Nigeria - National Core Welfare Indicators Survey, 2006, Third edition

Ver 1.2

2008-07-30

The first Pilot Survey on CWIQ was carried out in Lagos State in 1999. This was followed by that of Benue State in 2001. The pilot survey was later extended to cover six States, namely: Ekiti, Kogi, Kebbi, Enugu, Cross-river and Jigawa in 2002. Another six States: Gombe, Osun, Kebbi, Abia, Plateau and Cross-River States were covered in 2003 and 2004. It was possible through CWIQ to compare some poverty Indicators across States and Senatorial Districts. The combined CWIQ could not give poverty indicators at the National level. Consequent upon successful conduct of the Core Welfare Indicators Questionnaires (CWIQ) Survey in six States in the year 2003 and June 2004, there was the need to conduct a National Survey. The National Survey started with a regional workshop where the new CWIQ questionnaire was developed with major stakeholders from relevant Ministries, Agencies and International Development Partners.

Worldwide, the Core Welfare Indicator Questionnaire Survey (CWIQ) is designed to collect household data useful in quantitatively profiling the well-being of the population. The 2006 Nigerian CWIQ was a nationwide sample survey conducted to produce welfare indicators for the population at national and sub-national levels, particularly Zones, States and Senatorial Districts. The Survey compliments 2004 Nigerian Living Standards Survey (NLSS) by NBS which profiled poverty in the country.

The successful release of the 2006 report was a further proof of the capability of the CWIQ to provide data for monitoring poverty at very short notice and the premium the National Bureau of Statistics places on timely release of statistical data. Another significant experience in this survey was the Technology of scanning which helped in the shortening of data processing time and hence the timely release of the result within three months as against the traditional release of main report after one year of data collection.

Another innovation and experience is the outsourcing for qualified field staff for the field work which also helped in collection of high quality data. .

The CWIQ is another instrument for regular monitoring of welfare and social trends for different population groups of the society especially the poor. The survey gives a quicker method for monitoring poverty than the National Consumer Survey indicators for meaningful changes in welfare status. The indicators available in this CWIQ are of two types: Firstly, indicators of Living Standards for the household members - these indicators include land assets, home ownership, type of home construction, fuel for cooking, ownership of selected households goods, literacy level, employment, health and nutrition and secondly indicators of access, utilization and satisfaction - those include access to clean water, primary and secondary school services, utilization of facilities by way of enrolment rates by gender, satisfaction with school and medical facilities.

Preliminary reports in form of flyers have been released after three months of data collection. The flyers were issued at National and zonal levels which quickly met the needs of users.

The main report will be useful to the Federal Government of Nigeria, States, Non Government Organization (NGO), International Development Partners (World Bank, UNDP, UNICEF etc.) and other Institutions in monitoring poverty and other related social problems.

National Bureau of Statistics welcomes observations and comments that will help to improve the conduct of CWIQ Surveys in Nigeria.

Dr. V. O. Akinyosoye,

Director General

National Bureau of Statistics, (NBS)

Abuja

Sample survey data [ssd]

The sample design employed for National Core Welfare Indicator Questionnaire Survey 2005 is a 2-stage cluster sample design in which Enumeration Areas (EAs) or Primary Sampling Units (PSUs) constitute the 1st stage sample while the Housing Units (HUs) from EAs make up the 2nd stage sample or the ultimate sampling units.

Sampling Frame:

The enumeration Areas (EAs) as demarcated by the National Population Commission (NPopC) for the 1991 population census served as the sampling frame for the National Core Welfare Indicator Questionnaire (CWIQ) survey 2005. Although the frame was however deficient in two main areas, viz;

(i) No measure of size was readily available for the EAs which served as the primary sampling units (psus) and

(ii) The EAs were not stratified into urban and rural during the census exercises in 1991.

Nevertheless, the solution provided for the first problem was to take the PSU's with equal probability while the envisaged solution for the second deficient was the promise by the national Population Commission (NPopC) to supply the information at a later date.

Sample Size:

Sample sizes must meet some minimal requirement in order to obtain reliable estimate. Hence, for National CWIQ survey 2005, the sample size varies from state to state depending on the number of Local Government Areas (LGAs) in each state. Ten (10) Enumeration Areas (EAs) were selected in each LGA making a total of 7,740 EAs to be canvassed for throughout the federation from the 774 LGAs including Federal Capital Territory (FCT) Abuja. (See Appendix 1)

Selection Procedure:

The 7,740 EAs were selected directly from the population of EAs in the National Population Commission (NPopC) with equal probability of selection. Prior to selection, all the contiguous EAs were arranged serpentinely in each Local Government Area (LGA) of the state. This arrangement ensured that there was no overlapping between the LGAs and the EAs within the LGAs. With the frame so constructed, the EAs were cumulatively numbered to get the EA population. Therefore, 10 random starts (RS) were taken for each LGA and systematic selection approach were adopted in selection 10 EAs in each LGA.

A listing/updating exercise was carried out in each EA from where 10 housing units (HUs) were selected systematically to form the 2nd stage Sample or Ultimate interviewed using the Core Welfare Indicator questionnaire.

Jackknife method which is the CWIQ software for Error Estimate was used. This required forming replicate from t he full sample by randomly eliminating one sample cluster [enumeration area (EA)] at a time from a state which is the reporting domain. Then a pseudo-estimate is formed from the retained EAs, which are re-weighted to adjust for the eliminated units. Thus, for a particular state containing k EAs, k replicated estimates are formed by eliminating one of these, at a time, and increasing the weight of the remaining (k-1) EAs by a factor of k/(k-1). This process is repeated for each EA.

For a given state or reporting domain, the estimate of the variance of a rate, r, is given by where is the standard error, k is the number of EAs in the state or reporting domain.

r is the weighted estimate calculated from the entire sample of EAs in the state or reporting domain.

To obtain an estimate of the variance at a higher level, say, at the national level, the process is repeated over all states, with k redefined to refer to the total number of EAs (as opposed to the number in the state).