Forced migration and displacement

In her article, Krynsky Baal (2021) highlights several persistent challenges to data on forcibly displaced persons, the first of which is the lack of comparability between statistics on forcibly displaced persons both within and across countries. The 2018 International Recommendations on Refugee Statistics (IRRS) and the 2020 International Recommendations on Internally Displaced Persons Statistics (IRIS), both developed by the Expert Groups on Refugee, IDP and Statelessness Statistics (EGRISS), have provided clear definitions for addressing this shortcoming (which are outlined in Part I, Chapter 4).

Second, the common perception of displacement as a temporary or short-lived phenomenon has resulted in forcibly displaced persons being excluded from national development planning and budgeting. This is because most data collection among forcibly displaced persons does not include socio-economic indicators aligned with existing statistical standards which facilitate comparisons with the host population and because refugees, IDPs and stateless individuals are largely omitted from the official numbers of national statistical systems. The Uganda Refugee and Host Communities 2018 Household Survey took measures to address the lack of socio-economic indicators by aligning itself with national survey instruments. The survey demonstrated that while poverty and unemployment rates are generally higher among refugees than among the host population, access to certain resources is sometimes lower among the host population, highlighting that the former should also be included in humanitarian programming (World Bank, 2018). 

Third, age-, gender-, disability- and other diversity criteria-disaggregated data needed to inform targeted interventions are often missing. Even the breakdown of refugee and IDP population data by age and sex in the UNHCR Global Trends is based on statistical modelling instead of actual numbers (UNHCR, 2020). Developing sampling strategies for producing overall statistics based on a smaller representative sample of refugees, IDPs or stateless persons (from registration, flow monitoring or geospatial sampling frames) is a way of generating more disaggregated data without having to conduct a complete census (Krynsky Baal, 2021).

Finally, the data landscape in humanitarian contexts is often characterized by challenging operational realities, limited coordination between agencies and poor technical skills. Access to potential respondents is restricted for reasons of insecurity or inadequate infrastructure. Limited coordination between agencies and duplication of data collection efforts may contribute to "survey fatigue” among affected populations or lead to contradictory data making decisionmakers more sceptical of the need to act. Furthermore, data literacy, or the ability to master a basic understanding of data sources and analytical methods, is not widespread among humanitarian actors (Krynsky Baal, 2021).

Data on forcibly displaced persons is primarily collected by national authorities and both national and international humanitarian organizations through instruments aimed to inform programs delivering protection and assistance to forcibly displaced persons (Krynsky Baal, 2021). The United Nations High Commissioner for Refugees (UNHCR) collects and compiles data on asylum seekers and refugees more specifically on asylum applications, refugee status determination, recognition rates, refugee populations and movements, demographic characteristics (age and sex) as well as major refugee locations (camps, centres, urban areas, etc.). However, the scope and detail of the data collected by UNHCR depends on the nature of situation (i.e., stable or rapidly evolving) and the collaboration between UNHCR and the host government (EGRISS, 2018). IOM’s Displacement Tracking Matrix (DTM) also documents the movement and needs of refugees and asylum seekers through mobility tracking, flow monitoring surveys and baseline assessments. For example, IOM’s DTM collected data on refugee children from Venezuela in long stay or street situations in Colombia (IOM, 2021b) and on the needs, intentions and integration challenges of Ukrainian refugees in Poland and other neighbouring countries (IOM, 2023b).

Figure 1:

The IRRS recommend collecting data on the total stock of refugee and refugee-like populations, which should be further disaggregated by the total stock of each of the three sub-components of this population: persons needing international protection, persons with a refugee background and persons returned from abroad after seeking international protection (see Part I, Chapter 4 on definitions). They also recommend collecting data on the total flows of refugee and refugee-like populations, which should be further disaggregated by a) flows of persons in need of international protection (including prospective asylum seekers, new asylum seekers, prima facie refugees, new resettled refugees, and reunified refugee family members) and b) flows of persons returned from abroad after seeking international protection (including repatriating refugees, repatriating asylum seekers, and persons returning after having received other forms of international protection) (EGRISS, 2018). The statistics outlined above should be further disaggregated by the following basic classificatory variables:

• Age or date of birth;
• Sex;
• Country of birth;
• Country of citizenship (including stateless, undetermined and multiple citizenship);
• Date of arrival in country;
• Reason for migration;
• Country of previous or last residence (for both refugees and returnee refugees);
• Date of first displacement from previous country of habitual residence;
• Parents' refugee status;
• Status as an unaccompanied minor;
• Legal residential/international protection status in country.

In most countries affected by forced displacement, the census rarely allows the identification and measurement of refugees and refugee-like populations. The IRRS recommends that ‘forced displacement’ be included among the reasons for migration in population censuses and household surveys and that, depending on the context, “returning home after seeking international protection” also be included as a separate category.

In contexts where the proportion of refugees and populations in a refugee-like situation among the general population is high, then adding questions identifying these groups to an existing census or survey is desirable. In contexts where the proportion of these groups among the general population is low, then designing a targeted survey with a sampling frame based on the geographic locations of any forcibly displaced persons identified in existing data sources is more desirable (EGRISS, 2018). If refugee and refugee-like populations can be identified in a census or survey, then this data can be used as a sampling frame for prospective surveys targeting these populations. Censuses and surveys might be complemented by additional questionnaires designed specifically for persons in refugee camps, reception centres and informal settlements. However, the sampling frame provided by censuses can become quickly outdated during humanitarian emergencies, therefore the use of spatial information from satellite imagery and remote sensing for sampling purposes in the contexts of new camps and settlements should also be explored. 

It might, however, be more effective to use refugee and asylum seeker application registers compiled by national authorities or UNHCR as a sampling frame for surveys targeting these populations as opposed to using population censuses. These registers provide more information about the exact status of forcibly displaced persons in the country as they follow the different steps of an application for asylum/international protection. However, appropriate databases may not exist in contexts with rapidly evolving migration patterns and the quality of registration may be poor in those same countries (EGRISS, 2018). Furthermore, it is important to bear in mind that the definitions employed by NSOs or other national compilers of data and statistics may differ from international standards, for example, in some countries refugees or IDPs may be counted as economic migrants.

Refugee and refugee-like populations are often under sampled in censuses and surveys due to language barriers in completing questionnaires and a tendency to avoid contact with authorities/a suspicion about the motives of data collection. This can be overcome by using enumerators and interpreters with appropriate language skills, ensuring that questionnaires can be translated into other languages depending on migration patterns and contacting refugee representatives to explain the purpose of the survey and reassure respondents (EGRISS, 2018).

For IDPs, the IRIS differentiates between three sub-groups: those in locations of displacement, those in locations of return, and those in other settlement locations (EGRISS, 2020). There are four conditions for being counted among the IDP population.

• First, the persons should have been forced or obliged to move from their habitual place of residence by a triggering event, including, but not limited to, armed conflict, generalized violence, violations of human rights, natural or human-made disasters, or other forced evictions or displacements. Preventative displacements also count (i.e., if the person moved in anticipation of and thus before the actual event). Trigger events are difficult to verify in an objective manner but should be based on the IDP’s subjective perception based on self-declaration. Individuals who moved for economic or domestic circumstances should not be included, but these circumstances may be a result of slow-onset disasters. Recognition of the linkages between economic motivations and conflict, disasters or other factors affecting livelihood opportunities is key to ensuring that IDPs are not “mislabelled” as economic migrants. In order to minimize uncertainty around the overarching cause of displacement, statistical bodies should identify areas of the country that have been affected by slow-onset disasters such as droughts, rising
• water levels and climate-change induced loss of livelihoods.
• Second, displaced persons must have been a usual resident (bearing in mind the criteria for establishing residence) in the location within the country where the event occurred. The usual place of residence corresponds to the place where the person has been living or intends to live for a period of 12 months at the time of data collection. The habitual place of residence is the place of residence at the time of their initial displacement.
• Third, the displaced person must flee from the location.  Experiencing harm or loss of property without fleeing the location does not qualify them as an IDP. The distance a person flees however, is irrelevant and can refer to movement within the same administrative unit. 
• Finally, the person must be within the internationally recognized borders of the country. However, this categorization changes with time. An individual who displaces within the country and later moves abroad is still considered an IDP for up to 12 months. Once their time abroad reaches 12 months or more, they are no longer counted among the IDP population.

Nationality is not a defining characteristic of IDP statistics, as this population may include citizens, foreigners or stateless persons, but might still be useful for understanding the situation of IDPs. Family members of households that are displaced but who remain behind are also included among the IDP population. 

Figure 2: Persons included among the stocks and flows of IDPs

Although persons returning after seeking international protection may share characteristics with IDPs, they are not counted among IDPs but among people in a refugee-like situation (even if they return to the different location within their origin country). However, persons returning after seeking international protection can be counted among IDPs if they are displaced again as a result of a new event following their return.

In addition to collecting data on IDPs, the IRIS also urges collecting data on IDP-related populations, or the those born to IDPs after displacement occurred, as they are relevant for policy and programmatic reasons (EGRISS, 2020).

Also relevant for policy and programmatic reasons are non-displaced family members of IDPs and persons who have overcome key displacement vulnerabilities. While they should not be included among IDP statistics, IRIS suggests collecting data on these populations as well. The IRIS also provides the first authoritative criteria for determining the end of displacement. The Inter-Agency Standing Committee Framework on Durable Solutions for IDPs specifies that displacement ends when “IDPs no longer have any specific assistance and protection needs that are linked to their displacement and can enjoy their human rights without discrimination on account of their displacement (IASC, 2010: 5).” The framework outlines eight criteria that should be used to measure and determine progress towards durable solutions, namely, safety and security, adequate standard of living, access to livelihoods, restoration of housing, land and property, access to documentation, family reunification, participation in public affairs and access to effective remedies and justice. The measurement of these eight criteria is complex and varies significantly across organizations and countries. A general recommendation for determining whether these criteria have been met and the person can be subsequently removed from the IDP population is by comparing them to the general population on the same eight criteria. EGRISS is currently developing a module of questions to be included in a muti-topic household survey, such as the Living Standards Measurement Study (LSMS) and Demographic and Health Survey (DHS), aimed at assessing whether or not IDPs have overcome key displacement-related vulnerabilities.

In addition to collecting data on stocks of IDPs, divided into the three location sub-groups detailed above, it is also desirable to collect data on:

• the inflow of IDPs (people who become displaced);
• the outflow of IDPs (persons whose displacement has ended because they found a durable solution, emigrated more than 12 months ago or passed away); 
• the stocks of persons who have overcome IDP-related vulnerabilities;
• the age, gender and disability of IDPs and persons who have overcome IDP-related vulnerabilities;
• the origin locations from where IDPs and persons who have overcome IDP-related vulnerabilities were displaced, in order to better design policies and programmes addressing the root causes of displacement or facilitating the return and reintegration of IDPs.

Optimizing the use of censuses and surveys

The IRRS extensively describes the advantages and limitations of different data sources for statistics of refugee and refugee-like populations, most of which also applies to IDPs. When it comes to censuses, the IRIS recommends adding a question on reasons for migration with “forced displacement (IDPs, refugees, asylum seekers, temporary protection, etc.)” as an answer category. However, adding a question on reasons for migration in addition to those measuring internal migration doesn’t necessarily capture the whole picture because the last move of an IDP is not always the most relevant one. Therefore, it is recommended that a specific question on forced displacement is added, i.e., “have you ever been forced or obliged to flee (potentially adding reason)”, with follow-up questions on the timing, origin and destination of those who answered affirmatively. Persons living in IDP camps, reception centres, temporary structures and collective accommodation should be covered in all census enumerations. Furthermore, the questions on forced displacement should be asked of all household members.

When it comes to household surveys, the involvement of NSOs can ensure the survey meets national and international data quality, data privacy and data protection standards. Sampling frames for surveys where information on IDPs is needed should include camps, reception centres, informal settlements and collective accommodation. The distribution of the target population may change rapidly in emergencies, especially for highly mobile populations such as IDPs. Sampling frames can be improved by using displacement tracking tools and satellite imagery (EGRISS, 2020).

Operational data are collected by a range of governmental and non-governmental organizations to inform decision-making, programming and interventions in humanitarian contexts (EGRISS, 2020). Although operational data may not meet the same scientific standards as official statistics, they are collected on a more regular basis and are likely to be the only available data on displacement in humanitarian contexts or in countries with a less developed national statistical system. However, operational data might be a potential source for official statistics on IDPs, especially if an assessment of the quality of the data is carried out by the NSO or other relevant statistical entities.

Population movement tracking systems, such as IOM’s DTM, are often the main source of operational data, especially in emergency situations characterized by widespread displacement. Surveys referred to as “needs assessments” are another type of data collection employed by humanitarian agencies. These surveys may be used to study the situation of IDP populations at a particular point in time or in comparison to the situation of the host population.  Finally, registration or enrolment systems used by humanitarian organization in the delivery of aid and assistance to affected populations, such as the World Food Programme’s SCOPE platform, may be a useful source of data on the situation of IDPs.