ESTIMATING MONTHLY HOUSEHOLD INCOME USING THE NIGERIA LIVING STANDARDS SURVEY
This document explains the derivation of an estimate of monthly household income, using data from the 2018 Nigeria Living Standards Survey (NLSS) published on the Nigeria Housing Data Dashboard.
Expenditure as a proxy for income
Accurately measuring household income is notoriously challenging as survey respondents typically underreport income (see Deaton, 2018:29 for more detail). In our analysis of the NLSS, we used an estimate of household expenditure as a proxy for income - a common approach that is generally seen as more accurate for countries with large informal sectors. Nevertheless, it is worth noting that even this expenditure-based estimate may be understated, particularly in more affluent segments of the market, as some expenditure items are omitted from the survey. Most notably, the survey does not include the purchase of vehicles, a major expenditure category in developed economies, and presumably likewise in more affluent segments of the Nigerian market.
The figure below presents the distribution of households by monthly expenditure as calculated using the NLSS. In the sections that follow, we describe our methodology to calculate monthly household expenditure as a proxy for monthly household income.
Measuring expenditure using the Nigeria Living Standards Survey
The NLSS explores expenditure using the recall method, requiring respondents to report expenditure on various items over a stated time period. These time periods differ depending on the item. In some categories, respondents are asked to recall their expenditure over the previous seven days while in others the survey question refers to monthly or annual frequencies.
The NLSS has specific sections that contain most of the expenditure-related questions (referred to here as the ‘General Expenditure’ sections). However, there are also expenditure-related questions in the Education, Health, Housing and Agriculture sections of the survey.
Our expenditure estimate aggregates data from all these sections and adjusts for different frequencies by standardising expenditure to equivalent 30-day estimates. It then applies household weights to derive a distribution of households by household income. Data for households in Borno state are excluded from all analysis due to incomplete data collection as a result of regional conflict.
Sections of the survey included in the derivation of total monthly household expenditure are explained in more detail below.
General expenditure (Sections 6A, 6B and 7)
These sections of the survey explore food and non-food expenditure at differing frequencies. The included sections are:
- Section 6A: Meals consumed away from home (past 7 days)
- Section 6B: Food consumption expenditure
- Section 7: Non-food expenditure (past 7-days, past 30 days, and past 12 months, depending on the item)
Education (Section 2)
The Education section contains questions on the amount that households spend both directly and indirectly on education. The phrasing of questions in this section depends on the time of year. If the survey is conducted before the commencement of the school year, parents are asked how much the household expects to spend for the upcoming school year. If the survey is conducted during the course of the school year, parents are asked how much the household has spent so far, and how much more the household expects to spend for the remainder of the school year.
Both past education expenditure and expected education expenditure are included in the calculation to derive a total annual education expenditure. The specific variables included in these calculations are listed in the table below. Note that multiple question numbers exist to accommodate for the variability in timeframe explained above.