INTRODUCTION

Aim of the Programme Notes
The present Programme Notes are meant to serve as a running commentary to the agenda of the present consultation and to flag issues that might be relevant for the discussions under the different programme items. They are findings and reflections that have emerged from the preparatory research prior to the consultation. They are not exhaustive and mainly address the question of inter-agency collaboration on field staff security, which is the focus of the consultation, - a limited area within the larger, complex picture of field security.

Background
It may be seen as a sign of the times that the UN Secretary-General in his recent report on the safety and security of United Nations personnel has found it necessary to address the need for guidelines on how to arrange appropriate memorial services and how to respond in a caring manner to families of staff members killed in the service of the United Nations¹. Non-UN agencies working in the frontlines of humanitarian action are faced with the same needs, being confronted with the same threats to their staff and loss of human lives.

Since the end of the Cold War and the subsequent global, as well as local, change of power relations, a change in the character of violent conflict has been noticeable. The emergence of "failed states" and parties to violent conflict that have no knowledge and/or respect for international agreements, are just some of the characteristics of the present humanitarian scene. The prime victims of these developments are the populations to whose needs humanitarian actors try to respond. But these actors themselves are increasingly victims to violent conflict.

The number of incidents involving death or severe maltreatment of UN and non-UN humanitarian workers has risen alarmingly in recent years. According to UNSECOORD statistics, 198 UN staff were killed between January 1, 1992 and September 18, 2000². The total number of NGO humanitarian workers killed or maltreated is unknown. At the same time, it has become evident that these incidents increasingly hit all humanitarian workers indiscriminately, irrespective of flags or emblems. As a consequence of these developments, staff security has become an urgent concern for all humanitarian agencies.

At the level of individual organizations, be they UN, Red Cross/Red Crescent movement, NGOs or other humanitarian actors that are operational in the field, security management and preparedness, as well as the necessary costs for security, are receiving still greater attention in terms of allocations of human and material resources, training, and general awareness of the importance of security. Judging from available sources, such as security manuals and policies and interviews with humanitarian field workers, the emphasis seems to be on the analysis, strategies and security techniques within the organizations ³. While these sources will express understanding for the need for agency inter-dependence in matters of security, support for inter-agency security collaboration, and concern for the lack thereof, they rarely address the issue in any detail.

Agency security collaboration in the field has happened over the years in many different forms, most of them informal and of a context-specific nature, some effective, some not. They present a wealth of experiences on possible mechanisms for practical field collaboration.

Attempts to institutionalize field security arrangements between UN or UN agencies and non-UN actors have had varying success. A formal, binding Memorandum of Understanding between the UN and NGOs from 1996/97 ⁴ has not been signed at the HQ level by any NGO. The reasons given are mainly that it is not based on a consultative, partnership approach, that it is mainly concerned with legal/liability issues, and that it does not respond to the practical needs of field workers in security matters⁵ .

Another relevant example of inter-agency collaboration, though limited to one UN agencyand its partners, is the UNHCR-NGO Framework Agreement for Operational Partnership. While general in its scope, it has a component of security collaboration. The positive reception which this agreement has received has been attributed to the consultative approach and the spirit of partnership in which the collaborative efforts leading to the framework agreement were conducted.

The need for greater staff security for humanitarian workers, especially at the field level, the realizations that security is costly an labour intensive, and the various formal and informal experiences of trying to meet the challenges of field security in collaborative ways, are among the factors that prompted the IASC Working Group to initiate a broad consultative process on inter-agency field security collaboration. A task force chaired by WFP was established at the IASC Working Group meeting in February 2000 "to explore mechanisms of collaboration, both formal and informal, on all relevant aspects of security systems affecting NGO and UN operations, and to make proposals to the IASC Working Group."

At its meeting in May 2000, the Task Force decided to begin the consultative process with two consultations, one in Geneva, and one in Washington DC. The report from these consultations will be submitted to the Task Force, whose recommendations will be submitted to the IASC Working Group meeting in February 2001 and will contribute to shaping the way forward.

One desired result of the present consultation in Geneva is that the consultative process will help to clear up misunderstandings about agency roles, responsibilities and accountability. The expected outcome of the consultative process is the eventual formulation of practical, informal mechanisms for inter-agency field security collaboration, consultation and partnership, which may serve to improve humanitarian field staff security.

DAY ONE - EVENING, DECEMBER 6, 2000
INTRODUCTIONS, PRESENTATIONS, EXPECTATIONS

An informal presentation will be held, which will outline the format and process of the consultation introduce expectations as to the outcome of the consultation.

DAY TWO - MORNING, DECEMBER 7, 2000
CHALLENGES AND OPPORTUNITIES FOR COLLABORATION ON FIELD SECURITY
AMONG HUMANITARIAN WORKERS

1. Strategic/technical and organizational issues

   During the first part of day two, the participants are invited to define the needs, challenges and opportunities for security collaboration at the field level based on lessons learnt from concrete cases of security incidents and contexts.

   The issues involved in security collaboration in the field may be divided into what may be termed strategic/technical and organisational themes. The former include the whole discussion on what constitutes security, questions of security strategies (e.g. the triangle of acceptance, protection, deterrence ⁶), context analysis, selection criteria for security managers and humanitarian workers, the sharing of human and material resources, security training and - not least - information exchange.

   These issues have been, and are increasingly, being dealt with in publications ⁷ and training courses and are known both to UN agencies and non-UN humanitarian actors. The organizational issues of collaboration have received less attention to date.

2. Focus on organizational issues

   The theme of the present consultation is primarily organisational, focusing on possible mechanisms for inter-agency collaboration on security in the field. Among the issues that have emerged, the following may illustrate some of the needs, challenges and opportunities that are at stake in inter-agency field collaboration.

1. Collaboration, partnership, leadership, control
   At its meeting on May 23, 2000, the IASC Security Task Force addressed the need for clarification of concepts such as partnership and leadership/control in the context of security collaboration. The minutes read, "The humanitarian community needs to move away from the concept of leadership in security matters towards a true partnership where resources, information and, what is most important, responsibility / accountability are shared."
   

2. Partnership, consultation, communication, and "in-communication"
   UN and non-UN humanitarian actors have different degrees of interaction, which influence their security collaboration. For example, the IOM has signed the 1996/97 MoU on security coordination with the UN. And, in a few cases⁸ , especially in particularly dangerous situations, NGOs have adopted a version of this MoU at the field level.

   Some have a partnership agreement like the UNHCR-NGO "Framework Agreement for Operational Partnership", which includes a section on security. Others have operational partnerships without the framework agreement but a formal or informal understanding about security collaboration.

   Agencies that for various reasons are not UN operational partners may still exchange security information with other humanitarian actors in a field situation in a kind of consultative relationship. Others again may communicate informally with other actors, while others again may not want to communicate, be it on security or on other issues. These latter may be international agencies, or perhaps local charities whose potential contribution to humanitarian efforts may not have been recognized and therefore may not have been invited to collaborate⁹ .
   

3. Collaboration: advantages and drawbacks
   The assumption underlying this consultation is that collaboration in security is an asset in humanitarian action in the field. The many advantages are obvious, as experience from numerous successful cases of field security collaboration shows. The sharing of human and material resources and of information and analysis, common training, and transparency of roles and obligations, all contribute to increasing staff security and protecting material assets and thereby to greater effectiveness and efficiency of humanitarian assistance.

   That collaboration in security matters (as in humanitarian assistance generally) is an asset is, however, not universally accepted. Some find it "a lesser evil" (perhaps imposed by donors), at least better than separate arrangements. Others would refrain from security (and other forms of) collaboration for context-specific reasons, for reasons of principle, or for a number of other reasons. It could be due to a unique mandate, as in the Red Cross/Red Crescent movement, or because an agency wanted to avoid identification by association. An NGO might, for example, not want to be identified with any, or with a given, military operation, as expressed by SCHR in its presentation to the IASC Working Group 14 November 2000. Commenting on the Brahimi report the paper states:

   "The ongoing debate among humanitarian agencies about using armed peace keeper escorts for humanitarian assistance highlights some of the dilemmas faced by independent agencies. There are also times when NGOs need to emphasise an image of impartiality for reasons of access and security, and are forced to distance themselves from the UN and its co-ordination mechanisms.¹⁰"

   Often these differences in the perception of drawbacks and advantages of collaboration relate to differences in the concept of security and security strategies, as seen in the discussion about acceptance, protection and deterrence ¹¹.

   While it is recognised that inter-agency collaboration may indeed at times be a security liability in certain situations - some agencies would say in all - it rarely appears in agency security management manuals as a potential security risk. Mostly, manuals would mention the necessity to collaborate on security, but would not go into details about the forms such collaboration might take, or what its consequences, positive or negative, might be ¹².

   One of the challenges to this consultation, and the consultative process it is part of, is to try to identify collaborative mechanisms that maximise the positive aspects of inter-agency security collaboration and minimise the negative aspects. Even when collaboration is not deemed desirable, some form of mechanism might be found that makes it possible to avoid agencies' separate security arrangements conflicting or creating security hazards to either the populations they assist or to the humanitarian workers that who assist them.
   

4. Individual agency identity versus perceived collective identity
   The very co-existence and social inter-dependence of humanitarian actors in the same place and addressing the same crisis, with or without collaboration, may at times constitute a security risk.

   Whether out of ignorance or out of lack of respect for emblems or flags, antagonists often do not seem to make any significant distinction between humanitarian actors. They may perceive them collectively, be it as allies of their opponents, as potential sources of enrichment, targets for revenge, or instruments to make a political statement.

   One important consequence of this collective perception relevant to security collaboration are the - positive or negative - effects on the whole humanitarian community in a given field situation of the behaviour or activities of its individual actors ¹³.

5. Impartiality - real and perceived
   The lack of recognition of agency individuality may also imply a disregard of an agency's claims of impartiality.

   Impartiality, as part of an agency's mission statement, may appear very differently in the perception of outsiders. For example, UN humanitarian agencies are, according to the UN Charter, impartial. But, being United Nations, they will necessarily work with the government (if there is one) and may therefore be perceived as partial in a conflict involving the Government. So may non-UN actors working in partnership with the UN.

   This "perception by association" may place considerable strain on collaborative relations and may at times make them appear as a security risk rather than a security asset, and make organizations refrain from joining collaboration mechanisms in the field.

   Some would therefore claim that impartiality in practice is impossible, and that it is more transparent as well as less of a security risk for all those collaborating on security to openly accept this as a fact. Others might claim that impartiality is possible, but that it would require a both practised and perceived non-alignment with any actor in a conflict, thus making, for example, security collaboration problematic.

6. Multitude of non-UN humanitarian actors
   The "overcrowding of humanitarian space" in the sense of the proliferation of non-UN actors is often used as an argument against realistic collaboration between UN and non-UN agencies.

   It might be helpful for a discussion on humanitarian inter-agency collaboration to make distinctions, both among the UN agencies and among the non-UN agencies. The term humanitarian NGOs is often, incorrectly, used to cover all non-UN actors, although one counts among these the inter-governmental organisation IOM, the Red Cross/Red Crescent movement, which has a unique humanitarian mandate, as well as NGOs properly speaking ¹⁴. Among the NGOs, only a relatively restricted number would have a serious interest in inter-agency collaboration, consultation or communication on security.

   It might therefore also be helpful to distinguish between the group of non-UN agencies that are linked with the UN security system, formally or informally, and humanitarian actors that have their own security structure, though they might still interact with other actors, not least in security matters, such as ICRC and NGOs like MSF.

   Outside these are the humanitarian "non-communicados" which for various reasons choose not to interrelate with the mainstream humanitarian network. Some call them Wild Cards, but again that hides the fact that many may have legitimate arguments for working alone. Others, such as local NGOs and charities or religious institutions, may never have been invited into the network or recognized as potential partners.

7. Differences in organisational mandates and cultures
   The term "NGO" is problematic in that it denotes only what these agencies are not (governmental). It does not reflect the fact that many, if not most, are organisations of civil society which have been created with a particular purpose or cause and a particular constituency. Not only do they have different mandates/missions, but also different governance cultures, and organizational policies and strategies, e.g. regarding practices on decision-making competence at the field level. For these objective reasons, many humanitarian NGOs would not see themselves, for example, as UN sub-contractors or channels of government funds but maintain that they are independent partners in their own right. When it comes to security collaboration these questions of organizational culture play a significant role. It is not that they "will not follow instructions…"; mostly there are valid reasons why.

   UN agencies likewise have different organizational and governance cultures, both between them and inside the agencies. For example, some have a more developed partnership culture than others, which is significant when it comes to formulating mechanisms of security collaboration with non-UN organizations.

   The mutual respect for differences in organizational "cultures" may have the consequence that security (and other forms of) collaboration between some agencies would have to be practised with certain mutually agreed reservations. This is different from maintaining organisational identity or independence in ways that may be perceived as unpredictability and/or lack of transparency.

8. UN and NGO myths and realities
   How UN and non-UN actors perceive each other, and how far those perceptions may be from reality, has been very helpfully illustrated by Koenraad van Brabant in his review of aid agency security practices¹⁵.

9. "Humanitarian imperative" and thresholds:
   At a recent meeting of the UNHCR staff, the questions of thresholds and threat indicators were formulated in this way: "The security threshold - When to withdraw? At which point do humanitarians cease to deliver aid to avoid being martyrs to the cause?"

   The Red Cross - NGO Code of Conduct, rule no. 1, reads:

   "The humanitarian imperative comes first". However, humanitarian workers increasingly ask: but at what price? ¹⁶ The questions of when to withdraw from, and when to re-enter, a crisis situation and what degree of risk to human and material resources is acceptable, are among the central issues in humanitarian inter-agency discussions on field security.

   These dilemmas seem to overshadow the fact that the different security thresholds between humanitarian actors may, at times, be an asset rather than a liability, and may also be seen as a practical division of labour. Being able and willing to go where the UN is not advised/allowed to go may also be seen as one of the comparative advantages of NGO implementing partners.

10. Interface between civil and armed actors
    Humanitarian actors have widely differing policies and practices regarding the question of collaboration with the military, e.g. armed escorts. The ICRC mandate, for example, only in extreme situations allows ICRC workers to accept armed escorts. Some will seek their protection in a humanitarian practice leading to "acceptance" by the community and the parties to a conflict. Others will, out of experience, avoid being perceived as associated with the military out of fear of being targeted. Even others will argue pragmatically that they are accountable towards both donors and recipients, and only arms can protect their human and material resources.

    ---

    DAY TWO - AFTERNOON, DECEMBER 7, 2000
    IDENTIFY ELEMENTS OF INFORMAL INTER-AGENCY COLLABORATION
    MECHANISMS ON FIELD SECURITY FOR HUMANITARIAN WORKERS

    1. Introduction

       On the basis of the morning's findings the participants are invited to go through the exercise of outlining the elements of collaborative field security mechanisms. Such outlines would respond to the needs and wishes identified during the morning's findings, in light of the realities of the field that may help or hinder collaboration - or might speak against security collaboration or certain forms of it altogether. The focus is still on humanitarian action in the field.

       It might be helpful for the exercise to compare existing models for security collaboration either of a general or of a specific type. Each field situation develops its own model, if only because humanitarian actors are together in the same space and addressing the same crisis. At one end of a continuum one might have the absence of any formal or formalized collaboration on security, which is also a model - by design, by default, or by negligence. Still there will be some form of interfase, if only through information exchange in social contexts or in the context of other forms of agency relations. At the other end of the continuum is the integrated, collaborative security model, with transparent roles and formalized agreements.

       The following examples may serve as a starting point for discussions, supplemented with relevant field experiences and texts of different existing mechanisms for security collaboration.

    2. UNSECOORD - one model of inter-agency security co-ordination

       The security framework of the UN family - excluding peacekeeping operations - is a centralized and uniform structure, consisting of Designated Officials (DOs) and Field Security Officers (FSOs) deployed in areas of crisis. Decisions up to a certain level of security measures can be de-centralized, but the most severe measures (evacuation) have to be referred to the Secretary-General.

       The UNSECOORD field security personnel profile is traditionally military people or police officers. The DOs will traditionally be Resident/Humanitarian Coordinators and may therefore not necessarily be from humanitarian UN agencies like UNHCR, UNICEF and WFP, though presently the majority of them are. Until recently they had not always received security training, but major efforts are now undertaken by UNSECOORD and the humanitarian UN agencies to provide all field personnel with security training.

       The humanitarian UN agencies have their own security officers and also second staff to UNSECOORD. The system is severely understaffed. The reasons are partly that security has been given a low priority and, therefore, has not received adequate funding when allocations were insufficient. Only staff and FSOs are core funded according to a negotiated cost-sharing arrangement among the UN agencies. The rest has, so far, been part of programmes and has had to compete for programme funds. The budgetary process to allocate security officers is time-consuming and inflexible. The procedure within the different UN agencies takes less time. However, their security officers have no formal UN system responsibility. The only fast way to deploy security staff, one security officer noted, has been through professional security companies.

       The drawbacks of the UNSECOORD system are mostly said to be its inflexibility and its centralized decision-making procedure. Added to this is its present lack of resources in people and material. Among its advantages are mentioned, for example, its transparency, its uniformity for all UN agencies and the discipline in security matters it entails. It is also seen as an advantage that decision-makers have a certain distance from the crises they monitor. Provided that they are well informed this distance could make them better suited to make objective assessments than field staff who, after long exposure to risk, may have become immune or complacent in the face of security threats.

       The security system of the UN was subject to in-depth internal review in late 1999 and early 2000. The Secretary-General's report of 18 October 2000, "Safety and Security of United Nations personnel ¹⁷" is the first step towards the implementation of the recommendations emerging from the review.

       Relevant to the field relations between UN and non-UN security collaboration is the Secretary-General's mention (section 16) of the shared security threat to all humanitarian actors working in the same environment. "Threats against non-governmental organization staff can also directly affect the United Nations humanitarian assistance programmes, especially since conflicting parties often do not distinguish between United Nations and non-governmental personnel." The implication of this statement is that threats against UN agencies can directly affect NGOs.

    3. The 1996/97 UN-NGO Memorandum of Understanding (UNSECOORD)

       This MoU was meant as a mechanism to extend the UN security umbrella to implementing partners, such as IOM and NGOs ¹⁸. It is a service oriented instrument focusing on practical concerns, such as guards and other security measures, evacuation, and cost-sharing.

       The only organization that has so far fully adopted the MoU is IOM. No NGOs have subscribed to it at the Headquarters level. In 14 cases, NGOs are known to have signed the MoU at the field level ¹⁹. The way in which UNSECOORD presented the MoU to the potential NGO parties to the agreement was described by those parties as a one-way, top-down process without prior consultation, incongruent with what is perceived as partnership relationships. The MoU includes, for example, provisions for the case of partner non-compliance, but none for situations where the UN does not live up to its part of the agreement ²⁰.

       The proposed participation in the cost of UNSECOORD security staff network and in evacuation costs was not found to be consistent with the fact that non-UN staff would not be covered by insurance, nor could they be sure to be evacuated in the same order of priority as UN staff. The gain for NGOs was seen to be too little compared, for example, with the obligation to follow UN security rulings and, thereby, perhaps go against their own judgement or agency mandate, i.e. their independence and their own interpretation of what humanitarian response would be required in any given situation.

       The field reality often proves to be different from whatever formal arrangements are or are not in place. Recent examples are the evacuations from Sierra Leone and West Timor. Thus, after the murder of the three UNHCR staff in West Timor, no distinction was made between signatories and others, or between local staff and international staff. This appears to be the norm more than the exception.

    4. Security as part of a UN-NGO "Framework Agreement"

       In the UNHCR-NGO "Framework agreement for operational partnership," (FAOP)²¹ the security elements have been integrated into the general partnership relations. (This is an internal arrangement that does not carry any responsibility within the official UN security system²² ). The following elements are part of the "pact" between collaborating humanitarian actors:

       • Shared recognition of the mutual dependence of all humanitarian workers at the field level for their security and communications strategy;
       • Recognition of the potentially adverse impact of the actions of one partner on another
       • Information sharing;
       • Full briefing on the security situation, the security plan, restrictions and other security measures in force;
       • Responsibility for the behavior and action of the partners' own staff;
       • Ensuring the compliance with agreed security procedures;
       • Access by NGO partners to a UN agency telecommunications equipment;
       • Access by NGO partners to a UN agency security training.

       The FAOP has been in existence for about one year and is being adhered to by both implementing partners and for example advocacy agencies, some at the HQ level, others at the field level. Some organizations have not been able to adhere to the whole document, e.g. the sections referring to the Red Cross-NGO Code of Conduct, but have then added to their signature to a Letter of Understanding indicating their reservations.

       The document is not legally binding, but is based on trust. However, break of trust may have subsequent consequences for any party, for example the non-extention of a contract, or the removal or re-deployment of a humanitarian worker. It may eventually also have consequences for (continued) donor support.

    5. Informal security arrangements as a function of inter-agency field relations

       Most humanitarian operations develop practical, local (= context specific) security arrangements between UN and non-UN actors which vary considerably from situation to situation. Some of them may have many features in common with the MoU model, but contrary to this will not be legally binding. In these local arrangements, security will often be seen not as an isolated phenomenon but as an integrated component of the overall structure of inter-agency relationships.

       It has sometimes been claimed that the NGOs want "a free ride". This seems to imply that they want the protection of the UN system in terms of security material and the protection of security staff (including evacuation with all its related costs), while still wanting to keep the independence of decision-making on when to leave, when to go back, and where to go.

       While this may be true in some situations, most other situations may be more accurately described as give-and-take. The majority of the major NGO humanitarian actors have their own independent security policies and strategies, and their staff is insured according to the risk category of the humanitarian environment. They will often be able to provide unique information and valid analysis related to the security context thanks to their presence in the field beyond the range of the UN (and therefore more exposed to risk).

       Rather than emphasizing the differences in risk assessments and risk taking between UN and NGOs, one might stress their complementarity. It might for example be seen as an asset, not least for the populations that need assistance, that some NGOs consider it a humanitarian imperative to remain even if the UN withdraws, provided that reasonable security precautions are taken. On the other hand, the UN, having normally (though not always) much larger operations than NGOs, will have the capacity to offer a security umbrella. Sometimes this is expected by the donors. Generally speaking, in dealing with authorities as part of their ongoing activities, UN agencies will often voice the concerns of the whole humanitarian community vis-à-vis these authorities.

    6. A case of security collaboration in the field:
       CODE OF CONDUCT for Humanitarian Agencies in Sierra Leone

       The case of Sierra Leone may serve as an example of concrete, country-specific agency collaboration with a security component. Over a period of about one year, a mixed committee worked out the text of what became the "Code of Conduct for Humanitarian Agencies in Sierra Leone" ²³.

       It spells out general behavioral and ethical principles for inter-agency humanitarian collaboration and establishes a Code of Conduct Committee which "will monitor implementation of the Code of Conduct and advocate on its behalf with the parties to the conflict." It has no formal powers. But "[S]hould any departure from the principles occur, a timely and full disclosure of details of the extenuating circumstances should be provided to the Code of Conduct Committee and the relevant humanitarian forums, as it may impact on members of the humanitarian community."

       The Code of Conduct has a section on security collaboration, which states:

       1.1 "SECURITY

       Humanitarian action in favour of civilian population should not be seen as interference or an unfriendly act, and therefore, humanitarian personnel and assets must be respected and protected at all times. Safety in the Sierra Leonean context depends on the image and integrity of the humanitarian community as a whole, the manner in which agencies conduct themselves in the field, and the consent of armed groups to the principle of humanitarian access.

       Agencies agree to abide by the following in order to ensure consistency and professionalism:

       1.1.1 Humanitarian operations must be carried out under a "no-gun" policy, hence the transportation of armed personnel by humanitarian vehicles if forbidden.

       1.1.2 Payments at checkpoints should not be made.

       1.1.3 No payment should be made in order to access or bring relief to civilian population.

       1.1.4 Demands from any armed group for relief items or assets should be refused.

       1.1.5 Demands for payment under armed threat should be treated as armed robbery and the incident reported to the proper authority for action to be taken.

       The safety of staff is self-evident pre-requisite. It is understood that in the event of direct personnel threat to agency staff in times of crisis, agencies may be required to take the necessary steps to ensure personal safety without consultation."

    7. NGO-NGO field security collaboration

       Networks, families and alliances of non-UN organizations often include security as a component of collaboration both at the field and at the HQ level. Van Brabant notes that "much of the collaboration [within networks] so far seems to run on informal understandings."

       Van Brabant has found that frequent complications arise in NGO-NGO security relations over "different thresholds of acceptable risk, over positioning oneself as 'neutral' or not and what that means in practice, over adopting a high or low profile, over speaking out about abuses etc." He asks whether there is not "a need to formalize basic principles of collaboration and of decision-making in case of serious differences of opinion related to safety and security management."²⁴

       NGO-NGO field collaboration, including on security concerns, may also be found in the form of field-specific NGO fora of varying coherence and competence, for example the "Reseau" in Burundi or the NGO Working Group in Rwanda. They would normally group all or most non-UN organizations, including for example MSF and the ICRC. They would typically receive regular security briefings by the Humanitarian Coordinator and/or Peace-keeping forces.

       These NGO fora are also used to exchange relevant information from their own daily field experience, often in remote areas where the UN or Peace-keepers are not present, and to develop interpretations and analyses of the security situation to serve in their individual organizational security management. The collaboration mechanisms of these fora are informal, voluntary, and often depend on the commitment of individuals. In some cases, ICVA has seconded personnel to serve such NGO fora.

    8. Specific areas of security collaboration: Training & research

       a. The major UN humanitarian agencies as well as UNSECOORD provide security training²⁵, some for UN staff, some (e.g. UNHCR) involving also NGOs.

       b. InterAction, an American network of humanitarian agencies, has served as a platform for inter-agency collaboration to establish a curriculum for security training and to pilot it. The UN system was consulted during the process. With government funding (OFDA, DFID and others), RedR (Registered Engineers for Disaster Relief) is now presenting the curriculum, which had been piloted in the US and the UK, around the world for two years.

       Field personnel and managers have been the primary target groups, but lately InterAction has facilitated the development of a security course for Executive Directors and piloted it. The aim was to incorporate security consciousness into the cultures of the member agencies at decision-making levels.

       c. The Humanitarian Safety and Protection Network, HSPN, hosted by VOICE, funded by the EU, has on a pilot basis attempted to set up a database for security incidents. The international humanitarian organizations meeting within this network recognize that the usefulness of the HSPN project, "especially as regards to the HSPN database and the interactive sections of its website, is dependent of the members' participation as regards to submission of incidents reports and other security related information." The organizations commit themselves to the following points, among others:

       • "To recognise the indispensable character of the sharing of information with a view to the orderly management of safety problems.
       • To inform the Humanitarian Safety Network of any incident with a direct or indirect bearing on the organization's staff or actions in the forms determined, ensuring full transparency. They shall likewise pass on any information, which might be damaging to the integrity of their staff.²⁶"

       While the need for incident analysis is evident, the project has met with some practical difficulties. One indicated example is that the primary material from the field is not forthcoming as field workers often claim they do not have the time for filling in the forms and sending them to the database. Similarly, they need the kind of analysis that the HSPN might generate at a speed and of a quality and quantity to which the HSPN is not presently geared. Among other things, this is said to be due to limited HQ capacity, the limited number of members, and the quantity of the material they may produce not being sufficiently representative.

       One solution the Network is considering is to use the format for incident analysis developed by the Network locally or regionally, i.e. closer to the field. The Network is presently studying various alternatives for meeting the need for grouping relevant security information and for trend analysis.

       d. An EU Commission Working Document on Security of Relief Workers and humanitarian space from May 1998 ²⁷calls for donor support for a strengthening of inter-agency security capacity: "Action confined to the level of individual agencies is unlikely to be sufficient to deal with systemic and situational aspects of security. Umbrella organisations must therefore take on a responsibility for developing or disseminating standards, tools, guidelines and so forth…Donors have an important role to play in stimulating this process."

       The document also calls for active donor input in the process of reviewing the UN security system, which should lead to the improvement of security for both UN and non-UN humanitarian workers.

       e. Operational Security Management in Violent Environments by Koenraad van Brabant on security management good practices has recently been published by Overseas Development Institute, London ²⁸. It repeatedly mentions the need to collaborate on security but does not go into details about possible models for such collaboration.

       In a subsequent review of aid agency practices on organizational security management van Brabant addresses the issue of inter-agency relations on security. Having stated that "a fair degree of objective inter-dependence" exists between agencies he notes that his research has made it clear that there is a "significant scope for, and important cost-benefits to be had, from much more inter-agency collaboration" in security management.²⁹

       He goes on to state: "Yet the most emphatic message of aid organizations is their insistence on retaining full autonomy in security management." He finds that "acceptable, and legitimate inasmuch as they have the formal responsibility. But it should not blind people to areas where collaboration is possible and even actively required" (Chapter F).

    9. Specific areas of collaboration: Codes of Conduct
       1. The Code of Conduct for the International Red Cross and Red Crescent Movement and Non-Governmental Organizations (NGOs) in Disaster Relief ³⁰ is an attempt at codifying guidelines for humanitarian actors' ethical conduct in field operations. There are no compliance mechanisms. It is increasingly used either in its original form or adapted to specific field situations. It is in itself an inter-agency collaborative effort, and contains in its paragraph 6 the following commitment:

          "We will place a high priority on the proper co-ordination of our emergency responses. This is best done within the countries concerned by those most directly involved in the relief operations, and should include representatives of the relevant UN bodies."

          The assumption one might make is that a shared code of conduct would influence the perception of humanitarian actors positively towards greater "acceptance" and thereby fewer security risks.

       2. People in Aid is another inter-agency initiative which, while concerned with a code of practice for humanitarian assistance generally, also includes a component of security collaboration. Article 7 of its Seven Principles For Humanitarian Agencies says, "We take all reasonable steps to ensure staff security and well-being." This is further explained: "We recognize that the work of relief and development agencies often places great demands on staff in conditions of complexity and risk. We take all reasonable steps to ensure the security and well-being of staff and their families." ³¹

3RD DAY - MORNING, DECEMBER 8, 2000
LESSONS LEARNT FROM THE DRAFTING PROCESS AND
CONTINUATION OF THE PROCESS

1. Practical, field oriented collaborative mechanisms

   The meeting of the IASC Security Task Force on May 23 in Geneva suggested that the proposed consultations, as part of a broad consultative process, discuss what might be the "characteristics of a baseline document, to which all partners are agreeable and which can be "tailored" for specific country/field situations".

   These "characteristics" were seen as practical, informal, i.e. not legally binding, mechanisms that would respond to the perceived need for greater inter-agency collaboration on security at the field level. They would be complementary to existing more general and/or more formal collaboration agreements or arrangements.

   The participants in the present consultation are invited to discuss, envision and perhaps begin to formulate the characteristics of the envisaged collaborative mechanisms. It might help the envisioning process to address some of the following issues, adding others that are deemed necessary.

2. Who wants the collaborative mechanisms?

   The impetus for enhanced security for humanitarian workers, and efforts to improve their security through inter-agency collaboration, is concern for the needs of the populations to which the humanitarian actors are responding. If humanitarian actors' security is threatened so that they cannot perform their task, the lives of those populations become even more precarious.

   All actors, UN, non-UN, local agencies, local authorities etc, express concern for security and for improving it through a variety of means, one of them being improved inter-agency collaboration. UN Staff Councils or associations³² , for example, have pressed for a higher priority, i.e. more resources, for security, including security for implementing partners.

   Donors also have vested interests in collaboration, in the area of security as well as in other areas of humanitarian assistance, both to save lives and to protect material values. At the level of the ACC the need for UN-NGO security collaboration has been expressed recently, and it also appears in the UN Secretary-General's report on safety and security of United Nations personnel³³ .

   In the context of the IASC Working Group, the need for mechanisms for security collaboration in the field has especially been emphasized by non-UN actors.

3. What are their characteristics?

   The following characteristics of the envisaged collaborative mechanisms seem to emerge:

   • Local application of general mechanisms;
   • Flexible application, either fully or with reservations ("menu of options");
   • Immediately applicable in the field;
   • Informal, not legally binding, declaration of intents;
   • Based on partnership and trust;
   • Ethical code of collaborative conduct ("collaborative ethos");
   • Voluntary;
   • Transparent roles and responsibilities;
   • Complementary roles and responsibilities;
   • Complementary to other collaborative security arrangements.

   The collaborative mechanisms could address these - and no doubt other - practical issues:

   • Information exchange;
   • Security training;
   • Analysis and strategy development;
   • Security plans;
   • Human and material resources and services;
   • Cost sharing;
   • Security-specific selection criteria for humanitarian workers ³⁴.

4. What kind of mechanisms are envisaged?

   The term "Framework Agreement " has sometimes been used to describe the expected outcome of the IASC-WG consultative process on staff security. However, this is felt by some to connote legal implications, which would denounce the basic pragmatic and contextual character of collaborative mechanisms based on partnership, trust and ethical obligations. Some have suggested the term "agreement framework", others a "declaration of intent", or a "security code of conduct."

   Whatever the term, the function of the envisaged collaborative arrangements is generally seen as providing flexible mechanisms to be used and adapted to specific security situations by parties which have agreed to them. Thus, they serve as a basis for making context-specific, field-related agreements.

   The term "framework agreement" is inspired by the UNHCR-NGO model. This model, or a modified version of it, might be adopted by UN agencies and their partners (operational and others) in particular field locations. Whether or not it would be a separate, multilateral arrangement for all those interested to sign it, or an integrated element of bilateral agreements (such as operational (sub)agreements between a UN agency and an implementing partner) would perhaps depend on each particular field context and the degree of coherence of its special composition of humanitarian actors.

5. Should the collaborative mechanisms be legally binding?

   The assumption so far has been that the collaborative mechanisms would be of an informal nature, and would not be legally binding for the participants but would rely on trust and transparency.

   This need not prevent the parties from drawing other-than-legal conclusions vis-à-vis those humanitarian actors that do not comply with what has been agreed. That could be, for example, either an NGO or a UN agency ending an operational (sub)agreement, changes in personnel or management, exclusion from access to security material or, at least theoretically, exclusion from emergency evacuation. Practice often shows that in situations where life and death are at stake, most will be included irrespective of agreements or lack of them.

6. What is negotiable and what is not?

   Prior to any attempt at creating inter-agency collaborative mechanisms it is necessary to list which elements are negotiable and which ones are not. Thus, a non-changing element is the binding character of the UN security structure for UN agencies. This structure may still undergo changes in the wake of the UN Secretary-General's report on staff safety and security, but its binding character does not.

   Equally non-negotiable is, for example, the unique mandate of the International Red Cross and Red Crescent Movement to impartiality, neutrality and independence, and the limits this sets to any close association with other parties, while still allowing a degree of informal consultation.

   For NGOs their non-governmental status is non-negotiable³⁵ , in the sense that their identity is primarily related to their own global mandate and local mission ³⁶. They can however voluntarily and on a case-by-case basis choose to enter into agreements, partnerships, consortia or other forms of organizational collaboration, for example, related to particular field situations, or to particular concerns, such as security. A signed document, whichever form it may have, may help in making the partnership to which the parties have committed themselves both transparent and consistent.

   4 December 2000/AP. Rev. 4 January 2001/AP

   ---

   APPENDICES / ATTACHMENTS

   1. Report of the Secretary General on Safety and Security of United Nations personnel, 18 October 2000 (A/55/494). (To be distributed).
   2. Memorandum of Understanding between the United Nations and IGO/NGO implementing partners Regarding Coordination of Security Arrangements. United Nations Security Coordinator, Security Directive. Inclusion of International Staff of International and Nongovernmental Organizations in United Nations Security Arrangements, SD/1996/2 of 16 January 1996 and Addendum 1 of 16 December 1997. (To be distributed).
   3. Framework Agreement for Operational Partnership between the United Nations High Commissioner for Refugees and [Non Governmental Organizations] (14 November 2000).
   4. Code of Conduct for the International Red Cross and Red Crescent Movement and Non-Governmental Organizations (NGOs) in Disaster Relief.
   5. Code of Conduct for Humanitarian Agencies in Sierra Leone. Revised 21st November 1998.
   6. UNSECOORD: UN Civilian Staff Killed 1 Jan 92 - 18 Sep 00. Nationality of staff killed by Region.
   7. Statement by Ms. Catherine Bertini, Executive Director, World Food Programme, to the United Nations Security Programme Open Debate on Security of UN Humanitarian and Associated Personnel, 9 February 2000. (To be distributed).
   8. Joel McClellan on behalf of a number of NGOs, Statement on Field Security made to the UNHCR Standing Committee (Geneva, 27 September 2000). (To be distributed).
   9. Steering Committee for Humanitarian Response, Some NGO Views on the Humanitarian Implications of Implementing the Brahimi Report (7 November 2000).
   10. Koenraad van Brabant, Mainstreaming the Organisational Management of Safety and Security: A review of aid agency practices and a guide for management (ODI Humanitarian Policy Group, November 2000). Chapter F.
   11. Randolph Martin, Senior Director of Operations, International Rescue Committee, A More Proactive United Nations Role in the Security of NGO Staff? (October 23, 2000).
   12. Draft Questionnaire for Cases of Field Security Collaboration.

   ---

   Notes

   1. Report of the Secretary-General on Safety and Security of United Nations personnel, 18 October 2000 (A/55/494).

   2. According to WFP, since 1994, there have been 59 incidents of kidnapping and hostage taking affecting 228 UN staff. In 1999 alone, there were 292 violent robberies, assaults, rapes and vehicle hijackings, cf. statement by Mme Catherine Bertini, Executive Director, WFP, to the UN Security Council, 9 February 2000.
   3. Koenraad Van Brabant, Operational Security Management in Violent Environments (ODI, June 2000).
   4. United Nations Security Coordinator, Security Directive. Inclusion of International Staff of Intergovernmental and Nongovernmental Organizations in United nations Security Arrangements, (16 January 1996), SD/1996/2 and addendum 1 of 16 December 1997.
   5. Statement on Field Security made to the UNHCR Standing Committee by Joel McClellan on behalf of a number of NGOs (Geneva, 27September 2000).
   6. See the discussion in Koenraad Van Brabant, Operational Security Management in Violent Environments (ODI, June 2000).
   7. See for example Koenraad Van Brabant, Operational Security Management in Violent Environments (ODI, June 2000).
   8. According to Randolph Martin, IRC, NGOs have signed the MoU at country level in 14 cases. See his article on UNHCR-NGO collaboration from October 2000, A More Proactive United Nations Role in the Security of NGO staff? (to be published in Forced Migration Review, December 2000).
   9. One may imagine these differences in agency policy in a particular field situation as a number of circles, each with its own security focal point. The largest would normally be the one where UN and NGOs have found a mechanism for collaboration, which is then seen as decreasing in intensity: from collaboration, through consultation to communication. Other separate circles with their own security focal point (e.g. the Red Cross/Red Crescent Movement and MSF) might still touch the UN-NGO circle in the consultation or communication areas. Outside these circles, without touching them, are what may be termed the "in-communicados."
   10. SCHR, Some NGO Views on the Humanitarian Implications of Implementing the Brahimi Report, 7 November 2000.
   11. See the discussion in Koenraad Van Brabant, Operational Security Management in Violent Environments (ODI, June 2000).
   12. See for example the Danish Refugee Council Field Handbook chapter 7: "It is the policy of DRC to co-operate and share information with other relevant organisations both in the field and at headquarters level. It is in the interest of all organisations that close co-operation on security matters is maintained both at field and headquarters level…Therefore DRC staff responsible for security…should work actively for setting up common standards for security" (draft, September 2000).
   13. See for example the UN Secretary-General's report on Safety and security of United Nations personnel, 18 October 2000, A/55/494, section 16.
   14. The organizations behind the Code of Conduct use the term NGHAs, Non-Governmental Humanitarian Agencies. The term encompasses both the Red Cross Movement and NGOs. NGOs are defined as organizations, both national and international, which are constituted as separate from the government of the country in which they are founded.
   15. Koenraad van Brabant, Mainstreaming the Organisational Management of Safety and Security: A review of aid agency practices and a guide for management (ODI Humanitarian Policy Group, November 2000).
   16. See also statement by Ms Catherine Bertini, Executive Director of WFP, to the UN Security Council open debate on security of UN humanitarian and associated personnel, 9 February 2000: "When is the security risk for our staff so great that we cannot reach the victims of war - who then die for lack of food, shelter, water and medicine. Where do you draw the line?"
   17. A/55/494
   18. Memorandum of Understanding between the United Nations and IGO/NGO implementing partners Regarding Coordination of Security Arrangements. United Nations Security Coordinator, Security Directive. Inclusion of International Staff of International and Nongovernmental Organizations in United Nations Security Arrangements, SD/1996/2 of 16 January 1996 and Addendum 1 of 16 December 1997.
   19. Randolph Martin, Senior Director of Operations, IRC, A More Proactive United Nations' Role in the Security of NGO Staff? (October 2000, to be published December 2000 in Forced Migration Review).
   20. This may be the case for example when, perhaps exceptionally, UN personnel is evacuated by a non-UN humanitarian agency, as in the case of the IOM organized and pre-paid evacuation from Kupang, West Timor.
   21. Framework Agreement for Operational Partnership between the United Nations High Commissioner for Refugees and [Non Governmental Organizations] (14 November 2000).
   22. This is an internal arrangement that does not carry any responsibility within the official UN security system.
   23. Code of Conduct for Humanitarian Agencies in Sierra Leone. Revised 21st November 1998.
   25. Koenraad van Brabant, Mainstreaming the Organisational Management of Safety and Security: A review of agency practices and a guide for management (Unpublished, London, Overseas Development Institute, Humanitarian Policy Group, November 2000), chapter F.
   26. Cf. WFP's agency-wide security training programme for all WFP employees, which in 11 months trained about 5,400 staff, cf. WFP information, February 2000.
   27. HSPN Project Internal Regulations.
   29. Commission of the European Communities, Commission Working Document on Security of relief workers and humanitarian space, Brussels, 14.05.1998, SEC(1998) 797 final.
   30. Koenraad van Brabant, Operational SecurityManagement in Violent Environments (ODI Humanitarian Practice Network, Good Practice Review, June 2000).
   31. Koenraad van Brabant, Mainstreaming the Organisational Management of Safety and Security: A review of agency practices and a guide for management (Unpublished, London, Overseas Development Institute, Humanitarian Policy Group, November 2000).
   32. Code of Conduct for the International Red Cross and Red Crescent Movement and Non-Governmental Organizations (NGOs) in Disaster Relief.
   33. People in Aid, Statement of Principles.
   34. Interview with Association for the Security and Independence of International Civil Servants (ASIICS).
   35. Report of the Secretary-General, Safety and security of United Nations personnel, 18 October 2000 (A/55/494).
   36. Cf. For example Koenraad van Brabant, Mainstreaming the Organisational Management of Safety and Security: A review of agency practices and a guide for management (Unpublished, London, Overseas Development Institute, Humanitarian Policy Group, November 2000), on competencies of security officers: "There is a mainly informal 'debate' about whether security focal points/security officers should have a professional security background, ie be ex-military or ex-police. On the surface of it, this is one of the elements in the discussion about UN-NGO collaboration on security. Indeed, the UN makes it a formal requirement. Various NGO personnel have been sceptical if not outright critical of this. However most of this 'debate' is misplaced because it completely misses the point…" Among the real questions to ask, argues van Brabant, is this: "What specific competencies may have to be prioritised for a 'security officer' depending on the specifics of a field context?"
   37. See the definition used by the Red Cross-NGO Code of Conduct: "Organizations, both national and international, which are constituted separate from the government of the country in which they are founded."
   38. Randolph Martin, IRC Director of Operations, A more Proactive United Nations' Role in the Security of NGO staff? (October 23, 2000, to be published December 2000 in Forced Migration Review).