwork; and (v) back check interviews.
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Disclaimer and copyrights
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.