IDP and Returnees Profiling Exercise in Northern Yemen Methodology

IDPs and Returnees Profiling Exercise in Northern Yemen

June-August 2010

PROPOSED METHODOLOGY

Joint Inter-Agency profiling Service

June 2010

I Background and Justifications

As defined by the Memorandum of the RSG on the Human Rights of IDPs, following the ceasefire concluded between the Government of Yemen and Al-Houthi on 11 February 2010, Internally Displaced Persons (IDPs) in Yemen are in transition period characterized by either:

  1. Return and re-integration in safety and dignity within their own towns and villages of origin; or
  2. Protracted displacement.

Security in Sa’ada, Al-Jawf and Northern Amran Governorates remain volatile, with slow implementation of the ceasefire and continued low level clashes and roadblocks. Despite ongoing security concerns, increasing numbers of IDPs appear to be returning to their areas of origin[1]. In fact, according to an ongoing verification exercise of registered IDPs conducted by the Government with UNHCR support, a possible 32,000 IDPs of a total number of approximately 342,000 registered IDPs have returned. While the returnees are estimated to be more than that, and could amount to up to 60% of the total IDP population according to the Government, there is still no clear system to establish a more accurate number and information about the returnees and the return process. The UN planning figure is that 30% of IDPs will return home this year.

The main established mechanism for data collection so far has been the IDP registration conducted by the Government of Yemen with UNHCR support. While this available data gives very good information about the demographic characteristics of the IDPs and general trends, it suffers from some gaps:

  • Lack of consistent application of the agreed upon registration criteria in all the governorates
  • Non registration of IDPs that showed-up for registration after the deadline or without required documentation
  • The registration only extends to IDPs and not to other conflict-affected people who were unable to move.

In this context, as foreseen in the Yemen Humanitarian Response Plan 2010 and re-iterated by the Representative of the Secretary General on Human Rights of IDPs, it is important to “undertake a general profiling of the displaced population to assess the overall magnitude of the displacement, the location of the displaced, as well as their needs and prospects for their return. Leaving non-registered IDPs out of a durable solution strategy increases the risk of a protracted displacement situation”[2].

The profiling will build on existing registration/verification data and will be complemented by rapid assessments in Sa’ada as well as a planned mapping of districts in Sa’ada for return potential.

II Objectives

The objectives of the profiling exercise are:

  1. To provide a baseline of information about:
  2. Numbers and locations of IDPs disaggregated by sex and age, including the unregistered;
  3. Intentions and return movements of IDPs;
  4. Dynamics of the return movement
  5. Situation of those unwilling or unable to return and their ongoing needs.
  6. To provide a more credible basis for planning of durable solutions and protracted displacement.

III Expected outcomes

An IDPs and Returnees profiling would:

  1. Provide figures and a core dataset that is relevant to all clusters. This will ensure that all humanitarian actors and Government are working with a common understanding of the magnitude of the situation and to facilitate cross-sectoral needs assessment;
  2. Improve the availability and quality of quantitative and qualitative data;
  3. Provide a more accurate estimate of the IDP/ returnee and affected population, disaggregated by gender and age;
  4. Provide a dataset of information on the return process, dynamics and push- and pull-factors, disaggregated by gender and age;
  5. Assist humanitarian actors to prioritize their interventions and plans; and
  6. Create a baseline that could be used for future population movement tracking system;

IV Elements of methodology

1. Methodology

The proposed methodology is a combination of two methods:

  1. Household Survey based on stratified cluster sampling approach targeting 1900 households in the five governorates of Sana’a, Amran, Hajjah, Al Jawf and Saada. The questionnaire will be answered by the head of household, with a specific part directed to the wife, or oldest daughter (above 15 years old) in case it is not a female headed household. Such a differentiation will allow for gender based analysis by the end of the exercise.
  2. Community based focus group discussion targeting one male group and one female group in each of the selected survey sample districts. Within each focus group, the different age categories will be represented.

The combination of the two methods will allow us to have an adequate representation of the findings with in-depth discussion for better analysis and understanding of the intentions and dynamics of return.

1.1 Household survey: In order to ensure a sound representation of the sample, the selection of the statistical units should take into consideration the representation of the different residential characteristics of the IDPs and returnees. The table below maps the residential characteristics per Governorate.

Existent category

Table 1: Residential characteristics mapping

Residential. Characteristics / IDPs living with host families / IDPs in Camps / IDPs renting dwellings / Returnees
Sana’a
Amran
Hajjah
Al Jawf
Saada

Step 1: DEFINING THE SAMPLE SIZE: The first step to be done is to determine the sample size of the survey. The statistical unit in this exercise will be the governorate, thus, we have to decide at an initial stage the number of households to be interviewed in total and per Governorate. The used figures for this calculation are:

  1. The statistical yearbook 2009 issued by the Ministry of Planning and Int. Cooperation – Central Statistical organization to define the total number of population per governorate including IDPs and Non IDPs.
  2. Registration figure of the IDP executive units of the Government of Yemen supported by UNHCR to determine the total number of IDPs per governorate. Even though we expect that the numbers have changed due to the return, the registration figures provide an indicative basis on which to design the sample.

The details of the calculation of the sample size and its adjustments are self explanatory in Annex 1.

Step 2: DEFINING THE SURVEY BASELINE:In order to have a clear representation, the baseline(i.e the number of the IDPs in the surveyed area)should be determined. In the context of Yemen we have a baseline given through the registration exercise where we have the numbers of the registered IDPs per Governorate / per District both in camps and outside camps. While these figures are most probably changing on the ground due to the return process, in addition to the gaps in the registration system, they still representthe best availablebasis for drawing the sample.

In fact, the alternative could be to do a re-counting of IDPs per Governorate / per District.Unless we invest a considerable amount of time and resources in the counting phase (in a dynamic situation where the numbers could change quickly) we will anyhow get a rough estimate similar in accuracy to the figures provided by the registration exercise.

But, since one of the objectives of the exercise is to “provide baseline information about numbers and locations of IDPs disaggregated by sex and age, including the unregistered ones”, and since the profiling aims at taking a snapshot in a dynamic phase of return, we will have to do a counting and listing of the non-registered IDPs per Governorate / per district. Once we have a list of IDPs claiming that they are not registered, we will draw a sample from them. The questionnaire will allow us to determine if they are IDPs, the reasons of non-registration, and have an estimated number of non registered IDPs. The method of capturing this will be explained in details in the sample design.

Step 3: COUNTING UNREGISTERED IDPs: At this stage, each Governorate Field Coordinator will have to collect a list of Yemeni households claiming to be non registered IDPs per Governorate / Per District. Such information has been collected during the verification exercise that took place lately in all the Governorates. These lists are available at the IDP Executive Unit Level per Governorate as well as with NGOs working there.

Annex 2 could be used as a guide to compile the information. In general this information is already available under the same format. In case a need to reformat them has been identified, Annex 2 could be used for this purpose.

Each Field Coordinators should compile all the lists to make a Governorate / District / list of non-registered IDPs.

From these lists a sample will be drawn to be included in the survey.

Step 4: DRAWING THE SAMPLE: The different residential characteristics of IDP households require different sampling approaches during this second stage. Depending on the residential characteristic, a two- or multiple-stage cluster sampling approach is suggested, as to reduce fieldwork costs and time.

A-Non registered IDP households– For each governorate 10 % of the calculated sample size will go to the unregistered IDPs. This is due to a rough estimation that a possible 10% of the registered IDPs might have not registered for different reasons already mentioned. The 10% ratio could be subject to change in case it has proven that the collected lists have a significantly different proportion.

The sampling of non-registered households is the most straightforward procedure, given the known parameters of exact numbers of households provided by the counting and the listing done in step 3. The accumulated list per Governorate permits simple random sampling[3] to select the assigned number of households for interviews. The procedure is to randomly select 10% of the households.

B- IDPs living in camps: For each governorate that has camps, the size of the camp sample will be calculated proportional to the size of the camp population / total IDP size in the governorate after reducing the quota of the non registered IDPs.

  1. Calculate the camp sample size
  2. Multiply it X 2
  3. Generate the list of the registered IDPs in the camp
  4. Draw the sample through a simple random sampling selecting “sample size X 2”
  5. Generate the sample size*2 list

For the interview proceed in interviewing chronologically as per the sample list. In case selected households to be interviewed are not present in the camp. Mark them as absent and move to the next name in the list.

Keep on doing this until you have interviewed a number of households equal to the sample size.

C- IDPs / Returnees living outside camps:For the remaining of the registered IDPs living outside the camps, we divide the districts per governorate according to the level of density of IDPs, where:

  • High are the districts that have more than 4000 registered IDPs
  • Medium are the districts that have between 1000 and 4000 IDPs
  • Low are the districts that has less than 1000 IDPs

And then we proceed as follows:

  1. Calculate the sample size of each density group H, M and L proportional to the size of the IDP Population living in the districts of the group.
  2. Multiply it X 2 for, H, M and L
  3. Generate the list of the registered IDPs in H and in M and in L
  4. Draw the sample through a simple random sampling selecting “sample size X 2” for H and M and L

For the interview proceed in interviewing chronologically as per the 3 sample lists of H, M and L. In case selected households to be interviewed are not present, mark them as absent and move to the next name in the list. (In Annex 3 find attached the form that will be used for the lists of households to be interviewed outside Saada. In the form there is a column entitled status - under status, the enumerators should note, interviewed, or not checked or not available) for each householdKeep on doing this until you have interviewed a number of households equal to the sample size in H and In M and in L.

Attached in Annex 4 are the sample drawing for Sana’a, Amran, Al Jawf, Hajjah; Saada one will be done once we receive the final lists.

1.2 Focus Group Discussion: The FGD aims at providing more qualitative data at the community-based level, especially on the dynamics of return. The selection of where to conduct the FDG will be done as follows:

  1. Within each Governorate there are a number of different residential characteristic that are summarized in the table below.

Residential. Characteristics / IDPs living with host families / IDPs in Camps / IDPs Renting Dwellings / Returnees
Sana’a
Amran
Hajjah
Al Jawf
Saada
  1. The partners per governorate will have to mapzones where each category exists and conduct at a minimum 2 focus groups per category (1 for male / 1 for female), or more if they deem necessary following additional criteria that should developed at the governorate level.

FGD Male / 12-17 / 18-25 / 26-59 / 60+
FGD Female / 12-17 / 18-25 / 26-59 / 60+
  1. The selection of the members of FGD
    will be as per the following
    table.
  2. For each age cohort select 2 representatives.
  3. The selection of the representatives should be done randomly and not through strictly official channels in order to avoid a leaders meeting. Nevertheless, a preparatory meeting before the FGD will be held with the leadership of the community to explain the process. More details about the process will be given in the Enumerators Guide.
  4. At a minimum: 4 FGD will be conducted in Sana’a, 8 in Amran, 6 in Al Jawf and8 in Saada.

2. Target population and definitions

Target population: The Northern Yemen Profiling exercise targets the Internally Displaced and Returned Yemenis due to the 6th conflict between the Government of Yemen and Al Houthi. The Areas of displacement are the Governorates of Sana’a, Amran, Al Jawf, Hajjah and Saada. The exercise targets the IDPs in their areas of displacement and returnees in their areas of return.

The exercise will be based on focus groups and households as two units of measurements.

a. Focus Groupsare divided into two, one for male and one for female, as detailed in the following table. For each group cohort select two participants. The total number of the focus group participants will be 8 for male FG and 8 for female FG.

FGD Male / 12-17 / 18-25 / 26-59 / 60+
FGD Female / 12-17 / 18-25 / 26-59 / 60+

The selection of the focus group participants should be done randomly in order to avoid having a community leaders meeting.

Once in the selected location for FGD,

  • Prepare the ground for the work by meeting with the local leadership and explaining to them the process and the purpose of the exercise.
  • Select one IDP participant from every 10th dwelling. And Select one returnee participant from every 10th dwelling for returnees FGDs.

b. Households’definition is the same as adopted during the registration exercise:

One household is considered to be a nuclear family that includes extended blood-related family who live together in the same dwelling. (A nuclear family is a husband, wife or wives, and children.) For example, a single family could be a husband, two wives, their children and one of the husband’s cousins.

There are also exceptional circumstances where a non-blood related individual would be under the same household. For example, an adoption of a neighbour’s child who’s parents have now passed away or an unaccompanied adult female.

The household definition and its exception should also be assessed in line with the principle of family unity whereby relationship of social, emotional, economical dependency exists between an individual and the head of household.

Households can include more people than in the above definition if needed. It is acceptable for non-blood related persons to be joined to a family in exceptional circumstances, such as a sister-in-law whose husband has died and is now a dependent of her dead husband’s family.

When to Split a Household: When a member of the household who is not the Head of Household is married, that person should belong to a separate household. For example, two brothers who both have wives and children would be two households.

Polygamous Registration: For polygamous families, the husband should be registered only one time with the wife or wives with whom he lives. When multiple wives live together with the husband, they can all be registered as a single household. If one wife lives separately from the wife or wives with whom the husband habitually resides, then she should be registered as a separate female-headed household linked to the other household and the reason for the linkage should be listed as “polygamy.” For women-headed households, it is possible to appoint a designated person entitled to pick up their distribution entitlement. Please see the Standard Operating Procedures for Polygamous Family Registration for more detailed information.

3. Coordination

The cluster approach is adopted in Yemen.

The profiling exercise should be done in collaboration with the UNCT and Humanitarian actors that will be informed about the methodology and work progress.

The Profiling exercise will be done in close collaboration with the Protection Cluster, which will be informed about the methodology and work progress. The Protection Cluster could act as reviewing committee for the documents and tools if needed.

The operation in Yemen is coordinated by the Government of Yemen through the IDP Executive Unit. The IDPEU has representatives in all the affected Governorates. Working closely with them at the field level is crucial in order to:

  • Take advantage of their local knowledge
  • Take advantage of their role to introduce the project to the local community through the customary channels
  • Include them in the process as part of the team if possible

According to the governorates and districts, the involvement of some key players, such as the tribes and local leadership, is crucial.

On the more general level, the Profiling exercise is supported by Minister Kahlani. Nevertheless, periodic follow up should be done at the Ministerial level to make sure we have buy-in in terms of findings at the end of the process.

The project will be conducted by a number of different partners; establishing a solid consistent system is key for the credibility of the findings. That does not exclude the need to be flexible according to the operational requirements per governorate. Moreover, setting a system of information sharing between the different teams, especially if the survey if done in different dates, would give a good boost for the exercise.