Information-sharing and Coordination among NGOs
Working in the Refugee Camps of Ngara and Kibondo, Tanzania, 1994 – 1998(a)

Nan Buzard(b)
March 2000

Information sharing is one of the most basic activities of coordination. At its most rudimentary level, it allows organizations to know about one another’s activities. At higher levels, it may enable strategic planning and enhance multi-agency efficiency and program delivery. How non governmental organizations (NGOs) share information in refugee camp settings, and how that information is transmitted throughout NGO agency networks, United Nations (UN) bodies, and host governments, shape and determine the quality of emergency response.

The goal of this research project is to provide a descriptive and critical analysis of the collection and coordination of information regarding situational issues and refugee needs among the NGO communities working in the Ngara and Kibondo regions of Tanzania from 1994 through the summer of 1998. Although they were addressing very different types of refugee populations, these camps are often looked at together because they are in the same theatre of operation.(1)

This report provides suggestions for improving the normative standards of information flow, which, in turn, could lead to better services for refugees. It examines how sharing (or not) affects the efficacy of refugee services (such as camp management, food distribution, health services, vector control, shelter provision, water and sanitation, community services, self-reliance projects, and so forth). The report also explores the implications for communication when infrastructure and geography produce logistical hurdles.

Other issues, incidental to the research, but also bearing on information-sharing include:

• the obstacles logistics pose for information-sharing—for example, in the rainy season, it can take three hours for an NGO to reach a camp;
• questions regarding the management and control of technology access;
• neutral, technical data may obscure the political roots of an emergency;
• tensions between the local population and the refugees regarding the availability of services;
• tensions between the refugee population with professional skills and the rural, uneducated local population recommended by the host government for employment with NGOs;
• short-range institutional responses (for example, small agricultural plots, temporary hospital and schools buildings) to long-range problems;
• geo-political pressures in the Great Lakes region that produce recurring migration patterns.

Although not addressed in this paper, all these issues add to the complexity of providing aid and contribute to potentially ineffective coordination between agencies. This study also did not attempt to analyze the expenditures or any other financial data of various NGOs that may explain the different outcomes with regard to coordination and information-sharing.

There have been many arguments made against refugee camps;(2) the report does not take a stand on this issue, but it does identify some problems that arise in camp situations. Research in Tanzania for this report was conducted in both old-style camps (such as Kanembwa in Kibondo, where the refugee population had access to large plots of land and where refugees have generally replicated their home villages), as well as in larger, less-organized camps (such as Benaco, a large, spontaneous camp in Ngara).

II. Hypothesis 

Communication and information-sharing among NGOs is better (more robust, open, and detailed) during the crisis phase of a complex humanitarian emergency than during the care-and-maintenance phase. Regardless of the phase of an emergency (crisis or care-and-maintenance), refugee services in camps are highly inter-related, have mutual impacts, and require specific communication protocols to ensure efficient coordination of activities.(3)

Information was collected on the years 1994 to 1998 through extensive interviews with NGO staff, UN staff, and local officials of the Government of Tanzania. In addition, interviews were conducted with Red Cross and United Nations High Commissioner for Refugees (UNHCR) staff in Geneva, as well as humanitarian professionals in London. Secondary information, including situation reports, minutes of meetings, NGO internal memos, and UN mission briefs were read in the field. Direct observation of NGO activities in the Ngara and Kibondo camps was conducted over the course of five weeks during the summer of 1998.

The research sought to identify common weaknesses and strengths of agencies relative to sharing information. The following sections describe those features that exacerbate poor information-sharing, and those features that promote superior communications. The report also addresses the effects of information-sharing on the coordination of camp services. "Ultimately, my analyses are based in conviction derived from consistency, plausibility and rationality, rather than any formal measures of aggregated statistics."(4)

In the words of John Ryles,

This report may well join the many others that gather dust while pleading for better communications and coordination in emergency response. Part of the problem is that NGOs have not, until very recently, made a concerted effort to develop specific information-sharing protocols. Instead, agencies have relied on their policy units and technical research departments, without a fuller understanding of the need for more comprehensive analysis and the importance of developing and implementing high-quality communication mechanisms.(6)

There are various reasons—tired litanies—that explain the difficulties NGOs have in establishing good information-sharing practices. One reason is the short-term nature of most NGO funding; another is that international staff members often have only a rudimentary knowledge of the places in which they are working. Unfortunately, little academic research filters down to the field, even to the country director level. Moreover, the attention of aid agencies to specific areas is sporadic, and within agencies the high turnover of staff encourages institutional amnesia. Finally, there are often severe logistical and security constraints on information gathering in the field.(7) All these factors contribute to poor information-sharing practices.  

Part of the remedy lies in changing the culture of humanitarian response in such a way that information-sharing is valued. Only then can good communication practices be incorporated into emergency response. As an example of a change in the culture and values of the humanitarian community, the awareness and inclusion of programs that address the distinctive needs of women and children have, by now, become part of standard humanitarian response.  

Overall, it appears that the intensity, quality, and effectiveness of information-sharing were superior during the crisis phase of the emergency in Tanzania, particularly in Ngara. This is not to suggest that during the chaos of the crisis communication and coordination ran smoothly. However, there were measurable differences in resources and visibility between the crisis and the care-and-maintenance phases of the relief effort. In Ngara, in the late spring of 1994, substantial resources—including highly competent and mature field personnel—were generated by a confluence of factors. The intense media coverage exposed the international community’s guilt for its inaction during the Rwandan genocide and also graphically reported the great suffering in the Great Lakes region. In addition, the scale of the refugee influx into Ngara at that time (until over-shadowed by events in Goma a month later) also generated substantial attention. In Ngara, UNHCR, with the full support of the host government, was able to establish a highly selective pool of aid agencies to work in the camps. Initially, those agencies also sent their best people, those who could claim a piece of the pie. After an initial burst of attention to Ngara, but still during the crisis phase, the media left for Goma, thus providing a less competitive, "plant your flag" environment for the agencies in Ngara.

Qualitatively, relationships among NGO staff appeared to be more robust and open during the crisis, with numerous and substantive interactions at both the formal and informal levels. For example, in addition to daily meetings in the camps, all NGOs attended a weekly 3 to 4 hour meeting in Ngara town that has been described as informative, effective, and fun.(8) In contrast, cynicism, exhaustion, and the stressful competition for resources among NGOs in the latter stages of the emergency reduced transparency and appeared to restrict coordination of refugee services. By the summer of 1998 for example, the NGOs working in Lukole camp (the last remaining camp), had little interaction with UNHCR, apart from desultory weekly camp meetings in which each NGO presented a monologue and UNHCR field staff berated the agencies for minor problems.

Ultimately, NGO information flow must be examined in relation to UNHCR because the latter either controls the information itself or has the authority to shape response, at least within refugee-related operations. Whatever the particular nature of the relationship with UNHCR, no NGO can escape the need to interact with the agency, which confirms the need for better relations and communications between UNHCR and NGOs.

Key findings concern two particular issues in the information-sharing arena:

• the quality (and size and management) of coordination meetings themselves and the dissemination of the information shared at the meetings;  
• the effect of poor communication and coordination on highly linked camp services.  

Meetings 

Information provides the basis for program design and resource allocation and influences policy-making. There are many types of information (sectoral, numerical, epidemiological, nutritional, gender-based, security, and so forth) and various ways to share such information (such as camp meetings, regional interagency meetings, reports, minutes, and informal conversations in a bar). Although all modes of information-sharing were evident in the NGO community during both the crisis and the care-and-maintenance phases of the Ngara emergency, there were substantial differences in the frequency and quality of interactions.

Because meetings are the key mechanism for information-sharing and coordination, this research focused on their structure and content. Agency staff members disagreed about the relative merits of brief, focused meetings versus longer, more complex discussions; therefore any meeting had the potential to frustrate participants who preferred the other structure. From the reports read and interviews conducted, it appears that the meetings during the crisis phase were popular and efficient. By contrast, in later stages, meetings meandered and were full of details that were not meaningful or helpful to others. In both phases, the lack of rigor or lack of standards in the way information was recorded made determining most baseline parameters difficult.

Many of the international NGOs working in Ngara had worked together previously (either in Tanzania with the Burundian refugees or in the Dadaab camps in Kenya), which appears to be a central factor in the high level of cooperation achieved. There was willingness on the part of agencies in Ngara during the crisis to work collaboratively and not strictly within agency boundaries. Such inter-agency collaboration is usually made difficult by competition, roles, ego, and headquarter pressures. The leadership and quality of the UNHCR team also seemed to be an inspiration to the NGOs. UNHCR-led meetings were informative, encouraging, straightforward, and effective in distributing tasks to the most appropriate NGO. Many NGOs felt that the door was open for discussion and that their opinions and skills were valued. A World Food Program (WFP) staff member recalls highly coordinated discussion regarding rations among WFP, the Tanzania Christian Refugee Service (TCRS)—the primary agency for food distribution—and other NGO service providers (health and community services).(9)

NGOs living together at " K9,"(10) the compound located half way between Ngara town and the camps, exhibited a particularly strong sense of camaraderie, developed in part through social gatherings at the compound’s Hard Rock Benaco, a bar with satellite games and the site of many late night discussions. The sense of mission, the cordial relations with UNHCR, and UNHCR’s unprecedented ability to commit funds on the ground (instead of the usual one to three month-long approval process), generated openness and goodwill among the agencies. Overall, NGOs felt that there were good conditions for communication (well run, daily meetings, and living in close quarters). Another factor contributed to the good work environment: UNHCR dictated the number of agencies allowed to work in the camps, because they had, at that time, the support and trust of the host government. Of course, for the agencies that were excluded as implementing partners, communications were less positive.

What happened in the early phase of the emergency indicates that good communication at the onset can mitigate cultural and work-style conflicts. True, when UNHCR moved into a new compound (and out of poor accommodations that were similar to those of the NGOs), some NGOs perceived the agency as becoming exclusive and less available; yet this move did not precipitate the breakdown in communications that followed in the care and maintenance phase. Unfortunately, the level of cooperation that UNHCR had achieved with international NGOs was not attained with the local population. Instead, various misunderstandings, broken promises, and style differences led to resentment and mistrust between the UN and the host population.(11) This history contributed to the uneasy relationship that persisted between UNHCR and the Tanzanian Ministry of Home Affairs (MHA).

One reason information may be better shared in an emergency is that the needs are more clearly evident. As UNHCR representative Maureen Connolly said, "Move… away from immediate emergency response, and the more complicated the management of the operation becomes—as other priorities are defined."(12) Once a camp population has stabilized, there is room for more complexity in terms of services and management.

A combination of donor fatigue, a seemingly intractable refugee situation, discomfort about providing assistance to potential militia, and severe financial cutbacks characterized the Ngara camps in late 1998. Issues that should have been addressed and solved years before were still open—refugee women continued to suffer from sexual assault as they searched for firewood—and discussions regarding potential remedial actions continued with no resolution in sight. Part of the problem was that UNHCR staff experienced feelings of failure and guilt regarding the 1996 forced repatriation of Rwandans from the Ngara camps. One result was that the staff now had lowered protection standards. In turn, this situation created a breach of trust with NGO partners, because there was very little effort to explain or to involve these partners in the refugees’ return. To be precise, many NGOs were uncomfortable about UNHCR’s apparent collusion with the Tanzanian government’s December 1996 forced return of Rwandan refugees.

In Ngara as of late 1998, UNHCR was conducting 1 to 2 hour camp management meetings every two weeks. UNHCR perceived these meetings as a forum to discuss problems (and they did discuss linked services such as water quality and availability and related health problems). However, meetings seemed like a chore—there was too much technical detail (numbers of people trained in carpentry, attendance in schools, number of latrine roofs installed) and too little genuine interaction. Most NGO staff felt that UNHCR staff were more invested in the trappings of authority (chairing meetings and acting inaccessible) and holding onto their jobs than they were in serving the refugee population.

According to interviewees, there was much less informal interaction in 1998 than in 1994. By 1998, the staff of only two international NGOs lived at the "K9" compound. District-wide, there were many fewer agencies working in the camps, and these seemed to be consumed with financial survival and battling with UNHCR. National staff members were responsible for larger service areas, and, for the most part, national and international staff members did not spend much social time with one another. Considering how relevant a factor informal communications were during the crisis phase, it would be interesting to have quantitative data on how many such interactions occurred. Unfortunately, during this research it was not possible to measure systematically the volume and quality of social interaction among staff.

Kibondo did not follow the same trajectory as Ngara, in which care and maintenance followed a large crisis. Rather, during the crisis phase in Ngara, Kibondo was struggling along with the consequences of the less well-documented Burundian refugee crisis that had occurred in the autumn of 1993. In 1993 and early 1994, most refugee support came from local church and community groups, especially as the local community and the refugees shared a common language and culture, and UNHCR was not present. About 80 percent of this refugee population (about 85,000) returned home to Burundi within six months. The government of Tanzania then invited UNHCR to establish camps at sites (pre-selected by the government) as contingencies for future influxes. At that time, there were no international NGOs working in the district except TCRS, which was already conducting development work. Because of its presence, TCRS was asked to develop the camps (construct roads, clear and mark plots, and so forth). From 1994 to 1996, the two camps in the Kibondo district were considered to be providing care-and-maintenance for the remaining refugee population. A few other NGOs had joined TCRS in providing camp services, but Kibondo did not have the high profile or the resources of Ngara.

During 1995, meetings were held fortnightly or weekly, and the few agencies on the ground worked closely and well together. The UNHCR staff consisted of one international member (as compared with 40 in Ngara). During 1996, refugees fleeing continuing violence in Burundi began to fill up the Kibondo camps. Meanwhile, in December 1996, after the forced repatriation of the Rwandans from the Ngara camps, international NGOs came south to Kibondo seeking work. Also in December, the war in the Congo was generating increased instability throughout the region, and an additional 40,000 refugees crossed over the Tanzanian border into Kibondo. UNHCR increased its international staff to four and its national staff from 3 to 29. Meetings were held twice a day—at 7am and 7pm—to discuss issues and coordinate activities. For the most part, the NGOs cooperated well together, brainstorming ideas regarding malaria control, among other things, and gathering in the evenings at their shared quarters in town.

As the NGOs working in the Kibondo camps settled in for the long haul, they each built separate compounds. Although understandable from a comfort and security point of view, these compounds created a closed-off environment and increased the NGOs’ separation from one another, from the local population, and from the refugees.

Although the NGOs were supposed to meet weekly, UNHCR called meetings irregularly. The fact that meetings were usually organized by UNHCR was considered a problem, because if that agency did not convene them, they did not take place. In the camps, all service providers, refugee community leaders, and representatives of the MHA (Ministry of Home Affairs) attended the meetings. At the district level, a malaria task force met regularly during the malaria crisis of 1997. This task force was convened by the International Rescue Committee (IRC), which coordinated the four agencies involved in malaria prevention and care. This group met apart from UNHCR because of the feeling that meetings with the UN resulted only in "all talk and no action". For example, the UN staff spent considerable time debating about whether malaria was indeed present, and if so, how it was manifesting itself, while the NGOs felt that the real question centered on the practical costs and benefits of spraying. Even when UNHCR accepted the plan to spray, they wanted to hold a four-day workshop for sprayers, while NGOs argued that a workshop of that magnitude was unnecessary for the rudimentary technical demands of spraying.(13) This type of conflict seemed the norm during this researcher’s visit.

By 1997, monthly interagency meetings were taking place in Kigoma, the regional office to which the sub-offices of Ngara and Kibondo report. This half-day meeting was the most regular and comprehensive and was attended by the entire staff of sector specialists and coordinators of UNHCR, as well as by most NGO directors and team leaders. Unfortunately, because of transportation difficulties, not all NGOs attended regularly. In the view of some NGOs, these meetings seemed, overall, to perform a "social" function; although issues were discussed and data on mortality rates examined, clear conclusions, concrete actions, and deadlines were rarely forthcoming.(14) However, the validity of these claims is cast in some doubt by this researcher’s observations of the minutes of these meetings, which were of fairly high quality—comprehensive in substance but not overly detailed. The best minutes for the purpose of general information are not swamped in detail, but lay out the general concerns and actions by agency. These minutes circulate as monthly interagency memos and inform staff working in the region about other NGO initiatives and issues, as well as UNHCR activities.

Still, the general feeling among NGOs was that the meetings lacked in-depth discussion and decision-making. For example, interagency minutes from April to June 1997 reflected that the under-5 mortality rate had "reached alarming figures…"; yet, exactly a year later, the same mortality rates were observed. What explains this? It could indicate that experience acquired over previous years was not adequately utilized, or that even with institutional memory, the lack of resources made response difficult. Every year there is a rainy season (though El Niño brought heavier than usual rainfall in 1998), and every year there are increases in malaria and anemia, particularly in the under-five population. In any case, UNHCR’s perceived inability to translate this knowledge into preventative action created tension with NGOs.

Unfortunately, some NGOs also perceived the UN staff as cliquish and more interested in promoting and protecting one another than in addressing refugee problems. The region-wide negative feelings about the UN gave support to the observation that during an emergency, highly competent UN staff are deployed, but then, bit by bit, politics and favoritism enter, bringing down the caliber and capacities of the staff. In defense of the 1998 UNHCR staff, it should be said that multi-year posts in rural areas are isolating—although Tanzania is officially an accompanied post, there are no schools, thus limiting opportunities for families—and that the 1998 funding crisis had created an obsessive concern with the uncertainty of future employment.

There were many medical research initiatives and reports being conducted on malaria control, cholera prophylaxis, and underweight newborns (for example by the U.S. Centers for Disease Control (CDC), inter-governmental health ministries, the World Health Organization (WHO), and independent researchers). However, there was no central depository for and certainly no organized distribution of information on these efforts to the field. UNHCR’s regional medical coordinator, who ran medical coordination meetings every two weeks, apparently did little to pass information among health providers in the various camps. Further, there seemed to be very little interaction between IRC (the health provider for the four Kibondo camps) and the health providers in the other camps—the International Federation of the Red Cross in Kigoma and Norwegian Peoples Aid in Ngara. Obviously, this was in part the result of busy schedules and the difficult travel conditions, but also NGOs and UN agencies were not in the habit of explaining themselves or their actions to one another. All these actors—not to mention the refugees—could have benefited from participating in technical meetings and sharing best practices.

The subject of linking camp services generated many opinions and debates with little distinction between the crisis and the care and maintenance phases of an emergency. Services are always linked, regardless of the phase of the emergency, but when there are more resources available—a common characteristic of the crisis phase—linkages can sometimes be less crucial to service delivery. For example, if water distribution was low during the emergency phase in Benaco, there was money for trucking. During the care and maintenance phase, however, if water was low, refugees experienced cutbacks, and their health was adversely affected. In general, however, service delivery depends heavily upon linkages; where linkages are weak, questions of jurisdiction and responsibility arise. For example, there has been ambiguity about who is responsible for rising mortality rates in camps—the sanitation team or the health services providers.

An ongoing debate among aid providers centers on the pros and cons of vertical versus horizontal camp services. According to the former model, one agency performs all services in one camp; according to the latter, one agency provides the same services (such as health care or food distribution) in all camps in a district or region. There has been no final verdict as to which approach is preferable, but the need for better understandings of the relationship among services is abundantly clear to all agencies.

In the refugee camps in Tanzania, there were numerous examples of linkages and equally numerous examples of the tensions that arise when NGOs do not fully understand the implications of those linkages for service delivery. The following list outlines some of the linkages and the difficulties that arise when roles and responsibilities regarding services are not clear:

• Unaccompanied minors. This is an issue, for example, when more than one NGO is running the orphanages, as there is a need for one interactive database.
• Dead bodies. There is potential tension between the NGO responsible for public health and death certificates (which may take time to issue), and the NGO in charge of camp sanitation, which may be under pressure to remove bodies quickly.
• Underweight newborns. The NGO responsible for hospital and delivery services may be held responsible for underweight babies—although the cause may be the malnutrition status of the mothers, which in turn may be caused by inadequate food rations supplied by another NGO or WFP.
• Malaria control. The NGO in charge of shelter and the design of the refugee’s hut may not be communicating with the community services NGO responsible for design of mosquito nets, resulting in the nets not fitting the houses properly. Also, community services must interface properly with vendors and suppliers so that colored, not white nets are manufactured. (White nets tend to be washed more frequently, reducing the impregnated net’s effectiveness.) Frequently, community services and education are the last services to arrive in an emergency and the first to be discarded during a funding crisis—yet both services can play an important role in malaria prevention.
• Gender violence. The NGO that runs a gender violence program may come into conflict with the NGO that provides camp management, when the former advocates the provision of rations directly to women, while camp management prefers to distribute food to "community leaders."
• Skill training. If one NGO is providing skill training based on a self-reliance program (frequently referred to as community services), and another is providing micro-enterprise services, the two agencies may be in competition for projects. A camp can support only so many tailors and so many barbers before economic frustration undermines the services.
• Alternative energy. The NGO involved in community services needs to interface with the NGO that provides camp management. There is a need to integrate social with technical services or else the latter is prone to failure. For example, to reduce dependency on firewood, the camp management NGO introduced solar ovens and soaking beans—but the firewood was used predominantly for making beer, and soaking beans was a process foreign to the refugees. In other cases food was distributed, but there had not been prior distribution of cooking pots.
• Shelter construction. The NGO constructing tents and huts may provide structures that do not assure sufficient ventilation for cooking fires inside the huts, and the health NGO will find an increased incidence of respiratory diseases following the rainy season.
• Tsetse fly and other vector control programs. In one situation, before departing a camp, the NGO in charge of sanitation never handed over the tsetse fly eradication traps to the regional agricultural and livestock development office of the Tanzanian government, and now there is an increase in infection.
• Micro-enterprise programs. An NGO running a seedling project through a micro-enterprise program is constrained by a lack of water that is supplied and controlled by another NGO—while the faucets at the bore holes run open for hours at a time.
• Sanitation and health. These two services are continually and profoundly linked. Typhoid, cholera, and other water-borne diseases come from inadequate or poorly treated water. At worst, NGOs begin blaming one another when things go wrong. In the Tanzania camps, The International Federation of the Red Cross and Red Crescent Societies (IFRC) insists on conducting health, sanitation, and camp management (security, roads, planning of settlement), because these services are so inter-dependent.

A good example of linkages occurred during a malaria outbreak in Kibondo. The Dutch Relief and Rehabilitation Agency (DRA) (community services), IRC (health), TCRS (camp management), and Oxfam (vector control) all worked together to shape a comprehensive and complimentary malaria control program. Generally, however, this author observed little technical sharing among NGOs—each considers its problem unique and wants to solve it itself.

The NGOs’ preoccupation with uniqueness has roots in the stiff competition for funding. The general sentiment is, "if we pass around our technology and program expertise, we may lose ground and not be as competitive with donors." There is very little incentive to share, except to promote efficacy on the ground—a big value but not one that has proven to generate funds or attention.

Related to funding proposals is a concern with funds already spent. These "sunk costs" are additional obstacles to sharing, even if there is a genuine desire to share and coordinate. Most NGOs in Tanzania have poured enormous resources into compounds and now feel the additional pressure to make good on these investments. They need to stay in the compounds, therefore they need to maintain ongoing programs and services. Agencies also tend to hold onto their programs for psychological reasons. In an isolated environment your program is what you have to identify with and where you put your passion, and thus there is a tendency for NGO staff to hold onto their "babies".

During the course of research, this author read many reports, newsletters, and position papers regarding humanitarian aid. Much of what was observed, and the suggestions based upon these observations, is a reiteration of concerns already voiced throughout the NGO and UN communities. Many of these "best practices" are geared to diminishing the purported "uniqueness" of every emergency and to identifying cross-cutting issues. Best practices relate first to key findings, and second to general considerations.

• Clarification of the purpose of the meetings. There is a need to clarify the purpose of each meeting—is it for open discussion of ideas, or to make decisions, or to create action plans, or for general sharing of lessons learned? Is it a policy-generating group or an operationalizing group? Is it discussing what cargo should be given priority on flights or beneficiary participation?
• Management of the meetings. Rotating the chairing of meetings allows for capacity building and diversity. Meetings need to be run with respect for all participants. When appropriate, chairs should give meeting time over to real discussion and negotiation regarding issues.
• Attendance at meetings. There is a need for NGOs to develop and maintain continuity and consistency in their internal understanding and external promotion of each agency’s mission and role. The team leader or country director may go to an interagency meeting, while the sector-head may be the NGO representative at camp level meetings. This can be a good and appropriate separation of duties, as long as these representatives coordinate with one another. There is also a need to bring refugees into meetings to create stronger partnerships.
• Effective recording and distribution of minutes. Minutes are quite variable in quality—sometimes they are not taken, or poorly recorded, or not distributed, or not read. Better dissemination of memos and situation reports might encourage agencies to discuss technical and procedural issues with one another. Written reports can also play a crucial role in evaluation of projects.

• Balance between NGOs. Larger NGOs usually have a greater capacity to handle services, but smaller agencies can add specialized services (for example, HelpAge provides services for the elderly, something in which most NGOs do not specialize). Camp services require NGOs that have established reputations and the capacity to perform. On the other hand, there needs to be opportunity for new agencies that offer innovative services.
• Make explicit linkages between services. There is a need to establish a clear chain of command and a clear delineation of roles and responsibilities as camp services are awarded to various NGOs. UNHCR technical staff, or NGO working groups, could play a larger role in establishing the groundwork and an environment in which linkages would be understood, acknowledged, and respected.

First, and perhaps foremost, is the need for preparedness. This does not refer to pre-positioned food stocks or technical equipment, but to preparing the way information will be shared and coordinated. For good collaboration to take place, it needs to begin at the planning stage. NGOs can and should thrash out and adopt standards for operating in the field—before they get there. Once in the field, cooperation would be enhanced if NGOs and UNHCR had fuller understandings of their respective roles and responsibilities. Better introductions between agencies could help create a context for cooperation and more accurate expectations. There is some difficulty in standardizing operations between agencies at the headquarters level. (There are still too few models to follow, although initiatives like the Sphere Project provide concrete examples of interagency cooperation.) (15) However, agencies could pre-establish certain parameters, for example, the salary scales for national staff. There is enough experience now to set salary scales for all national (and refugee) staff members so that NGOs (and UNHCR) do not waste time squabbling in the field. Secondment could be worked out in advance between agencies to encourage better coordination.

Strategic advantage:

Collection and dissemination of information is not meaningful if it is not standardized, recorded, and presented in ways accessible to all the parties involved. UNHCR technical sectoral experts could provide leadership in this undertaking and could facilitate communication between NGOs. UNHCR is currently developing a database that analyzes the strengths, weaknesses, and capacities of implementing partners, so that their comparative advantages can be considered before selection. UNHCR should participate in clarifying NGO roles but not stifle creativity and innovation. The selection of NGOs should be based on their comparative advantages and complimentary programs—not on favoritism or political influence.

Documenting best practices:

There is a wealth of field reports located in the Programme and Technical Support Services (PTSS) at UNHCR in Geneva. Although at present these reports are not easily accessible or even known about by most NGOs, they are an important vehicle for institutional memory after technical experts leave the camps.

Aid agencies have varying definitions and expectations regarding the role of an information officer. Depending on the agency, the information officer could be responsible for public relations, or internal communications, or for liaison to UNHCR. Agencies need to understand what they require from information. Information-sharing mechanisms will work only when there is a strong belief and understanding that they are valuable.

Overall, very few NGOs are oriented toward sharing information. In Nairobi, the NGOs rarely coordinated their activities regarding logistical practices—even as they jockeyed for air cargo space to the Sudan, or earlier, to Goma. By working together, NGOs can stimulate discussion regarding the solution to specific logistical problems, as they did in the malaria response in Kibondo.

At the start, this research was concerned with the role of personality in the way work unfolded in the field, as it is common to read and hear accounts of field work that attribute the quality of the work entirely to the personalities involved. Concern about the potentially disproportionate role of individuals in these contexts drove the search for systems that might override, or at least neutralize, personality. While it is true that protocols and mechanisms for information-sharing are critical, this author also recognizes that people are the most crucial link, and that while frameworks for coordination can be vastly improved, it is staff who will need to understand, appreciate, and implement them for far-reaching efficacy.

Unfortunately, in the current political and financial climate, humanitarian agencies are likely to be operating in a world of diminishing resources. Although the competition for funds will probably work against NGO cooperation, increased accountability may encourage NGOs (as well as UN agencies) to share best practices. Especially in light of diminished funds, it is imperative to develop incentives for agencies to cooperate and to link services; it may also be salutary to establish penalties for failure to do so. Furthermore, NGOs should work to reduce staff turnover and to develop policy that reflects the real pressures of operations. These changes could create conditions for humanitarian culture and practice that would better serve refugees and other beneficiaries.