A recent survey showed that from the total adult population of 87.9 million, the income level has only 3.6 million Nigerian adults earning above N70, 000 per month. The income earning level for the country does not support affordability of high cost housing. In Nigeria, supply of housing is mainly targeted at the high income earners, and selling price which are termed ‘low’ are still unaffordable to the lower income bracket. House costs at the minimal can be estimated at N5,000,000, and this is in only a few locations in the country. Therefore, the high demand will remain if indices, statistics and features in the industry remain unchanged.
Quantitative Methodological Approach
The goal of the survey among other things is to provide the most comprehensive and up-to-date source of information on the demand for affordable housing and related affordable housing finance services in Lagos state. The survey was designed to cover the statutory recognized twenty (20) LGAs plus the thirty-seven (37) LCDAs and to provide a true representation of the housing demand situation in the State.
This section also covers important survey-related issues including sample design, questionnaire design, training of enumerators and supervisors, pre-testing of questionnaires, fieldwork and data processing. The methodological framework that guides the implementation of the quantitative component of Lagos Housing Demand Survey conducted by PISON/CMD is as follows;
Target Population and Coverage
The target population in this study is persons who reside in households within Lagos State that are economically active, aged 15-65 years. The primary unit of analysis is therefore the individuals within households. Based on 2013 Lagos state projected population estimates, the state had a population of 21.8 million out of a national estimate of 167 million. Out of the total population of Lagos, the projected population of 15-65 years in Lagos State is estimated at 14.3 million people constituting the universe of our study. The geographical boundary of the study area included the residential units within the Lagos state recognized 57 LCDAs embedded in the statutory 20 Local Government Areas.
Survey Sample Design
A random household survey of persons aged 15 – 65 years obtained through a multi-stage area based sampling approach was used. Given the diversity of Lagos state, the area sampling approach, (also known as Cluster sampling design) was adopted. This approach was preferred, not only for its appropriateness for a study of this magnitude, but also because of its economy and simplicity. This design will ensure the generation of a truly representative sample that is devoid of bias while guaranteeing precise projection of the survey result into the study population, within an acceptable confidence limit. Probability Proportional to Size (PPS) was adopted to determine the sample size per local government area while the ultimate samples were selected using Simple Random Sampling with Replacement.
For the purposes of this study, the economically active population is defined as persons who fall within the age bracket of 15 – 65 years and have a visible means of livelihood, and must have been residing in Lagos state in the last one year. While the Lagos State Bureau of Statistics provided technical support in drawing a representative sample for the study.
Survey Sample Frame
Obtaining a good sampling frame, and adopting methods to remedy any deficiencies present in it, is critical components of good survey design. The sample frame is a list(s) or resource(s) that contains the elements of the defined population. A desired sampling frame is one containing an actual list of elements to be sampled (e.g. households) where each element appears at least once and not more than once. Another desirable quality of a sampling frame is ease of identification of each element along a series of variables important in the study (e.g., housing tenure status of households).
The Lagos State Bureau of Statistics has been undertaking regular household surveys that focus on the economically active population in Lagos state. Therefore, it warehouses up-to-date and reliable sampling frame and fairly detailed maps to locate households in the field. The lists of households are routinely updated by the Lagos State Bureau of Statistics served as the sampling frame.
The sample frame was therefore based on the Lagos State 2013 Census population estimates and the lists of wards, streets and housing units within the state. The current view of the size of the population within the Local Government Areas by wards is as follows:
Sample Size and Selection Procedure
One of the basic decisions in developing survey plans is to determine the appropriate size of the sample. The main objective is to have a large enough sample to allow reaching reliable conclusions using survey results, and at the same time to have a sample as small as possible to economize on survey implication costs.
For 95% confidence, 5% precision and design effect of 1.5 (to compensate for the stratified/cluster effect) a total sample size of 6,016 households was drawn using a multistage stratified sampling procedure that cut across the entire local government areas in Lagos State. This will allow for reporting of the results at the Local Government Level.
A stratified multistage sample of households was drawn as follows:
- Target Population: households in Lagos State
- Strata: 20 LGAs + 37 LCDAs
- Primary sampling unit (PsU): Wards
- Secondary sampling unit (ssU): Streets
- Tertiary sampling unit (UsU): A Person 15-65 years within households
The first level of stratification comprised the Local Government Areas (20 LGAs + 37 LCDAs), with each of them divided into Political Wards (between 10 and 25). These wards formed the second level of stratification. All the streets in each ward were listed and all the housing units in selected streets were also listed, together with all the households living in them. Probability Proportional to Size (PPS) was adopted to determine the sample size per local government area while the ultimate samples were selected using Simple Random Sampling with Replacement.
Get Full Document CLICK HERE