Refugee and Host Household Survey in Nairobi, 2021
Kenya, 2021
Get MicrodataIdentification
KEN_2021_RHHS_v01_M
Refugee and Host Household Survey in Nairobi, 2021
RHHS 2021
| Name | Country code |
|---|---|
| Kenya | KEN |
1-2-3 Survey, phase 1 [hh/123-1]
This cross-sectional survey was conducted with refugee and host community households in Nairobi and conducted between May 22 to July 27, 2021.
The World Bank in collaboration with the Joint Data Center on Forced Displacement, Kenya National Bureau of Statistics (KNBS) and the United Nations High Commissioner for Refugees (UNHCR) conducted a cross-sectional survey on refugee and host populations living in Nairobi. The survey was based on the Kenya Continuous Household Survey (KCHS) and targets both host populations and refugees living in Nairobi. Through a participatory training format, enumerators learned how to collect quality data specific for refugees as well as nationals. Daily data quality monitoring dashboards were produced during the data collection periods to provide feedback to the field team and correct possible errors. The data was collected with CAPI technique through the World Bank developed Survey Solutions software; this ensured high standards of data storage, protection and pre-processing.
The sample is representative of refugees and other residents living in Nairobi. The refugee sample was drawn from UNHCR’s database of refugees and asylum seekers (proGres) using implicit stratification by sub-county and country of origin. The host community sampling frame was drawn using a two-stage cluster design. In the first stage, eligible enumeration areas (EAs) based on the 2019 Population and Housing Census were selected. In the second stage 12 households were sampled from each EA. The survey differentiates between two types of host communities: ‘core’ host communities were drawn from EAs located within the three areas with the largest number of refugee families: Kasarani, Eastleigh North and Kayole. At least 10 percent of the Nairobi refugee families reside in each of these areas. ‘Wider’ host communities cover the rest of the Nairobi population and were drawn from EAs which do not cover the three areas in which many refugees live.
For a subset of households, a women empowerment module was administered by a trained female enumerator to one randomly selected woman in each household aged 15 to 49.
The data set contains two files. hh.dta contains household level information. The ‘hhid’ variable uniquely identifies all households. hhm.dta contains data at the level of the individual for all household members. Each household member is uniquely identified by the variable ‘hhm_id’.
This cross-sectional survey was conducted between May 22 to July 27, 2021. It comprises a sample of 4,853 households in total, 2,420 of which are refugees and 2,433 are hosts.
Household, Individual
Version
Edited, anonymous dataset for public distribution.
Scope
This survey covers the following topics: Household roster, education, labor, housing, consumption and insecurity, vulnerabilities, social cohesion, coping mechanisms, trajectories, remittances, household air pollution (subset of households), women empowerment (1 female respondent per household).
Coverage
Nairobi county, Kenya
Producers and sponsors
| Name | Affiliation |
|---|---|
| Nistha Sinha | World Bank |
| Precious Zikhali | World Bank |
| Antonia Johanna Sophie Delius | World Bank |
| Nduati Maina Kariu | World Bank |
| Name | Abbreviation |
|---|---|
| Joint Data Center on Forced Displacement | JDC |
Sampling
The survey has two primary samples contained in the ‘sample’ variable: the refugee sample and the host community sample. The refugee sample used the UNHCR database of refugees and asylum seekers in Kenya (proGres) as the sampling frame. ProGres holds information on all registered refugees and asylum seekers in Kenya including their contact information and data on nationality and approximate location of living. We considered only refugees living in Nairobi and implicitly stratified by nationality and location. In total, the sample comprises 2,420 refugee families.
The host community sample differentiates between two types of communities. We consider ‘core’ host communities as residents who live in Eastleigh North, Kayole or Kasarani – at least 10 percent of the Nairobi refugee families reside in each of these areas. Nationals living outside these areas are considered part of the ‘wider’ host community in Nairobi. The samples for both host communities were drawn using a 2-stage cluster design. In the first stage, eligible enumeration areas (EA) were drawn from the list of EAs covering Nairobi taken from the 2019 Population and Housing Census. In the second stage a listing of all host community households was established through a household census within all selected EAs, ensuring that refugee households were excluded to prevent overlap with the refugee sampling frame. 12 households and 6 replacements were drawn per EA. Our total sample consists of 2,433 host community households, 1,221 core hosts and 1,212 wider hosts.
The three sub-samples – refugees, core hosts, and wider hosts – are reflected in the ‘strata’ variable. The EAs which form the primary sampling units for the two host samples are anonymized and included in the ‘psu’ variable. Please note that the ‘psu’ variable clusters refugees under one numeric code (888).
To make the sample representative of the refugee and host community living in Nairobi, cross-sectional weights are constructed in two steps:
Step 1: Construct raw weights combining the two national samples
For both the core and wider host community, the base selection probability of household k is calculated as the product of the probability of selecting EA h of household k and the probability of selecting household k among all the households listed in EA h. The raw weight is calculated as the inverse of the base selection probability. For refugees the raw weight is the inverse of the probability of being sampled from the proGres database.
Step 2: Apply re-weighting
(I) Non-response adjustment
(II) Post-stratification
(I) For refugees we adjust for non-response through a propensity score-based method following 5 steps:
- The predicted probability that a sample household was successfully interviewed is estimated through a propensity score estimation. The propensity score (PS) is modeled with a linear logistic model at the level of the household. The dependent variable is a dummy indicating whether a sampled household was interviewed. The following covariates were used in the linear logistic model: a set of dummies for languages spoken, location dummies and family-size.
- Rank households by PS and split into 10 equal groups.
- Calculate the non-response adjustment factor (af) as the reciprocal of the mean empirical response rate for the propensity score decile.
- Adjust base weights for non-response: W_2 = W_1 * af.
- Winsorize the top 1 percent of the weights distribution W_3 = winsorize(W_2).
(II) After the non-response adjustment is applied to the raw weights from step 1, we scale weights to population proportions in each stratum. Information on true population sizes for both host communities is obtained from the 2019 Kenya Population and Housing Census conducted by the KNBS (2019 Kenya Population and Housing Census, Volume II: Distribution of Population by Administrative Units, December 2019, Kenya National Bureau of Statistics, https://www.knbs.or.ke/?wpdmpro=2019-kenya-population-and-housing-census-volume-ii-distribution-of-population-by-administrative-units). Population totals for refugees are obtained from UNHCR’s proGres database. The household-level weight variable after post-stratification is ‘weight’ which should be used in the analysis.
Survey instrument
The Questionnaire is provided as external resources in pdf format. Questionnaires were produced through the World Bank developed Survey Solutions software. The survey was implemented in English,Swahili and Somali.
When variable names consist of a letter and a number, the letter indicates the questionnaire section and the number the order in which the question was administered. Variable names were kept consistent with the KCHS 2019 to allow for a straightforward comparison. An explanation of the variable contents is contained in the variable labels.
Extended missing values are used to indicate why a value is missing for all variables. The following extended missing values are used in the dataset:
• .a for ‘Don’t know’
• .b for ‘Refused to respond’
• .c for ‘Outliers set to missing’
• .e for ‘Field Skipped’ (where an error in the survey tool caused the question to be missed)
• .z for ‘Not administered’ (as the variable was not relevant to the observation)
Data collection
| Start | End | Cycle |
|---|---|---|
| 2021-05-22 | 2021-07-21 | Wave 1 |
- Computer Assisted Personal Interview [capi]
| Name | Abbreviation |
|---|---|
| Altai Consulting (Registered as Lhassa Consulting) | FZE |
| Trends and Insights for Africa | TIFA |
PRE-LOADED INFORMATION: Basic household information from the host community household listing (hosts) and UNHCR’s proGres database (refugees) was pre-loaded in survey solutions for each enumerator. The information, for example the household's location, household head name, phone numbers etc, was used to help enumerators identify the target households.
RESPONDENTS: The survey had one main respondent per household. The respondent was any knowledgeable adult aged 18 years or older. For the women empowerment module one woman aged 15-49 was randomly selected from each household and thus the respondent may differ from the main household respondent.
Depositor information
| Name |
|---|
| World Bank |
Data Access
| Name | Affiliation |
|---|---|
| Precious Zikhali | World Bank |
| Nistha Sinha | World Bank |
| Antonia Delius | World Bank |
| Nduati Maina Kariuki | World Bank |
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
Example:
Nistha Sinha (World Bank), Precious Zikhali (World Bank), Antonia Johanna Sophie Delius (World Bank), Nduati Maina Kariu (World Bank). Kenya - Refugee and Host Household Survey in Nairobi, 2021 (RHHS 2021). Ref: KEN_2021_RHHS_v01_M. Downloaded from [uri] on [date].
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.
Contacts
| Name | Affiliation | |
|---|---|---|
| Precious Zikhali | World Bank | pzikhali@worldbank.org |
| Nistha Sinha | World Bank | nsinha@worldbank.org |
| Antonia Delius | World Bank | adelius@worldbank.org |
| Nduati Maina Kariuki | World Bank | nkariuki@worldbank.org |
Metadata production
DDI_KEN_2021_RHHS_v01_M_WB
| Name | Abbreviation | Affiliation | Role |
|---|---|---|---|
| Development Data Group | DECDG | World Bank | Documentation of the study |
2023-10-10
Metadata version
Version 01 (2023-10-10)