Food Insecurity in Conflict Affected Regions in Nigeria 2017
Nigeria, 2017
Get Microdata
Reference ID
NGA_2017_FICAR_v01_EN_M_v01_A_OCS
Producer(s)
National Bureau of Statistics (NBS), The World Bank
Collections
Metadata
Created on
Jul 15, 2019
Last modified
Sep 06, 2019
Page views
398110
Downloads
2520
Identification
Survey ID Number
NGA_2017_FICAR_v01_EN_M_v01_A_OCS
Title
Food Insecurity in Conflict Affected Regions in Nigeria 2017
Country
| Name | Country code |
|---|---|
| Nigeria | NGA |
Study type
Other Household Survey [hh/oth]
Series Information
The food security survey was a telephone based survey conducted between August 15th and September 8th 2017. The interview was the second round of a telephone survey using a sub-set of the sample of GHS (General Household Survey) households. The first round of the telephone interview was administered during spring 2017 with 717 completed interviews with the following geographical distribution: 175 interviews in the North East, 276 in North Central and 266 in South South. The first round was focused on conflict exposure, while the second round discussed in this report focused on food insecurity in conflict affected regions.
This survey focuses on conflict in North East, North Central, and South South Nigeria. Each of these three geopolitical zones has a unique history and context of conflict.
North East Nigeria comprises six states: Adamawa, Bauchi, Borno, Gombe, Taraba, and Yobe.
North Central Nigeria consists of the states of Benue, Kogi, Kwara, Nasarawa, Niger, and Plateau, as well as the Federal Capital Territory (FCT).
South South Nigeria is made up of Akwa Ibom, Bayelsa, Cross River, Delta, Edo, and Rivers states.
This survey focuses on conflict in North East, North Central, and South South Nigeria. Each of these three geopolitical zones has a unique history and context of conflict.
North East Nigeria comprises six states: Adamawa, Bauchi, Borno, Gombe, Taraba, and Yobe.
North Central Nigeria consists of the states of Benue, Kogi, Kwara, Nasarawa, Niger, and Plateau, as well as the Federal Capital Territory (FCT).
South South Nigeria is made up of Akwa Ibom, Bayelsa, Cross River, Delta, Edo, and Rivers states.
Abstract
In this report, we present data from the emergency response survey conducted via telephone among households in three conflict affected regions of Nigeria, North East, North Central and South South between August-September 2017. This round is the second round of telephone data collected from a subsample of households in the Nigeria General Household Survey (GHS). The first round collected data on conflict exposure.
The purpose of this second round of data collection was to understand food insecurity in conflict affected regions. Armed conflict can have a detrimental effect on food security. This might be due to for example reduced agricultural production, or price increases due to malfunctioning markets. Food insecurity might be permanent, such that a household living below the poverty line has a constant struggle to acquire food from the market or produce food for their own use. In situations such as armed conflict, also better endowed households might be temporarily food insecure.
In this report, we find that food insecurity is a major concern in all the three regions studied:
· The mean household in all the three regions is “highly food insecure”
· North East of Nigeria is the most food insecure of the three regions
· Reducing meals or portion size is the most important coping strategy in all three regions
· Food prices are the most important source of food insecurity in all three regions
· A large majority of households rely on the market as the main source of food in all regions. Price concerns should therefore be taken very seriously by policy makers.
· Households in all three regions do not report there being an inadequate supply of food in the market.
The purpose of this second round of data collection was to understand food insecurity in conflict affected regions. Armed conflict can have a detrimental effect on food security. This might be due to for example reduced agricultural production, or price increases due to malfunctioning markets. Food insecurity might be permanent, such that a household living below the poverty line has a constant struggle to acquire food from the market or produce food for their own use. In situations such as armed conflict, also better endowed households might be temporarily food insecure.
In this report, we find that food insecurity is a major concern in all the three regions studied:
· The mean household in all the three regions is “highly food insecure”
· North East of Nigeria is the most food insecure of the three regions
· Reducing meals or portion size is the most important coping strategy in all three regions
· Food prices are the most important source of food insecurity in all three regions
· A large majority of households rely on the market as the main source of food in all regions. Price concerns should therefore be taken very seriously by policy makers.
· Households in all three regions do not report there being an inadequate supply of food in the market.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
Households
Scope
Notes
The questionnaire is divided into 9 sections including a household roster. Information on food insecurity (the coping strategy index, CSI), food and market access, water quality, employment, income, employment and assets was collected.
Topics
| Topic | Vocabulary |
|---|---|
| Agriculture & Rural Development | World Bank |
| Land (policy, resource management) | World Bank |
| Education | World Bank |
| Primary Education | World Bank |
| Secondary Education | World Bank |
| Tertiary Education | World Bank |
| Vocational Education | World Bank |
| Girls’ Education | World Bank |
| Environment | World Bank |
| Migration & Remittances | World Bank |
| Financial Market Integrity (Anti-Money Laundering) | World Bank |
| Transport | World Bank |
| Water | World Bank |
| Information & Communication Technologies | World Bank |
| Social Protection (includes Pensions, Safety Nets, Social Funds) | World Bank |
| Labor Markets | World Bank |
| Poverty | World Bank |
| Fragile & Conflict-affected States | World Bank |
| Financial Management | World Bank |
| Resettlement | World Bank |
| Gender | World Bank |
| Children & Youth | World Bank |
| Disaster Risk Management | World Bank |
Coverage
Geographic Coverage
National Coverage Households
Universe
The Survey covered all household members. The questionnaire was administered to only one respondent per household - most often a male household head.
Producers and sponsors
Primary investigators
| Name | Affiliation |
|---|---|
| National Bureau of Statistics (NBS) | Federal Government of Nigeria (FGN) |
| The World Bank | Federal Government of Nigeria (FGN) |
Producers
| Name | Affiliation | Role |
|---|---|---|
| World Bank | IDA | Technical Assistance in Questionnaire design, Sampling methodology, Data Processing and Analysis |
| National Bureau of Statistics | Federal Government of Nigeria (FGN) | Technical Assistance in Questionnaire design, Sampling methodology, Data Processing and Analysis |
Funding Agency/Sponsor
| Name | Abbreviation | Role |
|---|---|---|
| World Bank | WB | Funding |
Sampling
Sampling Procedure
The food security survey was a telephone based survey conducted between August 15th and September 8th 2017. The interview was the second round of a telephone survey using a sub-set of the sample of GHS (General Household Survey) households. The first round of the telephone interview was administered during spring 2017 with 717 completed interviews with the following geographical distribution: 175 interviews in the North East, 276 in North Central and 266 in South South. The first round was focused on conflict exposure, while the second round discussed in this report focused on food insecurity in conflict affected regions.
In the three conflict affected geographical zones comprising of 16 states of Nigeria, households from LGS's that had high conflict exposure were oversampled chosen for a pilot sample, conducted before the telephone surveys. These LGS's were chosen based on the following criteria: The oversampled LGS's needed to have over 10 conflict events during 2012-14 recorded in the Armed Conflict Location & Event Data Project (ACLED) database.
The first round of the telephone survey (which took place after the pilot) first attempted to reach 742 households from the GHS panel, of which 529 could be reached and interviewed. The rest did not have phone numbers or functioning phone numbers (only 2.7 percent refused to answer). In order to increase the sample size to a level that was considered adequate for the survey, an additional 288 replacement households were included in the sample also from the GHS panel. Out of these replacement households 188 could be interviewed. Therefore altogether 1030 households were attempted to be reached, with a final sample size of 717 completed interviews.
Conflict affected areas were oversampled in order to have a large enough sample of households that in fact experienced conflict events in order to shed light on the type of events that have happened. A random sample of the zones might have given too small sample of conflict affected households and therefore restricted the analysis of the various types of conflict events. Due to the oversampling however, the sample drawn was not representative at the level of the geographical zone, as is the case in the GHS. Therefore in the analysis we use sampling weights that adjust for the propensity of being in a conflict affected LGA in order to ensure that the sample is representative at the level of the geographical zone.
During the second round of the survey 582 of the 717 households were re-interviewed on food security related issues (only the 717 were attempted to be reached). Of the 582 households, 147 in the North East, 219 in North Central, and 216 in South South were interviewed. The attrition rates in our sample from round one to round two are hence 16 percent, 21 percent, and 19 percent for North East, North Central and South South, respectively. The attrition from the conflict survey round was mostly due to not being able to reach the respondents possibly due to non-functioning phone numbers. Only 3 percent of respondents refused to answer.
Similar telephone-based surveys are being conducted in six countries in Sub-Saharan Africa under the World Bank project "Listening to Africa". As a comparison, a mobile phone survey in Tanzania (see Croke et al. 2012 for details), had a high drop-out rate between the very first rounds from 550 to 458 respondents, but very low attrition for the subsequent rounds for the 458 respondents, who could reliably be reached by a mobile phone. In light of this reference point and also considering the fact that the households interviewed live in conflict affected regions, our attrition rates seem to be within reasonable limits.
In the three conflict affected geographical zones comprising of 16 states of Nigeria, households from LGS's that had high conflict exposure were oversampled chosen for a pilot sample, conducted before the telephone surveys. These LGS's were chosen based on the following criteria: The oversampled LGS's needed to have over 10 conflict events during 2012-14 recorded in the Armed Conflict Location & Event Data Project (ACLED) database.
The first round of the telephone survey (which took place after the pilot) first attempted to reach 742 households from the GHS panel, of which 529 could be reached and interviewed. The rest did not have phone numbers or functioning phone numbers (only 2.7 percent refused to answer). In order to increase the sample size to a level that was considered adequate for the survey, an additional 288 replacement households were included in the sample also from the GHS panel. Out of these replacement households 188 could be interviewed. Therefore altogether 1030 households were attempted to be reached, with a final sample size of 717 completed interviews.
Conflict affected areas were oversampled in order to have a large enough sample of households that in fact experienced conflict events in order to shed light on the type of events that have happened. A random sample of the zones might have given too small sample of conflict affected households and therefore restricted the analysis of the various types of conflict events. Due to the oversampling however, the sample drawn was not representative at the level of the geographical zone, as is the case in the GHS. Therefore in the analysis we use sampling weights that adjust for the propensity of being in a conflict affected LGA in order to ensure that the sample is representative at the level of the geographical zone.
During the second round of the survey 582 of the 717 households were re-interviewed on food security related issues (only the 717 were attempted to be reached). Of the 582 households, 147 in the North East, 219 in North Central, and 216 in South South were interviewed. The attrition rates in our sample from round one to round two are hence 16 percent, 21 percent, and 19 percent for North East, North Central and South South, respectively. The attrition from the conflict survey round was mostly due to not being able to reach the respondents possibly due to non-functioning phone numbers. Only 3 percent of respondents refused to answer.
Similar telephone-based surveys are being conducted in six countries in Sub-Saharan Africa under the World Bank project "Listening to Africa". As a comparison, a mobile phone survey in Tanzania (see Croke et al. 2012 for details), had a high drop-out rate between the very first rounds from 550 to 458 respondents, but very low attrition for the subsequent rounds for the 458 respondents, who could reliably be reached by a mobile phone. In light of this reference point and also considering the fact that the households interviewed live in conflict affected regions, our attrition rates seem to be within reasonable limits.
Response Rate
The first round of the telephone survey (which took place after the pilot), first attempted to reach 742 households from the GHS panel, of which 529 could be reached and interviewed. The rest did not have phone numbers or functioning phone numbers (only 2.7 per cent refused to answer). In order to increase the sample size to a level that was considered adequate for the survey, an additional 288 replacement households were included in the sample also from the GHS panel. Out of these replacement households 188 could be interviewed. Therefore altogether 1030 households were attempted to be reached, with a final sample size of 717 completed interviews. The response rate is 96%
Weighting
In the analysis, probability weights that adjust for the propensity of being in a conflict affected LGA in order to ensure that the sample is representative at the level of the geographical zone was used.
Data Collection
Dates of Data Collection
| Start | End | Cycle |
|---|---|---|
| 2017-08-15 | 2017-09-08 | Second round |
Data Collection Mode
Computer Assisted Telephone Interview [cati]
Data Appraisal
Data Appraisal
Limitations
Recall Bias
In the pilot data collection, respondents were asked to report on conflict events that had taken place in their family and their community over the last six years. This extremely long recall period must be considered when drawing inferences from the data. People are likely to under-report less severe (and therefore less memorable) events, particularly those that happened to community members in larger communities. Respondents are also more likely to recall events that happened to family members than those that happened to community members. Other biases may also be at play - for example, those who have been most highly affected by conflict over the last six years may have moved to another community. These factors demonstrate the importance of implementing a regular data collection schedule, which would allow far more accurate data to be collected.
Sampling Bias
The GHS is a panel survey taking place over multiple rounds through a period of time. Therefore, households that are more mobile or households that are nomadic are less likely to be represented in this sample. This may be particularly relevant in circumstances where nomadic groups are named as perpetrators of conflict events.
Power Dynamics
There are some disadvantages to the phone system, and for this reason it should be supplemented by additional types of data collection wherever possible. In a mobile phone survey, the respondent is the person who owns a mobile phone. In many areas, particularly those highly affected by poverty and those located in rural areas, only one family member owns a mobile phone. This is generally the household head, who is most likely male. Furthermore, in many of these communities, women are not allowed to have access to mobile phones and are forbidden from speaking to outsiders, which can prohibit mobile phone-based data collection.
Gender Dynamics
The questionnaire was administered to only one respondent per household - most often a male household head. This means that crimes that carry stigma, especially sexual violence, are less likely to be reported. In this dataset, no sexual assault was reported despite data collected elsewhere that indicate that rape was used as a weapon by Boko Haram and elsewhere. This also means that violence that affects members of the household with less power (such as women, children, and employees), is less likely to be reported. This may be particularly important when considering violence not related to ongoing external conflict, such as domestic violence.
Recall Bias
In the pilot data collection, respondents were asked to report on conflict events that had taken place in their family and their community over the last six years. This extremely long recall period must be considered when drawing inferences from the data. People are likely to under-report less severe (and therefore less memorable) events, particularly those that happened to community members in larger communities. Respondents are also more likely to recall events that happened to family members than those that happened to community members. Other biases may also be at play - for example, those who have been most highly affected by conflict over the last six years may have moved to another community. These factors demonstrate the importance of implementing a regular data collection schedule, which would allow far more accurate data to be collected.
Sampling Bias
The GHS is a panel survey taking place over multiple rounds through a period of time. Therefore, households that are more mobile or households that are nomadic are less likely to be represented in this sample. This may be particularly relevant in circumstances where nomadic groups are named as perpetrators of conflict events.
Power Dynamics
There are some disadvantages to the phone system, and for this reason it should be supplemented by additional types of data collection wherever possible. In a mobile phone survey, the respondent is the person who owns a mobile phone. In many areas, particularly those highly affected by poverty and those located in rural areas, only one family member owns a mobile phone. This is generally the household head, who is most likely male. Furthermore, in many of these communities, women are not allowed to have access to mobile phones and are forbidden from speaking to outsiders, which can prohibit mobile phone-based data collection.
Gender Dynamics
The questionnaire was administered to only one respondent per household - most often a male household head. This means that crimes that carry stigma, especially sexual violence, are less likely to be reported. In this dataset, no sexual assault was reported despite data collected elsewhere that indicate that rape was used as a weapon by Boko Haram and elsewhere. This also means that violence that affects members of the household with less power (such as women, children, and employees), is less likely to be reported. This may be particularly important when considering violence not related to ongoing external conflict, such as domestic violence.
Access policy
Contacts
| Name | Affiliation | URL | |
|---|---|---|---|
| Dr. Yemi Kale (Statistician-General) | National Bureau of Statistics (NBS) | [email protected] | Link |
| Mr. M Abul Kalam Azad | World Bank | [email protected] | |
| Mr. Fafunmi E.A (Head, ICT Department) | National Bureau of Statistics (NBS) | [email protected] | Link |
| Mr. Esiri Ojo | National Bureau of Statistics (NBS) | [email protected] | Link |
| Oke Florence | National Bureau of Statistics (NBS) | [email protected] | Link |
| Irenonse Victoria (Data Archivist) | National Bureau of Statistics (NBS) | [email protected] | Link |
Confidentiality
The confidentiality of the individual respondent is protected by law (Statistical Act 2007).
This is published in the Official Gazette of the Federal Republic of Nigeria No. 60 vol. 94 of 11th June 2007. See section 26 para.2. Punitive measures for breeches of confidentiality are outlined in section 28 of the same Act.
Access conditions
A comprehensive data access policy is been developed by NBS, however section 27 of the Statistical Act 2007 outlines the data access obligation of data producers which includes the realease of properly anonymized micro data.
Citation requirements
National Bureau of Statistics, Nigeria, Conflict and Violence in Nigeria-v1.0
Disclaimer and copyrights
Disclaimer
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.
Metadata production
DDI Document ID
DDI_NGA_2017_FICAR_v01_EN_M_v01_A_OCS_FAO
Producers
| Name | Abbreviation | Affiliation | Role |
|---|---|---|---|
| National Bureau of Statistics | NBS | Federal Government of Nigeria (FGN) | Metadata Producer |
| Office of Chief Statistician | OCS | Food and Agriculture Organization | Metadata adapted for FAM |
DDI Document version
NGA_2017_FICAR_v01_EN_M_v01_A_OCS_v01