Skip to content
Putting refugees in camps

Emerging research has demonstrated how, for social, economic, environmental and health reasons, the consequences of placing refugees in camps are often negative, not only for the refugees themselves but also frequently for the national populations and governments of receiving states. It is not the purpose of this article to reiterate these arguments. Rather, the aim is to take one step back from this debate and ask first: why have camps so often been preferred by both host governments and international assistance organisations? This would appear to be an essential step for those wishing to push policy in a different direction and translate our knowledge of the negative consequences of 'encampment’ into practical action.

A first essential point is to question what is meant by a 'camp’. There are various elements that constitute a standard view of a 'camp’, some of which might be regarded as more important than others in leading to adverse effects for refugees. For example, the most obvious example of 'camps’ described in this issue is that of the Saharawi refugee camps in Algeria. These are, effectively, tented cities, supplied wholly from the outside. In contrast, a wider definition is used by Edith Bowles in her article about the Thai- Burma border. She uses the word 'camp’ to describe both 'small, open settlements where the refugee communities have been able to maintain a village atmosphere’ and 'larger, more crowded camps’ where they are 'more dependent on assistance’. It is the latter that are more clearly recognisable as 'camps’, distinguished from what might be regarded as 'settlements’ or even 'villages’ by their size, density, dependence on external aid, and the level of control exercised over inhabitants by national or international authorities.

Each of these features of the 'camp’ has been cited in different reports as the main element that makes 'encampment’ a bad idea for refugees. But these classic elements of camps are not the only aspects of settlement policy that are problematic. For example, Tarig Misbah Yousif describes what might more accurately be considered an 'agricultural settlement’ rather than a camp. Yet, here too, there are clear practical and theoretical objections to this particular form of settlement policy. Such objections are backed up by Kibreab’s review(1) of over 100 such agricultural settlements across Africa, which is hardly more inspiring than the growing literature on 'camps’. Indeed, agricultural settlements that purport to reduce dependence and increase self-reliance can in certain circumstances be just as constraining and 'overcrowded’ (in relation to land resources) as camps. Nor are camps only used as a solution for refugees, as Stephanie Kleine-Ahlbrandt’s article on IDPs in Rwanda demonstrates.

In discussing the negative consequences of 'encampment’ of refugees, it is both common and quite important to be specific about the elements of 'camp life’ that are most problematic. For Bowles, it is the increased dependence of camp dwellers that is perhaps of most concern. For Yousif, while it is partly the association of 'camp-based settlements’ with overcrowding, also important is an apparent link with the withdrawal of international aid. For scholars concerned with environmental issues, it is again population densities that are often seen as important(2), although my own work has stressed social and organisational issues, especially the extent to which refugees are cut off from local populations(3). But if this specificity is important in examining the consequences of encampment, it is no less important in explaining its causes. At different times, and in different places, different elements of encampment may be more or less relevant in influencing settlement policy decisions. Indeed, justifications for a policy of encampment are rarely made in general terms by policy makers; rather, specific reasons are more often cited to justify its 'exceptional’ use in particular circumstances. Awareness of this specificity is crucial if a case is to be launched in any particular place to influence policy, or indeed if a general case against camps is to be seen to be watertight.

Why, then, do governments and international agencies so often seem to prefer the option of encampment? Some interesting clues are provided by an exchange in the medical journal The Lancet, prompted by van Damme’s observations of the negative consequences for human health of encampment policies in Zaire in 1994(4). For example, writing from the medical charity Medécins sans Frontières, van der Borght and Philips(5) agree that there is an 'all too common link between refugee ill-health and camps’. But they then go on to point out the 'practical advantages from the standpoints of service delivery, accountability, identification of individuals, physical access, cost effectiveness of the relief operations and monitoring of both the refugees’ status and received assistance’. They argue that refugees outside camps may become excluded from assistance, while warning against 'wishful thinking of easy integration’ outside the camp setting. Meanwhile, writing on behalf of the Programme and Technical Support Section of UNHCR, the section responsible for settlement planning, Dualeh(6) again agrees that 'refugees do not belong in camps’. Yet he then suggests that 'in Africa and elsewhere, huge influxes of refugees can completely overwhelm the capacity of the local population and the infrastructure, wreaking havoc on the environment and socioeconomic system.’ As a result, according to Dualeh, 'local authorities sometimes have no choice but to seek separate sites for the refugees’.

These two short contributions are characteristic of much of agency thinking on the problem of camps, and the detail of their arguments deserves to be dealt with, not least because both responses accept that, in principle, refugees should not be housed in camps. A first point concerns the perceived advantages of having refugees in camps from the point of view of efficiency and transparency of aid delivery. But the experience of camps set up for Rwandan refugees in Zaire and Tanzania from 1994-96 is hardly encouraging in terms of the ability of international agencies to identify individuals and ensure accountable distribution of aid. Some of the camps, at least, proved effectively to be 'no-go’ areas for international agencies beyond the distribution points for food delivery. Indeed, according to some agency accounts, they tended to function as zones in which those responsible for the Rwandan genocide were able to continue to intimidate camp populations and divert aid to military and para-military personnel. Physical access was hardly maximised either, with serious problems caused in Zaire as a result of the terrain. Numerous agency reports talk of vehicles being unable to leave roads, for fear of becoming bogged down. While this might be seen as independent of the question of whether refugees were in camps or not, it could be argued that the separation of refugees from local populations exacerbates their marginalisation onto the poorest quality, and probably least accessible, land. What is certainly true is that refugees were not placed in camps in Tanzania or Zaire in order to ensure their accessibility.

A second point raised above concerns the perceived negative consequences of allowing refugees to settle outside camps. But concerns about refugees’ 'exclusion’ from international aid are again perhaps a little misplaced in this context since surely, if refugees were free to settle where they wished, this could include settlement in areas concentrated around points of aid delivery. More serious is the question of relations between refugees and host populations – the notion, as van der Borght and Philips put it, that 'easy integration’ is 'wishful thinking’. Yet in some respects, the conflation of 'settlement outside camps’ with 'local integration’ is perhaps at the root of this lack of certainty about non-camp solutions. For example, just because refugees are not housed in camps, it does not mean that they can or should be 'integrated’ with local populations in every sense of the word. In the case of Guinea, cited by van Damme, there was an opportunity to integrate assistance to refugees and local populations where this was feasible and cost-effective – as in the case of health provision and joint use of educational facilities(7). But, socio-economically and politically, the two populations maintained a clear and distinct identity. Nor is it right to conflate 'settlement outside camps’ with 'unofficial’ settlement that is uncontrolled or uncontrollable. In countries where dispersed settlement has been achieved, from Guinea and Côte d’Ivoire to Senegal, Uganda, Sudan and Malawi, this has seldom been 'spontaneous’, involving at the very least negotiations between refugee and local leaders. More often than not, government and international agencies are involved too, with the latter especially adopting a facilitating and supportive role, helping to ensure that the refugees are not simply turned away.

This latter point is important, since it holds the key also to where any arguments against settlement in camps ought to be directed. To a certain extent, it is the international assistance agencies, and especially UNHCR, that have borne the brunt of criticism about the establishment of camps. These agencies are seen as favouring a policy that either helps them to carry out their mandate to assist refugees or, alternatively, strengthens their control of camp populations and accountability to donors, depending on one’s point of view. As Barbara Harrell-Bond puts it, 'Refugee policy in the South [is] largely driven by the demands of donors and humanitarian organisations’. However, donors and international agencies can easily deflect this criticism by pointing out – quite rightly in a technical sense – that it is not their job to decide where refugees will settle. Either refugees settle 'spontaneously’ outside official camps or they are settled in sites chosen or sanctioned by the government. In neither case do UNHCR or other agencies formally decide where they should go. The fact that camps appear in situations of internal displacement shows that it is not simply a problem of UNHCR.

By shifting the responsibility to host government authorities, however, international agencies can and do side-step the 'political’ issue of whether a policy of settlement in camps is a good one or not, and are able to confine their attention to more technical matters of camp layout, infrastructure and organisation.

Such a position is of course a caricature: many officers of UNHCR and other international agencies at both field and headquarters level would recognise their quasi-political role in encouraging governments to adopt certain policies. Some at least have used that role to argue against encampment. Nonetheless, it highlights how general arguments against settlement in camps might be better put to governments, who have ultimate responsibility for settlement policy, rather than to international organisations. This in turn requires us to be aware of host governments’ main reasons for preferring settlement in camps. For them, issues such as accessibility, efficiency and transparency of aid delivery are likely to be rather less important than potential conflict between refugees and locals; but uppermost on their minds is likely to be the political and security implications of the pattern of refugee settlement. No government will be happy with the dispersed settlement of refugees in a border area if it makes that area vulnerable to attack by parties to the conflict from which the refugees have fled in the first place. Nor is it reasonable to expect governments to condone the use of their territories by military groups, allied to the refugees, for launching attacks on neighbouring countries – although obviously their attitude will depend much on higher level relations between the two states. Indeed, in both such cases, resettlement of refugees (whether in camps or not) may also be in the interest of refugee protection.

If the issue of settlement policy is put in this context, the immediate policy question becomes less one of what is best for the refugees and more one of what is in the interest of the security of both refugee and host populations. In turn, to avoid there being a trade-off between security (in camps) and refugee well- being (in the absence of camps), those opposing encampment would need to demonstrate that camps do not necessarily enhance the security situation or reduce conflict between refugees and locals. Importantly, it is possible to make an argument along these lines, even if not in an absolute or universally applicable manner. Clearly there will be circumstances in which the security threat posed by certain groups of refugees does require effective measures of control. But, it could be argued that, in normal circumstances, control – and especially undue restrictions on refugees’ activities within camps – increases rather than decreases the security threat that is posed by refugees. For example, those working with refugees from Somalia have commonly portrayed Somalis as uncooperative and unruly populations, where the safety of humanitarian aid workers is often put into question. But, as Waldron and Hasci have pointed out(8), this lack of co-operation has much to do with the way Somalis have been forced into camps.

The point is that, while it is entirely legitimate for governments to seek to exercise control over refugee populations, there is also good evidence that encampment does not provide a particularly effective means of control. Indeed, especially in conditions where the state has a limited capacity, such control may be much better exercised by local (traditional) authorities than by international organisations with little direct experience of the area. An example is the case of rules on access to natural resources such as firewood. Here, the difference is striking between frequent non-compliance with agency-generated rules, which are seen by refugees as lacking legitimacy, and general compliance with rules established by local custom, tradition or edict(9). It is important not to romanticise about local capacity to control or regulate behaviour in conditions of forced displacement but, at the same time, it is reasonable to recognise that some capacity usually exists, and to seek to strengthen this, rather than impose a military-style camp regime.

Finally, to say that arguments against encampment might be better addressed to governments should not be seen as absolving international organisations or downplaying their role. In addition to the role that they can and do play in pressing governments to adopt strategies that ensure refugee protection, both implementing agencies and UNHCR continue to have a primary responsibility for a range of settlement issues, regardless of whether settlement is in camps or not. Some at least of the problematic aspects of camps highlighted by research are amenable to improvement through better planning. Site plans can limit settlement densities, while efficient provision of infrastructure such as water points, clinics and facilities for waste disposal can go a long way towards making life more 'sustainable’ at higher densities. Recent work by Tara Rao(10) has shown how better settlement design can reduce or eliminate features that have a negative impact on the situation of women.

Perhaps most important, whatever the settlement type, both governments and international agencies can promote policies that engage refugees and local populations in project design and implementation. Like the growing consensus against camps, such a participatory approach is increasingly part of the rhetoric of international agencies. Perhaps in both cases, the task for research is less to provide further evidence of the appropriateness of the policy but to help put that policy into practice.


Dr Richard Black is Lecturer in Human Geography in the School of African and Asian Studies, and co-Director of the Sussex Centre for Migration Research, both at the University of Sussex.



  1. Kibreab G 'Local settlements in Africa: a misconceived option?’, Journal of Refugee Studies, 2(4), pp 468-490, 1990.
  2. Jacobsen K 'Refugees’ environmental impact: the effect of patterns of settlement’, Journal of Refugee Studies, 10(1), pp 19-36, 1997.
  3. Black R and Sessay M 'Forced migration, land-use change and political economy in the forest region of Guinea’, African Affairs, 96(385), pp 587-605, 1997.
  4. Van Damme W 'Do refugees belong in camps? Experiences from Goma and Guinea’, The Lancet, 346, pp 360-362, 1995.
  5. Van der Borght S and Philips M 'Do refugees belong in camps?’, The Lancet, 346, pp 907-908, 1995.
  6. Dualeh M 'Do refugees belong in camps?’, The Lancet, 346, pp 1369-1370, 1995.
  7. Classes, however, remained separate, not least because the medium of instruction was French for Guineans, and English for the refugees.
  8. Waldron S and Hasci N Somali Refugees in the Horn of Africa. State of the Art Literature Review Studies on emergencies and disaster relief, Report no 3, Nordiska Afrikainstitutet, Uppsala, Sweden, 1995.
  9. Examples can be found in Kenya, Tanzania, Zambia, Guinea and Côte d’Ivoire. A review is provided by Richard Black, forthcoming, Refugees, Environment and Development, Longman, Harlow, UK.
  10. Rao T 'An unsettling settlement: the physical planning of refugee settlements: a gender perspective’, MA thesis, 1997, School of Development Studies, University of East Anglia, Norwich, UK.
This site is registered on as a development site. Switch to a production site key to remove this banner.