Background Paper for DFID Migration Team March 2005
Ann Whitehead1 and Iman Hashim2
Globally, high rates of adult migration in various forms means that the numbers of children affected by all kinds of migration world-wide is truly staggering. This paper is a preliminary effort to foreground the issue of children in DFID’s work on migration, through a review of the available literature and a discussion of the implications of their findings. The report concentrates only on children left behind and independent child migrants; a child being defined as anyone between the ages of 0 and 17 and ‘independent child migrants’ meaning any child who migrates separately from their parents.
The report finds that estimating the number of individuals and families that have migrated and/or are affected by migration is extremely difficult due to a range of methodological problems, including: the paucity and poor quality of the data sources; countries differing widely in the extent to which they document migration; figures not being collected in any standardised way and the illegal and/or undocumented nature of some movement. These problems are further compounded in the case of estimating the numbers of children affected, because: children often do not appear separately from adults in the statistics; reliable national level data about the incidence of all kinds of child migration are exceedingly rare; where these are available categories are used in different ways by different surveys and writers, and there is a lack of consensus regarding terms and their operationalisation. Additionally, the focus on particular types of child migrants – such as trafficked children – invisibilises child migrants who do not fall into these categories such that there are few estimates as to how many other children are moving and for what reasons. However, there does appear to be a general agreement in the literature that there is a substantial increase in the numbers of children leaving their home communities independently of their families.
The difficulties in estimating the numbers of child migrations makes it difficult also to assess the impact of migration on children’s health, mortality and education. In terms of children left behind, few studies tackle directly the implications that migration as a livelihood strategy may have for the well-being of children. Instead we have to infer the effects on children from studies of the effect on the household as a whole. However, we find that it is difficult to generalise about whether the impact on children left behind with parental labour migration is positive or negative. Whether the outcomes are positive or negative in strictly resource terms depends on a large number of factors, of which a key one is the extent to which the work that migrants are doing is well rewarded in local terms. Where fathers are migrants, and wives and children are cared for by other senior males so that they do not lose their social place in the community, then some of the potential detrimental effects may not emerge. Support from relatives may, however, come at a price. Where the rates of male absenteeism are very high and prolonged, fragmented family structures may result in profound effects on well being and on children because children have no father to bestow recognition on them. On the other hand, remittances by absent fathers can, in some instances, ensure food security, help repay debts and cover the costs of schooling and illness of relatives left behind. What little research has been done on the effects on children of mothers’ migrating suggests that this might have more adverse effects, particularly psycho-social ones, although there is growing discussion of scattered evidence that remittance behaviour differs between men and women migrants, with women being more concerned about children’s well being.
A major problem with assessing the harmful effects of independent child migration is the very large differences in the tenor and direction of the findings that come from the different kinds of literature. What research is available suggests that the effects of migration are likely to be context-specific, given the wide variety of migration flows involved. There is little direct material on the meaning and social context of children’s movement for work, or for other reasons. For example, the research on the kind of work independent child migrants do is really quite thin, and is skewed to those who work in the most harmful and abusive situations. Clearly the different kinds of migration have a profound effect on children. The medium term effects will depend, interalia, on what has been the trigger for migration, what kind of living situation they secure in their places of destination, whether they work or go to school, what kind of work they do, what kind of social support is in place for them, and whether they fall prey to the many hazards and dangers posed by intermediaries, bad employers, or bad working conditions, and so on.
A key issue in the literature that looks at the factors that affect children’s independent migration is the ambivalent treatment of the relative roles of the parent and the child in decision-making; the emphasis tends, instead, to be on the degree of compulsion or coercion from parents. As a result, a central motivation for children to migrate that is generally underplayed is their need or desire for income. Migrating for education is another insufficiently stressed aspect of children’s migration in a number of areas. What few sources there are suggest that the link between education and migration is also context specific, some finding that there is a statistical link between not going to school and the propensity of rural children to migrate to work, while others find that migration is clearly positively associated with access to education.
Areas where there does appear to be agreement in the literature are that; a good deal of contemporary independent migration is of children from districts where there are and have been high rates of adult migration; and that children rarely travel and seek work alone, and often work on the basis of deferred payment and/or payment in kind.
When considering the legislation relevant to children migrants, the report indicates that no international nor regional legislative frameworks exist that deal directly with this issue, although there a number that are directly or indirectly relevant to children’s accompanied and unaccompanied, forced or voluntary movement; most significantly those that deal with children’s welfare in general and those that are related to the protection of children from economic exploitation and harmful work. However, these protective measures are difficult to put into place in many developing countries. Many countries have enabling legislation, but it is rarely acted upon, often because of a lack of resources, and because measures put in place ostensibly to protect children, can also have negative effects. The report considers this in detail by discussing different assessments of the role of intermediaries in independent child migration; where some sources stress that these systems are open to abuse and may force children into conditions of servitude, while others suggest they may be highly protective.
The report concludes with a discussion of the policy implication of the findings of the literature review. We argue that the consequences of migration for children needs to have a much higher profile in policy discussions, although the contours of this concern will remain obscure until more research is available on all forms of migration flow that affect children, including those of independent child migrants. A key finding, though, is that the policy space to make recommendations with respect to independent child migrants is very narrow. It is squeezed by the international conventions and protocols which are key elements of child protection policies. It is also squeezed because the success of advocacy with respect to particularly abused and vulnerable children (bonded child labour, ‘street’ children, ‘trafficked’ children, etc.) has lead to this being a potential good source of development funding for national governments. International advocacy has focused much needed attention on exploited and abused child migrants, but has also made it difficult to address the very real needs of other child migrants.
The report highlights a number of policy areas for discussion. 1) The impact of poverty and underdevelopment, and the need for programmes that alleviate the regional and rural poverty that trigger high levels of adult and child migration. 2) As much child migration is migration for work, the need for open and sensitive national and regional debates to establish what is locally acceptable and unacceptable child labour, and mobilise discussions about young people’s working conditions and rights, as well as of the causes of child work and migration. 3) The need for systems of support and recourse to be built for all workingchildren in hardship, not simply those who have been trafficked. 4) The need to stress that education is a universal right for all children, regardless of work status and/or migrant status, and to institute measures to allow working/migrant children to access school, non-formal education and/or training. 5) The need to reassess international definitions of trafficking and the dominance of this category in the debates about and the interventions around children’s migration, in the light of studies indicating they increase child migrants’ risk of harm and exploitation, by ignoring the cultural context of migration and the institutional capacity to implement them.
Children and Migration
This paper is a preliminary effort to foreground the issue of children in DFID’s work on migration. The numbers of children3 affected by contemporary migration flows world-wide is very high, since they can be affected as children left behind, when either father, mother or both parents migrate; as children in families that have migrated, and when they migrate themselves independently of their families. Rates of adult migration are currently high and growing, and this is associated with a number of processes that are affecting many countries. These include urbanisation, in which particular forms of employment are spatially concentrated; diversification of livelihoods, in which migration is one set of diversification options; globalisation, which has created new forms of international divisions of labour that produce areas and countries of huge labour demand; conflict and environmental stress, which displace populations and produce refuges and internally displaced persons; and, finally, high rates of HIV/AIDS, which produce fragmented households incapable of maintaining rural livelihoods, and whose members move to cities and towns.
Rates of migration of persons under the age of 18, the internationally recognised age at which children become adults, are also growing, although we have few reliable estimates of the contemporary migration flows of children. As with adults, these are likely to be extremely varied in terms of who goes, to where, why, for how long, and with what effects. Given the high rates of all these forms of migration, the numbers of children affected world-wide is truly staggering.
The typology overleaf lays out the key categories of children affected by migration, with some broad indication of the regions in which particular kinds of migration flow predominate. It also gives a rough indication of the kinds of issues that have predominated in the literature. The report does not consider the effects of migration on the very many children world-wide that have experienced migration as young dependants when their parents have moved. It does not deal with forced migration, or family re-unification. It deals mainly with migration for work and focuses on two particular categories - the children left behind when one or both of their parents migrate and independent child migrants.4
Typology 1: Categories of Children Affected by Migration
Main Research Themes Relevant to Children
Migrating as Family Members
National Migration and Regional Migration to Urban and Rural Areas
S and S.E. Asia; Africa; Latin America; USA; Europe
Health/educational benefits and disadvantages
Children as Family Labour
Poor and hazardous conditions;
Children seek informal employment
At risk on the street
Poor and hazardous conditions
International Migration Asylum Seekers/ Refugees/ Economic Migrants
Estimating the numbers of children who belong to each category world-wide is highly problematic. There are four key factors here: 1) estimating the number of international migrants is intrinsically difficult; 2) there are many variations in the way in which national migration rates are arrived at; 3) routine national statistics do not identify households with one or more members away on migration; 4) these problems are compounded in the case of children where variable age definitions, lack of attention and problems of method all serve to make migrant children statistically invisible. Many sources do not disaggregate by age (see for example some of the better estimates of certain kinds of international migrants, summarised in Box 2).
However, sources that do pay attention to age are often unclear about the definition of age used to specify who a child is and their relation to international conventions; which generally define children as those below the aged of 18. It is often not clear in published estimates of numbers whether this has been the age cut-off used to define children. It is also not uncommon to distinguish between children up to the ages of 14, 15 and 16 from those in the older category of 16 to 18. This is particularly the case with respect to children’s work, where, in any case, ILO guidelines give different cut-off points at which children should be allowed to engage in work, which are then reflected in country legislation (see section 5).
The issue of numbers is also obscured because several different kinds of independent child migration are the subject of highly active advocacy lobbies. Children who are refugees, who live or work on the ‘street’, or who have been ‘trafficked’, are the subject of specialist literature which routinely estimates the numbers of children affected. Some of the limitations of these numbers that make comparisons difficult are discussed below. Importantly, as well, singling out particular kinds of child migrants for policy attention invisibilises child migrants who do not fall into these categories. This further means that there are few estimates as to how many other children are moving, for what reasons and, of course, the relation between hazardous child migration and more benign forms.
This report first considers some of the difficulties of establishing overall numbers of migrants in greater detail (section 2) and then goes on to discuss the children affected by migration as children left behind (section 3), including here a discussion of problems in estimating how many are affected and independent child migrants (section 4), along with a discussion of estimates of the numbers involved. Section 4 reviews what we know about the motives for and effects of migration for children who move without their families. It also discusses the differences in perspective and interpretation apparent from different kinds of sources, using the example of accounts of girls' domestic work. This section also discusses the differences in perspective and interpretation apparent from different kinds of sources, using the example of accounts of girls’ domestic work. In Section 5 we turn to look more closely at the link between some child protection issues in the developing world and child migration. We review the relevant international legislation and then focus on current anti-trafficking measures and their appropriateness in specific contexts. Section 6 of the report is a concluding discussion, containing recommendations.
2. ESTIMATING MIGRANT NUMBERS
Estimating the number of individuals and families that have migrated and/or are affected by migration is extremely difficult. The variety of migration flows, intrinsic complexities in identifying who are migrants and the illegal or undocumented nature of some movements, all contribute to the paucity and poor quality of the data sources (Black 2004, Hayase 2003). Countries differ widely in the extent to which they document internal migration and the figures are not collected in any standardised way. Box 1 identifies some of the difficulties of using national census data.
Box 1 Estimating Migration: Problems with National Census
Censuses are held at long intervals and rather irregularly in many developing countries. National estimates of the number of migrants are unreliable and difficult to compare. Some countries still do not include direct migration related questions in their censuses (for examples see Bell, Rees and Wilson 2003, Black 2004). Censuses are particularly likely to miss or underestimate certain types of migration, such as seasonal, circulatory or temporary migration (Edmonds 2003, Srivastava 2003).
Additionally, there are no standard practices with respect to the collection of migration data in national censuses. The most commonly asked questions are: 1) place of birth, 2) place of permanent residence (1, 5 or 10) years ago and 3) last permanent place of residence before coming to present place. Using a change in residence definition gives a figure for population mobility, but does not indicate reasons for migration and cannot distinguish worker mobility from other types of mobility, such as moving to marry (Srivastava 2003).
The raw census data can be made to yield very different results according to the definitions applied. Skeldon illustrates how migration rates data based on national censuses are affected by the spatial unit used to define migration. He points out that the 1981 Indian census counted only individuals who were living in a different state than previously as migrants. This produced a percentage of 3.6% of the total Indian born population as migrants. “When the much smaller district is used the proportion becomes 29.1% which suggests a completely different level of migration” (Skeldon 1987: 1076). The kinds of variation in questions asked to collect the raw data and then the variables and indicators produced from national censuses are surveyed in Bell, Rees and Wilson (2003), who attempt to provide a global inventory of internal migration data.
Estimates of the numbers of regional international migrants involved in circular and seasonal labour migration across national borders are very poor, especially where these are well-established historical trends (as, for example, in West Africa or South Asia) so that migration is relatively easy and unremarked upon. Substantial numbers can be involved. For example, about one third of the population of Côte d’Ivoire are migrants, the majority of them from nearby West African countries (SCF Canada 2003: 5).
Countries with high rates of transnational international out-migration rarely collect comprehensive data about rates. In Bangladesh, for example, the Bureau of Manpower Employment and Training has estimates for the numbers of documented out-migrants to the Middle East, but does not have figures for migrants to industrialised countries (Siddiqui 2003: 4). They estimate that the number on short term contracts to the Middle East is more than 3 million, with many more travelling irregularly. Developed industrialised countries do make estimates of in-migrants and Siddiqui suggests that these number 1.1 millions. This is about 4 or 5% of the population, but the number of households affected is likely to be around 20%.
Box 2 Estimating International Migration: Problems with Other Sources
There are several problems with data sources often used for estimates of international migration: These include:
available data being focused on particular areas of concern, such as the “brain drain” in Africa, with the result that little attention is paid to other types of migration or their implications for poor migrants’ livelihoods (Black 2004).
data often being based on imputation or proxies of the numbers of foreign born; in particular, data on citizenship are used in the absence of data on place of birth (UN 2002)..
responsibility for the formulation, implementation and evaluation of migration data often being diffused among Government bodies, as well as among international organizations (UN 2002, IOM 2001).
as a result data are frequently not recorded according to the same categories and formats (UN 2002).
figures frequently do not include undocumented cross-border migration or illegal migration; for example, data on the Senegalese diaspora shows that over half a million Senegalese live outside Senegal, at least half of which are not officially registered in their country of destination (Diatta and Mbow 1999).
Many receiving countries do estimate rates of international in-migration, but collecting this data is often the remit of agencies that focus on particular kinds of migration to produce estimates of their target populations. Box 2 lists some of the problems with these data sources. There are also many international bodies charged with collecting data. These figures reflect specific areas of policy concern and often arise out of international protocols, and international and national legislative definitions. Some sources for these kinds of figures are illustrated in Box 3.